CA2249169A1 - Mechanism for dispatching packets via a telecommunications network - Google Patents
Mechanism for dispatching packets via a telecommunications network Download PDFInfo
- Publication number
- CA2249169A1 CA2249169A1 CA002249169A CA2249169A CA2249169A1 CA 2249169 A1 CA2249169 A1 CA 2249169A1 CA 002249169 A CA002249169 A CA 002249169A CA 2249169 A CA2249169 A CA 2249169A CA 2249169 A1 CA2249169 A1 CA 2249169A1
- Authority
- CA
- Canada
- Prior art keywords
- packet
- queued
- queue
- new
- new packet
- 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.)
- Abandoned
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
- H04L49/9021—Plurality of buffers per packet
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/50—Queue scheduling
- H04L47/62—Queue scheduling characterised by scheduling criteria
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
- H04L49/901—Buffering arrangements using storage descriptor, e.g. read or write pointers
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
- H04L49/9084—Reactions to storage capacity overflow
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L9/00—Cryptographic mechanisms or cryptographic arrangements for secret or secure communications; Network security protocols
- H04L9/40—Network security protocols
Abstract
A mechanism for dispatching a sequence of packets via a telecommunications network includes a queue for packets for transmission and a queue controller responsive to receipt of a new packet for transmission to compare parameters of the new packet to parameters of any packet already in the queue, the queue controller determining whether to queue or drop the new packet depending on the result of the comparison(s). The queue can be implemented as a linked list of packet entries with individual pointers to the respective packets concerned. The queue entries can include details relating to the packet including data relating to the information flow and also the packet identity. In a TCP environment, the flow information can include the source IP address and the source TCP port, as well as the destination IP address and the destination TCP port.
The identity information can include sequence numbers and acknowledgement numbers for the packet concerned. In order to optimize network usage, it can be useful to drop some packets at a routing node. A decision to drop a packet can be made if the new packet and a queued packet relate to the same information flow, the new packet sequence number equals the queued packet sequence number and the new packet acknowledgement number is less than the queued packet acknowledgement number. The new packet is dropped where the new packet is a retransmission of a queued packet and the length of the queued packet is greater than or equal to that of the new packet. A queued packet is replaced by a new packet when the new packet is determined to be a retransmission of the queued packet and the length of the new packet is greater than that of the queued packet.
The identity information can include sequence numbers and acknowledgement numbers for the packet concerned. In order to optimize network usage, it can be useful to drop some packets at a routing node. A decision to drop a packet can be made if the new packet and a queued packet relate to the same information flow, the new packet sequence number equals the queued packet sequence number and the new packet acknowledgement number is less than the queued packet acknowledgement number. The new packet is dropped where the new packet is a retransmission of a queued packet and the length of the queued packet is greater than or equal to that of the new packet. A queued packet is replaced by a new packet when the new packet is determined to be a retransmission of the queued packet and the length of the new packet is greater than that of the queued packet.
Description
.. ..
SUN REF: P2685 BACKGROUND OF THE INVENTION
This invention relates to a mechanism for dispatching packets via a telecommllnications network and to a network sender or router station, or node, incorporating such a mech~nism The invention finds particular application to tr~ncmi~cion of TCP information via an inter-or intra-network operating under an Internet Protocol.
Figure 1 is a schematic representation of an inct~n~e of an Inter- or Intra-net with a router 10 being provided in the path between a source 12 and a destination 14.
Between the source 12 (or sender node) and the router node 10, a net 16 is shownand between the router node 10 and the dçstin~tion node 14 a further net 18 is shown.
In practice, the net 16 and the net 18 can be one and the same and the router 10errt;~;Liv~ly forms a "staging post" between the source 12 and the destination 14. In the following, reference is made to a dispatch mech~ni~m It should be appreciated t_at the dispatch mechanism could, in the present context, equally form part of the source 12, as opposed to being a separate "staging post" as illustrated in Figure 1.
Figure 2 is a schematic representation of the configuration of a station for a router 10 or source or dçstin~tion 12, 14. These stations can be implemented using any a~roL,liate technology. However, as illustrated in Figure 2, the station is implemented by a server co~ uLer 20 coulplisillg a system unit 22, optionally with a display 38, keyboard 40 and other input devices 42. It should be noted that the router 10 need not include a keyboard, display, etc. Figure 2A is a schematic block representation of aspects of the contents of the system unit 22. As illustrated in Figure 2A, the system unit includes a processor 28, memory 30, disk drives 24 and 26, and a communications adaptor 32 for connection to one or more telecommunications lines 34 for connection to the telecollllllunications network 16/18.
As illustrated in Figure 2A, the components of the system unit are connected via a bus arrangement 36. It will be appreciated that Figures 2/2A are a general schematic representation of one possible configuration for a server co~ ul~l for forming arouter or sender or ~1~stin~tion station and that many alternative configurations could be provided.
Conceptually, the Internet provides three levels of services. At the lowest ~. t.
SUN REF: P2685 level, a connectionless delivery system provides a foundation on which everything rests. At the next level, a reliable transport service provides a high level platform.
At the third level, application services are provided which rely on the reliabletransport service.
S A filntl~mental Internet service consists of an unreliable, best-effort,connectionless, packet delivery system. The service is described as being "unreliable" because delivery is not guaranteed. A packet may be lost, duplicated, or delivered out of order, but the Internet will not detect such conditions, nor will it inform the sender or receiver. The service is described as being "connectionless"
because each packet is treated independently from all others. A sequence of packets sent from one machine to another may travel over different paths, or some may belost while others are delivered. The service may be described as "best-effort"
because the Internet makes an earnest attempt to deliver packets.
The protocol that defines the unreliable, connectionless, delivery mechanism is called the "Internet Protocol", and is usually referred to by its initials IP. IP
defines the formal specification of data formats, including a basic unit of data transfer and the exact format of all data passing across the Internet. IP also includes rules which specify how packets should be processed and how errors should be handled.
In particular, IP embodies the idea of unreliable delivery and packet routing.
Further details of aspects of the Internet and TCP/IP protocols may be found, for example, in the following US patents: 5,293,379; 5,307,347; 5,307,413;
5,309,437; 5,351,237; and 5,535,199.
The basic unit of data transfer via the IP is termed an "Internet datagram", or alternative "IP datagram", or simply "datagram". A datagram comprises header anddata areas, and source and flestin~tion addresses. There is no fixed size for a datagram. Bearing this in mind, and also the physical constraints of the underlying hardware services on which the Internet is based, it is n~cçss~ry to divide the datagram into portions called "fragments".
Figure SA illustrates the format of an Internet datagram. The same format is used for a fragment of a datagram.
The 4 bit version field (VERS) specifies the IP protocol version and is used SUN REF: P2685 to ensure that all of the nodes along the path of the datagram agree on the format.
The LEN field gives the datagram header length measured in 32 bit words.
The TOTAL LENGTH field gives the length of the IP datagram measured in octets including the length of the header and data.
The SERVICE TYPE field contains h~n-lling details for the datagram.
Three fields in the datagram header, IDENT, FLAGS, and FRAGMENT
OFFSET, control fragmentation and reassembly of datagrams. The field IDENT
contains a unique identifier that identifies the datagram.
In the FLAGS field, a first bit specifies whether the datagram may be 10 fragmented, and a second bit in~lic~t~s whether this is the last fragment in the datagram. The FRAGMENT OFFSET field specifies the offset of this fragment in the original datagram, measured in units of 8 octets, starting at offset zero.
As each fragment has the same basic header format as a complete datagram, the combination of the FLAGS and FRAGMENT OFFSET fields are used to indicate 15 that the headers relate to fragments, and to in-lic~t~- the position of the fragment within the original datagram. The FRAGMENT OFFSET field identifies the position within the datagram, and the second of the FLAGS bits mentioned above (which is sometimes called the MORE FRAGMENTS flag) is used to indicate whether there are any more fragments in the datagram, or conversely that the fragment concerned 20 is the last fragment of the datagram.
The field PROTO is a form of type field. The HEADER CHECK SUM
figure ensures integrity of header values.
SOURCE IP ADDRESS and DESTINATION IP ADDRESS contain 32 bit Internet addresses of the datagram's sender and intended recipient. The OPTIONS
25 field and the PADDING field are optional in the datagram. The field labelled DATA
~;pl~sell~ the beginning of the data field.
As mentioned above, above the IP layer of the Internet protocol structure one service which is provided is a reliable transport service which is typically called the "reliable stream transport service", defined by the Tr~n~mi~ion Control Protocol30 (TCP). Although TCP is provided over the Internet, it is in fact an independent general purpose protocol which can also be used with other delivery systems. TCP
SUN REF: P2685 makes very few as~ulup~ions regarding the underlying network, and it can also beused over a single network like Ethernet, as well as over a complex Internet, orIntranet.
TCP provides a reliable stream delivery service which can be contrasted with the unreliable datagram protocol (UDP) which is also provided over the Internet.Whereas UDP provides an unreliable delivery service because delivery is not guaranteed, TCP provides a more complex structure which does ensure reliable delivery in the form of a stream.
UDP provides unreliable packet delivering, whereby packets may be lost or 10 destroyed when k~n~mi~sion errors il,l~,rl ~; with data, when network hardware fails, or when networks become too heavily loaded to accommodate the load presented.
TCP on the other hand, operates by providing delivery by means of a stream of bits, divided into eight-bit octets or bytes.
Given that the underlying Internet protocol is unreliable, TCP tr~n~mi~ions 15 operate in accordance with a technique known as positive acknowledgement withreL,~ i",ic~ion. The technique requires a recipient to co"-",ll.~ic~te with the source, sending back an acknowledgement message every time it receives data. The sender keeps a record of each packet that it sends and waits for an acknowledgement before sending the next packet. The sender also starts a timer when it sends its packet and 20 retransmits a packet if the timer expires before the acknowledgement arrives. Figure 3A is a schematic representation of the tr~n~mi~ion and receipt of packets and acknowledgements. The left hand side of Figure 3A represents events at a sender side 50, the right hand side represents events at a receiver side 52 and the middle portion represents network messages passing between the sender and the receiver.
At 54, the sender 50 (eg, the router 10) sends a packet P1 to the receiver (eg, the destination 14) via the network and starts a timer for message P1. When the receiver 52 receives, 56, the packet P1; the receiver then sends, 58, an acknowledgement A1. When the acknowledgement A1 is received, 60, at the sender 50, the sender can cancel the timer and send 62, the next packet P2 to the receiver 30 52 setting a timer for the message P2. When the receiver 52 receives, 64, the packet P2, it sends 66, a second acknowledgment A2, to the sender 50. Once again the SUN REF: P2685 sender can cancel the timer. The process then continues with the tr~n~mic~ion offurther packets on receipt of the second acknowledgement A2.
The process illustrated in Figure 3A, is a l~pl~selllalion of the system operating properly with responses received within an expected time (RTT or round-5 trip-time). The RTT concept will be described later. However, Figure 3B illustrates what might happen when a packet is not received (for example because a packet islost).
In Figure 3B, a packet is sent at 70 and a timer (RTT timer) is started. A
packet Pl is lost in tr~n~mi~sion between sender 50 and receiver 52. Accordingly, the 10 packet is not received at the receiver when it should have been at time 72.
Accordingly, no acknowledgement is sent as should have occurred at 74. Likewise,an acknowledgement is not received at the sender 50 when it should have been at 76.
At 78 the RTT timer times out indicating that a packet has been lost. Accordingly, the sender retransmits packet 1 as Pl' at 80. This is then successfully received at the receiver 52 at 82, which returns at 84 the acknowledgment A'2 to the sender which is received at 86.
The basic transfer protocol described with reference to Figures 3A and 3B
above, has the disadvantage that an acknowledgement must be received before a further packet can be sent. In order to increase the dataflow, an Internet stream 20 service can employ a concept known as a "sliding window". The sliding window approach is to enable a sequence of packets to be tran~mitl~d before receiving an acknowledgement. The number of packets which can be tr~n~mi~ti~l before receiving an acknowledgement is defined by the number of packets within the "window".
Accordingly, for a sequence of packets 1-6, a window might extend from packet 1 25 to packet 3. Accordingly, all of the first three packets can be tr~n~mi~t~l without waiting for an acknowledgement. However, packet 4 can only be tr~ncmi~t.?~l whenan acknowledgement has been received for packet 1. On receipt of the acknowledgement for packet 1, packet 4 is then sent. At this stage packet 5 cannot be sent until an acknowledgement has been received from packet 2. It can be seen30 therefore that the window effectively slides along the sequence of packets asacknowledgements are received. A sliding window protocol remembers which packets ~ S
~ ~, SUN REF: P2685 have been acknowledged and keeps a separate timer for each unacknowledged packet.
If a packet is lost, the timer expires and the sender retransmits that packet. As the sender slides its window, it moves past an acknowledged packet. At the receivingend, a similar window is m:lint~in~d, for accepting and acknowledging packets as they 5 arrive. It will be appreciated that the protocol is relatively complex, but does provide for more efficient transfer. Figure 4 is a schematic representation of the exchange of packets for a sliding window of size 3. This shows how the window W slides along the list of packets.
The present invention finds application to a reliable stream service such as that 10 provided by the Internet. This service is defined by the Tr~ncmission ControlProtocol, or TCP. The combination of the TCP protocol and the underlying Internet protocol (IP) is often referred to as TCP/IP.
TCP specifies the format of the data and acknowledgements that two computers are to exchange to achieve reliable transfer, as well as the procedure to 15 ensure that data arrives correctly. The TCP Protocol assumes very little about the underlying communication system and can be used with a variety of packet delivery systems including the IP datagram delivery service. The TCP service resides above the IP layer which in turn lies above the network interface of the Internet.
Figure 5B represents the format of a segment used to coll"llullicate between 20 two nodes under the TCP. Each segment is divided into two parts, a header followed by data. The header comprises SOURCE PORT and DESTINATION PORT fields cont~ining the TCP PORT numbers that identify the application programs at the end of the connection. The SEQUENCE NO. identifies the position in the sender's bytestream of the data in the segment. The ACKNOWLEDGEMENT NO. field identifies 25 the position of the highest byte that the source has received. The SEQUENCE NO.
refers to the stream flowing in the same direction as the segment, while the ACKNOWLEDGEMENT NO. refers to the stream flowing in the opposite direction.
The OFF field contains an integer that specifies the offset of the data portion of the segment. This is needed because the OPTIONS field varies in length. The field RES
30 is reserved for future use. Segments can be used to carry an acknowledgement or data or requests to establish or close a cormection. The CODE field is used to determine SUN REF: P2685 the purpose and content of the segment. The WINDOW field specifies the buffer size that the destination is willing to accept every time it sends a segment. The CHECK
SUM field includes a TCP header check sum. The URGENT POINTER field is used for identifying urgent data.
The OPTIONS field is used to commllnicate information with the destination.
For example, the OPTIONS field can be used to specify a maximum segment size.
The DATA indication represents the start of the data field of the segment.
As the TCP sends data and variable length segments, acknowledgements necessarily refer to a position in the stream, and not to packets or segments. Each 10 acknowledgement specifies one greater than the highest byte position that has been received. Accordingly, acknowledgements specify the number of the next byte thatthe receiver expects to receive.
Reference has been made above to the round-trip time (RTT). This represents the average round time for the tr~n~mi~cion of a segment until receipt of the 15 corresponding acknowledgement. The RTT time needs to be set dynamically as the round-trip time can vary over time. Figure 6 is a schematic representation of the way in which RTT may vary in response to an event Hl . Although the RTT may increasedramatically the algorithm used to actually generate the response time within the system, can vary more slowly. As a result the RTT and response curves will diverge, 20 at least for a time.
A consequence of variations in network load and of the queuing of packets by routers and sending stations is that the actual RTT can increase, due to the time that the packet is held in the queue. As a result, it is possible that llnn~cess~ry sion of packets can occur where an acknowledgement has not been 25 received. This is represented schematically in Figure 7. It can be seen in Figure 7 that due to the delayed tr:~n~mi~ion of packet P2, message P2 is retr~n~mitt~-l before receipt of the acknowledgement A2. As a result of the llnn~ce~s~ry let~ icsion of the packet P2, this leads to an llnnt~cess~ry increase in the traffic capacity over the network which can aggravate congestion on the network.
In summary, therefore, the TCP layer includes a retr~n.smi~sion mechanism to recover from the loss of data on the underlying network. The interval between SUN REF: P2685 retr:~ncmiccions is dyn~mic~lly calculated by the TCP layers so as to adapt it to the response time of the network. However, when the load on the network increases, the TCP layer cannot adapt its retr~ncmiccion as fast as the response time of the network increase. As a result, the TCP layer retransmits packets when this is not actually 5 necessary because the lack of an acknowledgement is not due to non receipt of the packet, but merely to delayed receipt thereof. The effect of retransmissions is to cause yet more traffic on the network thereby once again increasing the response time of the network. This effect is well known in the Internet commllnity and is typically caused "congestion collapse".
Accordingly, it is an aim of the present invention to address this problem.
CA 02i49169 1998-09-30 ., :
SUN REF: P2685 SUMMARY OF THE INVENTION
In accordance with an embodiment of the invention, there is provided a mechanism for dispatching a sequence of packets via a telecommunications network, S which dispatching mechanism comprises a queue for packets for tr~n.cmiccion and a queue controller responsive to receipt of a new packet for tr~n.cmic.sion to compare parameters of the new packet to parameters of any packet already in the queue, the queue controller determining whether to queue or to drop the new packet depending on the result of the comparison(s).
By comparing a new packet to packets already queued for tr:~ncmiccion, unnecessary duplicated tr~n.smiccion of a packet can be avoided where packet tr~ncmiccion has been delayed, for example due to network congestion. Avoiding retr~n.cmiccion of the queued packet avoids aggravating the network congestion.
Where the new packet is a retr~n.cmiccion of the queued packet, then retr~n.cmi.ccion 15 would be unnecessary as it is known that the queued packet has not been lost, but has merely been delayed.
Preferably, the queue is implemented as a linked list structure as this providesa flexible mech~ni.cm for allowing changes in sizes to the queue and the addition and deletion of queue entries. Preferably the linked list comprises entries cont~ining 20 information relating to the packet flow as well as packet identity information and a separate pointer to the packet itself. This also allows the queue conkoller to readily traverse the queue to perform the comparison(s) referred to above Preferably, the queue controller is arranged to compare flow parameters of the new and the queued packet(s) including source and ~lestin~tion parameters to establish 25 whether the new and queued packets relate to the same packet flow. In a TCP
environment, the source parameters can comprise a source IP address and a sourceTCP port and the destin~tion parameters can comprise a ~lestin~tion IP address and a destination TCP port.
In a ~l~r~lled embodiment, the queue controller is further arranged to 30 compare packet sequence numbers and/or acknowledgement numbers for the new packet and the queued packet(s) to establish whether the new packet is a ., ' SUN REF: P2685 retr~n~mi~ion of a queued packet.
In a preferred embodiment for a TCP environment, the queue controller is arranged to determine that a new packet is a retr~n.cmi.~.sion of a queued packet if:
i) the new packet sequence number equals the queued packet sequence number;
5 and ii) the new packet acknowledgement number is less than the queued packet acknowledgement number.
The queue controller is arranged to add the new packet to the queue when it is determined that the new packet is not a retr:~n.~mi~ion of a queued packet.
In a preferred embodiment of the invention, the queue conkoller is arranged to drop a new packet when it is determined that the new packet is a retr~n.cmi~ion of a queued packet and the length of the queued packet is greater than or equal to that of the new packet. It is also arranged to replace a queued packet in the queue by the new packet when it is determined that the new packet is a retr~n~mi~ion of the 15 queued packet and the length of the new packet is greater than that of the queued packet.
The dispatch mechanism can be implemented by means of software operating on Co~ ul~L hardware.
In accordance with another aspect of the invention, there is provided a station 20 for sending a sequence of packets via a telecomlllullications network, the station including a dispatch controller comprising:
a dispatch queue for packets;
a queue controller arranged to compare flow and packet sequence parameters of a new packet for dispatch to flow and packet sequence parameters of queued 25 packets and arranged to respond to detection of the new packet being a rek~n~mi.~.cion of a queued packet relating to the same flow path to discard either the new packet or the queued packet. The station can, for example, be a router for routing a sequence of packets via the telecommllnications network.
In accordance with a further aspect of the invention, there is provided a 30 method of m~n~gin~ the dispatch of a sequence of packets via a teleco~ ic~tions network, the method comprising:
.. CA 02i49169 1998-09-30 SUN REF: P2685 queuing packets for tr~n.cmi~.sion;
comparing flow and packet sequence parameters of a new packet for tr~n~mi~.sion to flow and packet sequence parameters of queued packets; and responding to detection of the new packet being a retr~n~mi~ion of a queued 5 packet relating to the same flow path to discard either the new packet or the queued packet.
In accordance with a further aspect of the invention, there is provided a software dispatch mechanism on a storage medium for controlling the dispatch of a sequence of packets via a telecommunications network, the software dispatch 10 mechanism being configured to be operable to define:
a queue for packets for tr~n.~mi~ion; and a queue controller responsive to receipt of a new packet for tr~n.~mi.~ion to compare parameters of the new packet to parameters of a packet already in the queue, the queue controller determining whether to queue or to drop the new packet 15 depending on the result of the comparison(s).
. _ .
SUN REF: P2685 BRIEF DESCRIPTION OF THE DRAWINGS
Exemplary embodiments of the present invention will be described hereinafter, by way of example only, with reference to the accompanying drawings in which like 5 reference signs relate to like elements and in which:
Figure 1 is a schematic representation of a telecommllnications environment including source and destination locations in a router interconnected via network;
Figure 2 is a schematic representation of one possible implementation of a router; 3B
Figure 3A and~'is a schematic representation of data exchanged between the sender and receiver;
Figure 4 is a schematic representation of data exchanged using a sliding window;
Figures 5A and 5B are schematic representations of a possible datagram format and packet format, respectively, for use on the network;
Figure 6 is a diagram for illustrating changes in response time;
Figure 7 is a schematic representation of a further situation where retr~n.~mi.c.cion occurs;
Figure 8 is a schematic representation of a router;
Figure 9 is a schematic representation of a queuing mechanism for a routing mechanism in accordance with the present invention;
Figure 10 is a schematic representation of a queue control data structure for an embodiment to the invention; and Figure 11 is a flow diagram illustrating an example of the operation of a routing mechanism in accordance with the invention.
. ~,. .
SUN REF: P2685 DESCRIPTION OF THE PREFERRED EMBODIMENTS
Figure 8 is a schematic representation of a router 150, having four bi-directional connections to a network or networks 154-160. The router can be 5 implemented using conventional hardware, for example as described with reference to Figure 2, with appropriate software implementing logic 152 for routing functions.
Although represented separately, the networks 154-160 can effectively be part of the same network.
Figure 8 illustrates schematically an example where two packets are received from the network 154 and are routed to the network 156 and the network 158, respectively. The routing operations can be effected in a conventional manner byextracting destination information from received datagram fragments and by reference to routing tables 153, including mappings between rl~stin~tions and routes, held in the router as part of the routing logic 152.
Also shown schem~tic~lly in Figure 8 is a dispatch mechanism 110 in each output path from the routing logic 152.
Figure 9 is a schematic representation of an embodiment of a dispatch mechanism 110 in accordance with the invention, for incorporation in a node of the telecommunications network, for example in a router or sender (source station) as illustrated, for example, in Figure 1. An embodiment of the present invention may be implemented within the same overall structure as illustrated in Figures 1-3.
However, in accordance with the invention, the control of the buffer, or queue of packets to be tr~nsmitt~-~l to the network is controlled in a particular manner to take account of the possible duplicate tr~nsmiscions.
The dispatch mechanism 110 can be connected, for example, to receive packets for dispatch from conventional routing logic of a router or sender station, as represented schematically by block 155 and provides a queuing mech~nism, or structure, as further described in the following.
As represented schem~tic~lly in Figure 9, a queue controller 112 manages a queue 116 for packets to be tr~n~mitt~l at 58 to the network. The queue controller comprises or makes use of a queue control record 114 for the management of the .~., .
SUN REF: P2685 queue 116. It should be noted that Figure 9 is a schematic representation of oneembodiment of the invention, and that other embodiments of the invention may comprise a different structure. It will be appreciated that the structure illustrated in Figure 9 can be implemented by means of specific hardware, or alternatively by 5 means of software operating on the computing system used to implement the dispatch mechanism. The dispatch mechanism may be implemented in a router (routing station) forming a "staging post" in the network or alternatively could be part of the source of packets (sender station) to be tr:~n~mitt~d to the network.
An embodiment of the invention provides a mechanism to detect duplicated 10 TCP packets in a dispatch queue and to discard them. Each time a new packet is to be queued, the dispatch mechanism checks the packets already queued to detect if a duplicate of the new packet is contained in the queue. If it is, a decision is made to drop either the packet to be queued or the packet already queued depending on certain parameters. A further development of this basic concept is to monitor received 15 acknowledgements in order to elimin~te even more duplicated data.
Figure 10 is a schematic representation of an example of structure for a queue control record 114 which is managed by the queue controller 112 for m~n~gin~ thepackets in the queue 116. The data structure can be held, for example, in randomaccess memory of computer hardware in which the dispatch mechanism is 20 implemented. The data structure comprises a base register pair 120 comprising a head pointer H and a tail pointer T to the head and tail respectively of a linked list of packet entries 122. The data in the fields is extracted from the header of the TCP
packets and/or the associated IP datagrams used for message transfer. For example the port information is extracted from the TCP header and the address information 25 from the IP header (compare Figures SB and SA, respectively). Each of the packet entries 122 includes the following fields:
NEXT : a pointer to the next packet entry in the linked list;
PREVIOUS : a pointer to the previous packet entry in the linked list;
SRC IPADDR : the source IP address;
. . .
SUN REF: P2685 DST IPADDR : the clestin~tion IP address;
SRC TCP PORT : the source TCP port;
TCP SEQ : the TCP sequence number;
TCP ACK : the TCP acknowledgement;
LENGTH : the length of the packet;
DATA : a pointer to the TCP packet 126.1 in the queue.
As illustrated in Figure 10, a linked list comprising two entries 122.1 and 122.2 is shown. The packet entry 122.1 forms the head of the list and includes apointer N to the packet entry 122.2 which forms the tail of the list. Similarly, the 10 packet entry 122.2 includes a previous pointer P to the packet entry 122.1. Each of the packet entries 122.1 and 122.2 includes a respective pointer to the respective TCP
packet 126.1 and 126.2.
Figure 11 is a flow diagram illustrating the operation of the queue controller 112 of Figure 10.
In step 130, the queue controller waits for a new packet to be available for tr~nsmission. The new packet may be a packet being sent for the first time, or could be a packet retr~nsmittt?~l in accordance with the conventional of RTT al~ "n ofthe router or sender. When the queue controller 122 detects at 130 that a new packet is available for tr~nsmission, all of the information relating to the TCP flow are 20 extracted. These data are identified in Figure 10 at 124 and comprise the source IP
address, the clestin~tion IP address, the source TCP port and the destination TCP
port. This information is compared to the corresponding information at 124 in each of the entries in the linked list of packet entries 122.1-122.2. If the flow information parameters of the new packet correspond to those of one of the packets identified by 25 the packet entries 122.1-122.2, it is determined at step 134 that these packets relate to the same TCP flow. In this case, a check is made at step 136 as to whether the new TCP packet is a retr~nsmi~sion of the queued TCP packet represented by the TCP entry 122, concerned.
In accordance with the preferred embodiment of the invention, in step 136 the 30 queue controller checks whether the queued TCP sequence number is the same as the new TCP sequence number and whether the queued TCP acknowledgement number CA 02249l69 l998-09-30 . ~ ~
SUN REF: P2685 is less than or equal to the new TCP acknowledgement number. If both of these tests are positive, then it is determined in step 136 that the new packet does indeed relate to a retr:~n~mi.c.sion of an earlier packet. In this case, one of the packets is then dropped. In other words, the queue controller determines that a new packet is a 5 retr~n~mi.csion of an earlier packet if:
i) the new packet sequence number equals the queued packet sequence number;
and ii) the new packet acknowledgement number is less than the queued packet acknowledgement number.
A determination is made as to which of the packets to drop by comparison of the packet lengths. In step 138, the queue controller checks whether the queued packet's length is greater than or equal to the new packet. If this is the case, then the new packet is dropped. If the new packet is longer than the queued packet, the queued packet is removed from the queue and it is replaced by the new packet. The replacement of the packet in the queue is achieved by replacing the packet entry 122 concerned by a packet entry relating to the new packet. This replacement operation includes setting a~pLopliate next N and previous P pointers in the packet entry 122 as well as storing appropriate information in further fields of the packet entry 122 and, not least, a data pointer to the new packet 126.
Dropping of the new packet is achieved by simply not adding an ~p.ol?liate packet entry 122 to the linked list structure.
If the test in step 134 or the test in the step 136 is negative, (ie. the queuedpacket does not relate to the same flow or does not form a l~ ",i~ion of a packet within the same flow, then the new packet is queued at step 140 by adding an appropriate packet header 122 at the tail of the queue. The addition of the new packet header 122 will comprise including appropriate data in the various fields of the new packet header 122 corresponding to the new TCP packet concerned, and including a pointer to that TCP packet. The data will include the flow data including source and destination port information from the packet header and source and destination address information from the IP datagram header. A previous pointer to the previous tail of the queue 122.2 will be placed in the previous P field of the new ., SUN REF: P2685 packet header 122 and that previous tail field 122 will receive a next pointer N to the new packet header entry 122. The tail pointer T will also be amended to point to the new packet 122 rather than the previous tail packet entry 122.2.
When a packet is finally sent from the queue, the corresponding packet entry 5 122 is removed from the linked list structure. Accordingly, for example, when the packet 126.1 is sent, the corresponding packet entry 122.1 will be deleted from the list by ch~nging the head pointer H to point to the next entry in the list (eg. 122.2) and the previous pointer of that next packet entry 122.2 will be sent to null indicating that it is the head of the list.
Accordingly, there has been described a mechanism which will avoid mn~cess~ry retr~n~mi~sion of duplicates of packets where a delay in receipt of an acknowledgement results from delays in being able to transmit the packet concerned from the queue. A queue controller is arranged to compare flow and packet sequence parameters of a new packet for dispatch to flow and packet sequence parameters of 15 queued packets and arranged to respond to detection of said new packet being a retr~n~mi~ion of a queued packet relating to the same flow path to discard either said new packet or said queued packet. This avoids increasing the congestion over thenetwork and reduces the risk of congestion collapse.
The present invention is applicable irrespective of whether a sliding window 20 approach is used as described in the introduction, and irrespective of the size of the sliding window concerned where a sliding window approach is used.
Although the invention has been described in particular in the context of data tr~n.cmi~ion in accordance with a TCP protocol over a network such as the Internet, it will be appreciated that the invention is not limited thereto. Accordingly, the use 25 of Internet-f~m~ r terms does not mean that the invention is limited to use with the Internet. Accordingly, bearing in mind that terminology varies from technology to technology, the terms used in the present specification are intended to cover equivalent terms as might be used in any particular environment.
Accordingly, it will be appreciated that although particular embodiments of the 30 invention have been described, many modifications/additions and/or substitutions may be made within the spirit and scope of the present invention as defined in the . ~ CA 02249169 1998-09-30 SUN REF: P2685 appended claims. With reference to those claims, it is to be noted that combinations of features of the dependent claims other than those explicitly enumerated in the claims may be made with features of other dependent claims and/or independent claims, as apl.,opliate, within the spirit and scope of the present invention.
SUN REF: P2685 BACKGROUND OF THE INVENTION
This invention relates to a mechanism for dispatching packets via a telecommllnications network and to a network sender or router station, or node, incorporating such a mech~nism The invention finds particular application to tr~ncmi~cion of TCP information via an inter-or intra-network operating under an Internet Protocol.
Figure 1 is a schematic representation of an inct~n~e of an Inter- or Intra-net with a router 10 being provided in the path between a source 12 and a destination 14.
Between the source 12 (or sender node) and the router node 10, a net 16 is shownand between the router node 10 and the dçstin~tion node 14 a further net 18 is shown.
In practice, the net 16 and the net 18 can be one and the same and the router 10errt;~;Liv~ly forms a "staging post" between the source 12 and the destination 14. In the following, reference is made to a dispatch mech~ni~m It should be appreciated t_at the dispatch mechanism could, in the present context, equally form part of the source 12, as opposed to being a separate "staging post" as illustrated in Figure 1.
Figure 2 is a schematic representation of the configuration of a station for a router 10 or source or dçstin~tion 12, 14. These stations can be implemented using any a~roL,liate technology. However, as illustrated in Figure 2, the station is implemented by a server co~ uLer 20 coulplisillg a system unit 22, optionally with a display 38, keyboard 40 and other input devices 42. It should be noted that the router 10 need not include a keyboard, display, etc. Figure 2A is a schematic block representation of aspects of the contents of the system unit 22. As illustrated in Figure 2A, the system unit includes a processor 28, memory 30, disk drives 24 and 26, and a communications adaptor 32 for connection to one or more telecommunications lines 34 for connection to the telecollllllunications network 16/18.
As illustrated in Figure 2A, the components of the system unit are connected via a bus arrangement 36. It will be appreciated that Figures 2/2A are a general schematic representation of one possible configuration for a server co~ ul~l for forming arouter or sender or ~1~stin~tion station and that many alternative configurations could be provided.
Conceptually, the Internet provides three levels of services. At the lowest ~. t.
SUN REF: P2685 level, a connectionless delivery system provides a foundation on which everything rests. At the next level, a reliable transport service provides a high level platform.
At the third level, application services are provided which rely on the reliabletransport service.
S A filntl~mental Internet service consists of an unreliable, best-effort,connectionless, packet delivery system. The service is described as being "unreliable" because delivery is not guaranteed. A packet may be lost, duplicated, or delivered out of order, but the Internet will not detect such conditions, nor will it inform the sender or receiver. The service is described as being "connectionless"
because each packet is treated independently from all others. A sequence of packets sent from one machine to another may travel over different paths, or some may belost while others are delivered. The service may be described as "best-effort"
because the Internet makes an earnest attempt to deliver packets.
The protocol that defines the unreliable, connectionless, delivery mechanism is called the "Internet Protocol", and is usually referred to by its initials IP. IP
defines the formal specification of data formats, including a basic unit of data transfer and the exact format of all data passing across the Internet. IP also includes rules which specify how packets should be processed and how errors should be handled.
In particular, IP embodies the idea of unreliable delivery and packet routing.
Further details of aspects of the Internet and TCP/IP protocols may be found, for example, in the following US patents: 5,293,379; 5,307,347; 5,307,413;
5,309,437; 5,351,237; and 5,535,199.
The basic unit of data transfer via the IP is termed an "Internet datagram", or alternative "IP datagram", or simply "datagram". A datagram comprises header anddata areas, and source and flestin~tion addresses. There is no fixed size for a datagram. Bearing this in mind, and also the physical constraints of the underlying hardware services on which the Internet is based, it is n~cçss~ry to divide the datagram into portions called "fragments".
Figure SA illustrates the format of an Internet datagram. The same format is used for a fragment of a datagram.
The 4 bit version field (VERS) specifies the IP protocol version and is used SUN REF: P2685 to ensure that all of the nodes along the path of the datagram agree on the format.
The LEN field gives the datagram header length measured in 32 bit words.
The TOTAL LENGTH field gives the length of the IP datagram measured in octets including the length of the header and data.
The SERVICE TYPE field contains h~n-lling details for the datagram.
Three fields in the datagram header, IDENT, FLAGS, and FRAGMENT
OFFSET, control fragmentation and reassembly of datagrams. The field IDENT
contains a unique identifier that identifies the datagram.
In the FLAGS field, a first bit specifies whether the datagram may be 10 fragmented, and a second bit in~lic~t~s whether this is the last fragment in the datagram. The FRAGMENT OFFSET field specifies the offset of this fragment in the original datagram, measured in units of 8 octets, starting at offset zero.
As each fragment has the same basic header format as a complete datagram, the combination of the FLAGS and FRAGMENT OFFSET fields are used to indicate 15 that the headers relate to fragments, and to in-lic~t~- the position of the fragment within the original datagram. The FRAGMENT OFFSET field identifies the position within the datagram, and the second of the FLAGS bits mentioned above (which is sometimes called the MORE FRAGMENTS flag) is used to indicate whether there are any more fragments in the datagram, or conversely that the fragment concerned 20 is the last fragment of the datagram.
The field PROTO is a form of type field. The HEADER CHECK SUM
figure ensures integrity of header values.
SOURCE IP ADDRESS and DESTINATION IP ADDRESS contain 32 bit Internet addresses of the datagram's sender and intended recipient. The OPTIONS
25 field and the PADDING field are optional in the datagram. The field labelled DATA
~;pl~sell~ the beginning of the data field.
As mentioned above, above the IP layer of the Internet protocol structure one service which is provided is a reliable transport service which is typically called the "reliable stream transport service", defined by the Tr~n~mi~ion Control Protocol30 (TCP). Although TCP is provided over the Internet, it is in fact an independent general purpose protocol which can also be used with other delivery systems. TCP
SUN REF: P2685 makes very few as~ulup~ions regarding the underlying network, and it can also beused over a single network like Ethernet, as well as over a complex Internet, orIntranet.
TCP provides a reliable stream delivery service which can be contrasted with the unreliable datagram protocol (UDP) which is also provided over the Internet.Whereas UDP provides an unreliable delivery service because delivery is not guaranteed, TCP provides a more complex structure which does ensure reliable delivery in the form of a stream.
UDP provides unreliable packet delivering, whereby packets may be lost or 10 destroyed when k~n~mi~sion errors il,l~,rl ~; with data, when network hardware fails, or when networks become too heavily loaded to accommodate the load presented.
TCP on the other hand, operates by providing delivery by means of a stream of bits, divided into eight-bit octets or bytes.
Given that the underlying Internet protocol is unreliable, TCP tr~n~mi~ions 15 operate in accordance with a technique known as positive acknowledgement withreL,~ i",ic~ion. The technique requires a recipient to co"-",ll.~ic~te with the source, sending back an acknowledgement message every time it receives data. The sender keeps a record of each packet that it sends and waits for an acknowledgement before sending the next packet. The sender also starts a timer when it sends its packet and 20 retransmits a packet if the timer expires before the acknowledgement arrives. Figure 3A is a schematic representation of the tr~n~mi~ion and receipt of packets and acknowledgements. The left hand side of Figure 3A represents events at a sender side 50, the right hand side represents events at a receiver side 52 and the middle portion represents network messages passing between the sender and the receiver.
At 54, the sender 50 (eg, the router 10) sends a packet P1 to the receiver (eg, the destination 14) via the network and starts a timer for message P1. When the receiver 52 receives, 56, the packet P1; the receiver then sends, 58, an acknowledgement A1. When the acknowledgement A1 is received, 60, at the sender 50, the sender can cancel the timer and send 62, the next packet P2 to the receiver 30 52 setting a timer for the message P2. When the receiver 52 receives, 64, the packet P2, it sends 66, a second acknowledgment A2, to the sender 50. Once again the SUN REF: P2685 sender can cancel the timer. The process then continues with the tr~n~mic~ion offurther packets on receipt of the second acknowledgement A2.
The process illustrated in Figure 3A, is a l~pl~selllalion of the system operating properly with responses received within an expected time (RTT or round-5 trip-time). The RTT concept will be described later. However, Figure 3B illustrates what might happen when a packet is not received (for example because a packet islost).
In Figure 3B, a packet is sent at 70 and a timer (RTT timer) is started. A
packet Pl is lost in tr~n~mi~sion between sender 50 and receiver 52. Accordingly, the 10 packet is not received at the receiver when it should have been at time 72.
Accordingly, no acknowledgement is sent as should have occurred at 74. Likewise,an acknowledgement is not received at the sender 50 when it should have been at 76.
At 78 the RTT timer times out indicating that a packet has been lost. Accordingly, the sender retransmits packet 1 as Pl' at 80. This is then successfully received at the receiver 52 at 82, which returns at 84 the acknowledgment A'2 to the sender which is received at 86.
The basic transfer protocol described with reference to Figures 3A and 3B
above, has the disadvantage that an acknowledgement must be received before a further packet can be sent. In order to increase the dataflow, an Internet stream 20 service can employ a concept known as a "sliding window". The sliding window approach is to enable a sequence of packets to be tran~mitl~d before receiving an acknowledgement. The number of packets which can be tr~n~mi~ti~l before receiving an acknowledgement is defined by the number of packets within the "window".
Accordingly, for a sequence of packets 1-6, a window might extend from packet 1 25 to packet 3. Accordingly, all of the first three packets can be tr~n~mi~t~l without waiting for an acknowledgement. However, packet 4 can only be tr~ncmi~t.?~l whenan acknowledgement has been received for packet 1. On receipt of the acknowledgement for packet 1, packet 4 is then sent. At this stage packet 5 cannot be sent until an acknowledgement has been received from packet 2. It can be seen30 therefore that the window effectively slides along the sequence of packets asacknowledgements are received. A sliding window protocol remembers which packets ~ S
~ ~, SUN REF: P2685 have been acknowledged and keeps a separate timer for each unacknowledged packet.
If a packet is lost, the timer expires and the sender retransmits that packet. As the sender slides its window, it moves past an acknowledged packet. At the receivingend, a similar window is m:lint~in~d, for accepting and acknowledging packets as they 5 arrive. It will be appreciated that the protocol is relatively complex, but does provide for more efficient transfer. Figure 4 is a schematic representation of the exchange of packets for a sliding window of size 3. This shows how the window W slides along the list of packets.
The present invention finds application to a reliable stream service such as that 10 provided by the Internet. This service is defined by the Tr~ncmission ControlProtocol, or TCP. The combination of the TCP protocol and the underlying Internet protocol (IP) is often referred to as TCP/IP.
TCP specifies the format of the data and acknowledgements that two computers are to exchange to achieve reliable transfer, as well as the procedure to 15 ensure that data arrives correctly. The TCP Protocol assumes very little about the underlying communication system and can be used with a variety of packet delivery systems including the IP datagram delivery service. The TCP service resides above the IP layer which in turn lies above the network interface of the Internet.
Figure 5B represents the format of a segment used to coll"llullicate between 20 two nodes under the TCP. Each segment is divided into two parts, a header followed by data. The header comprises SOURCE PORT and DESTINATION PORT fields cont~ining the TCP PORT numbers that identify the application programs at the end of the connection. The SEQUENCE NO. identifies the position in the sender's bytestream of the data in the segment. The ACKNOWLEDGEMENT NO. field identifies 25 the position of the highest byte that the source has received. The SEQUENCE NO.
refers to the stream flowing in the same direction as the segment, while the ACKNOWLEDGEMENT NO. refers to the stream flowing in the opposite direction.
The OFF field contains an integer that specifies the offset of the data portion of the segment. This is needed because the OPTIONS field varies in length. The field RES
30 is reserved for future use. Segments can be used to carry an acknowledgement or data or requests to establish or close a cormection. The CODE field is used to determine SUN REF: P2685 the purpose and content of the segment. The WINDOW field specifies the buffer size that the destination is willing to accept every time it sends a segment. The CHECK
SUM field includes a TCP header check sum. The URGENT POINTER field is used for identifying urgent data.
The OPTIONS field is used to commllnicate information with the destination.
For example, the OPTIONS field can be used to specify a maximum segment size.
The DATA indication represents the start of the data field of the segment.
As the TCP sends data and variable length segments, acknowledgements necessarily refer to a position in the stream, and not to packets or segments. Each 10 acknowledgement specifies one greater than the highest byte position that has been received. Accordingly, acknowledgements specify the number of the next byte thatthe receiver expects to receive.
Reference has been made above to the round-trip time (RTT). This represents the average round time for the tr~n~mi~cion of a segment until receipt of the 15 corresponding acknowledgement. The RTT time needs to be set dynamically as the round-trip time can vary over time. Figure 6 is a schematic representation of the way in which RTT may vary in response to an event Hl . Although the RTT may increasedramatically the algorithm used to actually generate the response time within the system, can vary more slowly. As a result the RTT and response curves will diverge, 20 at least for a time.
A consequence of variations in network load and of the queuing of packets by routers and sending stations is that the actual RTT can increase, due to the time that the packet is held in the queue. As a result, it is possible that llnn~cess~ry sion of packets can occur where an acknowledgement has not been 25 received. This is represented schematically in Figure 7. It can be seen in Figure 7 that due to the delayed tr:~n~mi~ion of packet P2, message P2 is retr~n~mitt~-l before receipt of the acknowledgement A2. As a result of the llnn~ce~s~ry let~ icsion of the packet P2, this leads to an llnnt~cess~ry increase in the traffic capacity over the network which can aggravate congestion on the network.
In summary, therefore, the TCP layer includes a retr~n.smi~sion mechanism to recover from the loss of data on the underlying network. The interval between SUN REF: P2685 retr:~ncmiccions is dyn~mic~lly calculated by the TCP layers so as to adapt it to the response time of the network. However, when the load on the network increases, the TCP layer cannot adapt its retr~ncmiccion as fast as the response time of the network increase. As a result, the TCP layer retransmits packets when this is not actually 5 necessary because the lack of an acknowledgement is not due to non receipt of the packet, but merely to delayed receipt thereof. The effect of retransmissions is to cause yet more traffic on the network thereby once again increasing the response time of the network. This effect is well known in the Internet commllnity and is typically caused "congestion collapse".
Accordingly, it is an aim of the present invention to address this problem.
CA 02i49169 1998-09-30 ., :
SUN REF: P2685 SUMMARY OF THE INVENTION
In accordance with an embodiment of the invention, there is provided a mechanism for dispatching a sequence of packets via a telecommunications network, S which dispatching mechanism comprises a queue for packets for tr~n.cmiccion and a queue controller responsive to receipt of a new packet for tr~n.cmic.sion to compare parameters of the new packet to parameters of any packet already in the queue, the queue controller determining whether to queue or to drop the new packet depending on the result of the comparison(s).
By comparing a new packet to packets already queued for tr:~ncmiccion, unnecessary duplicated tr~n.smiccion of a packet can be avoided where packet tr~ncmiccion has been delayed, for example due to network congestion. Avoiding retr~n.cmiccion of the queued packet avoids aggravating the network congestion.
Where the new packet is a retr~n.cmiccion of the queued packet, then retr~n.cmi.ccion 15 would be unnecessary as it is known that the queued packet has not been lost, but has merely been delayed.
Preferably, the queue is implemented as a linked list structure as this providesa flexible mech~ni.cm for allowing changes in sizes to the queue and the addition and deletion of queue entries. Preferably the linked list comprises entries cont~ining 20 information relating to the packet flow as well as packet identity information and a separate pointer to the packet itself. This also allows the queue conkoller to readily traverse the queue to perform the comparison(s) referred to above Preferably, the queue controller is arranged to compare flow parameters of the new and the queued packet(s) including source and ~lestin~tion parameters to establish 25 whether the new and queued packets relate to the same packet flow. In a TCP
environment, the source parameters can comprise a source IP address and a sourceTCP port and the destin~tion parameters can comprise a ~lestin~tion IP address and a destination TCP port.
In a ~l~r~lled embodiment, the queue controller is further arranged to 30 compare packet sequence numbers and/or acknowledgement numbers for the new packet and the queued packet(s) to establish whether the new packet is a ., ' SUN REF: P2685 retr~n~mi~ion of a queued packet.
In a preferred embodiment for a TCP environment, the queue controller is arranged to determine that a new packet is a retr~n.cmi.~.sion of a queued packet if:
i) the new packet sequence number equals the queued packet sequence number;
5 and ii) the new packet acknowledgement number is less than the queued packet acknowledgement number.
The queue controller is arranged to add the new packet to the queue when it is determined that the new packet is not a retr:~n.~mi~ion of a queued packet.
In a preferred embodiment of the invention, the queue conkoller is arranged to drop a new packet when it is determined that the new packet is a retr~n.cmi~ion of a queued packet and the length of the queued packet is greater than or equal to that of the new packet. It is also arranged to replace a queued packet in the queue by the new packet when it is determined that the new packet is a retr~n~mi~ion of the 15 queued packet and the length of the new packet is greater than that of the queued packet.
The dispatch mechanism can be implemented by means of software operating on Co~ ul~L hardware.
In accordance with another aspect of the invention, there is provided a station 20 for sending a sequence of packets via a telecomlllullications network, the station including a dispatch controller comprising:
a dispatch queue for packets;
a queue controller arranged to compare flow and packet sequence parameters of a new packet for dispatch to flow and packet sequence parameters of queued 25 packets and arranged to respond to detection of the new packet being a rek~n~mi.~.cion of a queued packet relating to the same flow path to discard either the new packet or the queued packet. The station can, for example, be a router for routing a sequence of packets via the telecommllnications network.
In accordance with a further aspect of the invention, there is provided a 30 method of m~n~gin~ the dispatch of a sequence of packets via a teleco~ ic~tions network, the method comprising:
.. CA 02i49169 1998-09-30 SUN REF: P2685 queuing packets for tr~n.cmi~.sion;
comparing flow and packet sequence parameters of a new packet for tr~n~mi~.sion to flow and packet sequence parameters of queued packets; and responding to detection of the new packet being a retr~n~mi~ion of a queued 5 packet relating to the same flow path to discard either the new packet or the queued packet.
In accordance with a further aspect of the invention, there is provided a software dispatch mechanism on a storage medium for controlling the dispatch of a sequence of packets via a telecommunications network, the software dispatch 10 mechanism being configured to be operable to define:
a queue for packets for tr~n.~mi~ion; and a queue controller responsive to receipt of a new packet for tr~n.~mi.~ion to compare parameters of the new packet to parameters of a packet already in the queue, the queue controller determining whether to queue or to drop the new packet 15 depending on the result of the comparison(s).
. _ .
SUN REF: P2685 BRIEF DESCRIPTION OF THE DRAWINGS
Exemplary embodiments of the present invention will be described hereinafter, by way of example only, with reference to the accompanying drawings in which like 5 reference signs relate to like elements and in which:
Figure 1 is a schematic representation of a telecommllnications environment including source and destination locations in a router interconnected via network;
Figure 2 is a schematic representation of one possible implementation of a router; 3B
Figure 3A and~'is a schematic representation of data exchanged between the sender and receiver;
Figure 4 is a schematic representation of data exchanged using a sliding window;
Figures 5A and 5B are schematic representations of a possible datagram format and packet format, respectively, for use on the network;
Figure 6 is a diagram for illustrating changes in response time;
Figure 7 is a schematic representation of a further situation where retr~n.~mi.c.cion occurs;
Figure 8 is a schematic representation of a router;
Figure 9 is a schematic representation of a queuing mechanism for a routing mechanism in accordance with the present invention;
Figure 10 is a schematic representation of a queue control data structure for an embodiment to the invention; and Figure 11 is a flow diagram illustrating an example of the operation of a routing mechanism in accordance with the invention.
. ~,. .
SUN REF: P2685 DESCRIPTION OF THE PREFERRED EMBODIMENTS
Figure 8 is a schematic representation of a router 150, having four bi-directional connections to a network or networks 154-160. The router can be 5 implemented using conventional hardware, for example as described with reference to Figure 2, with appropriate software implementing logic 152 for routing functions.
Although represented separately, the networks 154-160 can effectively be part of the same network.
Figure 8 illustrates schematically an example where two packets are received from the network 154 and are routed to the network 156 and the network 158, respectively. The routing operations can be effected in a conventional manner byextracting destination information from received datagram fragments and by reference to routing tables 153, including mappings between rl~stin~tions and routes, held in the router as part of the routing logic 152.
Also shown schem~tic~lly in Figure 8 is a dispatch mechanism 110 in each output path from the routing logic 152.
Figure 9 is a schematic representation of an embodiment of a dispatch mechanism 110 in accordance with the invention, for incorporation in a node of the telecommunications network, for example in a router or sender (source station) as illustrated, for example, in Figure 1. An embodiment of the present invention may be implemented within the same overall structure as illustrated in Figures 1-3.
However, in accordance with the invention, the control of the buffer, or queue of packets to be tr~nsmitt~-~l to the network is controlled in a particular manner to take account of the possible duplicate tr~nsmiscions.
The dispatch mechanism 110 can be connected, for example, to receive packets for dispatch from conventional routing logic of a router or sender station, as represented schematically by block 155 and provides a queuing mech~nism, or structure, as further described in the following.
As represented schem~tic~lly in Figure 9, a queue controller 112 manages a queue 116 for packets to be tr~n~mitt~l at 58 to the network. The queue controller comprises or makes use of a queue control record 114 for the management of the .~., .
SUN REF: P2685 queue 116. It should be noted that Figure 9 is a schematic representation of oneembodiment of the invention, and that other embodiments of the invention may comprise a different structure. It will be appreciated that the structure illustrated in Figure 9 can be implemented by means of specific hardware, or alternatively by 5 means of software operating on the computing system used to implement the dispatch mechanism. The dispatch mechanism may be implemented in a router (routing station) forming a "staging post" in the network or alternatively could be part of the source of packets (sender station) to be tr:~n~mitt~d to the network.
An embodiment of the invention provides a mechanism to detect duplicated 10 TCP packets in a dispatch queue and to discard them. Each time a new packet is to be queued, the dispatch mechanism checks the packets already queued to detect if a duplicate of the new packet is contained in the queue. If it is, a decision is made to drop either the packet to be queued or the packet already queued depending on certain parameters. A further development of this basic concept is to monitor received 15 acknowledgements in order to elimin~te even more duplicated data.
Figure 10 is a schematic representation of an example of structure for a queue control record 114 which is managed by the queue controller 112 for m~n~gin~ thepackets in the queue 116. The data structure can be held, for example, in randomaccess memory of computer hardware in which the dispatch mechanism is 20 implemented. The data structure comprises a base register pair 120 comprising a head pointer H and a tail pointer T to the head and tail respectively of a linked list of packet entries 122. The data in the fields is extracted from the header of the TCP
packets and/or the associated IP datagrams used for message transfer. For example the port information is extracted from the TCP header and the address information 25 from the IP header (compare Figures SB and SA, respectively). Each of the packet entries 122 includes the following fields:
NEXT : a pointer to the next packet entry in the linked list;
PREVIOUS : a pointer to the previous packet entry in the linked list;
SRC IPADDR : the source IP address;
. . .
SUN REF: P2685 DST IPADDR : the clestin~tion IP address;
SRC TCP PORT : the source TCP port;
TCP SEQ : the TCP sequence number;
TCP ACK : the TCP acknowledgement;
LENGTH : the length of the packet;
DATA : a pointer to the TCP packet 126.1 in the queue.
As illustrated in Figure 10, a linked list comprising two entries 122.1 and 122.2 is shown. The packet entry 122.1 forms the head of the list and includes apointer N to the packet entry 122.2 which forms the tail of the list. Similarly, the 10 packet entry 122.2 includes a previous pointer P to the packet entry 122.1. Each of the packet entries 122.1 and 122.2 includes a respective pointer to the respective TCP
packet 126.1 and 126.2.
Figure 11 is a flow diagram illustrating the operation of the queue controller 112 of Figure 10.
In step 130, the queue controller waits for a new packet to be available for tr~nsmission. The new packet may be a packet being sent for the first time, or could be a packet retr~nsmittt?~l in accordance with the conventional of RTT al~ "n ofthe router or sender. When the queue controller 122 detects at 130 that a new packet is available for tr~nsmission, all of the information relating to the TCP flow are 20 extracted. These data are identified in Figure 10 at 124 and comprise the source IP
address, the clestin~tion IP address, the source TCP port and the destination TCP
port. This information is compared to the corresponding information at 124 in each of the entries in the linked list of packet entries 122.1-122.2. If the flow information parameters of the new packet correspond to those of one of the packets identified by 25 the packet entries 122.1-122.2, it is determined at step 134 that these packets relate to the same TCP flow. In this case, a check is made at step 136 as to whether the new TCP packet is a retr~nsmi~sion of the queued TCP packet represented by the TCP entry 122, concerned.
In accordance with the preferred embodiment of the invention, in step 136 the 30 queue controller checks whether the queued TCP sequence number is the same as the new TCP sequence number and whether the queued TCP acknowledgement number CA 02249l69 l998-09-30 . ~ ~
SUN REF: P2685 is less than or equal to the new TCP acknowledgement number. If both of these tests are positive, then it is determined in step 136 that the new packet does indeed relate to a retr:~n~mi.c.sion of an earlier packet. In this case, one of the packets is then dropped. In other words, the queue controller determines that a new packet is a 5 retr~n~mi.csion of an earlier packet if:
i) the new packet sequence number equals the queued packet sequence number;
and ii) the new packet acknowledgement number is less than the queued packet acknowledgement number.
A determination is made as to which of the packets to drop by comparison of the packet lengths. In step 138, the queue controller checks whether the queued packet's length is greater than or equal to the new packet. If this is the case, then the new packet is dropped. If the new packet is longer than the queued packet, the queued packet is removed from the queue and it is replaced by the new packet. The replacement of the packet in the queue is achieved by replacing the packet entry 122 concerned by a packet entry relating to the new packet. This replacement operation includes setting a~pLopliate next N and previous P pointers in the packet entry 122 as well as storing appropriate information in further fields of the packet entry 122 and, not least, a data pointer to the new packet 126.
Dropping of the new packet is achieved by simply not adding an ~p.ol?liate packet entry 122 to the linked list structure.
If the test in step 134 or the test in the step 136 is negative, (ie. the queuedpacket does not relate to the same flow or does not form a l~ ",i~ion of a packet within the same flow, then the new packet is queued at step 140 by adding an appropriate packet header 122 at the tail of the queue. The addition of the new packet header 122 will comprise including appropriate data in the various fields of the new packet header 122 corresponding to the new TCP packet concerned, and including a pointer to that TCP packet. The data will include the flow data including source and destination port information from the packet header and source and destination address information from the IP datagram header. A previous pointer to the previous tail of the queue 122.2 will be placed in the previous P field of the new ., SUN REF: P2685 packet header 122 and that previous tail field 122 will receive a next pointer N to the new packet header entry 122. The tail pointer T will also be amended to point to the new packet 122 rather than the previous tail packet entry 122.2.
When a packet is finally sent from the queue, the corresponding packet entry 5 122 is removed from the linked list structure. Accordingly, for example, when the packet 126.1 is sent, the corresponding packet entry 122.1 will be deleted from the list by ch~nging the head pointer H to point to the next entry in the list (eg. 122.2) and the previous pointer of that next packet entry 122.2 will be sent to null indicating that it is the head of the list.
Accordingly, there has been described a mechanism which will avoid mn~cess~ry retr~n~mi~sion of duplicates of packets where a delay in receipt of an acknowledgement results from delays in being able to transmit the packet concerned from the queue. A queue controller is arranged to compare flow and packet sequence parameters of a new packet for dispatch to flow and packet sequence parameters of 15 queued packets and arranged to respond to detection of said new packet being a retr~n~mi~ion of a queued packet relating to the same flow path to discard either said new packet or said queued packet. This avoids increasing the congestion over thenetwork and reduces the risk of congestion collapse.
The present invention is applicable irrespective of whether a sliding window 20 approach is used as described in the introduction, and irrespective of the size of the sliding window concerned where a sliding window approach is used.
Although the invention has been described in particular in the context of data tr~n.cmi~ion in accordance with a TCP protocol over a network such as the Internet, it will be appreciated that the invention is not limited thereto. Accordingly, the use 25 of Internet-f~m~ r terms does not mean that the invention is limited to use with the Internet. Accordingly, bearing in mind that terminology varies from technology to technology, the terms used in the present specification are intended to cover equivalent terms as might be used in any particular environment.
Accordingly, it will be appreciated that although particular embodiments of the 30 invention have been described, many modifications/additions and/or substitutions may be made within the spirit and scope of the present invention as defined in the . ~ CA 02249169 1998-09-30 SUN REF: P2685 appended claims. With reference to those claims, it is to be noted that combinations of features of the dependent claims other than those explicitly enumerated in the claims may be made with features of other dependent claims and/or independent claims, as apl.,opliate, within the spirit and scope of the present invention.
Claims (23)
1. A mechanism for dispatching a sequence of packets via a telecommunications network, said dispatch mechanism comprising:
a queue for packets for transmission;
a queue controller responsive to receipt of a new packet for transmission to compare parameters of said new packet to parameters of a packet already in said queue, said queue controller determining whether to queue or to drop said new packet depending on the result of said comparison(s).
a queue for packets for transmission;
a queue controller responsive to receipt of a new packet for transmission to compare parameters of said new packet to parameters of a packet already in said queue, said queue controller determining whether to queue or to drop said new packet depending on the result of said comparison(s).
2. The dispatch mechanism of Claim 1, wherein said queue comprises a linked listof packets.
3. The dispatch mechanism of Claim 1, wherein said queue comprises a linked listof pointers to said queued packets.
4. The dispatch mechanism according to Claim 3, wherein said linked list of pointers also includes said parameters of said queued packets.
5. The dispatch mechanism of Claim 1, wherein said queue controller is arranged to compare flow parameters of said new and said queued packet(s) including source and destination parameters.
6. The dispatch mechanism of Claim 5, wherein said source parameters comprise a source IP address and a source TCP port and said destination parameters comprise a destination IP address and a destination TCP port.
7. The dispatch mechanism of Claim 6, wherein said queue controller is further arranged to compare packet sequence numbers for said new packet and said queued packet(s).
8. The dispatch mechanism of Claim 7, wherein said queue controller is arranged to determine that a new packet is a retransmission of a queued packet if:
i) the new packet sequence number equals the queued packet sequence number; and ii) the new packet acknowledgement number is less than the queued packet acknowledgement number.
i) the new packet sequence number equals the queued packet sequence number; and ii) the new packet acknowledgement number is less than the queued packet acknowledgement number.
9. The dispatch mechanism of Claim 1, wherein said queue controller is arranged to add said new packet to said queue when said new packet is not a retransmission of a queued packet.
10. The dispatch mechanism of Claim 1, wherein said queue controller is arrangedto drop said new packet when said new packet is a retransmission of a queued packet and the length of said queued packet is greater than or equal to that of said new packet.
11. The dispatch mechanism of Claim 1, wherein said queue controller is arrangedto a replace a queued packet in said queue by said new packet when said new packet is a retransmission of said queued packet and the length of said new packet is greater than that of said queued packet.
12. The dispatch mechanism of Claim 1, wherein said mechanism is a software mechanism.
13. A station for sending a sequence of packets via a telecommunications network, said station including a dispatch controller comprising:
a dispatch queue for packets;
a queue controller arranged to compare flow and packet sequence parameters of a new packet for dispatch to flow and packet sequence parameters of queued packets and arranged to respond to detection of said new packet being a retransmission of a queued packet relating to the same flow path to discard either said new packet or said queued packet.
a dispatch queue for packets;
a queue controller arranged to compare flow and packet sequence parameters of a new packet for dispatch to flow and packet sequence parameters of queued packets and arranged to respond to detection of said new packet being a retransmission of a queued packet relating to the same flow path to discard either said new packet or said queued packet.
14. A station according to Claim 13, wherein said station is a router for routing a sequence of packets via a telecommunications network.
15. A method of managing the dispatch of a sequence of packets via a telecommunications network, the method comprising:
queuing packets for transmission;
comparing flow and packet sequence parameters of a new packet for transmission to flow and packet sequence parameters of queued packets; and responding to detection of said new packet being a retransmission of a queued packet relating to the same flow path to discard either said new packet or said queued packet.
queuing packets for transmission;
comparing flow and packet sequence parameters of a new packet for transmission to flow and packet sequence parameters of queued packets; and responding to detection of said new packet being a retransmission of a queued packet relating to the same flow path to discard either said new packet or said queued packet.
16. The method of Claim 15, wherein said comparing step comprises comparing source and destination flow parameters of said new and said queued packet(s).
17. The method of Claim 16, wherein said source parameters comprise a source IP
address and a source TCP port and said destination parameters comprise a destination IP address and a destination TCP port.
address and a source TCP port and said destination parameters comprise a destination IP address and a destination TCP port.
18. The method of Claim 17, wherein said comparing step further comprises comparing packet sequence numbers for said new packet and said queued packet(s).
19. The method of Claim 15, wherein a new packet is determined to be a retransmission of a queued packet if:
i) the new packet sequence number equals the queued packet sequence number; and ii) the new packet acknowledgement number is less than the queued packet acknowledgement number.
i) the new packet sequence number equals the queued packet sequence number; and ii) the new packet acknowledgement number is less than the queued packet acknowledgement number.
20. The method of Claim 15, wherein said new packet is added to said queue when said new packet is not a retransmission of a queued packet.
21. The method of Claim 15, wherein said new packet is dropped when said new packet is a retransmission of a queued packet and the length of said queued packet is greater than or equal to that of said new packet.
22. The method of Claim 15, wherein said new packet replaces a queued packet in said queue when said new packet is a retransmission of said queued packet and the length of said new packet is greater than that of said queued packet.
23. A software dispatch mechanism on a storage medium for controlling the dispatch of a sequence of packets via a telecommunications network, said software dispatch mechanism being configured to be operable to define:
a queue for packets for transmission; and a queue controller responsive to receipt of a new packet for transmission to compare parameters of said new packet to parameters of a packet already in said queue, said queue controller determining whether to queue or to drop said new packet depending on the result of said comparison(s).
a queue for packets for transmission; and a queue controller responsive to receipt of a new packet for transmission to compare parameters of said new packet to parameters of a packet already in said queue, said queue controller determining whether to queue or to drop said new packet depending on the result of said comparison(s).
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US08/942,809 | 1997-10-02 | ||
| US08/942,809 US6473425B1 (en) | 1997-10-02 | 1997-10-02 | Mechanism for dispatching packets via a telecommunications network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CA2249169A1 true CA2249169A1 (en) | 1999-04-02 |
Family
ID=25478631
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CA002249169A Abandoned CA2249169A1 (en) | 1997-10-02 | 1998-09-30 | Mechanism for dispatching packets via a telecommunications network |
Country Status (5)
| Country | Link |
|---|---|
| US (1) | US6473425B1 (en) |
| EP (1) | EP0912028B1 (en) |
| JP (1) | JPH11234339A (en) |
| CA (1) | CA2249169A1 (en) |
| DE (1) | DE69833631D1 (en) |
Families Citing this family (133)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9286294B2 (en) | 1992-12-09 | 2016-03-15 | Comcast Ip Holdings I, Llc | Video and digital multimedia aggregator content suggestion engine |
| US7168084B1 (en) | 1992-12-09 | 2007-01-23 | Sedna Patent Services, Llc | Method and apparatus for targeting virtual objects |
| US6697868B2 (en) | 2000-02-28 | 2004-02-24 | Alacritech, Inc. | Protocol processing stack for use with intelligent network interface device |
| US8539112B2 (en) | 1997-10-14 | 2013-09-17 | Alacritech, Inc. | TCP/IP offload device |
| US6757746B2 (en) | 1997-10-14 | 2004-06-29 | Alacritech, Inc. | Obtaining a destination address so that a network interface device can write network data without headers directly into host memory |
| US6226680B1 (en) | 1997-10-14 | 2001-05-01 | Alacritech, Inc. | Intelligent network interface system method for protocol processing |
| US6434620B1 (en) | 1998-08-27 | 2002-08-13 | Alacritech, Inc. | TCP/IP offload network interface device |
| US8782199B2 (en) | 1997-10-14 | 2014-07-15 | A-Tech Llc | Parsing a packet header |
| US7174393B2 (en) | 2000-12-26 | 2007-02-06 | Alacritech, Inc. | TCP/IP offload network interface device |
| US7167927B2 (en) | 1997-10-14 | 2007-01-23 | Alacritech, Inc. | TCP/IP offload device with fast-path TCP ACK generating and transmitting mechanism |
| US8621101B1 (en) | 2000-09-29 | 2013-12-31 | Alacritech, Inc. | Intelligent network storage interface device |
| US7237036B2 (en) | 1997-10-14 | 2007-06-26 | Alacritech, Inc. | Fast-path apparatus for receiving data corresponding a TCP connection |
| PT1034262E (en) * | 1997-11-18 | 2005-10-31 | Pioneer Hi Bred Int | COMPOSITIONS AND METHODS FOR THE GENETIC MODIFICATION OF PLANTS |
| US6754905B2 (en) | 1998-07-23 | 2004-06-22 | Diva Systems Corporation | Data structure and methods for providing an interactive program guide |
| US7091968B1 (en) * | 1998-07-23 | 2006-08-15 | Sedna Patent Services, Llc | Method and apparatus for encoding a user interface |
| US9924234B2 (en) | 1998-07-23 | 2018-03-20 | Comcast Ip Holdings I, Llc | Data structure and methods for providing an interactive program |
| EP1097587A1 (en) | 1998-07-23 | 2001-05-09 | Diva Systems Corporation | Interactive user interface |
| US7664883B2 (en) | 1998-08-28 | 2010-02-16 | Alacritech, Inc. | Network interface device that fast-path processes solicited session layer read commands |
| JP3230671B2 (en) * | 1999-01-14 | 2001-11-19 | 日本電気株式会社 | Packet billing device |
| EP1179000A4 (en) | 1999-02-26 | 2005-10-12 | Millennium Pharm Inc | Secreted proteins and uses thereof |
| US7096487B1 (en) | 1999-10-27 | 2006-08-22 | Sedna Patent Services, Llc | Apparatus and method for combining realtime and non-realtime encoded content |
| US6904610B1 (en) | 1999-04-15 | 2005-06-07 | Sedna Patent Services, Llc | Server-centric customized interactive program guide in an interactive television environment |
| US6754271B1 (en) | 1999-04-15 | 2004-06-22 | Diva Systems Corporation | Temporal slice persistence method and apparatus for delivery of interactive program guide |
| CA2299022A1 (en) * | 1999-04-30 | 2000-10-30 | Nortel Networks Corporation | Method and apparatus for bandwidth management of aggregate data flows |
| US6700871B1 (en) * | 1999-05-04 | 2004-03-02 | 3Com Corporation | Increased throughput across data network interface by dropping redundant packets |
| DE19921589C2 (en) * | 1999-05-05 | 2002-10-24 | Siemens Ag | Method for operating a data transmission system |
| US7313627B1 (en) | 1999-09-30 | 2007-12-25 | Data Expedition, Inc. | Flow control method and apparatus |
| US7404003B1 (en) | 1999-09-30 | 2008-07-22 | Data Expedition, Inc. | Method and apparatus for client side state management |
| US7158479B1 (en) * | 1999-09-30 | 2007-01-02 | Data Expedition, Inc. | Method and apparatus for non contiguous sliding window |
| US7046665B1 (en) * | 1999-10-26 | 2006-05-16 | Extreme Networks, Inc. | Provisional IP-aware virtual paths over networks |
| CA2388606C (en) | 1999-10-27 | 2009-12-29 | Diva Systems Corporation | Picture-in-picture and multiple video streams using slice-based encoding |
| US6608818B1 (en) | 1999-11-10 | 2003-08-19 | Qualcomm Incorporated | Radio link protocol enhancements to reduce setup time for data calls |
| JP2001142845A (en) * | 1999-11-17 | 2001-05-25 | Toshiba Corp | Computer system and data transfer control method |
| US8199646B1 (en) | 1999-12-07 | 2012-06-12 | Nortel Networks Limited | System, device, and method for distributing link state information in a communication network |
| US6779038B1 (en) * | 1999-12-31 | 2004-08-17 | Nortel Networks Limited | System and method for extending virtual synchrony to wide area networks |
| EP1122959A1 (en) * | 2000-02-03 | 2001-08-08 | Telefonaktiebolaget Lm Ericsson | Handling of circuit-switched data services in IP-based GSM networks |
| US7324442B1 (en) * | 2000-02-28 | 2008-01-29 | The Board Of Trustees Of The Leland Stanford Junior University | Active queue management toward fair bandwidth allocation |
| US6744765B1 (en) * | 2000-08-24 | 2004-06-01 | Sun Microsystems, Inc. | Mechanism for completing messages in memory |
| US6381242B1 (en) * | 2000-08-29 | 2002-04-30 | Netrake Corporation | Content processor |
| US6826152B1 (en) | 2000-09-13 | 2004-11-30 | Harris Corporation | System and method of conserving bandwidth in the transmission of message packets |
| US6816458B1 (en) * | 2000-09-13 | 2004-11-09 | Harris Corporation | System and method prioritizing message packets for transmission |
| US6856599B1 (en) * | 2000-09-13 | 2005-02-15 | Harris Corporation | System and method of reducing retransmission of messages in a TCP/IP environment |
| US6826153B1 (en) | 2000-09-13 | 2004-11-30 | Jeffrey Kroon | System and method of increasing the message throughput in a radio network |
| US8019901B2 (en) | 2000-09-29 | 2011-09-13 | Alacritech, Inc. | Intelligent network storage interface system |
| US7013346B1 (en) * | 2000-10-06 | 2006-03-14 | Apple Computer, Inc. | Connectionless protocol |
| US7760737B2 (en) * | 2000-11-30 | 2010-07-20 | Audiocodes, Inc. | Method for reordering and reassembling data packets in a network |
| US7421505B2 (en) * | 2000-12-21 | 2008-09-02 | Noatak Software Llc | Method and system for executing protocol stack instructions to form a packet for causing a computing device to perform an operation |
| US7287090B1 (en) * | 2000-12-21 | 2007-10-23 | Noatak Software, Llc | Method and system for identifying a computing device in response to a request packet |
| US7512686B2 (en) * | 2000-12-21 | 2009-03-31 | Berg Mitchell T | Method and system for establishing a data structure of a connection with a client |
| US20020116605A1 (en) * | 2000-12-21 | 2002-08-22 | Berg Mitchell T. | Method and system for initiating execution of software in response to a state |
| US20020116532A1 (en) * | 2000-12-21 | 2002-08-22 | Berg Mitchell T. | Method and system for communicating an information packet and identifying a data structure |
| US20020116397A1 (en) * | 2000-12-21 | 2002-08-22 | Berg Mitchell T. | Method and system for communicating an information packet through multiple router devices |
| US7546369B2 (en) * | 2000-12-21 | 2009-06-09 | Berg Mitchell T | Method and system for communicating a request packet in response to a state |
| US7418522B2 (en) * | 2000-12-21 | 2008-08-26 | Noatak Software Llc | Method and system for communicating an information packet through multiple networks |
| US6772375B1 (en) * | 2000-12-22 | 2004-08-03 | Network Appliance, Inc. | Auto-detection of limiting factors in a TCP connection |
| US8051212B2 (en) * | 2001-04-11 | 2011-11-01 | Mellanox Technologies Ltd. | Network interface adapter with shared data send resources |
| US7068595B2 (en) | 2001-04-13 | 2006-06-27 | Sun Microsystems, Inc. | Method and apparatus for facilitating instant failover during packet routing |
| US7853781B2 (en) * | 2001-07-06 | 2010-12-14 | Juniper Networks, Inc. | Load balancing secure sockets layer accelerator |
| US7908472B2 (en) * | 2001-07-06 | 2011-03-15 | Juniper Networks, Inc. | Secure sockets layer cut through architecture |
| US7149892B2 (en) * | 2001-07-06 | 2006-12-12 | Juniper Networks, Inc. | Secure sockets layer proxy architecture |
| US7228412B2 (en) * | 2001-07-06 | 2007-06-05 | Juniper Networks, Inc. | Bufferless secure sockets layer architecture |
| US7793326B2 (en) | 2001-08-03 | 2010-09-07 | Comcast Ip Holdings I, Llc | Video and digital multimedia aggregator |
| US7908628B2 (en) | 2001-08-03 | 2011-03-15 | Comcast Ip Holdings I, Llc | Video and digital multimedia aggregator content coding and formatting |
| US7249193B1 (en) * | 2001-08-28 | 2007-07-24 | Emc Corporation | SRDF assist |
| WO2003028289A2 (en) * | 2001-09-26 | 2003-04-03 | Siemens Aktiengesellschaft | Method for transmitting real time data messages in a cyclic communications system |
| JP4153201B2 (en) * | 2001-12-19 | 2008-09-24 | 富士通株式会社 | Communication control method, communication system, and computer program |
| US7543087B2 (en) | 2002-04-22 | 2009-06-02 | Alacritech, Inc. | Freeing transmit memory on a network interface device prior to receiving an acknowledgement that transmit data has been received by a remote device |
| US20040044796A1 (en) * | 2002-09-03 | 2004-03-04 | Vangal Sriram R. | Tracking out-of-order packets |
| US7181544B2 (en) * | 2002-09-03 | 2007-02-20 | Intel Corporation | Network protocol engine |
| US20040143592A1 (en) * | 2002-09-30 | 2004-07-22 | Philippe Jung | Method for processing redundant packets in computer network equipment |
| US7324540B2 (en) | 2002-12-31 | 2008-01-29 | Intel Corporation | Network protocol off-load engines |
| DE602004011904D1 (en) * | 2003-02-27 | 2008-04-03 | Koninkl Philips Electronics Nv | METHOD AND WIRELESS COMPONENT FOR AVOIDING TCP PACKET TRANSMISSION DURING TRANSMISSION OF A MOBILE DEVICE |
| US20040213278A1 (en) * | 2003-04-24 | 2004-10-28 | Broadcom Corporation | System, method, and computer program product for in-place, lightweight Ack promotion in a cable modem environment |
| JP2005039726A (en) * | 2003-07-18 | 2005-02-10 | Matsushita Electric Ind Co Ltd | Base station system and transmitting method |
| US20050122977A1 (en) * | 2003-12-05 | 2005-06-09 | Microsoft Corporation | Efficient download mechanism for devices with limited local storage |
| US7917649B2 (en) * | 2003-12-19 | 2011-03-29 | Nortel Networks Limited | Technique for monitoring source addresses through statistical clustering of packets |
| US7394813B2 (en) * | 2004-05-05 | 2008-07-01 | Sharp Laboratories Of America, Inc. | Systems and methods for implementing an acknowledgement mechanism for transmission of a real-time data stream |
| US7570636B2 (en) * | 2004-06-29 | 2009-08-04 | Damaka, Inc. | System and method for traversing a NAT device for peer-to-peer hybrid communications |
| US7623476B2 (en) * | 2004-06-29 | 2009-11-24 | Damaka, Inc. | System and method for conferencing in a peer-to-peer hybrid communications network |
| US7656870B2 (en) | 2004-06-29 | 2010-02-02 | Damaka, Inc. | System and method for peer-to-peer hybrid communications |
| US8437307B2 (en) | 2007-09-03 | 2013-05-07 | Damaka, Inc. | Device and method for maintaining a communication session during a network transition |
| US7778187B2 (en) * | 2004-06-29 | 2010-08-17 | Damaka, Inc. | System and method for dynamic stability in a peer-to-peer hybrid communications network |
| US7623516B2 (en) * | 2004-06-29 | 2009-11-24 | Damaka, Inc. | System and method for deterministic routing in a peer-to-peer hybrid communications network |
| US20070078720A1 (en) * | 2004-06-29 | 2007-04-05 | Damaka, Inc. | System and method for advertising in a peer-to-peer hybrid communications network |
| US7933260B2 (en) * | 2004-06-29 | 2011-04-26 | Damaka, Inc. | System and method for routing and communicating in a heterogeneous network environment |
| US8009586B2 (en) | 2004-06-29 | 2011-08-30 | Damaka, Inc. | System and method for data transfer in a peer-to peer hybrid communication network |
| US8050272B2 (en) * | 2004-06-29 | 2011-11-01 | Damaka, Inc. | System and method for concurrent sessions in a peer-to-peer hybrid communications network |
| US8248939B1 (en) | 2004-10-08 | 2012-08-21 | Alacritech, Inc. | Transferring control of TCP connections between hierarchy of processing mechanisms |
| US7673060B2 (en) * | 2005-02-01 | 2010-03-02 | Hewlett-Packard Development Company, L.P. | Systems and methods for providing reliable multicast messaging in a multi-node graphics system |
| US7738500B1 (en) | 2005-12-14 | 2010-06-15 | Alacritech, Inc. | TCP timestamp synchronization for network connections that are offloaded to network interface devices |
| SG163590A1 (en) * | 2006-03-31 | 2010-08-30 | Qualcomm Inc | Memory management for high speed media access control |
| JP4882106B2 (en) * | 2007-05-10 | 2012-02-22 | Ｎｅｃカシオモバイルコミュニケーションズ株式会社 | Communication apparatus and program |
| JP2008306551A (en) * | 2007-06-08 | 2008-12-18 | Sanden Corp | Connection apparatus for communication device |
| US8862164B2 (en) * | 2007-09-28 | 2014-10-14 | Damaka, Inc. | System and method for transitioning a communication session between networks that are not commonly controlled |
| WO2009070718A1 (en) * | 2007-11-28 | 2009-06-04 | Damaka, Inc. | System and method for endpoint handoff in a hybrid peer-to-peer networking environment |
| US8369348B2 (en) * | 2008-01-28 | 2013-02-05 | Broadcom Corporation | Method, and system, and computer program product for dynamically adjusting acknowledgement filtering for high-latency environments |
| US8015313B2 (en) * | 2008-03-04 | 2011-09-06 | Sony Corporation | Method and apparatus for managing transmission of TCP data segments |
| US8539513B1 (en) | 2008-04-01 | 2013-09-17 | Alacritech, Inc. | Accelerating data transfer in a virtual computer system with tightly coupled TCP connections |
| US7551621B1 (en) * | 2008-07-21 | 2009-06-23 | International Business Machines Corporation | Method for detecting and reducing packet drops |
| US8341286B1 (en) | 2008-07-31 | 2012-12-25 | Alacritech, Inc. | TCP offload send optimization |
| US9306793B1 (en) | 2008-10-22 | 2016-04-05 | Alacritech, Inc. | TCP offload device that batches session layer headers to reduce interrupts as well as CPU copies |
| US8874785B2 (en) | 2010-02-15 | 2014-10-28 | Damaka, Inc. | System and method for signaling and data tunneling in a peer-to-peer environment |
| US8892646B2 (en) | 2010-08-25 | 2014-11-18 | Damaka, Inc. | System and method for shared session appearance in a hybrid peer-to-peer environment |
| US8725895B2 (en) | 2010-02-15 | 2014-05-13 | Damaka, Inc. | NAT traversal by concurrently probing multiple candidates |
| US8325623B1 (en) | 2010-02-16 | 2012-12-04 | Google Inc. | System and method for reducing latency during data transmissions over a network |
| US8689307B2 (en) * | 2010-03-19 | 2014-04-01 | Damaka, Inc. | System and method for providing a virtual peer-to-peer environment |
| US9043488B2 (en) | 2010-03-29 | 2015-05-26 | Damaka, Inc. | System and method for session sweeping between devices |
| US9191416B2 (en) | 2010-04-16 | 2015-11-17 | Damaka, Inc. | System and method for providing enterprise voice call continuity |
| US8352563B2 (en) | 2010-04-29 | 2013-01-08 | Damaka, Inc. | System and method for peer-to-peer media routing using a third party instant messaging system for signaling |
| US8446900B2 (en) | 2010-06-18 | 2013-05-21 | Damaka, Inc. | System and method for transferring a call between endpoints in a hybrid peer-to-peer network |
| US8611540B2 (en) | 2010-06-23 | 2013-12-17 | Damaka, Inc. | System and method for secure messaging in a hybrid peer-to-peer network |
| US8468010B2 (en) | 2010-09-24 | 2013-06-18 | Damaka, Inc. | System and method for language translation in a hybrid peer-to-peer environment |
| US8743781B2 (en) | 2010-10-11 | 2014-06-03 | Damaka, Inc. | System and method for a reverse invitation in a hybrid peer-to-peer environment |
| US8407314B2 (en) | 2011-04-04 | 2013-03-26 | Damaka, Inc. | System and method for sharing unsupported document types between communication devices |
| US8694587B2 (en) | 2011-05-17 | 2014-04-08 | Damaka, Inc. | System and method for transferring a call bridge between communication devices |
| US9154813B2 (en) | 2011-06-09 | 2015-10-06 | Comcast Cable Communications, Llc | Multiple video content in a composite video stream |
| US8478890B2 (en) * | 2011-07-15 | 2013-07-02 | Damaka, Inc. | System and method for reliable virtual bi-directional data stream communications with single socket point-to-multipoint capability |
| US9027032B2 (en) | 2013-07-16 | 2015-05-05 | Damaka, Inc. | System and method for providing additional functionality to existing software in an integrated manner |
| US9357016B2 (en) | 2013-10-18 | 2016-05-31 | Damaka, Inc. | System and method for virtual parallel resource management |
| US9240939B2 (en) * | 2013-10-22 | 2016-01-19 | Cisco Technology, Inc. | Detecting packet loss and retransmission in a network environment |
| WO2016022574A1 (en) | 2014-08-05 | 2016-02-11 | Damaka, Inc. | System and method for providing unified communications and collaboration (ucc) connectivity between incompatible systems |
| US10298494B2 (en) * | 2014-11-19 | 2019-05-21 | Strato Scale Ltd. | Reducing short-packet overhead in computer clusters |
| WO2016079626A1 (en) * | 2014-11-19 | 2016-05-26 | Strato Scale Ltd. | Reducing short-packet overhead in computer clusters |
| US9843530B2 (en) * | 2015-12-15 | 2017-12-12 | International Business Machines Corporation | System, method, and recording medium for queue management in a forwarder |
| US10091025B2 (en) | 2016-03-31 | 2018-10-02 | Damaka, Inc. | System and method for enabling use of a single user identifier across incompatible networks for UCC functionality |
| US10089339B2 (en) * | 2016-07-18 | 2018-10-02 | Arm Limited | Datagram reassembly |
| US10805434B2 (en) | 2017-06-08 | 2020-10-13 | Hyannis Port Research, Inc. | Dynamic TCP stream processing with modification notification |
| CN109274530B (en) * | 2018-09-05 | 2022-03-22 | 杭州安恒信息技术股份有限公司 | TCP (Transmission control protocol) -based network data packet zero-error scene reproduction method and device |
| CN114125940A (en) * | 2020-08-31 | 2022-03-01 | Oppo广东移动通信有限公司 | Data message sending method, data message processing method, data message sending device, data message processing device, data message sending equipment and data message |
| CN112583936B (en) * | 2020-12-29 | 2022-09-09 | 上海阅维科技股份有限公司 | Method for recombining transmission conversation flow |
| US11902343B1 (en) | 2021-04-19 | 2024-02-13 | Damaka, Inc. | System and method for highly scalable browser-based audio/video conferencing |
| US11770584B1 (en) | 2021-05-23 | 2023-09-26 | Damaka, Inc. | System and method for optimizing video communications based on device capabilities |
| CN113992363B (en) * | 2021-10-11 | 2024-02-27 | 杭州迪普科技股份有限公司 | IEC104 protocol communication-based method and device |
Family Cites Families (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5175765A (en) * | 1989-05-09 | 1992-12-29 | Digital Equipment Corporation | Robust data broadcast over a distributed network with malicious failures |
| US5309437A (en) | 1990-06-29 | 1994-05-03 | Digital Equipment Corporation | Bridge-like internet protocol router |
| US5231633A (en) * | 1990-07-11 | 1993-07-27 | Codex Corporation | Method for prioritizing, selectively discarding, and multiplexing differing traffic type fast packets |
| CA2065578C (en) | 1991-04-22 | 1999-02-23 | David W. Carr | Packet-based data compression method |
| US5307413A (en) | 1991-07-19 | 1994-04-26 | Process Software Corporation | Method and apparatus for adding data compression and other services in a computer network |
| US5379297A (en) * | 1992-04-09 | 1995-01-03 | Network Equipment Technologies, Inc. | Concurrent multi-channel segmentation and reassembly processors for asynchronous transfer mode |
| US5307347A (en) | 1992-04-10 | 1994-04-26 | International Business Machines Corporation | Method and apparatus for sharing a telecommunications channel among multiple users |
| JP2826416B2 (en) | 1992-06-05 | 1998-11-18 | 日本電気株式会社 | Connection router between local area networks |
| GB9326276D0 (en) * | 1993-12-23 | 1994-02-23 | Newbridge Network Corp | Frame relay interface |
| US5535199A (en) | 1994-09-06 | 1996-07-09 | Sun Microsystems, Inc. | TCP/IP header compression X.25 networks |
| US5650993A (en) * | 1995-03-20 | 1997-07-22 | Bell Communications Research, Inc. | Drop from front of buffer policy in feedback networks |
| US5621798A (en) * | 1995-04-18 | 1997-04-15 | Intel Corporation | Method and apparatus for cooperative messaging |
| US5586121A (en) * | 1995-04-21 | 1996-12-17 | Hybrid Networks, Inc. | Asymmetric hybrid access system and method |
| US5734865A (en) * | 1995-06-07 | 1998-03-31 | Bull Hn Information Systems Inc. | Virtual local area network well-known port routing mechanism for mult--emulators in an open system environment |
| US5754754A (en) * | 1995-07-26 | 1998-05-19 | International Business Machines Corporation | Transmission order based selective repeat data transmission error recovery system and method |
| GB2304210B (en) * | 1995-08-11 | 2000-02-16 | Fujitsu Ltd | Data receiving devices |
| US5664091A (en) * | 1995-08-31 | 1997-09-02 | Ncr Corporation | Method and system for a voiding unnecessary retransmissions using a selective rejection data link protocol |
| US6006268A (en) * | 1997-07-31 | 1999-12-21 | Cisco Technology, Inc. | Method and apparatus for reducing overhead on a proxied connection |
| US5960178A (en) * | 1997-08-08 | 1999-09-28 | Bell Communications Research, Inc. | Queue system and method for point-to-point message passing having a separate table for storing message state and identifier of processor assigned to process the message |
-
1997
- 1997-10-02 US US08/942,809 patent/US6473425B1/en not_active Expired - Lifetime
-
1998
- 1998-09-24 EP EP98402355A patent/EP0912028B1/en not_active Expired - Lifetime
- 1998-09-24 DE DE69833631T patent/DE69833631D1/en not_active Expired - Lifetime
- 1998-09-30 CA CA002249169A patent/CA2249169A1/en not_active Abandoned
- 1998-10-02 JP JP10280640A patent/JPH11234339A/en active Pending
Also Published As
| Publication number | Publication date |
|---|---|
| EP0912028A3 (en) | 2003-12-17 |
| EP0912028B1 (en) | 2006-03-01 |
| EP0912028A2 (en) | 1999-04-28 |
| DE69833631D1 (en) | 2006-04-27 |
| US6473425B1 (en) | 2002-10-29 |
| JPH11234339A (en) | 1999-08-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6473425B1 (en) | Mechanism for dispatching packets via a telecommunications network | |
| US6249530B1 (en) | Network bandwidth control | |
| US7483376B2 (en) | Method and apparatus for discovering path maximum transmission unit (PMTU) | |
| US7804780B2 (en) | Receiving and transmitting devices for providing fragmentation at a transport level along a transmission path | |
| US8190960B1 (en) | Guaranteed inter-process communication | |
| CN104025525B (en) | For sending the method and apparatus and exchange apparatus of packet | |
| US7742454B2 (en) | Network performance by dynamically setting a reassembly timer based on network interface | |
| US7103674B2 (en) | Apparatus and method of reducing dataflow distruption when detecting path maximum transmission unit (PMTU) | |
| JP5038425B2 (en) | Optimization process of traffic control in packet telecommunications network | |
| US6026093A (en) | Mechanism for dispatching data units via a telecommunications network | |
| US20110206059A1 (en) | Methods and devices for transmitting data between storage area networks | |
| JP2000253048A (en) | Data communication method, terminal, router, data communication system and its recording medium | |
| US20090080332A1 (en) | Method and System for a Fast Drop Recovery for a TCP Connection | |
| US7480301B2 (en) | Method, system and article for improved TCP performance during retransmission in response to selective acknowledgement | |
| US20070291782A1 (en) | Acknowledgement filtering | |
| JP2002217988A (en) | Device and method for controlling data delivery | |
| US6996105B1 (en) | Method for processing data packet headers | |
| EP1586182B1 (en) | Methods and devices for transmitting data between storage area networks | |
| US8578040B2 (en) | Method, system and article for client application control of network transmission loss tolerance | |
| US7593318B2 (en) | Method and apparatus for header updating | |
| US20030126192A1 (en) | Protocol processing | |
| US7738493B2 (en) | Methods and devices for transmitting data between storage area networks | |
| CA2246134C (en) | Enhanced network protocol | |
| JP2000341333A (en) | Network packet transmission/reception method and network adaptor | |
| Velten et al. | RFC0908: Reliable Data Protocol |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| FZDE | Discontinued |