WO2004021635A1

WO2004021635A1 - Variable puncturing for arq systems

Info

Publication number: WO2004021635A1
Application number: PCT/EP2002/009553
Authority: WO
Inventors: Benoist Sebire; Guillaume Sebire; Tommy Bysted Kristensen
Original assignee: Nokia Corporation
Priority date: 2002-08-27
Filing date: 2002-08-27
Publication date: 2004-03-11
Also published as: AU2002333724A1

Abstract

A physical layer (403) of a transmitter device, for example a mobile telephone or a base station of a mobile telecommunications network, allows for the transmission of multiple transport channels (405, 406, 407) simultaneously. Each channel is individually configurable as to CRC attachment (405a), channel coding (405b) and rate matching (405c). Blocks are punctured if required by the rate matching process (405c) before transmission. If a block is not received correctly, an algorithm in the rate matching process (405c) prepares a new data block for transmission using a different puncturing pattern. Each retransmission involves a different puncturing pattern until all possible puncturing patterns have been used, when the first used pattern is again used. This incremental redundancy results in improved reliability of data reception. A receiver is also disclosed, the receiver being arranged to read a transport format combination indicator forming part of the received block, from which the receiver can calculate an input to the rate matching algorithm in the transmitter, and from that calculate the puncturing pattern used and thus decode the data block.

Description

Title: Radio Devices

Field of the Invention

The present invention relates to a radio transmitting device, to a radio receiving device and to a method of generating a modulating signal. The invention relates also to a mobile device, and to a base transceiver station.

Background to the Invention

In GERAN (GSM/EDGE Radio Area Network) Iu mode at present, the MAC (medium access control) layer is responsible for the mapping between the logical channels (traffic or control channels) and the basic physical subchannels (Dedicated Basic Physical SubCHannel or Shared Basic Physical SubCHannel). The logical channels are the channels the physical layer offers to the MAC layer. These logical channels and the mapping to the basic physical subchannels are fully specified in GSM/EDGE standards, allowing the functionality in the MAC layer to be relatively simple.

A different approach is taken in UTRAN (UTMS Terrestrial Radio Access Network) where, instead of providing logical channels, the physical layer offers Transport Channels (TrCH), which can be used by the MAC layer. A transport channel can be used to transmit one flow over the air interface. A number of transport channels can be active at the same time and are multiplexed at the physical layer. The transport channels are configured at call set-up by the network.

The concept of transport channels is proposed to be used in GERAN. Each of these transport channels can carry one flow having a certain Quality of Service (QoS). A number of transport channels can be multiplexed and sent on the same dedicated physical subchannel thereby making it possible to have different protection on different classes of bits, for instance. The configuration used on a transport channel i.e. the number of bits, coding, interleaving etc. is denoted the Transport Format (TF). As in UTRAN, a number of transport formats can be associated with one transport channel. For instance, in adaptive multirate encoding (AMR), the class l bits have their own TrCH, with one transport format configured per AMR mode. The configuration of the transport formats can be controlled by the network and signalled to the mobile on call set-up. In both the mobile and the BTS, the transport formats can be used to configure the encoder and decoder units. When configuring a transport format, the network can choose between a number of predefined CRC (cyclic redundancy check) lengths and code types. For each of the transport channels, a given number of transport formats can be configured on call set- up.

Transport blocks (TB) are proposed to be exchanged between the MAC layer and the physical layer on a transport time interval (TTi) basis (e.g. 20ms). For each transport block a transport format is chosen and indicated through the transport format indicator (TFI). In other words, the TFI tells which channel coding to use for that particular transport block on that particular TrCH during the TTI.

Only some combinations of the transport formats of the different TrCH are allowed. A valid combination is called a Transport Format Combination (TFC). When transport formats are combined in a TFC the sum of the output bits adds up to the total number of available bits in a radio block on the basic physical sub-channel e.g. 464 bits for Gaussian minimum shift keying (GMSK) full rate channels. The set of valid TFCs on a physical subchannel is called the Transport Format Combination Set (TFCS).

In order to decode a received sequence, the receiver needs to know the active TFC for a radio block. This information is transmitted in the Transport Format Combination Indicator (TFCI) field. This field is a layer 1 header, and has the same function as the stealing bits commonly used at present. Each of the TFC within a TFCS is assigned a unique TFCI value, which is the first thing to be decoded by the receiver when a radio block is received. From the decoded TFCI value, the transport formats for the different transport channels can be found, allowing decoding to start.

Figure 1 shows the proposed architecture for a GERAN flexible layer one. Although it is inspired by the architecture that was standardised for the UL in UTRAN, it is significantly more simple.

Referring to Figure 1, a physical layer includes the following processes in sequence in respect of each TrCH provided by a layer two above: CRC attachment, channel coding, radio segment equalisation, first interleaving, segmentation, rate matching, transport channel multiplexing, TFCI mapping and second interleaving. In the CRC attachment step, error detection is provided on each transport block through a CRC. The size of the CRC to be used is fixed on each TrCH and is configured by a radio resource layer (RRC) higher than the layer one, and is a semi-static attribute of the transport format. The entire transport block is used to calculate the parity bits. Code blocks are output from the CRC attachment process.

Code blocks are then processed by the channel coding process, producing encoded blocks. The channel coding to be used is chosen by the RRC and can only be changed through higher layer signalling. The channel coding used is a semi-static attribute of the transport format, although in practise it will probably be fixed for each TrCH. Thus, for AMR, the same channel coding is used for all the modes, and rate matching simply adjusts the code rate by puncturing or repetition. In the radio segment equalisation step, radio segment size equalisation adjusts (by padding) the input bit sequence to ensure that the encoded block can be segmented into S_; data segments of same size. The first interleaver is a simple block i terleaver with inter-column permutation. Its task is to ensure that no consecutive coded bits are transmitted in the same radio block.

When the TTI is longer than the radio block duration, the input bit sequence is segmented by the segmentation process, and each S_; radio segment is mapped onto one radio block (S_; = Transmission time / radio block duration). As a result, the input bit sequence is mapped onto S; consecutive radio blocks.

The three last described processes (equalisation, first interleaving and segmentation) are only used when the TTI is longer than the radio block duration, and are transparent otherwise. For each encoded block, they produce S_; radio segments.

The rate matching process is the core of the flexible layer one. It causes bits of a radio segment on a transport channel to be repeated or punctured. Layers above the layer one assign a rate matching attribute for each transport channel. This attribute is semi-static and can only be changed through higher layer signalling. Once the number of bits to be repeated or removed is calculated, rate-matching attribute can begin. The higher the value of the attribute, the more important the bits (more repetition / less puncturing). Since the block size is a dynamic attribute, the number of bits on a transport channel can vary between different transmission times. When this happens, bits are repeated or punctured to ensure that the total bit rate after TrCH multiplexing is identical to the total channel bit rate of the allocated dedicated physical channels. Data output from the rate matching process is termed a radio frame. For every radio block to be transmitted, the rate matching produces one radio frame per radio segment, e.g. per TrCH.

In the TrCH multiplexing step, one radio frame from each TrCH is delivered to the TrCH multiplexing, for every radio block to be transmitted, according to the TFC. These radio frames are serially multiplexed into a coded composite transport channel (CCTrCH). For every radio block to be transmitted, the coded TFCI is attached at the beginning of the CCTrCH by the TFCI mapping process before interleaving. The coded TFCI and the CCTrCH are interleaved together by the second interleaving step on radio blocks. The interleaving can be either diagonal or block rectangular, and is configured on call set-up.

The input bit sequences before rate matching (radio segments) are denoted by

^U hi,l > h^υi,2 ' h^υi,3 >-~^υ hi,N, where i is the TrCH number and N_; ' is the number of bits. Only ^} one radio segment per TrCH is delivered to the rate matching block. For each radio block using transport format combination/^ the number of bits to be repeated or punctured ΔN_;j within one radio segment for each TrCH i is calculated with the following equations:

Z_0J = 0

∑RM_m x N_m x N. data

\m=l ^zu = for all i — 1 .. . I

ΫR xN m,j m=\

ΔN. . = Z. . - Z. , . - N; for all i = 1 ... I

For the calculation of the rate matching pattern of each TrCH i, the following relations are defined: e,„: - 1

^e minus *^{• X} ^u

where: [ xj Round x towards -co, i.e. integer such that x — 1 < |_JCJ < x .

\x\ Absolute value of x. I Number of TrCHs in the CCTrCH

N_data Total number of bits that are available in a radio block for the CCTrCH. N_t Number of bits in a radio segment before rate matching on TrCH i with transport format combination/.

AN,- • If positive, then is number of bits that have to be repeated in a radio segment on TrCH / with transport format combination ^" in order to produce a radio frame; if negative, then is number of bits that have to be punctured in a radio segment on TrCH / with transport format combination/ in order to produce a radio frame; if null, then no puncturing or repeating is required, i.e. the rate matching is transparent and the content of the radio frame is identical to the content of the radio segment on TrCH with transport format combinationy.

RM_t Semi-static rate matching attribute for transport channel i. e_lm- Initial value of variable e. e_pks Increment of variable e. ^e _mmm Decrement value of variable e.

Z₍. . Intermediate calculation variable.

This algorithm could be expressed as follows: if ΔV,. . < 0 — puncturing is to be performed e - e_jui — initial error between current and desired puncturing ratio m = 1 — index of current bit do while m ≤ N_; . — for each bit of the radio segment of TrCH/ e = e - e_minus - update error if e < 0 then — check if bit number m should be punctured puncture bit b_im — bit is punctured e — e + e_pllls — update error end if m — m + 1 — next bit end do else if AN,- • > 0 — repetition is to be performed e — e_ini — initial error between current and desired puncturing ratio m — \ — index of current bit do while m ≤ N_t • — for each bit of the radio segment of TrCH/ ^{e e} ^minus — update error do while e < 0 — check if bit number m should be repeated repeat bit b_{; m} - repeat bit e - e + e_plm — update error end do m — m + 1 — next bit end do else - ΔN. . = 0 do nothing — no repetition or puncturing end if.

For each TrCHi, the bit sequences output from the rate matching are denoted fi _{\ >}fi _{2 >}fi _{3 >} - _>fi v _> wh^el:e is the TrCH number and V,- is the number of bits in the radio frame of TrCH i (V, = N_t + ΔN,. _;).

Summary of the Invention

According to the present invention, there is provided a radio transmitting device comprising radio transmitter circuitry and processing means for processing digital signals to produce a modulating signal for supply to the radio transmitter circuitry, wherein the processing means is configured to implement a protocol stack including a flexible layer one with incremental redundancy. Also according to the present invention, there is provided a method of generating a modulating signal for supply to radio transmitter circuitry, the method comprising configuring a processor to implement a protocol stack including a flexible layer one with incremental redundancy.

According to a third aspect of the present invention, there is provided a radio receiver device comprising radio receiver circuitry and processing means for processing demodulated signals provided by the radio receiver circuitry, characterised in that the processing means is arranged to read a format indicator from a received block, to calculate from the indicator a value used to puncture the block of the transmitter side, and to decode the block accordingly.

The term 'flexible layer one' will be understood to mean a physical layer which can support plural active independently-configurable transport channels simultaneously.

Brief Description of the Drawings

Figure 1 shows a physical layer or flexible layer one architecture proposed for use in

GERAN.

Figure 2 shows a mobile communication system incorporating components according to the present invention;

Figure 3 is a block diagram of a mobile station of the Figure 1 system;

Figure 4 is a block diagram of a base transceiver station of the Figure 1 system;

Figure 5 illustrates the lower levels of a protocol stack used in an embodiment of the present invention; Figure 6 illustrates the generation of a radio signal by the present invention; and

Figure 7 illustrates the puncturing of an uncoded block of data.

Detailed Description of the Preferred Embodiment

Preferred embodiments of the present invention will now be described, by way of example only, with reference to the accompanying drawings.

Referring to Figure 2, a mobile phone network 1 comprises a plurality of switching centres including first and second switching centres 2a, 2b. The first switching centre 2a is connected to a plurality of base station controllers including first and second base station controllers 3a, 3b. The second switching centre 2b is similarly connected to a plurality of base station controllers (not shown).

The first base station controller 3a is connected to and controls a base transceiver station 4 and a plurality of other base transceiver stations. The second base station controller 3b is similarly connected to and controls a plurality of base transceiver stations (not shown).

In the present example, each base transceiver station services a respective cell. Thus, the base transceiver station 4 services a cell 5. Alternatively, a plurality of cells could be serviced by one base transceiver station by means of directional antennas. A plurality of mobile stations 6a, 6b are located in the cell 5. The number and identities of mobile stations in any given cell varies with time.

The mobile phone network 1 is connected to a public switched telephone network 7 by a gateway switching centre 8.

A packet service aspect of the network includes a plurality of packet service support nodes (one shown) 9 which are connected to respective pluralities of base station controllers 3a, 3b. At least one packet service support gateway node 10 connects the or each packet service support node 10 to the Internet 11.

The switching centres 3a, 3b and the packet service support nodes 9 have access to a home location register 12.

Communication between the mobile stations 6a, 6b and the base transceiver station 4 employs a time-division multiple access (TDMA) scheme.

Referring to Figure 3, the first mobile station 6a comprises an antenna 101, an rf subsystem 102, a baseband DSP (digital signal processing) subsystem 103, an analogue audio subsystem 104, a loudspeaker 105, a microphone 106, a controller 107, a liquid crystal display 108, a keypad 109, memory 110, a battery 111 and a power supply circuit 112.

The rf subsystem 102 contains if and rf circuits of the mobile telephone's transmitter and receiver and a frequency synthesizer for tuning the mobile station's transmitter and receiver. The antenna 101 is coupled to the rf subsystem 102 for the reception and transmission of radio waves.

The baseband DSP subsystem 103 is coupled to the rf subsystem 102 for receiving baseband signals therefrom and for sending baseband modulation signals thereto. , The baseband DSP subsystems 103 includes codec functions which are well-known in the art.

The analogue audio subsystem 104 is coupled to the baseband DSP subsystem 103 and receives demodulated audio therefrom. The analogue audio subsystem 104 amplifies the demodulated audio and applies it to the loudspeaker 105. Acoustic signals, detected by the microphone 106, are pre-amplified by the analogue audio subsystem 104 and sent to the baseband DSP subsystem 4 for coding.

The controller 107 controls the operation of the mobile telephone. It is coupled to the rf subsystem 102 for supplying tuning instructions to the frequency synthesizer and to the baseband DSP subsystem 103 for supplying control data and management data for transmission. The controller 107 operates according to a program stored in the memory 110. The memory 110 is shown separately from the controller 107. However, it may be integrated with the controller 107.

The display device 108 is connected to the controller 107 for receiving control data and the keypad 109 is connected to the controller 107 for supplying user input data signals thereto.

The battery 111 is connected to the power supply circuit 112 which provides regulated power at the various voltages used by the components of the mobile telephone. The controller 107 is programmed to control the mobile station for speech and data communication and with application programs, e.g. a WAP browser, which make use of the mobile station's data communication capabilities.

The second mobile station 6b is similarly configured.

Referring to Figure 4, greatly simplified, the base transceiver station 4 comprises an antenna 201, an rf subsystem 202, a baseband DSP (digital signal processing) subsystem 203, a base station controller interface 204 and a controller 207.

The rf subsystem 202 contains the if and rf circuits of the base transceiver station's transmitter and receiver and a frequency synthesizer for tuning the base transceiver station's transmitter and receiver. The antenna 201 is coupled to the rf subsystem 202 for the reception and transmission of radio waves.

The baseband DSP subsystem 203 is coupled to the rf subsystem 202 to receive baseband signals therefrom and for sending baseband modulation signals thereto. The baseband DSP subsystems 203 includes codec functions which are well-known in the art.

The base station controller interface 204 interfaces the base transceiver station 4 to its controlling base station controller 3a.

The controller 207 controls the operation of the base transceiver station 4. It is coupled to the rf subsystem 202 for supplying tuning instructions to the frequency synthesizer and to the baseband DSP subsystem for supplying control data and management data for transmission. The controller 207 operates according to a program stored in the memory 210.

When used for circuit-switched speech traffic, the channelisation scheme is as employed in GSM. The baseband DSP subsystems 103, 203 and controllers 107, 207 of the mobile stations 6a, 6b and the base transceiver stations 4 are configured to implement two protocol stacks. The first protocol stack is for circuit switched traffic and is substantially the same as employed in conventional GSM systems. The second protocol stack is for packet switched traffic.

Referring to Figure 5, the layers relevant to the radio link between a mobile station

6a, 6b and a base station controller 4 are the radio link control (RLC) layer 401, the medium access control (MAC) layer 402 and the physical layer or flexible layer one 403. Other layers exist above the shown layers, but these are not shown for clarity.

The radio link control layer 401 has two modes: transparent and non-transparent. In transparent mode, data is merely passed up or down through the radio link control layer without modification.

In non-transparent mode, the radio link control layer 401 provides link adaptation and constructs data blocks from data units received from higher levels by segmenting or concatenating the data units as necessary and performs the reciprocal process for data being passed up the stack. It is also responsible for detecting lost data blocks or reordering data block for upward transfer of their contents, depending on whether acknowledged mode is being used. This layer may also provide backward error correction in acknowledged mode.

The medium access control layer 402 is responsible for allocating data blocks from the radio link control layer 401 to appropriate transport channels and passing received radio blocks from transport channels to the radio link control layer 401.

The physical layer 403 is responsible for creating transmitted radio signals from the data passing through the transport channels, and for passing received data up through the correct transport channel to the medium access control layer 402. The physical layer 403 includes the architecture shown in Figure 1. Referring to Figure 6, data produced by applications 404a, 404b, 404c propagates down the protocol stack to the physical layer 403a, 403b. The physical layer 403a, 403b carries data from the applications 404a, 404b, 404c on different transport channels 405, 406, 407 according to the class to which the data belongs. Each transport channel 405, 406, 407 can be configured to process signals according to a plurality of processing schemes 405a, 405b, 405c, 406a, 406b, 406c, 407a, 407b, 407c. The configuration of the transport channels 405, 406, 407 is established during call set-up on the basis of the capabilities of the mobile station 6a, 6b and the network and the nature of the application or applications 404a, 404b, 404c being tun.

The processing schemes 405a, 405b, 405c, 406a, 406b, 406c, 407a, 407b, 407c are unique combinations of cyclic redundancy check 405a, 406a, 407a , channel coding 405b, 406b, 407b and rate matching 405c, 406c, 407c. These unique processing schemes are the TFCs referred to above. The other processing steps shown in the physical layer of Figure 1 are omitted from Figure 6 for clarity. Steps 405d, 406d and 407d are optional interleaving steps, which are omitted from Figure 1.

The combined data rate produced for the transport channels 405, 406, 407 must not exceed that of physical channel or channels allocated to the mobile station 6a, 6b. This places a hmit on the transport format combinations that can be permitted. For instance, if there are three transport formats TFI, TF2, TF3 for each transport channel, the following combinations might be valid and thus constitute TFCIs:-

TF1 TFI TF2 TFI TF3 TF3 but the following combinations might not be valid and thus not constitute TFCIs:-

TF1 TF2 TF2

TFI TFI TF3

The data output by the transport channel interleaving processes are multiplexed by a multiplexing process 410 and then subject to further interleaving 411. A TFCI is generated by a TFCI generating process 412 from information from the medium access control layer and coded by a coding process 413. The TFCI is inserted into the data stream by a TFCI insertion process after the further interleaving 411. The TFCI is spread across one radio block with portions placed in fixed positions in each burst. For example, on either side of the training symbols. The complete TFCI therefore occurs at fixed intervals, i.e. the block length 20ms. This makes it possible to ensure TFCI detection when different interleaving types are used e.g. 8 burst diagonal and 4 burst rectangular interleaving. Since the TFCI is not subject to variable interleaving, it can be readily located by the receiving station and used to control processing of the received data.

The physical layer 403 is provided with a facility for incremental redundancy (IR), also known as Hybrid Type 2 ARQ. When a transmitter determines that a block has not been received correctly by the receiver to which is was sent, the data is retransmitted with a different puncturing pattern. This feature increases the probability of successful decoding at the receiver side, since the prevailing unsuccessful (re)transmissions of a given block are re-used and combined in the receiver with the most recent retransmission of this block. The incremental redundancy process is illustrated schematically in Figure 6.

Referring to Figure 6, an uncoded block having M bits is encoded with a mother code at a rate of 1/k to provide an encoded block having M x k bits. A first puncturing step provides a first encoded block having Nl bits. A second puncturing step provides a second encoded block having N2 bits. Further puncturing steps provide further encoded blocks. Each of the steps utilises a different puncturing pattern.

The channel encoder forming part of the physical layer 402 on the transmitter side can generate different puncturing patterns achieving the same code rate M/N, based on the same mother code having a rate 1 / k, with:

N. k where: k is a positive integer, M is the length in bits of the uncoded block to encode and Ν_; is the length in bits of the i' encoded block The channel decoder on the receiver side needs to be able to cope with any of the possible puncturing patterns the channel encoder may generate for the same mother code, so that it can decode any encoded block. In addition, since received blocks are combined before decoding, it must be able to decode the combination of all the encoded blocks received for a given block.

Using these principles, incremental redundancy allows the transmission of differently encoded versions of the same block. If the initial transmission (which is of encoded block

1) fails, a retransmission is carried out with encoded block 2. If the block still cannot be correctly decoded even using the first and second encoded blocks together, a third encoded block is transmitted. The receiver combines the (re)transmissions for correcting errors.

The global coding rate (covering the initial transmission and subsequent retransmissions) after combining is adjusted incrementally after each retransmission. After n retransmissions, this global coding rate is:

M global _ coding _ rate = T N,. - amount of overlapping bits

where Ν₀ is the number of bits of the encoded block of the initial transmission, N_j the amount of bits of the encoded block of the 1^st retransmission and so on.

Soft combining of the overlapping bits can improve the decoding performance resulting in the coding rate:

global _ coding _ rate «

It will be appreciated by those skilled in the art how to perform soft decoding at the receiver side.

The rate matchers 405c, 406c, 407c of the physical layer 403 each include an algorithm similar to that described above. However, the algorithms include an additional input to allow for the generation of different puncturing patterns. The other inputs to this algorithm are unchanged. The additional input is termed retrans. Retrans is controlled by the RLC layer 401 and its value is notified to the physical layer 403 for operating IR (incremental redundancy). The value of retrans follows the rules :in unacknowledged and transparent RLC modes, retrans has a fixed value of zero, i.e. IR is not used. In acknowledged RLC mode when IR is used, the value of retrans is incremented by 1 after every transmission or retransmission of the same data block. The parameter retrans can be seen as a counter of the number of retransmissions of the same data block. Retrans equals 0 at the initial transmission, 1 at the first retransmission, and so on until it is ή at the n* retransmission. The physical layer 403 uses the parameter retrans to generate different puncturing patterns, if different patterns are possible.

The number of possible different puncturing patterns depends on factors that are specific to the physical layer 403, such as the mother code, the size of the transport block and the activity of other transport channels. Thus the physical layer 403 may not always be able to generate a different puncturing pattern for each value of retrans. When different patterns are available, the physical layer cycles through them in turn in a loop fashion. For instance, assuming that 3 puncturing patterns are possible (pO, pi and p2), pO is used for the 1^st transmission (retrans^O), pi for the 1^st retransmission (retrans=l), p2 for the 2^nd retransmission (retrans=2), and pO is used again for the 3^ld one (retrans=3), pi for the 4^th one (retrans=4) and so on. Thus, the RLC layer 401 decides how many retransmissions are needed and, for each of them, the physical layer 403 attempts to generate a different puncturing pattern.

For the same transport block, each value of retrans is associated with a different Transport Format i.e. it is a dynamic parameter. Two retransmissions of the same block use two different TFCIs. Thus, by decoding the TFCI, the receiver knows the value of retrans and therefore the puncturing to use.

According to the invention, to support IR when bits are punctured (i.e. when ΔN,- • < 0), the initial value of the error (e_; is modified according to the following:

'plus N;

'plus ^■ e rai ■n us ^Nu - ΔN ',< plus e_ini = 1 + (retrans mod d ) x -1

^~d^~X. The remainder of the algorithm given above is unchanged.

Here, dis the average distance between transmitted bits. For instance, if 2 bits are regularly punctured every 3, the average distance between transmitted bits is 3. d also reveals how many transmissions are possible with different puncturing patterns. For instance if the distance is 3, we obviously have 3 different possible puncturing patterns: one starting from bit 1, another one starting from bit 2 and one starting from bit 3. When the value of retrans is equal to 0, the algorithm is equivalent to the prior art algorithm described above.

'Retrans' gives the transmission number, which is '0' for a first transmission of a block, "V for a first retransmission and so on. The value of 'retrans' is multiplied by the modulus of the value 'd'.

The advantages of using "retrans mod d' instead of "retrans' only are twofold Firstly, the rate matching algorithm is protected from any misuse of retrans. For example, if d was equal to 3, that would mean that a maximum value of 2 is possible for retrans. Any value beyond 2 could not be accepted using just 'mod d without adverse effects. Using "retrans mod d any value can be accepted. Also, there is no need for higher layers (e.g. the RLC layer 401) to know how many retransmissions are possible. As a result, the higher layers must store fewer details that are specific to the physical layer 403. For instance, because of the flexibility allowed by the flexible layer one 403, some other transport channels can become active even when retransmissions are ongoing, changing datid therefore the management of retrans. If higher layers needed to be aware of and manage this kind of detail, it would greatly increase their complexity, and possibly lead to that kind of flexibility being impermissible. In that case, without "retrans mod d', the flexible layer one 403 could not accept other transport channels for as long as retransmissions are being sent for one transport channel. In a second embodiment, the number of bits to be repeated or punctured ΔN_{; j} within one radio segment for each TrCH / is calculated with the following equations:

^Zo = 0

ΔN,, : - Z._j - Z^_j -N^ for all / = 1

The rate matching rule is as follows:

if ΔN_; ,. < 0 puncturing is to be performed for p=l to ΔN,. J

puncture bit b_im where m (p — ϊ) x

end for else if ΔN. , > 0 repetition is to be performed fo p=l to ΔN,.,

repeat bit b_im where m =

end for else - ΔN„ = 0 do nothing no repetition nor puncturing end if.

In the above,

For each TrCHi, the bit sequences output from the rate matching are denoted fi _{\ >}fi _{2 >} fi - _>^* >fi r _> ^wnere i is the TrCH number and V_t is the number of bits in the radio frame of TrCH / (V, = N„ + ΔN, ).

The same advantages arise from using 'retrans mod d as is described above in relation to the first embodiment.

Results obtained using the first described algorithm are shown below. However, the punctured bits are exacdy the same whichever of the above embodied algorithms is used. In the example, only one transport channel is active (I — 1), and the size of the CCTrCH is limited to 10 bits (Ν_ώtø = 10) always. It will be noted that the value of the rate matching attribute is not important in the examples since only one transport channel is active. Bits are numbered starting from 0.

First example N, = 30 ΔN_; '•J -20

40

d = ?>

Second example N_; — 50 ΔN,, = -40

*,„,-„„, = 80

^ = ¹⁰⁰ d- 5

The above described algorithms are implemented in the physical layers of transmitters in both the mobile station 6a and the base transceiver station 4.

Each of the mobile station 6a and the base transceiver station 4 also include a respective receiver device, which is arranged to receive signals from the transmitter device of the other station. In a conventional IR receiver, the puncturing patterns are fixed, so the receiver knows what puncturing pattern will be used with a block even before that block is transmitted. In this invention, however, the puncturing pattern cannot be predicted by the receiver since it is susceptible to influence by a number of factors. Instead, each receiver is arranged, on receiving a block, to decode the TFCI, to calculate from the TFCI what value of retrans was used by the transmitter, and to calculate from this what puncturing was used with the block. The block receiver is then able to decode the received block.

Claims

1. A radio transmitting device comprising radio transmitter circuitry and processing means for processing digital signals to produce a modulating signal for supply to the radio transmitter circuitry, wherein the processing means is configured to implement a protocol stack including a flexible layer one with incremental redundancy.

2. A radio transmitting device as claimed in claim 1, in which an algorithm which the processing means runs to form part of a rate matching process of the flexible layer one is arranged to produce a different puncturing pattern on each retransmission of a given block of data.

3. A radio transmitting device as claimed in claim 2, in which the algorithm includes an input for controlling the puncturing pattern used with a transmitted block of data, a signal supplied to the input being changed for each retransmission of the block.

4. A radio transmitting device as claimed in claim 3, in which the input signal is changed by a predetermined amount, preferably incremented, for each retransmission of the block of data.

5. A radio transmitting device as claimed in any of claims 2 to 4, in which the algorithm includes instructions for multiplying the input signal by the modulus of a variable

6. A radio transmitting device as claimed in any of claims 2 to 5, in which the different puncturing patterns are used on a cyclic basis if the number of retransmissions is such that the number of possible puncturing patterns is insufficient.

7. A method of generating a modulating signal for supply to radio transmitter circuitry, the method comprising configuring a processor to implement a protocol stack including a flexible layer one with incremental redundancy.

8. A radio receiver device comprising radio receiver circuitry and processing means for processing demodulated signals provided by the radio receiver circuitry, characterised in that the processing means is arranged to read a format indicator from a received block, to calculate from the indicator a value used to puncture the block of the transmitter side, and to decode the block accordingly.

9. A mobile telephone including a radio transmitting device as claimed in any of claims 1 to 6 and/or a radio receiving device as claimed in claim 8.

10. A base transceiver station including a radio transmitting device as claimed in any of claims 1 to 6 and/or a radio receiving device as claimed in claim 8.