WO2002037700A2

WO2002037700A2 - Dynamic wireless link adaptation

Info

Publication number: WO2002037700A2
Application number: PCT/US2001/051094
Authority: WO
Inventors: Jean Gosselin; Donald J. Harper
Original assignee: Airnet Communications Corporation
Priority date: 2000-11-01
Filing date: 2001-10-29
Publication date: 2002-05-10
Also published as: AU2002235273A1; WO2002037700A3

Abstract

The instant invention provides a method and apparatus for efficiently selecting an optimal channel coding scheme from a plurality of successively higher order channel coding schemens (500) that are utilized over a packetized radio link in a wireless communication system. In a wireless system that utilizes packet switching, a more efficient and robust link can be maintained by dynamically selecting an optimal channel coding scheme best suited for the instantaneous conditions that exist on the radio link. The minimum data rate (500b-1) corresponding to a specified maximum data rate (500a-1) for a given channel coding scheme is determined. A normalized minimum throughput (500c-1) and its corresponding normalized maximum throughput (500d-1) for each channel coding scheme are determined based on the maximum data rate for the specified coding scheme. The optimal channel coding then determined based on whether an instantaneous measured throughput falls within the range of permitted throughput specified by the normalized minimum and maximum throughout for the current channel coding scheme used on the channel.

Description

DYNAMIC WIRELESS LINK ADAPTATION

BACKGROUND OF THE INVENTION

Field of the Invention

The instant invention pertains generally to the field of wireless cellular

communication. More particularly, the invention describes a method and

apparatus for dynamically selecting an optimal coding scheme dependent on the

channel conditions.

Description of Related Art

In the field of wireless cellular communication, such as analog cellular

telephony for example, Advanced Mobile Phone Service (AMPS), digital cellular

telephony, for example, Time Division Multiple Access (TDMA) or Code Division

Multiple Access (CDMA), innovative ways have been created to transmit voice

using various coding schemes associated with each access methodology. While

these schemes are acceptable for coding voice information for transmission over

a radio communications channel, they have not always provided a satisfactory

approach for dealing with non-voice data transmission. As the demand for

communicating high bandwidth information increases in order to accommodate

mobile data computing traffic, the coding scheme used in those traditional

access technologies have become inadequate. To address this inherent

inadequacy, new methods have to be devised to more efficiently transmit packet

based information over these radio channels. General Packet Radio Service (GPRS) is a high-speed packet based

technology for GSM networks that supports TCP/IP and X.25 connectivity. The

GPRS packet based air interface is overlaid on the current GSM circuit switched

network, and packet switching ensures that scarce radio resources are used

only when a subscriber needs to transmit or receive information. This virtual

connectivity allows concurrent use of radio resources by a plurality of users

within a single cell and eliminates the need for a dedicated communication path

for each user. By utilizing all 8 time slots in a TDMA frame, GPRS facilitates

communication at speeds up to a theoretical maximum of 171 .2 Kbps. When

compared to today's circuit switched mobile data networks that require a dialup

modem connection, GPRS, being packet based, allows communication that is

"always on" without having to dialup every time a communication session is

required. Hence, the need to maintain a dedicated transmission path as is

typical with circuit switched networks is eliminated. Furthermore, the time

required to establish a connection is orders of magnitude less than that which is

required for circuit switched communication.

While today's GSM based communication systems allow point-to-point

transmission of data through Short Message Service (SMS), the number of

characters that can be sent is typically limited to 160 characters (Octets).

However, with GPRS there is no such limitation, since information is split into

packets before being transmitted and then reassembled at the receiving end. In

order to realize these advantages gained by employing a packet based radio interface channel coding schemes are required to code the packetized

information over the air interface. Accordingly, there is a need to provide an

efficient solution for coding packetized information over an air interface. GPRS

provides four (4) basic channel coding schemes. Each successively higher order

channel coding scheme provides a relatively higher level of performance under

varying channel conditions and provides a relatively lower level of protection

against the effects of noise and interference. Data blocks received with errors

due to noise and interference are typically retransmitted which significantly

reduces the system throughput. As the conditions on the channel vary, the

channel coding scheme can be dynamically changed to a different level to

mitigate adverse effects and/or maximize throughput.

Although GPRS provides four coding schemes, there is no provision that

dictates the criteria for dynamically selecting an appropriate channel coding

scheme given the instantaneous conditions that exist on the communications

channel. Consequently, there exists a need to define a criterion for selecting an

optimal channel coding scheme dependent on the instantaneous conditions that

exist on the communication channel/link.

With prior art wireless communication systems, several factors are used

for estimating a channel to determine an appropriate coding scheme best suited

for conditions that exist on a communications channel. To be effective, an

indicator must be able to provide some representation of the relationship

between the desired signal strength as compared to interference and noise, that is, the carrier-to-interference ratio, C/l. For example, a receive signal strength

indicator (RSSI), although a measure of the received power of the signal, is not

by itself an acceptable indicator since there is nothing to distinguish between

the desired signal and the noise or interference component.

¹ By comparison, bit error rates (BER) are commonly used as an indication

of channel conditions. Since bit error rates are correlated to the noise and

interference that occurs on a communication channel, BER can provide some

estimate of channel conditions. However, BER estimates must be taken over

large periods of time to be a fair estimator. As a result of this large estimation

time, substantial fluctuations in signal quality, noise and interference can occur

over the estimation period. These fluctuations may not be accurately reflected

in the BER. Consequently, the BER estimate for an acceptable carrier quality

and an unacceptable carrier might be similar.

While the drawbacks of channel indicators such as RSSI and BER are

known, some prior art systems attempt to minimize the effects by using a

combination of indicators to generate a channel estimate. However, such

methods are complicated and require excessive computational power. As a

result, in wireless communication systems that utilize a plurality of coding

schemes that are adaptive to real time conditions, the aforementioned methods

for channel estimation are unsuitable. Therefore, there exists a need to provide

a criteria for selecting an optimal channel coding scheme from a plurality of successively higher order channel coding scheme that best suits the

instantaneous channel conditions.

SUMMARY OF THE INVENTION

The invention discloses a method for efficiently and dynamically selecting

an optimal channel coding scheme from a plurality of channel coding schemes

each having a successively higher order coding scheme. The channel coding

schemes are utilized in a wireless cellular communication system having a

packet based radio link between a base station located within a home cell and a

plurality of subscriber transceiver units proximately located. The method

comprises, measuring based on signals communicated over the radio link, a data

throughput value and determining exclusively from the measured data

throughput, an optimal channel coding scheme. The channel coding scheme

over the radio link is then dynamically changed to the optimal channel coding

scheme.

According to one aspect of the method disclosed in the invention, the

method of determining the optimal channel coding scheme comprises comparing

a measured data throughput value to a set of predetermined limits, wherein the

limits have a minimum and a maximum throughput value. The measured

throughput, along with the maximum and the minimum throughput are

normalized with respect to a data rate for a selected channel coding scheme.

The maximum data rate for the selected channel coding scheme is used for the

normalization. In accordance with the method disclosed in the invention, the normalized

minimum throughput is a ratio of a minimum data rate to a corresponding

maximum data rate for a selected coding scheme. The maximum data rate is an

upper limitation on the rate at which data can be transmitted on a channel for a

selected coding scheme. The normalized maximum throughput value is a

predefined parameter that can be changed by the system operator or an

automated means.

In a further aspect of the method disclosed in the invention, the minimum

data rate corresponding to a maximum data rate for each of the plurality of

channel coding schemes is determined by assigning to a minimum data rate, a

maximum data rate from an immediately previous lower order channel coding

scheme.

In a accordance with the in the instant invention, a system for efficiently

and dynamically selecting an optimal channel coding scheme from a plurality of

channel coding schemes each having a successively higher order coding scheme

is disclosed. The channel coding scheme can be used in a wireless

communication system having a packet based radio link between a base station

located within a home cell and a plurality of subscriber transceiver units

proximately located. The system comprises a processing means for measuring

based on signals communicated over the radio link, a data throughput value.

The system also has a means for determining exclusively from the measured

data throughput, an optimal channel coding scheme. A channel coder is used for dynamically changing the channel coding scheme over the radio link, to an

optimal channel coding scheme.

In a further aspect of the system disclosed in the invention, the optimal

channel coding scheme comprises comparing the measured data throughput

value to a set of predetermined limits, the limits having a minimum and a

maximum throughput value. The measured throughput along with the maximum

and the minimum throughput are normalized with respect to a data rate for a

selected channel coding scheme. The maximum data rate for the selected

channel coding scheme is used for normalization.

In accordance with the system disclosed in the invention, the normalized

minimum throughput is a ratio of the minimum data rate to the corresponding

maximum data rate for a selected coding scheme. The maximum data rate

defines an upper limitation to the rate at which data can be transmitted on a

channel for a selected coding scheme. Moreover, the normalized maximum

throughput value is a predefined parameter.

In a further aspect of the system disclosed in the invention, the minimum

data rate corresponding to the maximum data rate for each of the plurality of

channel coding scheme is determined by assigning to the minimum data rate, a

maximum data rate from an immediately previous lower order channel coding

scheme. BRIEF DESCRIPTION OF THE DRAWINGS

There are shown in the drawings embodiments which are presently

preferred, it being understood, however, that the invention is not limited to the

precise arrangements and instrumentalities shown, wherein:

FIG. 1 is a diagram illustrating an exemplary conventional wireless

communication system.

FIG. 2 is a simplified architecture diagram of the exemplary conventional

wireless communication system illustrated in FIG. 1 , showing the

communication channels.

FIG. 3 illustrates an improved wireless communication system having a

packetized radio interface.

FIG. 4 illustrates an general process by which a normalized maximum and

a normalized minimum throughput can be derived from a given maximum data

rate for each corresponding channel coding scheme.

FIGs. 5a, 5b, 5c and 5d illustrates the steps of an exemplary process by

which a normalized maximum and a normalized minimum throughput can be

derived from a given maximum data rate for each corresponding channel coding

scheme.

FIG. 6 is a flow chart illustrating the algorithm used to change the channel

coding scheme in accordance with the invention. DETAILED DESCRIPTION OF THE INVENTION

Referring now to FIG. 1 and FIG. 2, a conventional wireless

communication system architecture is disclosed therein. The system is

comprised of one or more base stations 1 10-1 & 1 10-4 connected to a serving

base station controller (BSC) 100. Each base station, alternately referred to as

the base transceiver station (BTS), handle speech encoding and decoding and

data rate adaptation in the case of data. Radio signals 120-1 & 120-4 from

subscriber transceiver units 1 15-1 & 1 15-4 are received at a serving base

station 1 10-1 & 1 10-4 respectively. The radio signals 120 have an uplink

channel 200-1 and a downlink channel 200-2.

To save on transmission facilities, a mobile switching center (MSC) 105 is

preferably co-located with the base station controller 100. Although not

shown, MSC 105 can control a plurality of BSCs 100. In addition to being

connected to the BSC 100, MSC 105 can be connected to one or more other

networks such as the Public Switched Telephone network (PSTN) 125. MSC

105 acts as a wireless switch providing switching capability between the

cellular system and other networks such as the PSTN 125. In addition to

coordinating the setting up and tearing down of calls, the MSC also handles

switching between the various BSC under its control.

The BSC 100 controls the radio interface management through the remote

command of the base station and the subscriber transceiver units. This can also

be achieved in conjunction with the base transceiver station (BTS). Some of these management tasks include handover management and the allocation and

release of radio channels. The BSC can also handles intra-BSC or inter-BSC

switching in some systems. Hence, calls handled by base stations under the

control of the same base station controller can be switched by the BSC. For

example, calls originating from subscriber transceiver station 1 15-3 and destined

for. subscriber transceiver station 1 15-4 can be switched by BSC 100 rather

than by the MSC 105.

Referring now to FIG. 2, a simplified architecture of the radio interface

between the subscriber transceiver station 1 15 and a base station 1 10 is

illustrated. The radio interface consists of a plurality of uplink and downlink

channel pairs 200-1 and 200-2. The channels are arranged in such a manner

that each uplink and each downlink RF carrier contains 8 sub-channels or

timeslots. As a result, 8 subscriber stations/units may share a single uplink RF

carrier. For example, a single mobile station 1 15 can be assigned to one of the

8 timeslots of the uplink RF channel 200-1 , which can be used to communicate

signals from the subscriber station/unit 1 15 to the base station 1 10.

Referring now to FIG. 3, an improved wireless communication system

having a packetized radio interface is illustrated. Consistent with the

conventional wireless network of FIG. 1 , additional communications network

entities overlaid onto the conventional wireless communications system are

shown. A GPRS network 130 comprising one or more serving GPRS support

nodes (SGSN) 135-1 , 135-2 and a gateway GPRS support node (GGSN) 140 is incorporated within the conventional system. Serving GPRS support nodes

interconnects one or more base stations via a packet control unit (PCU) 160,

which is typically co-located with the base station controller. For example,

SGSN 135-1 is interconnected to BS 1 10-1 and BS 1 10-2 through PCU 160

which is co-located with BSC 100. Although the PCU is typically co-located

with the BSC, since it forms part of the base station system (BSS), it could be

located elsewhere, such as with the BTS. In addition to being the standard

interface to the SGSN, the packet control unit is in charge of packet control

processing on the base station system side. In addition to managing the radio

channels and radio link control, the PCU can also convert frames from the radio

interface to packets and vice versa. Those skilled in the art will appreciate that

these functions can alternately be performed elsewhere in the system. For

example, these functions may be performed at the BTS.

The SGSN provides packet routing functionality for the SGSN service

area, and functions as a node that is used to send and receive data to and from

the subscriber transceiver stations. In order to accomplish this task, the SGSN

must keep track of the mobile stations being served by the base station within it

service area.

The GGSN 140 interconnects the SGSNs located within the GPRS

network 130. For example, GGSN 140 interconnects SGSN 135-1 and SGSN

135-2. The GGSN provides gateway functionality and connectivity to networks

external to the GPRS network. For example, GGSN 140 provides connectivity to the PSTN 125 through a packet-to-circuit translator (PCT) 126 (used to

translate packet switched data to circuit switched data and vice versa), to public

data networks (PDN), X.25 network 145 and IP network 150. The PCT

translates packet switched data to circuit switched data and vice versa. The

GGSN functions as a network edge device providing connectivity and routing of

data to and from other networks.

Packetized information from a subscriber transceiver station, for example

1 15-1 , is formatted for transmission in one or more of a plurality of uplink

timeslots. A timeslot may be shared, wherein data and voice traffic are

periodically bursted into the same slot. Alternately, a timeslot may be a

dedicated timeslot, wherein only data destined for the GPRS network is

transmitted within that timeslot. The formatted packetized data is then

transmitted in the appropriate timeslot of an uplink channel, for example 200-1 ,

over the air interface and received at a serving base station 1 10. Packet control

units (PCU), for example 160, typically co-located with the base station

controller, receive the packetized data from the base station and dispatches the

packets to the serving GPRS support node (SGSN). The SGSN, which provides

packet routing for the SGSN service area, then routes the packets to the

appropriate GGSN where it is forwarded to the appropriate public data network,

for example X.25 network 145. As an illustration, packets from subscriber

station 1 15-2 arriving at base station 1 10-2 are routed to the serving SGSN

135-1 via PCU 160 co-located with BSC 100. The information is then de- encapsulated using the GPRS tunneling protocol (GTP) before being routed by

the GGSN 140 to its appropriate destination. The GGSN 140 can also

encapsulate information using the GPRS tunneling protocol. For information

going from the public data networks 140 and 150, the data is first encapsulated

using the GPRS tunneling protocol in order to protect the information during

transmission on the radio link.

The instant invention provides a link adaptation algorithm that can

dynamically select an optimal channel coding scheme suitable for instantaneous

channel/link conditions. In one embodiment of the invention, four (4) different

coding schemes can be utilized, with each successive channel coding scheme

having a higher order coding. It should readily be understood by one skilled in

the art that this is not a limitation on the instant invention, since more than four

or less than four different coding schemes of varying orders may be used to

practice the invention. These coding schemes can be defined for both the uplink

and downlink packet data transfer. Each of the coding schemes offer varying

degrees of data encoding to assist in the recovery from adverse effects such as

Rayleigh fading and interference. For illustrative purposes, CS1 , CS2, CS3 and

CS4 reference each coding scheme (CS) respectively. In addition, the link

adaptation algorithm could support forced release or forced handover if the data

rate while in CS1 , drops below a configured level.

Typically, higher order (less robust) channel coding schemes are suitable

for low noise environments wherein the carrier/interference (C/l) ratio is high. Under such conditions, the block error rate (BLER) or corresponding bad data

block ratio is expected to be low. Similarly, lower order (more robust) channel

coding schemes are typically employed in high noise environments wherein the

carrier-to-noise C/N and the C/l ratio is low and the block error rate (BLER) or

corresponding bad data block ratio is expected to be high. The more complex

the degree of coding used on the channel, the lower the actual throughput of

the channel since redundancy has to be added during the coding process.

Likewise, the less complex the degree of coding used on the channel, the

greater the actual throughput of the channel. Since it is always desirable to

have the highest possible throughput, the coding scheme used on the

communication channel can be dynamically changed to minimize adverse

effects. Additionally, since interference is dynamic, rapid adaptation must be

made to utilize an optimal coding scheme. An optimal coding is achieved when

the highest possible throughput is realized to mitigate the corresponding adverse

channel effects.

To facilitate rapid switching between the various coding schemes, a high¬

speed digital signal processor can be used to change the coding algorithm

associated with each coding scheme. In one aspect of the invention, the coding

algorithm could be stored in a high-speed memory and dynamically downloaded

to the DSP processing the channel, whenever there is a need to change the

channel coding scheme. In another aspect of the invention, in order to maximize the efficiency of

dynamically selecting the channel coding scheme, timers can be used to ensure

that the coding scheme is not changed too often. For example, a host process

or CPU can start a timer whenever the coding scheme is switched. Before this

new coding scheme is switched to another scheme, the timer is first checked to

ensure that some minimal predefined time has elapsed before allowing the

coding scheme to be switched. This process ensures that an acceptable steady

state is achieved before a decision is made to select a different channel coding

scheme.

Referring now to FIG. 4, the process of deriving the maximum and

minimum throughput for each corresponding channel coding scheme is now

described. Each of the channel coding schemes CS1 , CS2, CS3, and CS4 can

be assigned a corresponding maximum date rate (DR _max), CS1 (DR1_max),

CS2(DR2_max), CS3(DR3_max), and CS4(DR4_max) respectively. These corresponding

data rates represent the maximum data rate that is permitted for the

corresponding coding scheme. The maximum data rate (DR _max) can be an

assigned parameter. For example, in the GSM specification, a maximum data

rate (DR _max) is defined for each of the four channel coding schemes. Each

coding scheme can also have a corresponding minimum permitted data rate,

CS1 (DR1 _min), CS2(DR2_min), CS3(DR3_min) and CS4(DR4_min).

In accordance with one aspect of the invention, the maximum data rate

permitted for a corresponding coding scheme is used as the minimum data rate for the immediately successive higher order coding scheme. For example,

CS1 (DR1 _max) which is the maximum data rate for coding scheme CS1 , can be

used as the corresponding minimum data rate for the immediately successive

coding scheme CS2, with CS2 being a higher order coding scheme than CS1 .

Similarly, CS2(DR2_max) which is the maximum data rate for coding scheme CS2,

can be used as the corresponding minimum data rate for the immediately

successive coding scheme CS3, with CS3 being a higher order coding scheme

than CS2.

The maximum and minimum data rates for each coding scheme

corresponds to switching points, which indicate the point at which it is

appropriate to switch to a different coding scheme or maintain the current

coding scheme. Hence, the minimum data rate DR_mιn for each coding scheme

CS1 , CS2, CS3 and CS4, defines the lowest rate that should be expected

before the coding scheme should be changed to a more robust coding scheme,

having a lower throughput. Should the data rate fall below the minimum data

rate for CS1 , CS1 (DR1 _mιn), then the BTS can initiate a handover using a

procedure such as the BSS initiated handover for GPRS as defined in the GPRS

standard. Similarly, the maximum data rate DR_max, defines the highest data rate

allowable for a given coding scheme, at which point the coding scheme should

be switched to a coding scheme with a higher throughput.

The optimum switching points corresponding to each coding scheme can

be a function of several different parameters. These include, but are not limited to, the carrier-to-interference ratio (C/l), the channel model, the length of the

message and the type of traffic on the channel. The C/l is the ratio of the

carrier signal to that of any interfering signals, namely, co-channel interference.

Other interference can be due to adjacent channel interference. The channel

model is the mathematical representation or scheme that can be used to

estimate the behavior of the channel. For example, a transfer function for the

channel can be used to describe the effect of the channel on signals as a

function of their frequency. In addition, an impulse response for the channel can

be used to show the kind of time domain signal that can be convolved with any

input signal during channel propagation. The message length affects how the

data can be coded on the channel. For example, voice traffic is very bursty and

contains large gaps of silence. On the other hand, data traffic is non-bursty

without gaps.

In accordance with a further aspect of the instant invention, a normalized

minimum throughput (S_min) and a normalized maximum throughput (S_max) can be

defined. The minimum throughput (S_min) can be used to indicate the actual

graduation point for changing to a lower throughput, which is less robust. The

normalized maximum throughput (S_max) can be used to indicate the actual

graduation point for changing to a higher throughput, which is more robust. The

normalized minimum and maximum throughputs correspond to normalization of

the throughputs (S) for each of the coding schemes by the corresponding data

rate (DR). The normalized maximum throughput (S_max) is a theoretical or ideal value that can be set as a parameter. A typical value can be 95%. The

normalized minimum throughput for a given coding scheme is derived by

expressing the minimum data rate (DR_min) for that coding scheme as a

percentage of the maximum data rate (DR_max) for that coding scheme.

The maximum theoretically expected throughput S_max is the same

regardless of the coding scheme, since it is expected that the coding algorithm

for each coding scheme will mitigate adverse effects affecting the channel at

any given point. It should be understood that this is just an ideal value. The

normalized throughput value could also be set based on measurements taken at

various points within the system. Such measurements can be dynamically taken

and the maximum throughput accordingly set. Preferably, the maximum

throughput is a parameter set by the system operator. An exemplary value

could be 95%.

In accordance with the invention, FIGs. 5a, 5b, 5c and 5d shows an

exemplary embodiment of the instant invention that illustrates a manner for

deriving the switching points at which the coding schemes should be changed.

Given the maximum data rate for a given channel coding scheme, the

corresponding normalized minimum throughput and normalized maximum

throughput are derived. It should readily be understood by one skilled in the art

that the illustration is not intended to constitute a limitation on the invention.

For GSM, the specification defines four channel coding schemes, CS1 , CS2,

CS3 and CS4. The specification assigns a maximum data rate to each of the four channel coding schemes. CS1 is assigned a maximum data rate (DR_max) of

9.05 Kbps. CS2 is assigned a maximum data rate (DR_max) of 13.4 Kbps. CS3 is

assigned a maximum data rate of 15.6 Kbps. CS4 is assigned a maximum data

rate of 21 .4 Kbps. These assignments are illustrated in FIG. 500a at 500a-1.

Given the foregoing maximum data rates, each of the channel coding

schemes CS1 , CS2, CS3 and CS4 can then be assigned a corresponding

minimum data rate (DR_min). The minimum data rate DR_min is derived from the

maximum data rate DR_max for the successive coding scheme, as illustrated in

FIG. 5b at 500b-1 . For example, the maximum data rate DR_max of 9.05 Kbps for

CS1 can be assigned to CS2 as the minimum data rate DR_min of 9.05 Kbps for

CS2. The maximum data rate DR_max of 13.4 Kbps for CS2 can be assigned to

CS3 as the minimum data rate DR_min of 13.4 Kbps for CS3. The maximum data

rate DR_max of 15.6 Kbps for CS3 can be assigned to CS4 as the minimum data

rate DR_min of 15.6 Kbps for CS4. The minimum data rate for CS1 , CS1 (DR1 _min),

is arbitrarily chosen to be 1.0 Kbps. This is an exemplary value and is not

intended to be a limitation. This value can be used to determine the normalized

minimum allowable throughput for CS1. Hence, if the normalized throughput

falls below this value, then the system can force a release of resources or

initiate a handover.

For each of the channel coding schemes, a maximum and a minimum

throughput corresponding to a maximum and minimum data rate respectively, can be determined. The throughput (S) is related to the block error rate (BLER)

by the following equation:

S = DR * (1 BLER_C/I) (1 )

where:

DR is the Data Rate;

S is the throughput; and

BLER_C/| is the block error rate at a given C/l ratio.

The block error rate is a function of the carrier-to-interference (C/l) ratio.

As interference on the channel increases, the C/l ratio will fall and the

corresponding block error rate for that specific C/l ratio will increase. With

reference to the equation defining the throughput (S), it can be seen that for a

constant data rate, as the block error rate increases, the parameter (1 -BLER_C/1)

will decrease, thereby causing the throughput (S) to decrease. Hence, in order

to maximize the throughput, the coding scheme must be robust enough to

minimize the adverse effects that will cause the block error rate to increase.

Since the data rate (DR) is a multiplying factor, it is desirable to normalize the

throughput (S) by the data rate (DR).

Referring to equation (1 ), since the throughput (S) is normalized by the

data rate (DR), the normalized throughput is now equivalent to the parameter (1

BLER_C/I). The normalized minimum throughput (S_min) can then derived by

expressing the minimum data rate (DR_min) for each of the coding scheme as a percentage of the corresponding maximum data rate (DR_max) as illustrated in FIG.

5c at 500c- 1 .

The corresponding normalized maximum throughput (S_max) for the each of

the coding scheme can be an assigned parameter. A typical exemplary

assignment for the normalized maximum throughput (S_max) for the each of the

coding scheme can be 95%, as illustrated in FIG. 5d at 500d-1 .

Referring now to FIG. 6, an exemplary algorithm that can be used to

change the channel coding scheme in accordance with the principles of the

invention is described. The algorithm starts with step 600, followed by step

602, which includes the allocation of channel resources and the setting up of a

data call on the channel. In an effort to choose a starting channel coding

scheme, the system can take a measurement of the carrier-to-interference (C/l)

ratio, step 604, and then assign a channel coding scheme as in step 606, based

on the measured C/l ratio. In one aspect of the invention, the initial assignment

of the channel coding scheme based on the measured C/l can be done by a

lookup table. Alternatively, a default channel coding scheme could be selected

for startup without having to measure the C/l or any other parameter. It should

be recognized by one skilled in the art that the initial measurement of a C/l ratio

is not intended to be a limitation on the invention. Other measurements such as

the block error rate, the carrier-to-noise (C/N) ratio, the throughput, and/or any

combination thereof, can be used. Notwithstanding, it might be preferably advantageous to measure the instantaneous C/l since other parameters can take

more time to accumulate the pertinent statistics.

Once the channel coding scheme is assigned and data is transmitted over

the channel as in step 608, throughput statistics of the channel are measured in

step 610. A determination is made in step 612 as to whether the measured

normalized throughput is greater than the predetermined normalized maximum

throughput limit (S_max). If the measured normalized throughput is greater than

the predetermined normalized maximum throughput limit (S_max), then the coding

scheme can be changed as in step 614, to the next higher order channel coding

scheme. Step 614 can then be followed by steps 608 and 610 to continue

making adjustments dynamically.

If the measured normalized throughput is not greater than the

predetermined normalized maximum throughput limit (S_max) as in step 612, then

a determination is made in step 616 as to whether the measured normalized

throughput is less than the predetermined normalized minimum throughput limit

(S_min). If the measured throughput is less than the normalized minimum

throughput limit (S_min), then the channel coding scheme is checked in step 620

to determine if the current coding scheme is CS1 . If the current channel coding

scheme is not CS1 , then the coding scheme can be changed in step 618, to the

next lower order channel coding scheme. In the event that the channel coding

scheme is CS1 , then the resources for the call are released in step 622 and the

algorithm ends in step 624. Alternately, the BTS could initiate a handover using the procedure for BSS initiated handover as defined in the GPRS standard.

Although steps 620, 622 and 624 are illustrated in the FIG 6., they can be

practiced as optional steps.

There are several factors that might dictate whether it is permissible to

change to a higher order channel coding scheme or to a lower order channel

coding scheme as in steps 614 and 618. As previously stated, a timer could be

used to ensure that a steady state could be reached on the channel. Hence, if a

minimum time has not elapsed since the last change, then the change would not

be allowed. The timer can therefore, be consulted before the change is made.

After the change, the algorithm is restarted with steps 608 and 610 respectively

followed.. Returning to step 616, if the measured throughput is not less than

the normalized minimum throughput limit (S_min), then steps 608 and 610 are

respectively followed to continue monitoring the channel and dynamically make

changes to the coding scheme.

While exemplary systems and methods embodying the present invention

are shown by way of example, it will be understood that the invention is not

limited to these embodiments. Modifications may be made by those skilled in

the art, particularly in light of the foregoing teachings. For example, each of the

elements of the aforementioned embodiments may be utilized alone or in

combination with elements of the other embodiments. Additionally, the

algorithm disclosed in FIG. 6 could easily be modified without departing from the

true spirit of the invention. For example, in addition to making changes to the coding scheme based solely on the normalized throughput, during

implementation, other factors such as the C/l could be considered to ensure that

instantaneous incremental changes in the C/l are accounted for, such as with

another control loop utilizing C/l measurements, or others mentioned herein.

Claims

CLAIMSWhat is claimed is:

1 . In a wireless cellular communication system having a packet based radio

link between a base station located within a home cell and a plurality of

subscriber transceiver units proximately located, a method for efficiently and

dynamically selecting an optimal channel coding scheme from a plurality of

channel coding schemes each having a successively higher order coding

scheme, comprising:

measuring based on signals communicated over said radio link, a data

throughput value;

determining exclusively from said measured data throughput, an optimal

channel coding scheme; and

changing dynamically, the channel coding scheme over the radio link to

said optimal channel coding scheme.

2. The method according to claim 1 , wherein determining said optimal

channel coding scheme comprises comparing said measured data throughput

value to a set of predetermined limits, said limits having a minimum and a

maximum throughput value.

3. The method according to claim 2, wherein said measured throughput and said

maximum and said minimum throughput are normalized with respect to a data

rate for a selected channel coding scheme.

4. The method according to claim 3, wherein said data rate is a maximum

data rate for said selected channel coding scheme.

5. The method according to claim 4, wherein said normalized minimum

throughput is a ratio of a minimum data rate to a corresponding said maximum

data rate for a selected coding scheme.

6. The method according to claim 5, wherein said maximum data rate is an

upper limitation to the rate at which data can be transmitted on a channel for a

selected coding scheme.

7. The method according to claim 6, wherein said normalized maximum

throughput value is a predefined parameter.

8. The method according to claim 6, wherein said minimum data rate

corresponding to said maximum data rate for each of said plurality of channel

coding scheme is determined by assigning to said minimum data rate, a maximum data rate from an immediately previous lower order channel coding

scheme.

9. The method according to claim 8, wherein said step of changing

dynamically said channel coding scheme comprises one of handing-over and

releasing resources if said measured data throughput is less than said minimum

throughput corresponding to a lowest order channel coding scheme of said

plurality of channel coding schemes.

10. In a wireless cellular communication system having a packet based radio

link between a base station located within a home cell and a plurality of

subscriber transceiver units proximately located, a system for efficiently and

dynamically selecting an optimal channel coding scheme from a plurality of

channel coding schemes each having a successively higher order coding

scheme, the system comprising:

processing means for measuring based on signals communicated

over said radio link, a data throughput value;

means for determining exclusively from said measured data

throughput, an optimal channel coding scheme; and

a channel coder for changing dynamically, the channel coding

scheme over the radio link to said optimal channel coding scheme.

1 1 . The system according to claim 10, wherein determining said optimal

channel coding scheme comprises comparing said measured data throughput

value to a set of predetermined limits, said limits having a minimum and a

maximum throughput value.

12. The system according to claim 1 1 , wherein said measured throughput and

said maximum and said minimum throughput are normalized with respect to a

data rate for a selected channel coding scheme.

13. The system according to claim 12, wherein said data rate is a maximum

data rate for said selected channel coding scheme.

14. The system according to claim 13, wherein said normalized minimum

throughput is a ratio of a minimum data rate to a corresponding said maximum

data rate for a selected coding scheme.

15. The system according to claim 14, wherein said maximum data rate is an

selected coding scheme.

16. The system according to claim 15, wherein said normalized maximum

throughput value is a predefined parameter.

17. The system according to claim 15, wherein said minimum data rate

corresponding to said maximum data rate for each of said plurality of channel

coding scheme is determined by assigning to said minimum data rate, a

maximum data rate from an immediately previous lower order channel coding

scheme.

18. The system according to claim 17, wherein dynamic changes to said

channel coding scheme comprises one of handover and release of resources if

said measured data throughput is less than said minimum throughput

corresponding to a lowest order channel coding scheme of said plurality of

channel coding schemes.