CA2123449C - Method and apparatus to speed up the path selection in a packet switching network - Google Patents
Method and apparatus to speed up the path selection in a packet switching networkInfo
- Publication number
- CA2123449C CA2123449C CA002123449A CA2123449A CA2123449C CA 2123449 C CA2123449 C CA 2123449C CA 002123449 A CA002123449 A CA 002123449A CA 2123449 A CA2123449 A CA 2123449A CA 2123449 C CA2123449 C CA 2123449C
- Authority
- CA
- Canada
- Prior art keywords
- network
- node
- nodes
- links
- routing
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Lifetime
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/02—Topology update or discovery
- H04L45/04—Interdomain routing, e.g. hierarchical routing
Abstract
Currently, the routing algorithms compute all the available paths in the network, from the source node to the destination node before to select the optimal route. No assumption is made on the network topology and the route computation is often time and resource consuming. Some paths which are calculated are not acceptable due to the particular geographical configuration of the network.
In the real world, large transport networks are not fully meshed. The present invention is based on the observation that networks are usually built around a hierarchical structure. A set of nodes, interconnected by high throughput lines, are used to build a 'Backbone' (401) with a high degree of meshing to allow the redundancy and reliability required by the user. The other nodes or 'local' nodes (404) are attached to one or several backbone nodes. It is the network designer responsibility, at the configuration time to define for each node what is its attribution: backbone (402) or local node (404). The list of the node attributions appears in the topology table (306) and is updated each time a node is added to or dropped from the network. The routing algorithm can take advantage of the particular network topology to drastically reduce the complexity of paths computation. For a given connection, only a limited number of nodes are eligible and are taken in account by the algorithm in the optimal route search. The object of the invention is to split the network in backbone and local nodes to speed up the path selection.
In the real world, large transport networks are not fully meshed. The present invention is based on the observation that networks are usually built around a hierarchical structure. A set of nodes, interconnected by high throughput lines, are used to build a 'Backbone' (401) with a high degree of meshing to allow the redundancy and reliability required by the user. The other nodes or 'local' nodes (404) are attached to one or several backbone nodes. It is the network designer responsibility, at the configuration time to define for each node what is its attribution: backbone (402) or local node (404). The list of the node attributions appears in the topology table (306) and is updated each time a node is added to or dropped from the network. The routing algorithm can take advantage of the particular network topology to drastically reduce the complexity of paths computation. For a given connection, only a limited number of nodes are eligible and are taken in account by the algorithm in the optimal route search. The object of the invention is to split the network in backbone and local nodes to speed up the path selection.
Description
::' '~23~9 ' ' ' I~IE,'I'IIOD ANI) ~I'I~RA'I'US '1'0 Sl'E~,I,I) Ul''l'IIE l'A'I'II SEl,r~"1'10N
IN A P~C5~1E'l' SWI'I'CIIINC NE'I'WORK
.
, Tec~mic(llJiel~l .
The present invention relates to a high speed transmission system in a large ;packet switching network, more particularly, to an apparatus and method for speeding up the path selection between a source and a destination node and specially a method to decompose the network into backbone and local nodes.
: Background ~IYt .
It has become increasingly useful to interconnect a plurality of data processingelements by means of a packet switching network in which data is transmitted as data assemblages called "packets". Such networks include a plurality of interconnected switching nodes which, in turn, are connected to end nodes supporting data processing equipments. Such packet ne~works can become quite large with an extensive geographical distribution. In such a situation, the selection of an efficient path between two end nodes which wish to eommuni-5'' cate with eaeh other becomes of paramount irnportance. Different methods are summarized by H. Nussbaumer in Telei~lJoimntiqlle ~1 ~pages ~2 to 117) Pr~sses Polytec~miql~es Romnndes 1987.
I
Pnth Selectio~l -I he general problem which must be solved at the network Icvcl is to rlnd a path between a source and a destination node. When the network is using datagrams, the path selection must be done individually for each packet. For virtual eireuits, the path decision is done onee only at the time of the cireuitestablishment. In both eases, the choice of a routing algorithm is not easy be-cause it must satisfy a large number of often conflicting requirements. This algorithm must be simple to implement to not complicate the nodes realization, ,, '';
~- rR993014 .
., - .
,, . .
. .
, ., ~, ~
,,... ~ , ~ .,. , :
IN A P~C5~1E'l' SWI'I'CIIINC NE'I'WORK
.
, Tec~mic(llJiel~l .
The present invention relates to a high speed transmission system in a large ;packet switching network, more particularly, to an apparatus and method for speeding up the path selection between a source and a destination node and specially a method to decompose the network into backbone and local nodes.
: Background ~IYt .
It has become increasingly useful to interconnect a plurality of data processingelements by means of a packet switching network in which data is transmitted as data assemblages called "packets". Such networks include a plurality of interconnected switching nodes which, in turn, are connected to end nodes supporting data processing equipments. Such packet ne~works can become quite large with an extensive geographical distribution. In such a situation, the selection of an efficient path between two end nodes which wish to eommuni-5'' cate with eaeh other becomes of paramount irnportance. Different methods are summarized by H. Nussbaumer in Telei~lJoimntiqlle ~1 ~pages ~2 to 117) Pr~sses Polytec~miql~es Romnndes 1987.
I
Pnth Selectio~l -I he general problem which must be solved at the network Icvcl is to rlnd a path between a source and a destination node. When the network is using datagrams, the path selection must be done individually for each packet. For virtual eireuits, the path decision is done onee only at the time of the cireuitestablishment. In both eases, the choice of a routing algorithm is not easy be-cause it must satisfy a large number of often conflicting requirements. This algorithm must be simple to implement to not complicate the nodes realization, ,, '';
~- rR993014 .
., - .
,, . .
. .
, ., ~, ~
,,... ~ , ~ .,. , :
2~1 23~'19 it must ensure a correct progress Or the r)ackets whatevcl the disturbance On the network. The algorithm must be able to provide satisfying results despite of the traffie variations and the network topology. It must also ensure the eq-uity between the different end users and provide an equal distribution of the rights to aceess the network. At last, the routing algorithm must allow, when possible, to exploit the network in an optimal way, according to a criterion which can vary with the utilization type. In most of the cases, the network is realized in a way to min;mize the paeket transit time and to transfer the max-imum number of paekets. The principal objectives are respectively the re-duction of the transit time and the increase of the throughput. In other cases, the objective is to decrease the communication cost, or to develop a reliable network able to operate correctly (without too sensible performance degrada-tion) either in case of catastrophic linc, node failure or peaks of traffic.
Beeause of the variety of the eonstraints, there are a large number of differentrouting types. Some eorrespond to deterministic or adaptive polieies according to their eapacity of integrating the traffie variations and the network topology.
Routing algorithms can be centralized if the paths are determined from a par-ticular node. Others are distributed between all nodes: this is favorable in a reliability point of view but the algorithm is rnore complicated and the path optimization is rnore difficult to implement. ~ome Algorithms are difricult to c]assify: they are using special techniques called Flooding Rollting or l~andom Routing.
. .
FIOOdillg ~OUting The Flooding Routing is based on a very simple principle consisting for each node to retransmit packets on all output links of the node, e~cept, but the input link. A node connected to K other nodes retransmit K-l eopies of the packet which has been just reeeived. This method allows the delivery to the destina-tion node of at least one eopy of the packet with one condition: it must exist at Ieast one path between the source to the destination node. This routing is ensured even when the network topology is changing for example after a cat-astrophie failure of some components in the network. The Flooding Routing ~';
~ IR993014 . . .
, .....
, .
,.~
.~.
:' 3 212~4~9 allows a very good robustness in thc nctwolk cxploitation~ Furthcrltlolc, as :
all possibles links between the source and the destination node are tried in an i~ exhaustive way, the method ensures that at Ieast one copy of the packet will reach the recipient through the shol-ter path with the minimum delay ir the network is not overloaded. This technique is simple to implement since neither routing tables to route the messages, nor knowledge of the geogràphical posi-:~ tion of the destination node in the network are required. The recipient must only know that the paeket is addressed to it.
i The price to pay ifor these qual;ties of robustness, simplicity and rapidity is a bad utilization of the network resources and a tendency to the congestion.
Usually, it is advantageous to place in the packet header a ~leld indicating the.7 maximum node number that a packet is authorized to pass. This ~leld is dec-remented each time the packet go through a node, and packets with an empty field are discarded.
~' :!, Ralldom or Stochastic Routing i,:
This technique, like the Flooding Routing, does not demand to tlle nodes to have the complete knowledge of the network structure and of the trafi~lc to takea routing decision at their own Ievel. However, here, to limit the generation ofa too important ghost traffic, the nodes avoid to sencl systematically over all ::
output links copies ~f the packets they receive. Thc Random Routing consists in emitting one or more counterparts of the received packet on output links selected at random. This method is also called Selective Flooding. This policy Ieads to a very simple routing at the node Ievel and limits the saturation of the network due to the packet duplication. In return, the average routing delay is longer than this resulting from deterministic techniques. Packets are taking sinuous routes instead of taking the most direct path toward the destination node.
. .
:; ..
i 1 i.~ I R 9 93 0 14 , .
, , .
;, ~ .
;, , . .
I, 2 1 2 3 ~
~ 4 . .
Determillislic Routing , Fixed Roll tin g .... .
The Fixed Routing def'ines paths that must takc thc differcnt packcts accord-ing to the gencral network characteristics such as the network topology and the '~ mean traffic expected on the communication lines. Thc routing rulcs are es-tablished once and aims for optimizing the performance criterion privileged by ~ the user. In most of the cases, the average packet transmission time through ~!, the nctwork must be minimized. Thc rnethod consists in building a routing ta-ble at the level of each node. Its consultation allows the nodc to determinc on ::' which output link the packet it has just received must be transfcrred. The Fixcd Routing is vcry simple in its principles. The only processing done in the ,.i nodes consists in queucs managcment and tables reading and the optimization ' algorithm is initiated only once at the network dcsign Ievel. The Fixed Routing . aims for a long term and a global optimization of the network, but in com-parison with thc random routing, this technique allows to consi(Jerably speed up thc packct transmission through the nctwork.
., , ~:' Adaptive Rotlting . .
Contrary to lhe Fixed Routing, the purpose of the Adaptive Routing is to satisfy at any time thc optimization critcrion. Thc tablcs arc permancntly up-dated according to, for examplc, thc instantancous statc of thc traffic on the lines.
,~, ~ Centralized Routing . .
When the characteristics of the network fluctuate, it is ~ossible to a(lapt the routing by assigning to one node the responsibility to updatc pcriod-ically the routing tables according to the traffic variations and the topology. This method, simple in its principles, is callcd Centralized Rout-ing. Its principal disadvantage is to generate an important auxiliary traffic and to subordinate the good functioning of the net~vork to only one node.
:
. I R 9 93 0 14 '''' 'i; .
, ~,...
;., ., . .
~ ~~23~9 s The differenl nodes must send to ~he rouling cen~cr the .~tate of the net-work as they perceive it (operational acljacent no(les, trarric .. ), and the routing center must, in its turn, disp.ltcll to the no(Jes their routing tables.Tlle auxiliary traffic is concentrated around the routing center and this has for result to saturate this part of the network. Furthermore, the Central-ized Routing can generate some problems at the time the tables are re-freshed because said tables cannot be received at the same time by all the nodes. A solution is to decentralize the tables establishment at the level of each node.
o Local Routing . .
The local and distributed routing methods are both based on the estab-Iishment by each node of its own routing table according to information locally collected. With the local routing technique, each node builds its ta-~ ble without exchanging information with the neighboring nodcs. In its most simple form, the method consists in placing the packet just received in tlle shorter output queue and in transmitting it as rapidly as possible. Thc lo-cal algorithm tends in its principle to circulate the packcts very quickly in the network. However, as the routes are selected in some way or other the '~ mean paths Iength is far to be minimal.
~ .
i .
'~ Q Distributed Routing The Distributed Routing is a local method in which neighboring nodes are exchanging messages concerning the trafric and the network conditi(>ll to . update their routing and delay tables.
ilicr.llcllical l~outing , .
': With the non Stochatic Routing methods, each node requires a table with as many rows as nodes in the network and a number of columns equal to ~' the number of output links. Furthermore, when the routing is adaptive, . nodes must exchange periodically messages to update their routing tables.
;., ~ IR993014 ..
. ., .:
~.2 ~ 2 ~
The tables size an(J the impollance of tlle conlrol messages are increa.sing rapidly with lhc number of no~les ancl become unacceptable when the net-work con~ains more than ten nodes. To solve this problem, the network is . .
divided in a hierarchy of sub networks as illustrate(l in Figure 0. In par-ticular, optimal clustering structures (003) are determined so as to mini-mize the length of the routing tables required. Said tables are taking into account only the nodes in a sub nctwork (0()3) and a packet a~klresse(l to a node in another sub network will necessarily go through the access node (005) of this rermote sub network. Packets transit through some mandatory .
nodes (005) and the price to pay for this simplification is that smaller routing tables give less precise routing information whieh then results in Ionger path lengths for the message traffic. Hierarchieal Routing proee-dures are partieularly effective for large packet switehed networks (001).
With smaller routing tables, they require less storage and proeessing in the nodes (002) as well as less eommunications overhead in the lines (Oû4).
The problem of the optimization of the hiera1chical levels has been stu(lie(l by L. Kleinrock and IF. I~amoun - Hicralcfiicnl RolltingJor l,a~ge IVetworks, Cornputer NetwoYks, Vol. 1, pp. 155-174, January 1977.
The main idea for reducing the routing table Iength is to keep, at any node, eomplete routing information about nodes which are close to it and Iess information about nodes located further away rrom it. This can be re-alized by providing one entry per (Iestination for the eloser nodes and one entry per set of destinations for the remote nodes.
The reduction of routing table Iength is achieved througll a hierarchical partitioning of the network. Basically, an m-level hierarcl-ieal elustelillg of a set of nodes (Figure 0) consists in grouping the nodes (002) into a Ist Ievel clusters (003), which in turn, are grouped into 2nd Ievel cluster e~c...
Tllis operation eontinues in a bottom up fashion. The mth level cluster is the highest level cluster (3rd level cluster) and as such it includes all the nodes of the network (001).
Unfortuncltely, the gains in table length are accompanied with an increase of the message path Iength in the network. This result is a degradation of the network performanee (delay, throughput) due to the excess internal . ,.
:;
-: .
.
J' , ~ , J
''''"
; : ~
~" 7 212.~449 trarric caused by longer pa~h Iength. I lowever, in vcry large networks, an ;; enormous table reduction may be achieved with no significant increase in ': net~ork path Iength.
High Pe~ forn&a1ace l~et~vorks , The data transmission is now evolving with a speci~lc focus on applications and by integrating a fundamental shift in the customer traffic profile. Driven by the growth of workstations, the local area networks (LAN) interconnection, the distributed processing between workstations and super computers, the new applications and the integration of various and often conflicting structures -hierarchieal versus peer to peer, wide (WAN) versus local (LAN) area net-works, voice versus data - the data pro~lle has become higher in bandwidth, bursting, non deterministic and requires rnore connectivity. Based on the above, it is clear that ~here is strong requirement for supporting distributed , ~'. computing applications across high speed networks that can earry LAN com-'- munications, voice, video, and traffie among ehannel attached hosts, business, .~ engineering workstations, terminals, and small to intermediate file servers.
This vision of a high speed multiprotocol network is the driver for the emer-gence of fast packet switching networks architectures in which data, voice, and video information is digitally encoded, chopped into small packets and trans-' mitted through a common set of nodes and links.
., (~
Throllghput ~;~; The key requirement of these new architectures is to reduce the end-to-end -!J delay in order to satisfy real t;me delivery cons~raints and to achicve the nec-~';. essary high nodal throughput for the transport of voice and video. Increases o in link speeds have not been matched by proportionate increases in the proc-. essing speeds of communication nodes and the fundamental challenge for high speed networks is to minimize the packet processing time within each node.
r ~
~R 9 93 014 : ,;
~ .
;1'.!, ''.:i ~-~ 2123~'3 R
As example, for meeting a typical 1()() ms (Ielay to ~Ieliver a voice r)acket be-tWCCIl two end users:
:
o A lotal of 36 ms might be nee(led rO, the paclcelizalion and play-out func-tions at the end points.
. ~
o About 20 ms is the unalterahle propagation delay needed, say, to cross the i United States.
,, There remains 44 ms for all the intra-node processing time as the packet moves through the network. In a ~ nodes network, each node would have about 8 ms for all processing time ineluding any queueing time. In a 10 nodes network, each node would have about 4 ms.
Another way of looking the same constraint is illustrated in Figure 1: taking a node with an effective processing rate of I MIPS (Millions of Instructions Per Second), it is possible to rlll a 9.6 kbps line with 1000 byte packets evcn if a network node must execute ~33 000 instructions pcr packct processc(l. For a 64 kbps line the node can afford 125 000 instructions per packet. In order to fill an OC24 link, however, our 1 MIPS node could only exeeute 7 instructions per packet. In the latter case even an effective rate of 10-30 MIPS would allow only 70-200 instructions per packet.
In order to minimize the processing time and to take full advantage of the high speed/low error rate technologies, most of the transport and control functions provided by the ncw high bandwidth network architectures are performed on an end-to-end basis. The flow control and particularly the path selcction are managed by the access points of the network which reduces both ~he awareness and the function of the intermediate nodes.
.,'~
;, , . . .
Rollting Mlodes ..
- The routing within the network presents two aspects:
. .
:,, ' 1. Determining what the route for a given connection shall be, ~' 2. Actually switching the packet within a switching node.
, ~, ~,;"
.. I~R 9 93 014 .
' , .
. .
. ,.
:~i ,:'.! 2 ~ ~
., ., 1 here are many methods of (Ie~erminin~ a rOu~e throu~h a ne~volk. For very high throughput, the eritical item is that the switehing elements must be able to decide where to route an incoming packet in a very sh()rt portion of time.
As described in the document l~ ll Speed Ne~ oskillg T~chnolo~y"~ln Intro-~luc~ory S~lrvey (palges ~8 to g6) - CC24-3816-0() ITSC Raleigll Marc~l 19~2, the routing modes, the most widely used, are the Source Routing and the Label Swapping.
~, ,~ .
~i !".
~' ~ Source Routing .: -, !.
The Source Routing is a particular implementation of the Distributed Routing for eonnectionless networks. The souree node (or aecess node) is responsible for ealeulating the route the paeket must take through the net-work. A routing field is appended to every paeket sent and that field is used by the intermediate nodes to direet the paeket towards its destination.
In this teehnique, the sending node must know the network topology and it must be able to find the optimal route. ~ut onee the route is determined, intermediate nodes do not need to refer to any tables or parameters to ....
make the routing deeision. The next stage of the route is right there in the paeket header. A drawback of this metho(l is that the routing reld in lhe packet header takes some storage and is an overhead. But this is quite small and the benerlts of being able to make a fast routing decision out-weigh the small increase in bandwidth overhead. Since all routing infor-mation is contained in eacll packet a ~set up is not required along the patll to provide an end to end communieation. Thus, data applications ~vhich ., benefit from a datagram service ean be effeetively supl-orted by this teeh-nique. However, the datagram traffic is ehalacterize(l by the fact that each datagram can be viewed as a connection and requires the computation of a path. Each packet is routed through the network as a self containcd unit and is independent of all other packets. The key point for the source node is to determine for each datagram the optimal route in the shorter lapse of time.
.
,:
; t:? ~R. 9 93 014 !~ 1 ., ~'' '~
~,':' .
' ;.' .':
~ :' '"'~' . ~ Label Swar)ping . .
Ihe Label Sv~appin~, is a r)articul.ll im~-lelllcllt.ltion Of Ihe L)istlibu~cd Routing for connection oriented networks. These networks typically multi-plex many connections (or sessions) on a link using some form of logical "channel". Each packet sent on the link has a header which includes an ,, arbitrary number identifying which logical connection that this packet be-longs to. For example, each packet can contain in its routing field a label that will be used by the intermediate nodes to identify the transmission link , the packet should be routed over. When a packet arrives at a no(le, the label is extracted from its header, and used to index a routing table that provides a new label along with a link address. The new label is over writ-ten on the old label, and the packet is forwarded using the link address.
The information in the routing table is updated at the connection set up time. Each node on a path selects a label for each direction of the con-nection, and sends it to the neighboring node.
Tlle call set up and the resource reservatioll plOCCSS COIIlpliSCs tllC follow-ing steps:
.
~'~ a A connectiorl reqllest is speci~led by the user via a ~set of p.lr.lmeter.s including origin and destination network address and data flow char-~' acteristics, :
The p~th determinatioll is realized by the source node from its local routing topology database . . , ~ ~ The route reserv~tion is sent out in a special message along the specified - nodes. Said message, which is usually routed according to the previ-i ously described Source Routing technique, signals the nodes to set up their connection tables and to reselve their resourccs to providc the Ievel of service required by the traffic type (for examplc, a bandwidth , .
~~~ reservation on each of the links of the path).
"
~ The tables upd.ate reflects the changes in the availability of the network '~' resources. The information is sent to every node in the network.
:
, . .
I:R 9 93 014 . .
~. ~
:~, :
: .
-~ 11 212~4~L!~
The Labcl Swapping tecllnique re(!uires ~hclt the conneclion tables be set up and main~aine(l dynamically. Tha~ mcans ~hat whcn a new connec~ion is established or an okl one is terminate~l the tables are updated (~he da-tabase Of network ~opology can of course be maintaine(l quile separately).
This is possible if the connection sct up is rclatively infrequcnt and is not too time critical. This restriction makes datagram transport quite ineffi-cient. Howcver, connections that requirc vcry short packcts, like real-timc voice conncctions, can bc cffectively supportcd by this tcchnique, bccause of the low packet overhead. Once the connection is cstablishcd, thcre is no need to place a destination address in the headcr cvcry time a packet is , .
sent. All is needed is a rlcld to specify which conncction is to be uscd for this paekct.
.,..~
, .
~,.;
Path Selection Criterion Onc of thc major critcrion for sclecting paths bctwccn nodcs in packct net-works is the minimum hop count and minimum path Iength. The hop count is the number of links used to construct the path betwecn the two cnd nodes.
The path length is a function of the ovcrall transmission delay imposed by the path between the two end nodes. In most high speed networks, the delay (path Iength) is not a major consideration since the worst-case delay through such networks is nearly always acceptable. Thc hop count~ howcvcr, is a dircct measure of the amount of resources rcquired to implcment a givcn path and hencc is of considcrable importancc in selecting paths. It is to bc noted that aselected path need not be a minimum hop count path since congestion on the nctwork links may force the choicc of a larger hop count path. However, such longer alternate paths cannot be allowcd to grow without limit since inordinate amounts of network rcsources might be committed to one path, resulting in further congestion for other paths and forcing yet longer hop count paths to be selected for yct other connections. The long term network throughput could thereby be adversely affccted. The problem, then, is to select a path between an origin node and a destination node which has a minimum hop count, a ., .
~R 9 93 014 ;.~
i:,~, ",,~
i ~
. . .
: ., ~. .
~.. . . .. . . .. . .
~ 1 2 2 1 2 3 ~1 4 9 minimum path Iength, which dl)es no~ utilize an inor(lillate am()unt of network resou rces.
:
.:
~ Key requirements .:, The Distributed Routing mechanism in large and high speed packed switching networks supporting both connection oriented and connectionless routing modes implies some requirements in terms of performance and resource con-~~sumption which can be summarized as follows:
.~
-. ~ the source node (or the node providing the route calculation for the source~' node) must be able to decide where to route an incoming packet in a very ' short portion of time (the computation must be sufficiently rapid to com-pute an optimum path for each connection request) ~', 4 the switching time in the intermediate nodes must be minimized, (minimum processing time) ~. 0 the network resources along the selected path must be optimized according to the criterion of the minimum hop count.
o Control messages must be as much as possible limited not to overload the network.
:
In high speed networks the cost for updating the routing tables generated by the exchange of control messages is not critical so long as this can be performed before the connection set up. The very high line throughput authorizes the interehanging of routing inrormation without creatillg a signirleant burcJen on the eomrnunication lines themselves. These considerations sugges~ a better ac-eess and use of the routing tables maintained within the nodes without saeri-ficing the criterion Or optimal path contrary to the Hierarchieal Routing method proposed by L. Kleinroek and F. Kamoun.
,, .~Summary ~ the invcntion ""
- ., ;.Currently, the routing algorithms eompute all the available paths in the net-~-work, from the souree node to the destination node before to select the optimal .;
~I R 9 93 014 ' ';'~
..
i : ~ , . . .
2123~49 rOu~e. No as~umption is made on the nelwork topolo~y and ~he route cOmpU-tation i.s often time and resouree consuming. For example, some paths wllicll arc calcukltcd are no~ acceptable due to the particular geographical conl'igllra-tion of the network.
In the real world, large transport networks are not fully meshed. The present invention is based on the simple observation that networks are u.sually built around a hierarchical structure. A set of nodes, interconnected by high throughput lines, are used to build a 'Backbone' with a high degree of meshing to allow the redundancy and reliability required by the user. The other nodes or 'local' nodes are attached to one or several backbone nodes. It is the net-work designer responsibility, at the configuration time to define for each node what is its attribution: backbone or local node. The list of the node attri-butions appears in the topology table and is updated each time a node is added to or dropped from the network. The routing algorithm can take advantage of the particular network topology to drastically reduce the eomplexity of pathscomputation. For a given connection, only a limited number of nocles are eli-gible and are taken in account by the algorithm in the optimal route search.
The object of the invention is to split the network in backbone and local nodes to speed up the path selection.
Descri,~tion of the clrawin~,~s Figure O represent a network topology for Hierarcllical Routing according to the method sllggested in the prior art by L. Kleinrock an(J F. Kamoun ,, Figure I sllows the processing times (or numbel Of instluc~iolls pCI secs)nd) required in function of the different line throughpu~s supported by the high speed networks.
., Figure 2 shows a typical model of high speed packet switching network in-cluding the access and transit nodes claimed in the present invention.
''.'.', ., Figure 3 describes a high speed Routing Point according to the present in-- vention r '.
:~ ' ' . ,.
~.',' .
''"
' :
Il 2 123A49 .,.
Figurc 4 illustrates the decomposition of the network in a plurality of back-bone no(lcs and local nodes accolding to the prcscnt invcntion.
'.'.
~- Figure 5 illustrates the addition in the nctwork of ncw nodes and links and their classirlcation in the topology database according to the present inventiom ~:' F;igure 6 reprcsents the optimum path trce of node A
''' Figure 7 represents the optimum path tree of node A according to thc present invention .
, Figure 8 summarizes the path selection methodology according to the present :
invention.
;
Dc.~cription of tfie prcfcrY~d cmbo(liol~nt oJ t~lC~ inven~io~l ., :
As illustratcd in Figure 2, a typical model of communication system is made of several user networks (212) communicating through a high performance network (200) using private lines, carrier provided services, or public data nctworks. Each user nctwork can bc dcscribed as a ~sct of communication processors and links (211) intcrconnceting large computers used as cnterprise servers (213), user groups using workstations or personnel computers attached on LAN (Local Area Networks 214), applications servers (215), PBX ~Privatc Branch cXchangc 216) Or vi<leo servers (217). Tllese user nctwolks, dispelse(l in differcnt establishmcnts, nccd to bc interconnected through wide area transport facilities and different approaches can be used ~or organizing thc , data transfer. Some architect-lres involve the checking for data integrity at - cach network node, thus slowing clown the transmission. Others are cssentially looking for a high speed data transfer and to that end the transmission, routing;~ and switching techniqucs within thc nodcs are optimized to process the flowing packets towards their final destination at the highest possible rate. The present ~'' invention bclongs essentially to the latter category and more particularly to the ~.
:
, ., :
,~
~, .;:
~'', ,~ ' 15 2~23~
~,~; ra~St packet switclling nelwork architectule detaile(J in the following para-graphs.
:
IYigh speedpacket s~vitching ~tet)vo~ks .
The general view in Figure 2 shows a fast packet switching transmission sys-tem comprising eight nodes (201 to 208) each node being interconnected by means of high speed communication lines called Trunks (209). The access (210) to the high speed network by the users is realized through Access Nodes (202 to 205) located at the periphery. These Access Nodes comprise one or more Ports, each one providing an access point for attaching external devices supporting standard interfaces to the network and performing the conversions required to transport the users data flow across the network from and to other external devices. As example, the Access Node 202 interfaces respectively a Private Branch eXchange (P13X), an application server and a hub through three Ports and communicates through the network by means of lhe a(Jjacent Transit Nodes 201, 208 and 205.
,.
S~vitchillg No~les , ~s Each network node (201 to 208) includes a Routing Point where the incoming data packets are selectively routed on the outgoing Trunks to-vards the neigh-boring Transit Nodes. Such routing decisions are made according to the in-;~ formation contained in the hea(ier o~ the data packets. In additioll to the basic packet routing function, the network nodes also provide ancillary services such ~' as:
~ . .., the determination of routing paths for packets originated in the node, directory services like retrieving and updating information about network ~i ~ users and resources, the maintaining of a consistent view of the physical network topology, in-' ~ cluding link utilization information, and ... .
.
.
'':.
,'',~.',' .i .
. ~,i . . .
. ., ~: .
-, ~
r ' 2123~1~L9 ~ thc rc~crvation Or rcsourccs al acccss ilOillts 01' l.llc nct\vork.
' Each Port is connected to a plurality of u~er processing equipments, each user equipment comprising either a soulce of di~ital data to be transmitted to an-other user system, or a data sink for consuming digital data received rrom an-other user system, or, typically, both. Tlle interpretation of the users protocols~
~. the translation of the users data into packe~s formatted appropriately for their - transmission on the packet network (200) and the generation of a header to route these packets are executed by an Access Agent running in the Port. This ~ header is made of Control and Routing Fiekls.
, . .
o The Routing Fiekls contain all the information necessary to route the packet through the network (200) to the destination node to which it is ". addresscd.
.', 9 The Control Fields include, among other things, an encoded identirlcationof the protocol to be used in interpreting the Routing Field (Source Rout-ing, Labcl Swapping, ....).
~, llti~ltg Poi1tts Figure 3 shows a general block diagram of a typical Routing Point (300) such as it can be found in the network nodes (201 to 208) illustrated in Figure 2. A
Routing Point comprises a high speed packet Switch (302) onto which packets . arriving at the Routing Point are entered. Such packets are received:
., 0 from other nodes over high speed transmission links ~303) via Trunk ;~ Adapters (304).
from users via appiication adapters callc i Ports (301).
Using information in the packet header, the adapters (304, 301) determine which packets are to be routed by means of the Switch (302) towards a local user network ~307) or towards a transmission link (303) Ieaving the node. The adapters (301 and 304) include queuing circuits for queuing packets prior to .' or subsequent to their launch on the Switch (302).
,.
i ..
~ rR993014 ::~
:.
; :
.'. ', "' .
., ~
~ i .
:
17 21234~9 The Route Controller (305) calculates the optimum routes through the net-work (200) S(3 as to satixry a given set of (lualiLy of service specifie(l by the user, and to minimize the amount of network resources use(l to complete a commu-nication path. Then, it builds the header of the packets generated in the Routing Point. The optimization criterion includes the number of intermediate nodes, the characteristics of the connection request, the capabilities and the utilization of the Trunks in the path ... All the information necessary for the routing, about the nodes and transmission links connected to the no(Jes, are contained in a Network Topology Database (3~)6). Under stea~y state condi-tions, every Routing Point has the same view of the network. The network topology information is updated when new links are activated, new nodes added to the network, when links or nocles are (Iropped or when link loa(ls change signifieantly. Such information is originated at the network node to whieh the resources are attaehed and is exehanged by means of eontrol mes-sages with all other Route Controllers to provide the up-to-date topological information needed for route ealculation (such database up(lates are carried on paekets very similar to the data paekets exehanged between end users of the network). The fact that the network topology is kept current in every node through continuous updates allow~s dynamic ne~work reconfigulaLions ~vitho disrupting end users logical connections (sessions).
The incoming transmission links to the packet Routing Point may comprise links from external devices in the local user networks (210) or links (Trunks) ,., from adjacent network nodes (209). In any case, the Routing Point operates in ~~ the same manner to receive each data packet an(l forwar(l it on to another ;~ Routing Point as dietated by the information in the packet hea(ler. The fast ~' packet switching network operates to enable a communicati()n betwcen any two end user applications without dedicating any transmission or node faeili-ties to that communieation path except for the duration of a single packet. In ... ~
this way, the utilization of the communication facilities of the packet network ~0u is optimized to carry significantly more lraffic than would be possible wilh dedicated transmission links for each communication path.
:
.~ ~
, . . .
:
1~ 993014 :, -~ ,.
., ,:
.
''' ~~'' 18 2123~/19 ~. .
Net~vc))k D~c~--7positio~t . . .
Modified Bellmann-Ford Algoritllm There are several possible algorithms to compute an optimal route in a net-work. The modified Bellmann-Ford algorithm is one of the most currently used in fast packet packet switching networks. As described in European Patent Application 9348()030.1 - MethcJds and Apparafusfor Optimllm Pa~h Se~ection in Packet Transmission Networks, this one is invoked for each connection with the following parameters:
~ the source node address, ~ the destination node address, ~ thc quality of scrvice required for thc connection. For examplc:
~ maximum call set up dcl~y (vcry important paramcter for connec~ions ....
serviccd in real time), . ~ maxirnum end to cnd transit dclay, ~ maximum information loss, crror probability, ~ ...
Thc algorithm uscs the local copy,in the sourcc node (or acccss nodc), of thc network topology to determinc the minimum hop and path lcngtll with the dcstinatil)n nodc ~hich mcets the ~luality of ~servicc re(luilcmellts. As men-tioned, the modi~led Bellman-Ford algorithm makes no assumption on the nctwork geographical con~lguration and it requires the samc complc~ity whcther the network is fully meshcd or hicrarchized. The purposc of the pres-cnt invention is, for a gi~en connection, to simplify the nctwork topology by reducing thc number of eligible nodes for the path calculation.
.;.
. ~
.
' '' ., :
:~ ~
~ 19 2123~49 Algoritlllll Comple.~ity s The total hop count neeessary from a givcn no(le to reach all thc other nodes gives a measure of the complexity C of the pa~h ~selection accor(ling to the minimum hop criterion. C is proportional to the average number of links I per node, to the total number of nodes N in the network and to the average num-ber of optimal hops.
.
C = N x O
1 . 1 (N-1)/ N) + (first hop) .. 1 i 2 . 1 (N-1)/ N) + (second hop) 3 . 1 (N-1)/ N) + (third hop) t i.:'~';~ i : (i+1). 1 (N-1)/ N) (i+1 th hop) . . . ~
:, . C = hopt . N . 1 = hopt . N . 2.L/N = hopt . 2L with :
.
hopt = average hop count for an optimal path average number of links per node ~'. N = total number of nodes in the network . L = total number of links in the network , The complexity is directly related to the total number of links in the network.
~' For a given path search between two nodes, the complexity can be reduced in :
a very large proportion by pre-seleeting the links that will be used for the computation of the route. This pre-selection is emcient if the minimum hop .~.
number criterion is not degrade(l in the same proportion, that means if ilOpt ~:' remains eonstant.
. ~r ~''''' .', ! .
:'.. ~.
:,'.'.
~R 993014 :.
., ..',~
', ;~
2() 2123~49 Nod~s an~l Links Attriblltes .., :, The basic idea of the prcscnt invention is to pre-select in a given no(le thc physical links which must be con~,i(lered for building a path to anothcr nodc ::
and then to use said pre-selected links at path selection time. -I his invcntiondetails also, how to initiatc and maintain the link table.
Some nodes are selected to built a backbone. The other nodes called local nodes are attached to at least one node of this backbone. Both local and backbone nodes are abie to support Ports and Trunks, without any restriction.
The node attributes are recorcied in Ihe topology database and updated for each change in the network, node addition or node suppression. Based on the node attributes, each link is qualificd by a new attribute, according to the fol-Iowing rules:
!.. ", . .
.',~
Node_i Nodej Link_ij Type ,:
.
Backbone Backbone Backbone Backbone Local Local :
Local Backbone Local .. Local Local Local ;,i, These link attributes are not defined by the network designer. They are dy-:
namically built by each node, according to the Node_i and Nodej types.
~: The classification of the nodcs can be done accor(ling to one of the following ~' methods:
: .
.~
~ by hand in case of small networks. The search of the path from the source to the des~ination node is limi~ed to the backbone Ievcl, all possible searcl through the local network are eliminated.
by using off-line a path selection algorithm for each possible source-, destination.
.
;'-; I~R993014 :.
.:
. ., , .
:' ....
','' :
21 2~23~
'I'ol)ology Datab.lsc The topology database is replicate(l in evcry node and contains information on the network no(les and link.s. A topology algorithm i.s used to maintain a consistent view of the network in all nodes. The topology database contains two classes of information:
The physical topology of the network including static information like ~' physical characteristics Or nodes and links, ~ The link utilization.
Only the first class of parameters are relate(l to the present invention.
;, ."' Assllmptions .
The network comprises N nodes and L, links, ~ Every node has the same knowledge of the network topology.
, ~
~ Def initions ,.
In a given node (Node i), it must be built a simple structure able to deter-mine which links have to be used to reach any other node (Nodej). This matrix structure can be defined as follows:~. :
o There is a line for each link 1, and a column for each node of the network.
- The so defined structure is a L x N matrix.
The matrix element ,Eln, is a Boolean value:
. Eln = I (TRUE) means: Linl; I can be used to reach Node_n from Node i ' ~ Eln = 0 (FALSE) means: Link I cannot be used to reach Node n from Node_i ~;, . ' I R 9 93 014 ~~o ' ':
., ~ .
~.!
~2 2 1 2 t~ 9 ,. Alatrix Init(ation ' .
The matrix is derlned, at the conrlguration time, independently in each nocle, according to the following principles:
....
1~ Links (or matrix lines) i' The links going from Node_i to the backbone nodes are all usable to . reach other nodes (except Node_i itself), ,; a The links between any backbone nodes are all usable to reach any ' destination in the network.
The other local links (not directly attached to Node_i) are usable only to reach the local node(s) they are attached to.
.
..
~; o Nodes (or matrix columns). In Nodej column, :'.
~ :;
All backbone links are usable, ; ~ All links attached to Node_i or/and Nodej are usable - ~ All other links are not usable.
, :: Matrix lJpdate ~.
.; Each time a new node or link is added, the topology database located in every node of the network must be updated. This is done by means of control mes-~'''L, sages containing the new topology and particularly the attributes of the new : node. The matrix is updated according to the following rules:
The addition of a new local Node_k involves the addition of a new column with a:
value I (TRUE) for backbone links and links attached to Node_i and/or Node_k.
value 0 (FALSE) for all other local links :.
;;" .
; .
..:.
~,....
:'..
: .
:
.: :,.
-:
~: .
;: ~
s 2~ 2~23~49 I hc ad(li~ion ol' a ne~v l~ackbolle No(lc k in~olvcs tlle .Id(lition ol' a colulnn .~ _ ' with a:
i~, ,~, a valuc I (TRUE) for the backbone link.s, and links attachcd to No(Jc_i.
,. value O (FALSE) for all other local links ~, ,s o The addition of a new local link Ljk involves the ad(lition of a new row ,:, with a:
,'' .
~'~ . VALUE 1 (TRUE) for thc local node j or k if that link is not directly '' attached to Node i, value I (TRUE) for all nodes (except Node_i) if that link is directly attached to Node i, .,j _ value O (FALSE) for all other nodcs.
~ . . .
'~ o the addition of a new backbone link Ljk involves the addition of a new row ., - in the matrix with a I value (TRUE) for all nodes (except Node_i).
~ ., ,. .
- .., .
Optimal Path Search Metllo(lology . The method according to the prescnt invention can be summarized as follows ; ( Figure 8):
o Step 1: The node is characterized with an attribute: the node is define(l as backbone or local node.
.-' ~ Step 2: The node stores the network configuration in its topology data-Si base. This database is initiated and maintained by means of control mes-sages exchanges between all the nodes in thc network.
, "
' ~ Step 3: From the information stored in the topology database the node ;;. identi~les the backbone and the local nodes.
o Step 4: The node determines the attribute of each link according to the A'''','' node attri butes.
~. .
,:~
i".
;;
":
. .
,~, ~"
r.;
:
:
~ 2123~49 i 24 o S~cp 5: 'I''he node selects the usablc links ror cach destinalion nodc in thc ''~ network by building a L, x N matrix.
~,.,. :
.,, Step 6: At each connec~ion request, the routing algorithm is computed "' among the set of pre-selected usable links.
,.~.
'' o Step 7: During all the time of thc connection, thc dat,l packcts arc scnt to the destination node along the path previous]y computed.
,~.
-'' :~ E~ample :
'.
.r" As illustrated in Figure 4, the network is decomposed at the configuration ' time, in a plurality of backbone (401) and local nodes:
. - . .
. ., Backbone nodes (402): W,X,Y,Z
Local nodes (404): A,B,C,D,E
'' and ., ''- Backbone links (403): Lwx,Lwz,Lxy,Lxz,Lyz, ~'' Local Links (405): Law,Lax,Lbw,Lbx,Lcw,Lcx, ~'' Ldy,Ldz,Lde,Ley,Lez ~ ., .
' The corresponding matrix for the node A is:
:.
. .
... .
.-.
.
.
. ~;, ,, .
.
, : . . .
,,, . I~R993014 .
....
' '!
. .
,' . .
",~, .....
:
2 ~ 2 ~ 9 ~ Nodes A B C D E W X Y %
~ r!
~; Li~ks Law O
Lax O
Lbw 0 1 0 0 0 0 0 0 0 .:~
Lbx 0 1 0 0 0 0 0 0 0 Lcw O 0 1 0 0 0 0 0 0 Lcx O 0 1 0 0 0 0 0 0 Ldy O O 0 1 0 0 0 0 0 Ldz O O 0 1 0 0 0 0 0 Lde O O O O O O O O O
Ley O O O 0 1 0 0 0 0 Lez O O O 0 1 0 0 0 0 Lwx 0 Lwz O
:':
. Lxy O
Lxz O
Lyz O
":
Figure S shows the example of the addition in the network of two nodes: a local node F and a backbone node T with the:
i' new backbone links: Ltx ,Lty,Ltw ~.;.
and new local links: Lfx,Lfz,Lat,Lab The matrix is updated accordingly as illustrated hereunder:
;'' j~:
., .
, , , .~
. . , :
,.',:.'' .:..
:
,,,:~ .
' :~
''',' ~ R 993014 :::
1, ! ,. .;
:~, , : .
~- 2~23~19 . .. ~
. 2fi :
Nodes A B C D E W X Y Z F T
Links i'i"
Law O
Lax O
Lbw 0 1 0 0 0 0 0 0 0 0 0 Lbx 0 1 0 0 0 0 0 0 0 0 0 Lcw O 0 1 0 0 0 0 0 0 0 0 Lcx O 0 1 0 0 0 0 0 0 0 0 Ldy O O 0 1 0 0 0 0 0 0 0 Ldz O O 0 1 0 0 0 0 0 0 0 Lde O O O O O O O O O O O
Ley O O O 0 1 0 0 0 0 0 0 Lez O O O 0 1 0 0 0 0 0 0 Lwx O
Lwz O
Lxy O
Lxz O
Ly~ O
Lfx O O O O O O O O 0 1 0 Lfz O O O O O O O O 0 1 0 Ltx O
Lty O
Ltw O
Lat O
.
Lab o .
. ., ~
To simplify the notation, the matrix in Node_i is represented as a L entry ta-;:. ble Ti(k) (k = I ,.. ,L) with:
, .
Ti(k) = I when link k is usable for any path, '. ~ Ti(k) = O when link k is not usable for any path, Ti(k) = j when the link k is usable only on the path from Node i to Node;
(j = I ,...,N except i)).
~, .
'',; ~R993014 ., . :.: .
:, ~
,.. ::, .
~, .
'~
~ ~'' 27 2123~49 In our example, the tables Ti in nodes A to Z have the following valucs:
, '.~
~ Links TA TB TC TD TE TF TT TW TX TY TZ
-'~
~ Law 1 A A A A A A A A A A
~-.
Lax 1 A A A A A A A A A A
Lbw B 1 B B B B B B B B B
Lbx B 1 B B B B B B B B B
Lcw C C 1 C C C C C C C C
Lcx C C 1 C C C C C C C C
Ldy D D D 1 D D D D D D D
Ld~ D D D 1 D D D D D D D
Lde O O O E D O O O O O O
Ley E E E E 1 E E E E E E
Lez E E E E I E E E E E E
Lw~ 1 1 1 1 1 1 1 1 1 1 1 Lwz Lxy Lxz Lyz Lfx F F F F F 1 F F F F F
Lf~ F F F F F 1 F F F F F
Ltx Lty Ltw Lat 1 A A A A A A A A A A
Lab B A O O O O O O O O O
. ., .::
.... .
- The path between two nodes in the network is considered as optimum if the '~~ number of hops is minimum. All optimum paths create a tree of which root is the source node. Figure 6 shows all the optimum paths established from node A. At the first level are placed all nodes adjacent to the source node, then ,;, at the second level all the adjacent nodes to the first level nodes and not al-i,'!',~ ready placed and so on until exhaustion. Figure 7 illustrates the path search ~, :':
~ . .
.
"
: ,;
1:;
....
. i.
., '~ "
~ '' .
~;:
,.~
2123~9 ,.
frorn no(lc A to no(lc E accor(lillg to thc prcscnt invcntion. As prcviou.~ly dc-scribcd, the cli~ible links to rcach node E arc dcfinc(l in thc tablc TA.
'::1 .,, Eligible Non-Eligible Links TA Links TA
;
.' Law 1 Lbw B
:Y Lax 1 Lbx B
.: Lat 1 Lcw C
~' ';
"!. Ley E Lcx C
; Lez E Ldy D
Lwx 1 Ldz D
Lwz 1 Lde O
~ Lxy 1 Lfx F
.:. Lxz 1 Lfz F
. Lyz 1 Lab B
. Ltx Lty Ltw : '~
Only the backbone links, and the local links attaching the source node and the destination node to the backbone (701) are taken into account in the search process. The other links (702) are not taking part to the path sf lection which . reduces the cornplexity of the routing algorithm computation (for example the Bellman-Ford Algorithm). In our example, four paths - AWZE, ATYE, " ~
AXZE, AXYE - are satisfying the minimal hop constraint with a number of :
. three hops.
The complexity of the network represented in Figure S can be approximated .~ as follows:
j .
. .
'' ;
.....
.,.
. ~
:, ,~, i .., ~
:,, ,, :.
~:~
. .
: ~"
''' " 29 2123~9 ~.,.
.
: C = hopt . 2L = 1.69 x (2x23) = 77.7 with :
'.'~
. hopt = 1.69 hops . L = 23 links . . N = 11 nodes . 1 = 4.18 (46/11) links per node . .
~, The average hop number is determined from all the optimum (minimum) paths between the nodes:
L '~
.
~ I hop: AW, AT, AX, AB, BW, BX, CW, CX, FX, FZ, DE, DZ, DY, EZ, EY, WT, WX, WZ, XZ, XY, XT, TY, ZY.
o 2 hops: AC, AZ, AY, AF, BC, BZ, BY, BF, BT, CZ, CY, CF, CT, FD, FE, FY, FT, FW, DX, DT, DW, EX, ET, EW, ZT, YW.
~, o 3 hops: AE, AD, BD, BE, CD, CE
!i hopt = (lx23 ~ 2x26 + 3x6) / (23 + 26 + 6) = 93/55 -~ = 1.69 hops per optimum path :
.
~ With the assumption that the pre-selection of the usable links according to the ~..
present invention does not degrade the minimum hop constraint, the algorithm complexity can, in our example, be nearly reduced in a factor two:
.
:, C = hopt . 2L' = 1.69 x (2x13) = 43.9 with :
:i i . L' = 13 links i~:
~,....................... . hopt = 1.69 hops -~ This reduction of complexity can be much important typical networks.
~.
':i ~ R 9 93 014 ~, I!r ' , .
:, , ~..
::.
Beeause of the variety of the eonstraints, there are a large number of differentrouting types. Some eorrespond to deterministic or adaptive polieies according to their eapacity of integrating the traffie variations and the network topology.
Routing algorithms can be centralized if the paths are determined from a par-ticular node. Others are distributed between all nodes: this is favorable in a reliability point of view but the algorithm is rnore complicated and the path optimization is rnore difficult to implement. ~ome Algorithms are difricult to c]assify: they are using special techniques called Flooding Rollting or l~andom Routing.
. .
FIOOdillg ~OUting The Flooding Routing is based on a very simple principle consisting for each node to retransmit packets on all output links of the node, e~cept, but the input link. A node connected to K other nodes retransmit K-l eopies of the packet which has been just reeeived. This method allows the delivery to the destina-tion node of at least one eopy of the packet with one condition: it must exist at Ieast one path between the source to the destination node. This routing is ensured even when the network topology is changing for example after a cat-astrophie failure of some components in the network. The Flooding Routing ~';
~ IR993014 . . .
, .....
, .
,.~
.~.
:' 3 212~4~9 allows a very good robustness in thc nctwolk cxploitation~ Furthcrltlolc, as :
all possibles links between the source and the destination node are tried in an i~ exhaustive way, the method ensures that at Ieast one copy of the packet will reach the recipient through the shol-ter path with the minimum delay ir the network is not overloaded. This technique is simple to implement since neither routing tables to route the messages, nor knowledge of the geogràphical posi-:~ tion of the destination node in the network are required. The recipient must only know that the paeket is addressed to it.
i The price to pay ifor these qual;ties of robustness, simplicity and rapidity is a bad utilization of the network resources and a tendency to the congestion.
Usually, it is advantageous to place in the packet header a ~leld indicating the.7 maximum node number that a packet is authorized to pass. This ~leld is dec-remented each time the packet go through a node, and packets with an empty field are discarded.
~' :!, Ralldom or Stochastic Routing i,:
This technique, like the Flooding Routing, does not demand to tlle nodes to have the complete knowledge of the network structure and of the trafi~lc to takea routing decision at their own Ievel. However, here, to limit the generation ofa too important ghost traffic, the nodes avoid to sencl systematically over all ::
output links copies ~f the packets they receive. Thc Random Routing consists in emitting one or more counterparts of the received packet on output links selected at random. This method is also called Selective Flooding. This policy Ieads to a very simple routing at the node Ievel and limits the saturation of the network due to the packet duplication. In return, the average routing delay is longer than this resulting from deterministic techniques. Packets are taking sinuous routes instead of taking the most direct path toward the destination node.
. .
:; ..
i 1 i.~ I R 9 93 0 14 , .
, , .
;, ~ .
;, , . .
I, 2 1 2 3 ~
~ 4 . .
Determillislic Routing , Fixed Roll tin g .... .
The Fixed Routing def'ines paths that must takc thc differcnt packcts accord-ing to the gencral network characteristics such as the network topology and the '~ mean traffic expected on the communication lines. Thc routing rulcs are es-tablished once and aims for optimizing the performance criterion privileged by ~ the user. In most of the cases, the average packet transmission time through ~!, the nctwork must be minimized. Thc rnethod consists in building a routing ta-ble at the level of each node. Its consultation allows the nodc to determinc on ::' which output link the packet it has just received must be transfcrred. The Fixcd Routing is vcry simple in its principles. The only processing done in the ,.i nodes consists in queucs managcment and tables reading and the optimization ' algorithm is initiated only once at the network dcsign Ievel. The Fixed Routing . aims for a long term and a global optimization of the network, but in com-parison with thc random routing, this technique allows to consi(Jerably speed up thc packct transmission through the nctwork.
., , ~:' Adaptive Rotlting . .
Contrary to lhe Fixed Routing, the purpose of the Adaptive Routing is to satisfy at any time thc optimization critcrion. Thc tablcs arc permancntly up-dated according to, for examplc, thc instantancous statc of thc traffic on the lines.
,~, ~ Centralized Routing . .
When the characteristics of the network fluctuate, it is ~ossible to a(lapt the routing by assigning to one node the responsibility to updatc pcriod-ically the routing tables according to the traffic variations and the topology. This method, simple in its principles, is callcd Centralized Rout-ing. Its principal disadvantage is to generate an important auxiliary traffic and to subordinate the good functioning of the net~vork to only one node.
:
. I R 9 93 0 14 '''' 'i; .
, ~,...
;., ., . .
~ ~~23~9 s The differenl nodes must send to ~he rouling cen~cr the .~tate of the net-work as they perceive it (operational acljacent no(les, trarric .. ), and the routing center must, in its turn, disp.ltcll to the no(Jes their routing tables.Tlle auxiliary traffic is concentrated around the routing center and this has for result to saturate this part of the network. Furthermore, the Central-ized Routing can generate some problems at the time the tables are re-freshed because said tables cannot be received at the same time by all the nodes. A solution is to decentralize the tables establishment at the level of each node.
o Local Routing . .
The local and distributed routing methods are both based on the estab-Iishment by each node of its own routing table according to information locally collected. With the local routing technique, each node builds its ta-~ ble without exchanging information with the neighboring nodcs. In its most simple form, the method consists in placing the packet just received in tlle shorter output queue and in transmitting it as rapidly as possible. Thc lo-cal algorithm tends in its principle to circulate the packcts very quickly in the network. However, as the routes are selected in some way or other the '~ mean paths Iength is far to be minimal.
~ .
i .
'~ Q Distributed Routing The Distributed Routing is a local method in which neighboring nodes are exchanging messages concerning the trafric and the network conditi(>ll to . update their routing and delay tables.
ilicr.llcllical l~outing , .
': With the non Stochatic Routing methods, each node requires a table with as many rows as nodes in the network and a number of columns equal to ~' the number of output links. Furthermore, when the routing is adaptive, . nodes must exchange periodically messages to update their routing tables.
;., ~ IR993014 ..
. ., .:
~.2 ~ 2 ~
The tables size an(J the impollance of tlle conlrol messages are increa.sing rapidly with lhc number of no~les ancl become unacceptable when the net-work con~ains more than ten nodes. To solve this problem, the network is . .
divided in a hierarchy of sub networks as illustrate(l in Figure 0. In par-ticular, optimal clustering structures (003) are determined so as to mini-mize the length of the routing tables required. Said tables are taking into account only the nodes in a sub nctwork (0()3) and a packet a~klresse(l to a node in another sub network will necessarily go through the access node (005) of this rermote sub network. Packets transit through some mandatory .
nodes (005) and the price to pay for this simplification is that smaller routing tables give less precise routing information whieh then results in Ionger path lengths for the message traffic. Hierarchieal Routing proee-dures are partieularly effective for large packet switehed networks (001).
With smaller routing tables, they require less storage and proeessing in the nodes (002) as well as less eommunications overhead in the lines (Oû4).
The problem of the optimization of the hiera1chical levels has been stu(lie(l by L. Kleinrock and IF. I~amoun - Hicralcfiicnl RolltingJor l,a~ge IVetworks, Cornputer NetwoYks, Vol. 1, pp. 155-174, January 1977.
The main idea for reducing the routing table Iength is to keep, at any node, eomplete routing information about nodes which are close to it and Iess information about nodes located further away rrom it. This can be re-alized by providing one entry per (Iestination for the eloser nodes and one entry per set of destinations for the remote nodes.
The reduction of routing table Iength is achieved througll a hierarchical partitioning of the network. Basically, an m-level hierarcl-ieal elustelillg of a set of nodes (Figure 0) consists in grouping the nodes (002) into a Ist Ievel clusters (003), which in turn, are grouped into 2nd Ievel cluster e~c...
Tllis operation eontinues in a bottom up fashion. The mth level cluster is the highest level cluster (3rd level cluster) and as such it includes all the nodes of the network (001).
Unfortuncltely, the gains in table length are accompanied with an increase of the message path Iength in the network. This result is a degradation of the network performanee (delay, throughput) due to the excess internal . ,.
:;
-: .
.
J' , ~ , J
''''"
; : ~
~" 7 212.~449 trarric caused by longer pa~h Iength. I lowever, in vcry large networks, an ;; enormous table reduction may be achieved with no significant increase in ': net~ork path Iength.
High Pe~ forn&a1ace l~et~vorks , The data transmission is now evolving with a speci~lc focus on applications and by integrating a fundamental shift in the customer traffic profile. Driven by the growth of workstations, the local area networks (LAN) interconnection, the distributed processing between workstations and super computers, the new applications and the integration of various and often conflicting structures -hierarchieal versus peer to peer, wide (WAN) versus local (LAN) area net-works, voice versus data - the data pro~lle has become higher in bandwidth, bursting, non deterministic and requires rnore connectivity. Based on the above, it is clear that ~here is strong requirement for supporting distributed , ~'. computing applications across high speed networks that can earry LAN com-'- munications, voice, video, and traffie among ehannel attached hosts, business, .~ engineering workstations, terminals, and small to intermediate file servers.
This vision of a high speed multiprotocol network is the driver for the emer-gence of fast packet switching networks architectures in which data, voice, and video information is digitally encoded, chopped into small packets and trans-' mitted through a common set of nodes and links.
., (~
Throllghput ~;~; The key requirement of these new architectures is to reduce the end-to-end -!J delay in order to satisfy real t;me delivery cons~raints and to achicve the nec-~';. essary high nodal throughput for the transport of voice and video. Increases o in link speeds have not been matched by proportionate increases in the proc-. essing speeds of communication nodes and the fundamental challenge for high speed networks is to minimize the packet processing time within each node.
r ~
~R 9 93 014 : ,;
~ .
;1'.!, ''.:i ~-~ 2123~'3 R
As example, for meeting a typical 1()() ms (Ielay to ~Ieliver a voice r)acket be-tWCCIl two end users:
:
o A lotal of 36 ms might be nee(led rO, the paclcelizalion and play-out func-tions at the end points.
. ~
o About 20 ms is the unalterahle propagation delay needed, say, to cross the i United States.
,, There remains 44 ms for all the intra-node processing time as the packet moves through the network. In a ~ nodes network, each node would have about 8 ms for all processing time ineluding any queueing time. In a 10 nodes network, each node would have about 4 ms.
Another way of looking the same constraint is illustrated in Figure 1: taking a node with an effective processing rate of I MIPS (Millions of Instructions Per Second), it is possible to rlll a 9.6 kbps line with 1000 byte packets evcn if a network node must execute ~33 000 instructions pcr packct processc(l. For a 64 kbps line the node can afford 125 000 instructions per packet. In order to fill an OC24 link, however, our 1 MIPS node could only exeeute 7 instructions per packet. In the latter case even an effective rate of 10-30 MIPS would allow only 70-200 instructions per packet.
In order to minimize the processing time and to take full advantage of the high speed/low error rate technologies, most of the transport and control functions provided by the ncw high bandwidth network architectures are performed on an end-to-end basis. The flow control and particularly the path selcction are managed by the access points of the network which reduces both ~he awareness and the function of the intermediate nodes.
.,'~
;, , . . .
Rollting Mlodes ..
- The routing within the network presents two aspects:
. .
:,, ' 1. Determining what the route for a given connection shall be, ~' 2. Actually switching the packet within a switching node.
, ~, ~,;"
.. I~R 9 93 014 .
' , .
. .
. ,.
:~i ,:'.! 2 ~ ~
., ., 1 here are many methods of (Ie~erminin~ a rOu~e throu~h a ne~volk. For very high throughput, the eritical item is that the switehing elements must be able to decide where to route an incoming packet in a very sh()rt portion of time.
As described in the document l~ ll Speed Ne~ oskillg T~chnolo~y"~ln Intro-~luc~ory S~lrvey (palges ~8 to g6) - CC24-3816-0() ITSC Raleigll Marc~l 19~2, the routing modes, the most widely used, are the Source Routing and the Label Swapping.
~, ,~ .
~i !".
~' ~ Source Routing .: -, !.
The Source Routing is a particular implementation of the Distributed Routing for eonnectionless networks. The souree node (or aecess node) is responsible for ealeulating the route the paeket must take through the net-work. A routing field is appended to every paeket sent and that field is used by the intermediate nodes to direet the paeket towards its destination.
In this teehnique, the sending node must know the network topology and it must be able to find the optimal route. ~ut onee the route is determined, intermediate nodes do not need to refer to any tables or parameters to ....
make the routing deeision. The next stage of the route is right there in the paeket header. A drawback of this metho(l is that the routing reld in lhe packet header takes some storage and is an overhead. But this is quite small and the benerlts of being able to make a fast routing decision out-weigh the small increase in bandwidth overhead. Since all routing infor-mation is contained in eacll packet a ~set up is not required along the patll to provide an end to end communieation. Thus, data applications ~vhich ., benefit from a datagram service ean be effeetively supl-orted by this teeh-nique. However, the datagram traffic is ehalacterize(l by the fact that each datagram can be viewed as a connection and requires the computation of a path. Each packet is routed through the network as a self containcd unit and is independent of all other packets. The key point for the source node is to determine for each datagram the optimal route in the shorter lapse of time.
.
,:
; t:? ~R. 9 93 014 !~ 1 ., ~'' '~
~,':' .
' ;.' .':
~ :' '"'~' . ~ Label Swar)ping . .
Ihe Label Sv~appin~, is a r)articul.ll im~-lelllcllt.ltion Of Ihe L)istlibu~cd Routing for connection oriented networks. These networks typically multi-plex many connections (or sessions) on a link using some form of logical "channel". Each packet sent on the link has a header which includes an ,, arbitrary number identifying which logical connection that this packet be-longs to. For example, each packet can contain in its routing field a label that will be used by the intermediate nodes to identify the transmission link , the packet should be routed over. When a packet arrives at a no(le, the label is extracted from its header, and used to index a routing table that provides a new label along with a link address. The new label is over writ-ten on the old label, and the packet is forwarded using the link address.
The information in the routing table is updated at the connection set up time. Each node on a path selects a label for each direction of the con-nection, and sends it to the neighboring node.
Tlle call set up and the resource reservatioll plOCCSS COIIlpliSCs tllC follow-ing steps:
.
~'~ a A connectiorl reqllest is speci~led by the user via a ~set of p.lr.lmeter.s including origin and destination network address and data flow char-~' acteristics, :
The p~th determinatioll is realized by the source node from its local routing topology database . . , ~ ~ The route reserv~tion is sent out in a special message along the specified - nodes. Said message, which is usually routed according to the previ-i ously described Source Routing technique, signals the nodes to set up their connection tables and to reselve their resourccs to providc the Ievel of service required by the traffic type (for examplc, a bandwidth , .
~~~ reservation on each of the links of the path).
"
~ The tables upd.ate reflects the changes in the availability of the network '~' resources. The information is sent to every node in the network.
:
, . .
I:R 9 93 014 . .
~. ~
:~, :
: .
-~ 11 212~4~L!~
The Labcl Swapping tecllnique re(!uires ~hclt the conneclion tables be set up and main~aine(l dynamically. Tha~ mcans ~hat whcn a new connec~ion is established or an okl one is terminate~l the tables are updated (~he da-tabase Of network ~opology can of course be maintaine(l quile separately).
This is possible if the connection sct up is rclatively infrequcnt and is not too time critical. This restriction makes datagram transport quite ineffi-cient. Howcver, connections that requirc vcry short packcts, like real-timc voice conncctions, can bc cffectively supportcd by this tcchnique, bccause of the low packet overhead. Once the connection is cstablishcd, thcre is no need to place a destination address in the headcr cvcry time a packet is , .
sent. All is needed is a rlcld to specify which conncction is to be uscd for this paekct.
.,..~
, .
~,.;
Path Selection Criterion Onc of thc major critcrion for sclecting paths bctwccn nodcs in packct net-works is the minimum hop count and minimum path Iength. The hop count is the number of links used to construct the path betwecn the two cnd nodes.
The path length is a function of the ovcrall transmission delay imposed by the path between the two end nodes. In most high speed networks, the delay (path Iength) is not a major consideration since the worst-case delay through such networks is nearly always acceptable. Thc hop count~ howcvcr, is a dircct measure of the amount of resources rcquired to implcment a givcn path and hencc is of considcrable importancc in selecting paths. It is to bc noted that aselected path need not be a minimum hop count path since congestion on the nctwork links may force the choicc of a larger hop count path. However, such longer alternate paths cannot be allowcd to grow without limit since inordinate amounts of network rcsources might be committed to one path, resulting in further congestion for other paths and forcing yet longer hop count paths to be selected for yct other connections. The long term network throughput could thereby be adversely affccted. The problem, then, is to select a path between an origin node and a destination node which has a minimum hop count, a ., .
~R 9 93 014 ;.~
i:,~, ",,~
i ~
. . .
: ., ~. .
~.. . . .. . . .. . .
~ 1 2 2 1 2 3 ~1 4 9 minimum path Iength, which dl)es no~ utilize an inor(lillate am()unt of network resou rces.
:
.:
~ Key requirements .:, The Distributed Routing mechanism in large and high speed packed switching networks supporting both connection oriented and connectionless routing modes implies some requirements in terms of performance and resource con-~~sumption which can be summarized as follows:
.~
-. ~ the source node (or the node providing the route calculation for the source~' node) must be able to decide where to route an incoming packet in a very ' short portion of time (the computation must be sufficiently rapid to com-pute an optimum path for each connection request) ~', 4 the switching time in the intermediate nodes must be minimized, (minimum processing time) ~. 0 the network resources along the selected path must be optimized according to the criterion of the minimum hop count.
o Control messages must be as much as possible limited not to overload the network.
:
In high speed networks the cost for updating the routing tables generated by the exchange of control messages is not critical so long as this can be performed before the connection set up. The very high line throughput authorizes the interehanging of routing inrormation without creatillg a signirleant burcJen on the eomrnunication lines themselves. These considerations sugges~ a better ac-eess and use of the routing tables maintained within the nodes without saeri-ficing the criterion Or optimal path contrary to the Hierarchieal Routing method proposed by L. Kleinroek and F. Kamoun.
,, .~Summary ~ the invcntion ""
- ., ;.Currently, the routing algorithms eompute all the available paths in the net-~-work, from the souree node to the destination node before to select the optimal .;
~I R 9 93 014 ' ';'~
..
i : ~ , . . .
2123~49 rOu~e. No as~umption is made on the nelwork topolo~y and ~he route cOmpU-tation i.s often time and resouree consuming. For example, some paths wllicll arc calcukltcd are no~ acceptable due to the particular geographical conl'igllra-tion of the network.
In the real world, large transport networks are not fully meshed. The present invention is based on the simple observation that networks are u.sually built around a hierarchical structure. A set of nodes, interconnected by high throughput lines, are used to build a 'Backbone' with a high degree of meshing to allow the redundancy and reliability required by the user. The other nodes or 'local' nodes are attached to one or several backbone nodes. It is the net-work designer responsibility, at the configuration time to define for each node what is its attribution: backbone or local node. The list of the node attri-butions appears in the topology table and is updated each time a node is added to or dropped from the network. The routing algorithm can take advantage of the particular network topology to drastically reduce the eomplexity of pathscomputation. For a given connection, only a limited number of nocles are eli-gible and are taken in account by the algorithm in the optimal route search.
The object of the invention is to split the network in backbone and local nodes to speed up the path selection.
Descri,~tion of the clrawin~,~s Figure O represent a network topology for Hierarcllical Routing according to the method sllggested in the prior art by L. Kleinrock an(J F. Kamoun ,, Figure I sllows the processing times (or numbel Of instluc~iolls pCI secs)nd) required in function of the different line throughpu~s supported by the high speed networks.
., Figure 2 shows a typical model of high speed packet switching network in-cluding the access and transit nodes claimed in the present invention.
''.'.', ., Figure 3 describes a high speed Routing Point according to the present in-- vention r '.
:~ ' ' . ,.
~.',' .
''"
' :
Il 2 123A49 .,.
Figurc 4 illustrates the decomposition of the network in a plurality of back-bone no(lcs and local nodes accolding to the prcscnt invcntion.
'.'.
~- Figure 5 illustrates the addition in the nctwork of ncw nodes and links and their classirlcation in the topology database according to the present inventiom ~:' F;igure 6 reprcsents the optimum path trce of node A
''' Figure 7 represents the optimum path tree of node A according to thc present invention .
, Figure 8 summarizes the path selection methodology according to the present :
invention.
;
Dc.~cription of tfie prcfcrY~d cmbo(liol~nt oJ t~lC~ inven~io~l ., :
As illustratcd in Figure 2, a typical model of communication system is made of several user networks (212) communicating through a high performance network (200) using private lines, carrier provided services, or public data nctworks. Each user nctwork can bc dcscribed as a ~sct of communication processors and links (211) intcrconnceting large computers used as cnterprise servers (213), user groups using workstations or personnel computers attached on LAN (Local Area Networks 214), applications servers (215), PBX ~Privatc Branch cXchangc 216) Or vi<leo servers (217). Tllese user nctwolks, dispelse(l in differcnt establishmcnts, nccd to bc interconnected through wide area transport facilities and different approaches can be used ~or organizing thc , data transfer. Some architect-lres involve the checking for data integrity at - cach network node, thus slowing clown the transmission. Others are cssentially looking for a high speed data transfer and to that end the transmission, routing;~ and switching techniqucs within thc nodcs are optimized to process the flowing packets towards their final destination at the highest possible rate. The present ~'' invention bclongs essentially to the latter category and more particularly to the ~.
:
, ., :
,~
~, .;:
~'', ,~ ' 15 2~23~
~,~; ra~St packet switclling nelwork architectule detaile(J in the following para-graphs.
:
IYigh speedpacket s~vitching ~tet)vo~ks .
The general view in Figure 2 shows a fast packet switching transmission sys-tem comprising eight nodes (201 to 208) each node being interconnected by means of high speed communication lines called Trunks (209). The access (210) to the high speed network by the users is realized through Access Nodes (202 to 205) located at the periphery. These Access Nodes comprise one or more Ports, each one providing an access point for attaching external devices supporting standard interfaces to the network and performing the conversions required to transport the users data flow across the network from and to other external devices. As example, the Access Node 202 interfaces respectively a Private Branch eXchange (P13X), an application server and a hub through three Ports and communicates through the network by means of lhe a(Jjacent Transit Nodes 201, 208 and 205.
,.
S~vitchillg No~les , ~s Each network node (201 to 208) includes a Routing Point where the incoming data packets are selectively routed on the outgoing Trunks to-vards the neigh-boring Transit Nodes. Such routing decisions are made according to the in-;~ formation contained in the hea(ier o~ the data packets. In additioll to the basic packet routing function, the network nodes also provide ancillary services such ~' as:
~ . .., the determination of routing paths for packets originated in the node, directory services like retrieving and updating information about network ~i ~ users and resources, the maintaining of a consistent view of the physical network topology, in-' ~ cluding link utilization information, and ... .
.
.
'':.
,'',~.',' .i .
. ~,i . . .
. ., ~: .
-, ~
r ' 2123~1~L9 ~ thc rc~crvation Or rcsourccs al acccss ilOillts 01' l.llc nct\vork.
' Each Port is connected to a plurality of u~er processing equipments, each user equipment comprising either a soulce of di~ital data to be transmitted to an-other user system, or a data sink for consuming digital data received rrom an-other user system, or, typically, both. Tlle interpretation of the users protocols~
~. the translation of the users data into packe~s formatted appropriately for their - transmission on the packet network (200) and the generation of a header to route these packets are executed by an Access Agent running in the Port. This ~ header is made of Control and Routing Fiekls.
, . .
o The Routing Fiekls contain all the information necessary to route the packet through the network (200) to the destination node to which it is ". addresscd.
.', 9 The Control Fields include, among other things, an encoded identirlcationof the protocol to be used in interpreting the Routing Field (Source Rout-ing, Labcl Swapping, ....).
~, llti~ltg Poi1tts Figure 3 shows a general block diagram of a typical Routing Point (300) such as it can be found in the network nodes (201 to 208) illustrated in Figure 2. A
Routing Point comprises a high speed packet Switch (302) onto which packets . arriving at the Routing Point are entered. Such packets are received:
., 0 from other nodes over high speed transmission links ~303) via Trunk ;~ Adapters (304).
from users via appiication adapters callc i Ports (301).
Using information in the packet header, the adapters (304, 301) determine which packets are to be routed by means of the Switch (302) towards a local user network ~307) or towards a transmission link (303) Ieaving the node. The adapters (301 and 304) include queuing circuits for queuing packets prior to .' or subsequent to their launch on the Switch (302).
,.
i ..
~ rR993014 ::~
:.
; :
.'. ', "' .
., ~
~ i .
:
17 21234~9 The Route Controller (305) calculates the optimum routes through the net-work (200) S(3 as to satixry a given set of (lualiLy of service specifie(l by the user, and to minimize the amount of network resources use(l to complete a commu-nication path. Then, it builds the header of the packets generated in the Routing Point. The optimization criterion includes the number of intermediate nodes, the characteristics of the connection request, the capabilities and the utilization of the Trunks in the path ... All the information necessary for the routing, about the nodes and transmission links connected to the no(Jes, are contained in a Network Topology Database (3~)6). Under stea~y state condi-tions, every Routing Point has the same view of the network. The network topology information is updated when new links are activated, new nodes added to the network, when links or nocles are (Iropped or when link loa(ls change signifieantly. Such information is originated at the network node to whieh the resources are attaehed and is exehanged by means of eontrol mes-sages with all other Route Controllers to provide the up-to-date topological information needed for route ealculation (such database up(lates are carried on paekets very similar to the data paekets exehanged between end users of the network). The fact that the network topology is kept current in every node through continuous updates allow~s dynamic ne~work reconfigulaLions ~vitho disrupting end users logical connections (sessions).
The incoming transmission links to the packet Routing Point may comprise links from external devices in the local user networks (210) or links (Trunks) ,., from adjacent network nodes (209). In any case, the Routing Point operates in ~~ the same manner to receive each data packet an(l forwar(l it on to another ;~ Routing Point as dietated by the information in the packet hea(ler. The fast ~' packet switching network operates to enable a communicati()n betwcen any two end user applications without dedicating any transmission or node faeili-ties to that communieation path except for the duration of a single packet. In ... ~
this way, the utilization of the communication facilities of the packet network ~0u is optimized to carry significantly more lraffic than would be possible wilh dedicated transmission links for each communication path.
:
.~ ~
, . . .
:
1~ 993014 :, -~ ,.
., ,:
.
''' ~~'' 18 2123~/19 ~. .
Net~vc))k D~c~--7positio~t . . .
Modified Bellmann-Ford Algoritllm There are several possible algorithms to compute an optimal route in a net-work. The modified Bellmann-Ford algorithm is one of the most currently used in fast packet packet switching networks. As described in European Patent Application 9348()030.1 - MethcJds and Apparafusfor Optimllm Pa~h Se~ection in Packet Transmission Networks, this one is invoked for each connection with the following parameters:
~ the source node address, ~ the destination node address, ~ thc quality of scrvice required for thc connection. For examplc:
~ maximum call set up dcl~y (vcry important paramcter for connec~ions ....
serviccd in real time), . ~ maxirnum end to cnd transit dclay, ~ maximum information loss, crror probability, ~ ...
Thc algorithm uscs the local copy,in the sourcc node (or acccss nodc), of thc network topology to determinc the minimum hop and path lcngtll with the dcstinatil)n nodc ~hich mcets the ~luality of ~servicc re(luilcmellts. As men-tioned, the modi~led Bellman-Ford algorithm makes no assumption on the nctwork geographical con~lguration and it requires the samc complc~ity whcther the network is fully meshcd or hicrarchized. The purposc of the pres-cnt invention is, for a gi~en connection, to simplify the nctwork topology by reducing thc number of eligible nodes for the path calculation.
.;.
. ~
.
' '' ., :
:~ ~
~ 19 2123~49 Algoritlllll Comple.~ity s The total hop count neeessary from a givcn no(le to reach all thc other nodes gives a measure of the complexity C of the pa~h ~selection accor(ling to the minimum hop criterion. C is proportional to the average number of links I per node, to the total number of nodes N in the network and to the average num-ber of optimal hops.
.
C = N x O
1 . 1 (N-1)/ N) + (first hop) .. 1 i 2 . 1 (N-1)/ N) + (second hop) 3 . 1 (N-1)/ N) + (third hop) t i.:'~';~ i : (i+1). 1 (N-1)/ N) (i+1 th hop) . . . ~
:, . C = hopt . N . 1 = hopt . N . 2.L/N = hopt . 2L with :
.
hopt = average hop count for an optimal path average number of links per node ~'. N = total number of nodes in the network . L = total number of links in the network , The complexity is directly related to the total number of links in the network.
~' For a given path search between two nodes, the complexity can be reduced in :
a very large proportion by pre-seleeting the links that will be used for the computation of the route. This pre-selection is emcient if the minimum hop .~.
number criterion is not degrade(l in the same proportion, that means if ilOpt ~:' remains eonstant.
. ~r ~''''' .', ! .
:'.. ~.
:,'.'.
~R 993014 :.
., ..',~
', ;~
2() 2123~49 Nod~s an~l Links Attriblltes .., :, The basic idea of the prcscnt invention is to pre-select in a given no(le thc physical links which must be con~,i(lered for building a path to anothcr nodc ::
and then to use said pre-selected links at path selection time. -I his invcntiondetails also, how to initiatc and maintain the link table.
Some nodes are selected to built a backbone. The other nodes called local nodes are attached to at least one node of this backbone. Both local and backbone nodes are abie to support Ports and Trunks, without any restriction.
The node attributes are recorcied in Ihe topology database and updated for each change in the network, node addition or node suppression. Based on the node attributes, each link is qualificd by a new attribute, according to the fol-Iowing rules:
!.. ", . .
.',~
Node_i Nodej Link_ij Type ,:
.
Backbone Backbone Backbone Backbone Local Local :
Local Backbone Local .. Local Local Local ;,i, These link attributes are not defined by the network designer. They are dy-:
namically built by each node, according to the Node_i and Nodej types.
~: The classification of the nodcs can be done accor(ling to one of the following ~' methods:
: .
.~
~ by hand in case of small networks. The search of the path from the source to the des~ination node is limi~ed to the backbone Ievcl, all possible searcl through the local network are eliminated.
by using off-line a path selection algorithm for each possible source-, destination.
.
;'-; I~R993014 :.
.:
. ., , .
:' ....
','' :
21 2~23~
'I'ol)ology Datab.lsc The topology database is replicate(l in evcry node and contains information on the network no(les and link.s. A topology algorithm i.s used to maintain a consistent view of the network in all nodes. The topology database contains two classes of information:
The physical topology of the network including static information like ~' physical characteristics Or nodes and links, ~ The link utilization.
Only the first class of parameters are relate(l to the present invention.
;, ."' Assllmptions .
The network comprises N nodes and L, links, ~ Every node has the same knowledge of the network topology.
, ~
~ Def initions ,.
In a given node (Node i), it must be built a simple structure able to deter-mine which links have to be used to reach any other node (Nodej). This matrix structure can be defined as follows:~. :
o There is a line for each link 1, and a column for each node of the network.
- The so defined structure is a L x N matrix.
The matrix element ,Eln, is a Boolean value:
. Eln = I (TRUE) means: Linl; I can be used to reach Node_n from Node i ' ~ Eln = 0 (FALSE) means: Link I cannot be used to reach Node n from Node_i ~;, . ' I R 9 93 014 ~~o ' ':
., ~ .
~.!
~2 2 1 2 t~ 9 ,. Alatrix Init(ation ' .
The matrix is derlned, at the conrlguration time, independently in each nocle, according to the following principles:
....
1~ Links (or matrix lines) i' The links going from Node_i to the backbone nodes are all usable to . reach other nodes (except Node_i itself), ,; a The links between any backbone nodes are all usable to reach any ' destination in the network.
The other local links (not directly attached to Node_i) are usable only to reach the local node(s) they are attached to.
.
..
~; o Nodes (or matrix columns). In Nodej column, :'.
~ :;
All backbone links are usable, ; ~ All links attached to Node_i or/and Nodej are usable - ~ All other links are not usable.
, :: Matrix lJpdate ~.
.; Each time a new node or link is added, the topology database located in every node of the network must be updated. This is done by means of control mes-~'''L, sages containing the new topology and particularly the attributes of the new : node. The matrix is updated according to the following rules:
The addition of a new local Node_k involves the addition of a new column with a:
value I (TRUE) for backbone links and links attached to Node_i and/or Node_k.
value 0 (FALSE) for all other local links :.
;;" .
; .
..:.
~,....
:'..
: .
:
.: :,.
-:
~: .
;: ~
s 2~ 2~23~49 I hc ad(li~ion ol' a ne~v l~ackbolle No(lc k in~olvcs tlle .Id(lition ol' a colulnn .~ _ ' with a:
i~, ,~, a valuc I (TRUE) for the backbone link.s, and links attachcd to No(Jc_i.
,. value O (FALSE) for all other local links ~, ,s o The addition of a new local link Ljk involves the ad(lition of a new row ,:, with a:
,'' .
~'~ . VALUE 1 (TRUE) for thc local node j or k if that link is not directly '' attached to Node i, value I (TRUE) for all nodes (except Node_i) if that link is directly attached to Node i, .,j _ value O (FALSE) for all other nodcs.
~ . . .
'~ o the addition of a new backbone link Ljk involves the addition of a new row ., - in the matrix with a I value (TRUE) for all nodes (except Node_i).
~ ., ,. .
- .., .
Optimal Path Search Metllo(lology . The method according to the prescnt invention can be summarized as follows ; ( Figure 8):
o Step 1: The node is characterized with an attribute: the node is define(l as backbone or local node.
.-' ~ Step 2: The node stores the network configuration in its topology data-Si base. This database is initiated and maintained by means of control mes-sages exchanges between all the nodes in thc network.
, "
' ~ Step 3: From the information stored in the topology database the node ;;. identi~les the backbone and the local nodes.
o Step 4: The node determines the attribute of each link according to the A'''','' node attri butes.
~. .
,:~
i".
;;
":
. .
,~, ~"
r.;
:
:
~ 2123~49 i 24 o S~cp 5: 'I''he node selects the usablc links ror cach destinalion nodc in thc ''~ network by building a L, x N matrix.
~,.,. :
.,, Step 6: At each connec~ion request, the routing algorithm is computed "' among the set of pre-selected usable links.
,.~.
'' o Step 7: During all the time of thc connection, thc dat,l packcts arc scnt to the destination node along the path previous]y computed.
,~.
-'' :~ E~ample :
'.
.r" As illustrated in Figure 4, the network is decomposed at the configuration ' time, in a plurality of backbone (401) and local nodes:
. - . .
. ., Backbone nodes (402): W,X,Y,Z
Local nodes (404): A,B,C,D,E
'' and ., ''- Backbone links (403): Lwx,Lwz,Lxy,Lxz,Lyz, ~'' Local Links (405): Law,Lax,Lbw,Lbx,Lcw,Lcx, ~'' Ldy,Ldz,Lde,Ley,Lez ~ ., .
' The corresponding matrix for the node A is:
:.
. .
... .
.-.
.
.
. ~;, ,, .
.
, : . . .
,,, . I~R993014 .
....
' '!
. .
,' . .
",~, .....
:
2 ~ 2 ~ 9 ~ Nodes A B C D E W X Y %
~ r!
~; Li~ks Law O
Lax O
Lbw 0 1 0 0 0 0 0 0 0 .:~
Lbx 0 1 0 0 0 0 0 0 0 Lcw O 0 1 0 0 0 0 0 0 Lcx O 0 1 0 0 0 0 0 0 Ldy O O 0 1 0 0 0 0 0 Ldz O O 0 1 0 0 0 0 0 Lde O O O O O O O O O
Ley O O O 0 1 0 0 0 0 Lez O O O 0 1 0 0 0 0 Lwx 0 Lwz O
:':
. Lxy O
Lxz O
Lyz O
":
Figure S shows the example of the addition in the network of two nodes: a local node F and a backbone node T with the:
i' new backbone links: Ltx ,Lty,Ltw ~.;.
and new local links: Lfx,Lfz,Lat,Lab The matrix is updated accordingly as illustrated hereunder:
;'' j~:
., .
, , , .~
. . , :
,.',:.'' .:..
:
,,,:~ .
' :~
''',' ~ R 993014 :::
1, ! ,. .;
:~, , : .
~- 2~23~19 . .. ~
. 2fi :
Nodes A B C D E W X Y Z F T
Links i'i"
Law O
Lax O
Lbw 0 1 0 0 0 0 0 0 0 0 0 Lbx 0 1 0 0 0 0 0 0 0 0 0 Lcw O 0 1 0 0 0 0 0 0 0 0 Lcx O 0 1 0 0 0 0 0 0 0 0 Ldy O O 0 1 0 0 0 0 0 0 0 Ldz O O 0 1 0 0 0 0 0 0 0 Lde O O O O O O O O O O O
Ley O O O 0 1 0 0 0 0 0 0 Lez O O O 0 1 0 0 0 0 0 0 Lwx O
Lwz O
Lxy O
Lxz O
Ly~ O
Lfx O O O O O O O O 0 1 0 Lfz O O O O O O O O 0 1 0 Ltx O
Lty O
Ltw O
Lat O
.
Lab o .
. ., ~
To simplify the notation, the matrix in Node_i is represented as a L entry ta-;:. ble Ti(k) (k = I ,.. ,L) with:
, .
Ti(k) = I when link k is usable for any path, '. ~ Ti(k) = O when link k is not usable for any path, Ti(k) = j when the link k is usable only on the path from Node i to Node;
(j = I ,...,N except i)).
~, .
'',; ~R993014 ., . :.: .
:, ~
,.. ::, .
~, .
'~
~ ~'' 27 2123~49 In our example, the tables Ti in nodes A to Z have the following valucs:
, '.~
~ Links TA TB TC TD TE TF TT TW TX TY TZ
-'~
~ Law 1 A A A A A A A A A A
~-.
Lax 1 A A A A A A A A A A
Lbw B 1 B B B B B B B B B
Lbx B 1 B B B B B B B B B
Lcw C C 1 C C C C C C C C
Lcx C C 1 C C C C C C C C
Ldy D D D 1 D D D D D D D
Ld~ D D D 1 D D D D D D D
Lde O O O E D O O O O O O
Ley E E E E 1 E E E E E E
Lez E E E E I E E E E E E
Lw~ 1 1 1 1 1 1 1 1 1 1 1 Lwz Lxy Lxz Lyz Lfx F F F F F 1 F F F F F
Lf~ F F F F F 1 F F F F F
Ltx Lty Ltw Lat 1 A A A A A A A A A A
Lab B A O O O O O O O O O
. ., .::
.... .
- The path between two nodes in the network is considered as optimum if the '~~ number of hops is minimum. All optimum paths create a tree of which root is the source node. Figure 6 shows all the optimum paths established from node A. At the first level are placed all nodes adjacent to the source node, then ,;, at the second level all the adjacent nodes to the first level nodes and not al-i,'!',~ ready placed and so on until exhaustion. Figure 7 illustrates the path search ~, :':
~ . .
.
"
: ,;
1:;
....
. i.
., '~ "
~ '' .
~;:
,.~
2123~9 ,.
frorn no(lc A to no(lc E accor(lillg to thc prcscnt invcntion. As prcviou.~ly dc-scribcd, the cli~ible links to rcach node E arc dcfinc(l in thc tablc TA.
'::1 .,, Eligible Non-Eligible Links TA Links TA
;
.' Law 1 Lbw B
:Y Lax 1 Lbx B
.: Lat 1 Lcw C
~' ';
"!. Ley E Lcx C
; Lez E Ldy D
Lwx 1 Ldz D
Lwz 1 Lde O
~ Lxy 1 Lfx F
.:. Lxz 1 Lfz F
. Lyz 1 Lab B
. Ltx Lty Ltw : '~
Only the backbone links, and the local links attaching the source node and the destination node to the backbone (701) are taken into account in the search process. The other links (702) are not taking part to the path sf lection which . reduces the cornplexity of the routing algorithm computation (for example the Bellman-Ford Algorithm). In our example, four paths - AWZE, ATYE, " ~
AXZE, AXYE - are satisfying the minimal hop constraint with a number of :
. three hops.
The complexity of the network represented in Figure S can be approximated .~ as follows:
j .
. .
'' ;
.....
.,.
. ~
:, ,~, i .., ~
:,, ,, :.
~:~
. .
: ~"
''' " 29 2123~9 ~.,.
.
: C = hopt . 2L = 1.69 x (2x23) = 77.7 with :
'.'~
. hopt = 1.69 hops . L = 23 links . . N = 11 nodes . 1 = 4.18 (46/11) links per node . .
~, The average hop number is determined from all the optimum (minimum) paths between the nodes:
L '~
.
~ I hop: AW, AT, AX, AB, BW, BX, CW, CX, FX, FZ, DE, DZ, DY, EZ, EY, WT, WX, WZ, XZ, XY, XT, TY, ZY.
o 2 hops: AC, AZ, AY, AF, BC, BZ, BY, BF, BT, CZ, CY, CF, CT, FD, FE, FY, FT, FW, DX, DT, DW, EX, ET, EW, ZT, YW.
~, o 3 hops: AE, AD, BD, BE, CD, CE
!i hopt = (lx23 ~ 2x26 + 3x6) / (23 + 26 + 6) = 93/55 -~ = 1.69 hops per optimum path :
.
~ With the assumption that the pre-selection of the usable links according to the ~..
present invention does not degrade the minimum hop constraint, the algorithm complexity can, in our example, be nearly reduced in a factor two:
.
:, C = hopt . 2L' = 1.69 x (2x13) = 43.9 with :
:i i . L' = 13 links i~:
~,....................... . hopt = 1.69 hops -~ This reduction of complexity can be much important typical networks.
~.
':i ~ R 9 93 014 ~, I!r ' , .
:, , ~..
::.
Claims (4)
1. A network access node for a packet switching communication network having a plurality of nodes interconnected with transmission links, said network access node including:
means for receiving and transmitting data packets, a network topology database manager for storing and updating data representing the characteristics and attributes of nodes and transmission links, said data forming a topology database defining the network configuration, link selecting means responsive to data contained in said topology database for identifying usable links for forming data paths with a destination node located in the network, said link selection means further including means for identifying each link in the network as either a backbone link or a local link, means for selecting as usable for a path determination all backbone links, means for selecting as usable for a path determination all local links attached to the network access node and to the destination node, and means for discarding as not usable for a path determination all other links, and means for storing link identifiers identifying selected links in the topology database; and means for determining, for a connection request to the destination node, an optimal routing path from said access node to the destination node using only said identified selected links.
means for receiving and transmitting data packets, a network topology database manager for storing and updating data representing the characteristics and attributes of nodes and transmission links, said data forming a topology database defining the network configuration, link selecting means responsive to data contained in said topology database for identifying usable links for forming data paths with a destination node located in the network, said link selection means further including means for identifying each link in the network as either a backbone link or a local link, means for selecting as usable for a path determination all backbone links, means for selecting as usable for a path determination all local links attached to the network access node and to the destination node, and means for discarding as not usable for a path determination all other links, and means for storing link identifiers identifying selected links in the topology database; and means for determining, for a connection request to the destination node, an optimal routing path from said access node to the destination node using only said identified selected links.
2. A network access node according to claim 1 wherein said routing path determination further includes means for implementing a modified Bellmann-Ford algorithm in making the optimal routing path determination.
3. A method for selecting a routing path in a packet communication network comprising a plurality of nodes interconnected with transmission links, said method being performed in a network access node and including the steps of:
storing and updating data in a network topology database, said data representingcharacteristics and attributes of the nodes and transmission links and defining the network configuration, selecting links usable for forming data paths between the network access node and a destination node in the network, said selecting step including the further steps of using data stored in the network topology database to identify each link in the network as either a backbone link or a local link, selecting as usable for a path determination all backbone links, selecting as usable all local links attached to the network access node and to the destination node, discarding as not usable for a path determination all other links, and storing in the topology database link identifiers identifying all selected links, determining, for a connection request to the destination node, an optimal routing path from said access node to the destination node using only the identified selected links.
storing and updating data in a network topology database, said data representingcharacteristics and attributes of the nodes and transmission links and defining the network configuration, selecting links usable for forming data paths between the network access node and a destination node in the network, said selecting step including the further steps of using data stored in the network topology database to identify each link in the network as either a backbone link or a local link, selecting as usable for a path determination all backbone links, selecting as usable all local links attached to the network access node and to the destination node, discarding as not usable for a path determination all other links, and storing in the topology database link identifiers identifying all selected links, determining, for a connection request to the destination node, an optimal routing path from said access node to the destination node using only the identified selected links.
4. A method for selecting a routing path according to claim 3 wherein said routing path determining step includes the further step of determining the optimal routing path through application of a modified Bellmann-Ford algorithm.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| EP93480104.4 | 1993-07-30 | ||
| EP93480104A EP0637152A1 (en) | 1993-07-30 | 1993-07-30 | Method and apparatus to speed up the path selection in a packet switching network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CA2123449A1 CA2123449A1 (en) | 1995-01-31 |
| CA2123449C true CA2123449C (en) | 1999-02-16 |
Family
ID=8214840
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA002123449A Expired - Lifetime CA2123449C (en) | 1993-07-30 | 1994-05-12 | Method and apparatus to speed up the path selection in a packet switching network |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US5491690A (en) |
| EP (1) | EP0637152A1 (en) |
| JP (1) | JP2685069B2 (en) |
| CA (1) | CA2123449C (en) |
Families Citing this family (157)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| NL9301544A (en) * | 1993-09-07 | 1995-04-03 | Nederland Ptt | Method for selecting connections in networks, and networks in which the method is applied. |
| US6044075A (en) * | 1994-12-15 | 2000-03-28 | International Business Machines Corporation | Apparatus and method for routing a communication in a network |
| DE4445800C1 (en) * | 1994-12-21 | 1996-07-04 | Siemens Ag | Method for forming routing information intended for the subsequent switching of traffic relationships in a communication network |
| US5712712A (en) * | 1995-06-01 | 1998-01-27 | Rapidata Systems, Inc. | Rapid delivery of facsimile or other data sets to a massive number of recipients |
| KR970004518A (en) * | 1995-06-09 | 1997-01-29 | 김광호 | Finding a Path for Packet Transmission in a Broadband Telecommunication System |
| US6393486B1 (en) | 1995-06-23 | 2002-05-21 | Cisco Technology, Inc. | System and method using level three protocol information for network centric problem analysis and topology construction of actual or planned routed network |
| US6883034B1 (en) | 1995-06-23 | 2005-04-19 | Cisco Technology, Inc. | Method of resolving conflicts in access control lists in router by comparing elements in the lists based on subsumption relations |
| US5699347A (en) * | 1995-11-17 | 1997-12-16 | Bay Networks, Inc. | Method and apparatus for routing packets in networks having connection-oriented subnetworks |
| EP0781007B1 (en) * | 1995-12-21 | 2003-03-12 | Siemens Aktiengesellschaft | Method for creating routing information in an ATM communications network |
| US6237029B1 (en) | 1996-02-26 | 2001-05-22 | Argosystems, Inc. | Method and apparatus for adaptable digital protocol processing |
| US6279039B1 (en) * | 1996-04-03 | 2001-08-21 | Ncr Corporation | Resource management method and apparatus for maximizing multimedia performance of open systems |
| GB2312592A (en) * | 1996-04-24 | 1997-10-29 | Ibm | Quality of service parameters |
| US5854899A (en) * | 1996-05-09 | 1998-12-29 | Bay Networks, Inc. | Method and apparatus for managing virtual circuits and routing packets in a network/subnetwork environment |
| DE69720857T2 (en) * | 1996-05-31 | 2004-02-05 | Hewlett-Packard Co. (N.D.Ges.D.Staates Delaware), Palo Alto | Systems and methods for operating a network management station |
| US6400681B1 (en) | 1996-06-20 | 2002-06-04 | Cisco Technology, Inc. | Method and system for minimizing the connection set up time in high speed packet switching networks |
| US5787271A (en) * | 1996-06-26 | 1998-07-28 | Mci Corporation | Spare capacity allocation tool |
| US5940396A (en) * | 1996-08-21 | 1999-08-17 | 3Com Ltd. | Method of routing in an asynchronous transfer mode network |
| US6058419A (en) * | 1996-10-30 | 2000-05-02 | Nec Corporation | Communication network control system and control method therefor |
| US6016307A (en) | 1996-10-31 | 2000-01-18 | Connect One, Inc. | Multi-protocol telecommunications routing optimization |
| US6252856B1 (en) * | 1996-12-03 | 2001-06-26 | Nortel Networks Limited | Method and apparatus for minimizing calculations required to construct multicast trees |
| SE9604491L (en) | 1996-12-05 | 1998-06-06 | Ericsson Telefon Ab L M | Device and method of transmission system |
| US6038212A (en) * | 1996-12-13 | 2000-03-14 | International Business Machines Corporation | Method and system for optimizing the connection set up time in high speed communication networks for recovering from network failure |
| US5951644A (en) * | 1996-12-24 | 1999-09-14 | Apple Computer, Inc. | System for predicting and managing network performance by managing and monitoring resourse utilization and connection of network |
| US6122514A (en) * | 1997-01-03 | 2000-09-19 | Cellport Systems, Inc. | Communications channel selection |
| US6934249B1 (en) | 1997-04-01 | 2005-08-23 | Cisco Technology, Inc. | Method and system for minimizing the connection set up time in high speed packet switching networks |
| US6097727A (en) * | 1997-04-29 | 2000-08-01 | International Business Machines Corporation | Methods, systems and computer program products for end-to-end route selection in compound wide/local area networks |
| US6690681B1 (en) | 1997-05-19 | 2004-02-10 | Airbiquity Inc. | In-band signaling for data communications over digital wireless telecommunications network |
| US6493338B1 (en) | 1997-05-19 | 2002-12-10 | Airbiquity Inc. | Multichannel in-band signaling for data communications over digital wireless telecommunications networks |
| US5974460A (en) * | 1997-06-16 | 1999-10-26 | International Business Machines Corporation | Apparatus and method for selecting an optimum telecommunications link |
| US6151305A (en) * | 1997-11-14 | 2000-11-21 | Nortel Networks Limited | Method and apparatus for planning an ATM network for an area of concentrated demand for transport bandwidth |
| US6246669B1 (en) | 1997-11-28 | 2001-06-12 | Cisco Technology, Inc. | Method and system for optimizing connection set-up operations in a high speed digital network |
| SE510146C2 (en) * | 1997-12-19 | 1999-04-26 | Telia Ab | External road handler to complement distributed routing protocols |
| DE19802599C1 (en) | 1998-01-23 | 1999-05-27 | Siemens Ag | Connection path determination in communications network |
| DE19810559C2 (en) * | 1998-03-11 | 2000-05-11 | Siemens Ag | Switchboard |
| US6208623B1 (en) | 1998-04-13 | 2001-03-27 | 3Com Corporation | Method of combining PNNI and E-IISP in an asynchronous transfer mode network |
| US6130881A (en) * | 1998-04-20 | 2000-10-10 | Sarnoff Corporation | Traffic routing in small wireless data networks |
| US6639897B1 (en) * | 1998-04-22 | 2003-10-28 | Nippon Telegraph And Telephone Corporation | Communication network of linked nodes for selecting the shortest available route |
| US6396842B1 (en) | 1998-04-30 | 2002-05-28 | 3Com Corporation | Method of searching using longest match based Randix Search Trie with variable length keys and having prefix capability |
| US6212188B1 (en) | 1998-05-01 | 2001-04-03 | 3Com Corporation | Method of source routing in an asynchronous transfer mode network when a node is in an overload state |
| US6192043B1 (en) | 1998-05-01 | 2001-02-20 | 3Com Corporation | Method of caching routes in asynchronous transfer mode PNNI networks |
| US6262984B1 (en) | 1998-05-12 | 2001-07-17 | 3Com Corporation | Method of preventing overlapping branches in point to multipoint calls in PNNI networks |
| US6223149B1 (en) | 1998-05-28 | 2001-04-24 | 3Com Corporation | Non-distributed LAN emulation server redundancy method |
| US6205146B1 (en) | 1998-05-28 | 2001-03-20 | 3Com Corporation | Method of dynamically routing to a well known address in a network |
| US6466574B1 (en) * | 1998-06-05 | 2002-10-15 | International Business Machines Corporation | Quality of service improvement of internet real-time media transmission by transmitting redundant voice/media frames |
| GB2339108A (en) | 1998-07-01 | 2000-01-12 | Ericsson Telefon Ab L M | Call routing data management |
| US6147971A (en) * | 1998-11-18 | 2000-11-14 | 3Com Corporation | Optimized routing method based on minimal hop count for use in PNNI based asynchronous transfer mode networks |
| US6493317B1 (en) * | 1998-12-18 | 2002-12-10 | Cisco Technology, Inc. | Traffic engineering technique for routing inter-class traffic in a computer network |
| US6584102B1 (en) * | 1998-12-21 | 2003-06-24 | At&T Corp. | Communication network apparatus and method |
| US7012892B1 (en) * | 1999-04-16 | 2006-03-14 | Alcatel Canada Inc. | Method and apparatus for supporting connection type partitioning in a communications network |
| US6456600B1 (en) | 1999-04-28 | 2002-09-24 | 3Com Corporation | Complex node representation in an asynchronous transfer mode PNNI network |
| US6483808B1 (en) | 1999-04-28 | 2002-11-19 | 3Com Corporation | Method of optimizing routing decisions over multiple parameters utilizing fuzzy logic |
| US6594235B1 (en) | 1999-04-28 | 2003-07-15 | 3Com Corporation | Method of triggering reroutes in an asynchronous transfer mode network |
| US6577653B1 (en) | 1999-04-28 | 2003-06-10 | 3Com Corporation | Apparatus for and method of establishing a route utilizing multiple parallel segments in an asynchronous transfer mode network |
| US6473408B1 (en) | 1999-05-19 | 2002-10-29 | 3Com Corporation | Building a hierarchy in an asynchronous transfer mode PNNI network utilizing proxy SVCC-based RCC entities |
| US6614792B1 (en) | 1999-05-27 | 2003-09-02 | 3Com Corporation | Proxy MPC for providing MPOA services to legacy lane clients in an asynchronous transfer mode network |
| US6275470B1 (en) | 1999-06-18 | 2001-08-14 | Digital Island, Inc. | On-demand overlay routing for computer-based communication networks |
| AUPQ128699A0 (en) | 1999-06-30 | 1999-07-22 | Telefonaktiebolaget Lm Ericsson (Publ) | A scalable computer system |
| US6678241B1 (en) | 1999-11-30 | 2004-01-13 | Cisc Technology, Inc. | Fast convergence with topology switching |
| AU2072601A (en) * | 1999-12-09 | 2001-06-18 | Zephyr Media, Inc. | System and method for integration of a universally publicly accessible global network |
| JP2001217839A (en) * | 2000-01-31 | 2001-08-10 | Fujitsu Ltd | Node device |
| GB0008364D0 (en) | 2000-04-06 | 2000-05-24 | Hewlett Packard Co | Deriving location information about a communication entity |
| US6836465B2 (en) * | 2001-11-29 | 2004-12-28 | Ipsum Networks, Inc. | Method and system for path identification in packet networks |
| US8510468B2 (en) | 2000-04-17 | 2013-08-13 | Ciradence Corporation | Route aware network link acceleration |
| JP3501093B2 (en) | 2000-04-18 | 2004-02-23 | 日本電気株式会社 | QoS path calculator |
| US6944130B1 (en) * | 2000-05-31 | 2005-09-13 | Cisco Technology, Inc. | Method and apparatus for determining a layer 2 path in a switched network |
| US6781959B1 (en) | 2000-07-31 | 2004-08-24 | Cisco Technology, Inc. | Method and apparatus for determining troubleshooting information for completed calls in a telecommunications network |
| CA2463479A1 (en) | 2000-10-10 | 2002-04-18 | Radiant Networks Plc | Communications meshes |
| JP3729051B2 (en) * | 2000-10-18 | 2005-12-21 | 日本電気株式会社 | Inter-domain routing apparatus, system and method |
| US7003559B1 (en) | 2000-10-23 | 2006-02-21 | Hewlett-Packard Development Company, L.P. | System and method for determining probable network paths between nodes in a network topology |
| US20030128987A1 (en) * | 2000-11-08 | 2003-07-10 | Yaron Mayer | System and method for improving the efficiency of routers on the internet and/or cellular networks an/or other networks and alleviating bottlenecks and overloads on the network |
| WO2003060624A2 (en) * | 2001-12-28 | 2003-07-24 | Yaron Mayer | System and method for improving network router efficiency and alleviating network bottlenecks and overloads |
| US6917970B2 (en) * | 2001-03-09 | 2005-07-12 | Sun Microsystems, Inc. | Method and program code for identifying new nodes on a bus following a reset |
| DE10122422A1 (en) | 2001-05-09 | 2002-11-21 | Siemens Ag | Method for adjusting bandwidth in a connection between two communications terminals in a data network allocates a transmission channel to the connection for transmitting data. |
| US6992988B2 (en) * | 2001-08-20 | 2006-01-31 | Sun Microsystems, Inc. | System and method for deadlock-free routing on arbitrary network topologies |
| WO2003032676A1 (en) * | 2001-10-04 | 2003-04-17 | Siemens Aktiengesellschaft | Distributed transmission of traffic flows in communication networks |
| US7152113B2 (en) * | 2001-10-19 | 2006-12-19 | Sun Microsystems, Inc. | Efficient system and method of node and link insertion for deadlock-free routing on arbitrary topologies |
| US7215965B2 (en) | 2001-11-01 | 2007-05-08 | Airbiquity Inc. | Facility and method for wireless transmission of location data in a voice channel of a digital wireless telecommunications network |
| US7200117B2 (en) * | 2002-01-31 | 2007-04-03 | Sun Microsystems, Inc. | Method of optimizing network capacity and fault tolerance in deadlock-free routing |
| US7068610B2 (en) | 2002-02-26 | 2006-06-27 | Unruh Lincoln J | System and method for reliable communications over multiple packet RF networks |
| US7072304B2 (en) * | 2002-02-27 | 2006-07-04 | Nortel Networks Limited | Network path selection based on bandwidth |
| DE10238290A1 (en) * | 2002-08-21 | 2004-03-04 | Siemens Ag | Distribution compartments for efficient, fail-safe traffic distribution in a packet network |
| US7551634B2 (en) | 2002-11-12 | 2009-06-23 | Fujitsu Limited | Communication network system |
| JP2004206459A (en) * | 2002-12-25 | 2004-07-22 | Matsushita Electric Ind Co Ltd | Session management device |
| WO2004068424A2 (en) | 2003-01-28 | 2004-08-12 | Cellport Systems, Inc. | Secure telematics |
| GB0306855D0 (en) * | 2003-03-25 | 2003-04-30 | Ideas Network Ltd | Data communication network |
| US20040264503A1 (en) * | 2003-06-30 | 2004-12-30 | Microsoft Corporation | Method and system for providing a virtual protocol interlayer |
| US7693143B2 (en) * | 2003-08-15 | 2010-04-06 | Accton Technology Corporation | Forwarding and routing method for wireless transport service |
| US7975035B2 (en) * | 2003-12-01 | 2011-07-05 | International Business Machines Corporation | Method and apparatus to support application and network awareness of collaborative applications using multi-attribute clustering |
| US7392482B2 (en) | 2003-12-18 | 2008-06-24 | International Business Machines Corporation | Selection accelerator in topology views |
| US9160571B2 (en) * | 2004-03-11 | 2015-10-13 | Hewlett-Packard Development Company, L.P. | Requesting a service from a multicast network |
| US7450495B2 (en) * | 2004-03-15 | 2008-11-11 | Iptivia, Inc. | Method and system for path change root-cause identification in packet networks |
| JP2005266933A (en) * | 2004-03-16 | 2005-09-29 | Fujitsu Ltd | Storage management system and storage management method |
| US7346370B2 (en) * | 2004-04-29 | 2008-03-18 | Cellport Systems, Inc. | Enabling interoperability between distributed devices using different communication link technologies |
| US20060067337A1 (en) * | 2004-09-30 | 2006-03-30 | Netravali Arun N | Methods and devices for generating a hierarchical structure for the internet |
| DE102004063461B4 (en) * | 2004-12-23 | 2008-06-26 | Nec Europe Ltd. | A method of searching for services, resources and / or functionalities in a network |
| US7664037B2 (en) | 2005-01-04 | 2010-02-16 | Intel Corporation | Multichannel mesh network, multichannel mesh router and methods for routing using bottleneck channel identifiers |
| JP4571182B2 (en) * | 2005-01-27 | 2010-10-27 | 富士通株式会社 | NETWORK DEVICE MANAGEMENT DEVICE, NETWORK DEVICE MANAGEMENT METHOD, NETWORK DEVICE, PROGRAM USED FOR THEM, AND SYSTEM |
| US7508810B2 (en) * | 2005-01-31 | 2009-03-24 | Airbiquity Inc. | Voice channel control of wireless packet data communications |
| JP4073923B2 (en) * | 2005-03-30 | 2008-04-09 | 富士通株式会社 | Network device management apparatus, network device management program, and network device management method |
| US8014942B2 (en) * | 2005-06-15 | 2011-09-06 | Airbiquity, Inc. | Remote destination programming for vehicle navigation |
| US7474620B2 (en) * | 2005-07-10 | 2009-01-06 | Hewlett-Packard Development Company, L.P. | Communications network having transport diversity |
| US7688739B2 (en) * | 2005-08-02 | 2010-03-30 | Trilliant Networks, Inc. | Method and apparatus for maximizing data transmission capacity of a mesh network |
| JP4725382B2 (en) * | 2006-03-23 | 2011-07-13 | 富士通株式会社 | Packet relay method and packet relay node |
| US8300798B1 (en) | 2006-04-03 | 2012-10-30 | Wai Wu | Intelligent communication routing system and method |
| US7924934B2 (en) | 2006-04-07 | 2011-04-12 | Airbiquity, Inc. | Time diversity voice channel data communications |
| US7756063B2 (en) * | 2007-01-19 | 2010-07-13 | Lockheed Martin Corporation | Parallel wireless topologies |
| US7756064B2 (en) * | 2007-01-19 | 2010-07-13 | Lockheed Martin Corporation | Reduced complexity staged topologies for wireless communications |
| US7756062B2 (en) * | 2007-01-19 | 2010-07-13 | Lockheed Martin Corporation | Periodically reconfigured networks |
| US7962595B1 (en) * | 2007-03-20 | 2011-06-14 | Emc Corporation | Method and apparatus for diagnosing host to storage data path loss due to FibreChannel switch fabric splits |
| US8027293B2 (en) * | 2007-07-16 | 2011-09-27 | Cellport Systems, Inc. | Communication channel selection and use |
| EP2206328B1 (en) | 2007-10-20 | 2017-12-27 | Airbiquity Inc. | Wireless in-band signaling with in-vehicle systems |
| US7801028B2 (en) * | 2007-12-31 | 2010-09-21 | Schneider Automation Inc. | Method and apparatus for transparent auto-recovery in chain and ring networks |
| US8082290B2 (en) * | 2008-03-19 | 2011-12-20 | Verizon Patent And Licensing Inc. | Intelligent establishment of peer-to-peer communication |
| US8594138B2 (en) | 2008-09-15 | 2013-11-26 | Airbiquity Inc. | Methods for in-band signaling through enhanced variable-rate codecs |
| US7983310B2 (en) * | 2008-09-15 | 2011-07-19 | Airbiquity Inc. | Methods for in-band signaling through enhanced variable-rate codecs |
| KR100959077B1 (en) * | 2008-09-19 | 2010-05-20 | 한국전자통신연구원 | Method for gap analysis for network topology inspection in ethernet-based network |
| US7860100B2 (en) * | 2008-10-01 | 2010-12-28 | Cisco Technology, Inc. | Service path selection in a service network |
| US8442043B2 (en) * | 2008-12-29 | 2013-05-14 | Cisco Technology, Inc. | Service selection mechanism in service insertion architecture data plane |
| US8036600B2 (en) | 2009-04-27 | 2011-10-11 | Airbiquity, Inc. | Using a bluetooth capable mobile phone to access a remote network |
| US8418039B2 (en) | 2009-08-03 | 2013-04-09 | Airbiquity Inc. | Efficient error correction scheme for data transmission in a wireless in-band signaling system |
| US8249865B2 (en) | 2009-11-23 | 2012-08-21 | Airbiquity Inc. | Adaptive data transmission for a digital in-band modem operating over a voice channel |
| US8965291B2 (en) | 2010-07-13 | 2015-02-24 | United Technologies Corporation | Communication of avionic data |
| US8743885B2 (en) | 2011-05-03 | 2014-06-03 | Cisco Technology, Inc. | Mobile service routing in a network environment |
| US8848825B2 (en) | 2011-09-22 | 2014-09-30 | Airbiquity Inc. | Echo cancellation in wireless inband signaling modem |
| US9794379B2 (en) | 2013-04-26 | 2017-10-17 | Cisco Technology, Inc. | High-efficiency service chaining with agentless service nodes |
| US9525638B2 (en) | 2013-10-15 | 2016-12-20 | Internap Corporation | Routing system for internet traffic |
| US9379931B2 (en) | 2014-05-16 | 2016-06-28 | Cisco Technology, Inc. | System and method for transporting information to services in a network environment |
| US9479443B2 (en) | 2014-05-16 | 2016-10-25 | Cisco Technology, Inc. | System and method for transporting information to services in a network environment |
| US10417025B2 (en) | 2014-11-18 | 2019-09-17 | Cisco Technology, Inc. | System and method to chain distributed applications in a network environment |
| USRE48131E1 (en) | 2014-12-11 | 2020-07-28 | Cisco Technology, Inc. | Metadata augmentation in a service function chain |
| US9660909B2 (en) | 2014-12-11 | 2017-05-23 | Cisco Technology, Inc. | Network service header metadata for load balancing |
| US9762402B2 (en) | 2015-05-20 | 2017-09-12 | Cisco Technology, Inc. | System and method to facilitate the assignment of service functions for service chains in a network environment |
| US11044203B2 (en) | 2016-01-19 | 2021-06-22 | Cisco Technology, Inc. | System and method for hosting mobile packet core and value-added services using a software defined network and service chains |
| US10187306B2 (en) | 2016-03-24 | 2019-01-22 | Cisco Technology, Inc. | System and method for improved service chaining |
| US10931793B2 (en) | 2016-04-26 | 2021-02-23 | Cisco Technology, Inc. | System and method for automated rendering of service chaining |
| US10419550B2 (en) | 2016-07-06 | 2019-09-17 | Cisco Technology, Inc. | Automatic service function validation in a virtual network environment |
| US10320664B2 (en) | 2016-07-21 | 2019-06-11 | Cisco Technology, Inc. | Cloud overlay for operations administration and management |
| US10218616B2 (en) | 2016-07-21 | 2019-02-26 | Cisco Technology, Inc. | Link selection for communication with a service function cluster |
| US10225270B2 (en) | 2016-08-02 | 2019-03-05 | Cisco Technology, Inc. | Steering of cloned traffic in a service function chain |
| US10218593B2 (en) | 2016-08-23 | 2019-02-26 | Cisco Technology, Inc. | Identifying sources of packet drops in a service function chain environment |
| US10361969B2 (en) | 2016-08-30 | 2019-07-23 | Cisco Technology, Inc. | System and method for managing chained services in a network environment |
| US10225187B2 (en) | 2017-03-22 | 2019-03-05 | Cisco Technology, Inc. | System and method for providing a bit indexed service chain |
| US10257033B2 (en) | 2017-04-12 | 2019-04-09 | Cisco Technology, Inc. | Virtualized network functions and service chaining in serverless computing infrastructure |
| US10178646B2 (en) | 2017-04-12 | 2019-01-08 | Cisco Technology, Inc. | System and method to facilitate slice management in a network environment |
| US10884807B2 (en) | 2017-04-12 | 2021-01-05 | Cisco Technology, Inc. | Serverless computing and task scheduling |
| US10333855B2 (en) | 2017-04-19 | 2019-06-25 | Cisco Technology, Inc. | Latency reduction in service function paths |
| US10554689B2 (en) | 2017-04-28 | 2020-02-04 | Cisco Technology, Inc. | Secure communication session resumption in a service function chain |
| US10735275B2 (en) | 2017-06-16 | 2020-08-04 | Cisco Technology, Inc. | Releasing and retaining resources for use in a NFV environment |
| US10798187B2 (en) | 2017-06-19 | 2020-10-06 | Cisco Technology, Inc. | Secure service chaining |
| US10397271B2 (en) | 2017-07-11 | 2019-08-27 | Cisco Technology, Inc. | Distributed denial of service mitigation for web conferencing |
| US10673698B2 (en) | 2017-07-21 | 2020-06-02 | Cisco Technology, Inc. | Service function chain optimization using live testing |
| US11063856B2 (en) | 2017-08-24 | 2021-07-13 | Cisco Technology, Inc. | Virtual network function monitoring in a network function virtualization deployment |
| US10791065B2 (en) | 2017-09-19 | 2020-09-29 | Cisco Technology, Inc. | Systems and methods for providing container attributes as part of OAM techniques |
| US11018981B2 (en) | 2017-10-13 | 2021-05-25 | Cisco Technology, Inc. | System and method for replication container performance and policy validation using real time network traffic |
| US10541893B2 (en) | 2017-10-25 | 2020-01-21 | Cisco Technology, Inc. | System and method for obtaining micro-service telemetry data |
| US10666612B2 (en) | 2018-06-06 | 2020-05-26 | Cisco Technology, Inc. | Service chains for inter-cloud traffic |
| CN112804117A (en) * | 2020-09-04 | 2021-05-14 | 中国航空无线电电子研究所 | Avionic network time delay calculation method |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JPS62256547A (en) * | 1986-04-30 | 1987-11-09 | Toshiba Corp | Route selecting system |
| US4827411A (en) * | 1987-06-15 | 1989-05-02 | International Business Machines Corporation | Method of maintaining a topology database |
| FR2643532B1 (en) * | 1989-02-17 | 1991-05-10 | France Etat | METHOD FOR RESERVING RATES AND TIME SWITCHES OF ASYNCHRONOUS PACKETS |
| GB8911395D0 (en) * | 1989-05-18 | 1989-07-05 | British Telecomm | Data communications network |
| US5115433A (en) * | 1989-07-18 | 1992-05-19 | Metricom, Inc. | Method and system for routing packets in a packet communication network |
| US5216591A (en) * | 1990-02-06 | 1993-06-01 | Sprint International Communications Corp. | Method for efficient distributed data communications network backbone node location |
| US5265092A (en) * | 1992-03-18 | 1993-11-23 | Digital Equipment Corporation | Synchronization mechanism for link state packet routing |
| US5233604A (en) * | 1992-04-28 | 1993-08-03 | International Business Machines Corporation | Methods and apparatus for optimum path selection in packet transmission networks |
-
1993
- 1993-07-30 EP EP93480104A patent/EP0637152A1/en not_active Withdrawn
-
1994
- 1994-05-12 CA CA002123449A patent/CA2123449C/en not_active Expired - Lifetime
- 1994-07-07 JP JP15559994A patent/JP2685069B2/en not_active Expired - Lifetime
- 1994-07-22 US US08/279,373 patent/US5491690A/en not_active Expired - Lifetime
Also Published As
| Publication number | Publication date |
|---|---|
| US5491690A (en) | 1996-02-13 |
| CA2123449A1 (en) | 1995-01-31 |
| JP2685069B2 (en) | 1997-12-03 |
| EP0637152A1 (en) | 1995-02-01 |
| JPH0766835A (en) | 1995-03-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CA2123449C (en) | Method and apparatus to speed up the path selection in a packet switching network | |
| US5495479A (en) | Method and apparatus for an automatic decomposition of a network topology into a backbone and subareas | |
| US6038212A (en) | Method and system for optimizing the connection set up time in high speed communication networks for recovering from network failure | |
| US5101348A (en) | Method of reducing the amount of information included in topology database update messages in a data communications network | |
| JP2755344B2 (en) | Packet switched communication network, method of generating spanning tree, and method and apparatus for transmitting extended information packet | |
| JPH05199227A (en) | Communication network. | |
| US9397924B2 (en) | Method for applying macro-controls onto IP networks using intelligent route indexing | |
| US7843837B2 (en) | Management of protocol information in PNNI hierarchical networks | |
| US7414985B1 (en) | Link aggregation | |
| Gerstel et al. | The layout of virtual paths in ATM networks | |
| US8948178B2 (en) | Network clustering | |
| EP1009130A1 (en) | Distributed directory services for locating network resources in a very large packet switching network | |
| EP0954916A1 (en) | Connectionless group addressing for directory services in high speed packet switching networks | |
| WO2022269357A1 (en) | Deadlock-free multipath routing for direct interconnect networks | |
| JP3184822B2 (en) | Method, network apparatus and PNNI network for generating optimal PNNI compound node representation for limited cost | |
| US6515998B1 (en) | Table data retrieving apparatus retrieving table in which reference data is stored by using retrieval key | |
| US7532584B2 (en) | Implementation of constraints to ensure deadlock avoidance in networks | |
| CN1898903B (en) | Method and apparatus for performing routing operations in communications network | |
| Wang et al. | A survey and comparison of multi-ring techniques for scalable battlespace group communications | |
| US20220086085A1 (en) | Neural Network for Secure Data Transport, System and Method | |
| Bujewski et al. | The Supercomputer Supernet Testbed: A WDM-Based Supercomputer Interconnect | |
| Kos et al. | Topological planning of communication networks | |
| EP1413113B1 (en) | Routing of peer-to-peer messages in a communication network | |
| Katz et al. | SDH management network: Architecture, routing and addressing | |
| Tsuchiya | An architecture for network-layer routing in OSI |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| EEER | Examination request | ||
| MKEX | Expiry |
Effective date: 20140512 |