Method, data transmitter and computer program for adapting BICM schemes to amplitude shaping combined with puncturing

Abstract

The present invention relates to a method for transmitting symbols, wherein a first signal is obtained, the first signal comprising bit sequences representing a first data flow and a second data flow. The first data flow comprises shaped bits. The second data flow comprises unshaped bits, including quantification bits and / or sign bits. At least some of the unshaped bits of the second data flow are punctured, thereby obtaining an incomplete bit stream. Symbols are obtained based on at least the incomplete bit stream and the first data flow. The present invention further relates to a corresponding data transmitter and a corresponding computer program.

Description

[Technical Field]

[0001] The present disclosure relates to the field of telecommunications, and more particularly to a method for transmitting symbols, a data transmitter, and a computer program. [Background technology]

[0002] For a particular communication channel in a communication system, the channel capacity characterizes the highest information rate that can be achieved at a fixed average transmit power while maintaining a small error probability.

[0003] Generally, the channel input transmitted by the transmitter is a finite set of symbols. This finite set is called a constellation. The constellation can be, for example, a one-dimensional amplitude-shift keying (ASK) constellation or a two-dimensional quadrature-amplitude modulation (QAM) constellation. Two-dimensional quadrature amplitude modulation can be thought of as the Cartesian product of two ASK constellations.

[0004] In many cases, the channel capacity cannot be reached if each symbol in the constellation is transmitted with equal probability. As a result, the transmitter must process the data so that each symbol in the constellation is transmitted with a probability that approaches capacity. This operation is called shaping.

[0005] In addition to shaping, messages also need to be protected using error-correcting codes. The 5G new radio (NR) standard specifies a coding scheme called bit-interleaved coded modulation (BICM), which does not perform shaping.

[0006] Combining shaping and coding is not trivial and requires specific algorithms, especially if 5G NR coding schemes are to be adhered to.

[0007] Therefore, there is a need for a coding scheme that is suitable as a BICM scheme, preferably as a 5G NR coding scheme, and that can provide as output symbols that comply with the shaping performed by a distribution matcher (DM). Summary of the Invention [Problem to be solved by the invention]

[0008] The present disclosure improves this situation. [Means for solving the problem]

[0009] 1. A method in a data transmitter for transmitting k″ symbols, where k″ is a positive integer, the method comprising: obtaining an input bit sequence representing a first data flow and a second data flow; the first data flow includes k' x (mq) shaped bits representing at least mq input labeling bit streams, where k' is a positive integer greater than or equal to k", m is the number of bit streams output by the demultiplexer, and q is a positive integer; obtaining a second data flow including c' unshaped bits, c' being a positive integer less than or equal to (q-1)k', and the c' unshaped bits including at most q-1 quantification bits and / or sign bits; puncturing at least a portion of the unshaped bits of the second data flow, thereby obtaining at most q-1 incomplete bit streams; Obtaining k″ symbols based at least on at most q−1 incomplete bit streams and mq input labeling bit streams of the first data flow; A method is proposed, including:

[0010] In other words, a technique is proposed in which non-shaped bits are punctured, whereby when an input bit sequence contains c' non-shaped bits, c of the c' non-shaped bits can be punctured after encoding.

[0011] The proposed technique can be implemented as an improvement to existing 5G NR coding schemes, or more generally, any BICM scheme, with minor modifications to the specifications, so that the symbols at the output of the coding comply with the shaping performed by the distribution matcher. In particular, the proposed technique allows puncturing some systematic bits without changing the target shaping distribution, for example, during rate matching (RM) and hybrid automatic repeat request (HARQ).

[0012] Optionally, the method further includes ordering the input bit sequence by placing c of the c' unshaped bits at systematic puncturing positions, where c is a positive integer less than or equal to c', and puncturing at least some of the unshaped bits of the second data flow includes puncturing at least c unshaped bits that are at the systematic puncturing positions.

[0013] Alternatively, the c unshaped bits may be received at default positions in the input bit sequence, and the systematic puncturing positions may be chosen to correspond to these default positions. This eliminates the need to order the input bit sequence before puncturing. In other words, the method may further include designating positions of at least c of the c' unshaped bits in the input bit sequence as systematic puncturing positions, and puncturing at least some of the unshaped bits of the second data flow includes puncturing at least c unshaped bits that are at the systematic puncturing positions.

[0014] Optionally, the method further comprises: obtaining an output of an encoder that receives as input bits associated with an input bit sequence using a systematic error correction coding scheme, the encoder output comprising qk'-d parity bits, where d=c'-c, and d is a positive integer or 0; completing at most q-1 incomplete bit streams using (qk'-d)-k' of the qk'-d parity bits to obtain at most q-1 complete bit streams; and Obtaining k symbols is mq input labeling bit streams of a first data flow; at most q-1 completed bitstreams; parity bits output by the encoder that were not used to complete at most q-1 incomplete bit streams; can be based on

[0015] It is not necessary for the data transmitter itself to obtain the encoder output, complete the incomplete bit stream, and / or identify the k" symbols. According to an alternative possible form, the method may include transmitting the q-1 incomplete bit streams to an entity other than the data transmitter. Such another entity may be further adapted to determine k" symbols based on the q-1 incomplete bit streams and to transmit the determined k" symbols back to the data transmitter. Having thus obtained the k" symbols from the other entity, the data transmitter may then transmit those symbols over an appropriate channel.

[0016] Optionally, c'=k'.

[0017] Optionally, c=c'.

[0018] Optionally, obtaining k″ symbols based on m original labeling bit streams includes: ordering an output bit sequence comprising m original labeling bit streams by placing x bits into additional puncturing positions, where x is a positive integer and the x bits comprise x / m bits of each of the m original labeling bit streams; puncturing x bits located at additional puncturing positions to obtain m punctured labeling bit streams; Obtaining k″ symbols by labeling a corresponding input symbol using one bit of each punctured labeling bit stream; Includes:

[0019] In other words, a technique is further proposed in which shaped bits are punctured. An input bit sequence can be provided to an encoder, and then the input bit sequence can be concatenated with the encoder output to obtain an output bit sequence. Reordering this output bit sequence involves using a second interleaver to place all punctured bits in additional puncturing positions, for example, at the end of the output bit sequence. In a first example, the input bit sequence includes only shaped bits. In this case, a "first interleaver" is not required before encoding, and the punctured bits are x bits, including x / m bits of each of the m original labeling bit streams. In a second example, the input bit sequence includes both shaped bits and non-shaped bits. In this case, only shaped bits, only non-shaped bits, or both shaped bits and non-shaped bits can be punctured. In the case of puncturing both, both techniques can be combined, and the punctured bits can therefore include not only x bits comprising the x / m bits of each of the m original labeling bit streams, but also c of the c' unshaped bits.

[0020] A combined technique in which both shaped and unshaped bits are punctured allows for rate

number

number

[0021] Optionally, the additional puncturing positions correspond to the end of the output bit sequence after ordering the output bit sequence.

[0022] A rate matcher (similar to that in 5G NR LDPC coding schemes) that punctures the last coded bits can then discard the punctured bits.

[0023] In another example, obtaining k symbols is using one bit of each original labeling bit stream to label a corresponding input symbol, thereby obtaining k″ symbols and an additional symbol; discarding additional symbols using a rate matcher; Includes:

[0024] In that case, compared to the previous example, no interleaver is needed before the rate matcher.

[0025] Optionally, the method further includes, after transmitting the k″ symbols, separately transmitting a symbol having the c punctured unshaped bits and the x punctured bits as labeling bits.

[0026] 1. A data transmitter configured to transmit k″ symbols, where k″ is a positive integer, the transmission comprising at least obtaining an input bit sequence representing a first data flow and a second data flow; the first data flow includes k' x (mq) shaped bits representing at least mq input labeling bit streams, where k' is a positive integer greater than or equal to k", m is the number of bit streams output by the demultiplexer, and q is a positive integer; the second data flow includes c' unshaped bits, c' being a positive integer less than or equal to (q-1)k', and the c' unshaped bits including at most q-1 quantification bits and / or sign bits; puncturing at least a portion of the unshaped bits of the second data flow, thereby obtaining at most q-1 incomplete bit streams; Obtaining k″ symbols based at least on at most q−1 incomplete bit streams and mq input labeling bit streams of the first data flow; A data transmitter is further proposed, wherein the data transmitter is implemented by:

[0027] It is further proposed a computer program comprising instructions that cause a computer to carry out the above method when the program is executed by a computer.

[0028] Other features, details and advantages are set forth in the following detailed description and in the figures. [Brief explanation of the drawings]

[0029] [Figure 1] FIG. 1 illustrates a 5G NR LDPC coding scheme known in the prior art. [Figure 2] FIG. 1 illustrates the natural bit labeling of a 16-ASK constellation as known in the prior art. [Figure 3] FIG. 1 illustrates an implementation of a HARQ scheme in 5G known in the prior art. [Figure 4] FIG. 1 illustrates a stochastic amplitude shaping technique known in the prior art. [Figure 5] FIG. 1 illustrates a system in which a distributed matcher (DM) is placed in front of a 5G NR LDPC coding scheme. [Figure 6] FIG. 1 illustrates a puncturing scheme in one embodiment. [Figure 7]FIG. 1 shows an example of a quantized target shaping distribution for a 16-ASK constellation known in the prior art. [Figure 8] FIG. 1 shows two sub-constellations of a 16-ASK constellation known from the prior art, with natural labeling of the sub-constellations according to the main constellation with the less significant bit on the left (but without quantification bits). [Figure 9] FIG. 1 illustrates an adaptation of a 5G NR LDPC scheme in which systematic unshaped bits are punctured, in one embodiment. [Figure 10] FIG. 10 illustrates another adaptation of the 5G NR LDPC scheme in which systematic unshaped bits are punctured, in another embodiment. [Figure 11] FIG. 6 illustrates a system that differs from the system shown in FIG. 5 in that an interleaver is placed between the systematic parity encoder and the rate matcher, in one embodiment. [Figure 12] FIG. 12 illustrates bit ordering after the interleaver shown in FIG. 11 in one embodiment. [Figure 13] FIG. 6 illustrates a system that differs from the system shown in FIG. 5 in that the locations of the rate matchers and multiplexers have been swapped, in one embodiment. [Figure 14] FIG. 12 illustrates a system that differs from the system shown in FIG. 11 in that an additional interleaver is placed after the multiplexer, in one embodiment. [Figure 15] FIG. 13 illustrates an alternative bit ordering to that shown in FIG. 12 in one embodiment. [Figure 16] FIG. 1 illustrates an adaptation of a 5G NR LDPC scheme in which both systematically shaped bits and systematically unshaped bits are punctured, in one embodiment. [Figure 17] FIG. 17 illustrates an example of a sequence at the output of the parity coding in FIG. 16 before puncturing and interleaving, in one embodiment. [Figure 18] FIG. 17 illustrates coded bits at the output of the error correcting code of FIG. 16 in one embodiment. [Figure 19] FIG. 1 illustrates an adaptation of a HARQ scheme in which both systematically shaped bits and systematically unshaped bits are punctured in one embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0030] First, let us consider the 5G New Radio (NR) low-density parity-check (LDPC) coding scheme shown in Figure 1, as described in 3GPP™ TS38.212. An overview can also be found in [DPS2020].

[0031] The 5G NR LDPC coding scheme operates as follows: k information bits are used as input to a systematic parity encoder 100. Because the code is systematic, n+ε-k parity bits 102 are generated without changing the information bits. After encoding, ε=2Z systematic bits are punctured by a puncturer 104, where Z is a parameter that depends on the code length used. The resulting codeword of size n is called the mother codeword 106. The mother code has a rate R=k / n. A rate-matcher (RM) 108 then removes (punctures) some bits of the mother codeword so that only n′ bits are output. The effective rate of the system is R=k / n′. Downstream, the n′ bits are provided to a demultiplexer 110, which is combined with a symbol mapper to label the symbols to be transmitted over the channel.

[0032] Bits are transmitted over a channel when the modulation scheme used is binary phase shift keying (BPSK) or quadrature phase shift keying (QPSK).

number

[0033] first,

number

number

[0034] After the RM, a demultiplexer (called an interleaver in the standard) outputs m streams, one for each labeling bit of the modulation symbols (i.e., bit level, see Figure 2). The symbols at the output of the symbol mapper are then transmitted over the channel.

[0035] We define the labeling bit stream as the subset of coded bits before the demultiplexer that are used to label one bit level of a symbol. For example, B1 in Figure 1 is the labeling bit stream, and B1 = b1 1 ...b1 k’ where k' is the number of symbols transmitted on the channel. ik' shaped bits b represent a unique data flow among k input bits. i When the bitstream B contains i is said to be "original". Therefore, B1 in Figure 1 is an example of an original labeling bitstream. On the other hand, a bitstream with a size smaller than k' is not considered "original". Examples of non-original bitstreams are described later in this specification.

[0036] Next, we introduce the Hybrid Automatic Repeat Request (HARQ) scheme implemented in 5G. This scheme belongs to the category of incremental redundancy schemes. In each round, a subset of the mother codeword is selected (by the RM) and then transmitted.

[0037] Figure 3 shows how HARQ is implemented in 5G. First, the bits of a mother codeword are placed in a circular buffer 300. Then, in each of the four possible rounds (RV0, RV1, RV2, and RV3), a different subset of the coded bits is transmitted. For example, in round RV2 302, only parity bits are transmitted.

[0038] Next, we will introduce the probabilistic amplitude shaping (PAS) method disclosed in [BSS2015]. This method is a general technique that combines shaping and coding. The principle of PAS, shown in Figure 4, is as follows: A distribution matcher (DM) 400 outputs symbols according to the positive side of the target shaping distribution. Each symbol is labeled with several bits. A function b() 402 calculates the amplitude of a particular symbol x i The bits corresponding to the symbol labeling are used as input for a systematic error correcting code. These bits are called shaping bits. Block "P" 404 calculates and outputs one parity bit for each symbol. These parity bits determine the sign of the transmitted symbol. Function s() 406 calculates the parity of the bit b si These bits are called sign bits.

[0039] The key ideas underlying PAS are: - Since systematic coding is used, the distribution of the shaped bits is not changed by the error correcting code, so the shaped bits can have unequal probability. - It is widely accepted that the parity bits of error correcting codes have an equiprobability distribution, which is suitable for a symmetric shaped distribution, since the symbol has the same probability of being positive as it does of being negative, so that the sign bits remain equiprobable.

[0040] As a result, PAS successfully combines formatting and encoding.

[0041] Using this scheme, the baseline rate of the code is R = (m - 1) / m. This rate can be easily increased by using some sign bits as systematic bits. On the other hand, this rate cannot be decreased using this standard scheme.

[0042] As we will explain below, problems arise when trying to combine shaping schemes such as PAS with 5G NR LDPC coding schemes.

[0043] Regarding puncturing of shaped bits, ideally, a shaping scheme 500 is "plugged into" a 5G NR LDPC coding scheme 510, as shown in the system illustrated in Figure 5. In this example, m = 4 bits, including three shaped bits, are used per symbol. For simplicity, in this example, the puncturing of both systematic and non-systematic bits is given to the RM as a task.

[0044] However, such a combination of shaping and coding schemes cannot achieve the desired shaping distribution: in fact, if the shaping bits used as systematic bits are inadvertently punctured, the resulting symbol distribution will be significantly altered.

[0045] Regarding the puncturing of the HARQ scheme as shown in Figure 3, some rounds of the HARQ scheme do not include systematic bits (for example, RV2 in Figure 3), and therefore only include parity bits. Since the parity bits have equal probability, independent, and identical distributions, shaping cannot be performed using only these bits.

[0046] First, we clearly distinguish well-formed bits from unwell-formed bits. That is, a bit corresponding to a particular bit level (see Figure 2) is unwell-formed if it has an equiprobability distribution (taking values ​​0 and 1) and the distribution is independent of other bit levels. If a bit corresponding to a particular bit level is not unwell-formed, then it is well-formed.

[0047] We distinguish two cases for puncturing systematic bits: when the punctured systematic bits are shaped bits and when the systematic bits are non-shaped bits.

[0048] We then propose a technique that focuses on systematic non-shaped bit puncturing.

[0049] For this purpose, a quantized target shaping distribution is first evoked. Figure 7 shows an example of a quantized target shaping distribution 1200 for 16-ASK that meets the requirements set forth in European Patent Application No. 22305529.4. Simulation results show that this distribution achieves suboptimal performance, as shown in Figure 4 of [CG2022]. Therefore, this distribution can be considered as a target shaping distribution.

[0050] Figure 8 shows two subconstellations of 16-ASK. The natural labeling according to the main constellation is also shown, with the least significant bits on the left. The bits in this diagram correspond to bit levels 2, 3, and 4. The first subconstellation 1300 corresponds to bit level 1=0, and the second subconstellation 1302 corresponds to bit level 1=1.

[0051] Therefore, the constellation

number

number

number

[0052] Consider again the 16-ASK constellation as an example. The natural labeling of the symbols in this constellation is shown in Figure 2. It can be seen that the first bit level distinguishes between the two subconstellations shown in Figure 7.

[0053] As a result, using this distribution, the two bits, the shaping bit and the bit that distinguishes the two constellations, have equiprobable independent distributions. This second bit is called the quantification bit.

[0054] In this example, there is only one quantification bit. Instead of having only two adjacent symbols of equal probability, stronger quantification can be achieved by having four, eight, etc. adjacent symbols of equal probability. Let q be the number of quantification bits and sign bits. In the example above, q=2.

[0055] Furthermore, techniques have been proposed that use quantization bits and / or sign bits as punctured systematic bits (independently from other punctured systematic bits).

[0056] The additional bit level with equal probability distribution allows both the quantization bit and / or the sign bit to be used as parity bits, which allows the rate of the code to be reduced to R=(mq) / m.

[0057] Furthermore, since these bits have independent equal probability distributions, they can be placed in systematic puncturing positions and replaced by parity bits in symbol mapping. Note that there is no constraint on puncturing other bits in order to puncture these bits.

[0058] FIG. 9 shows a first example adaptation of the 5G NR LDPC scheme. In this example, quantification bits b1 are used as (unshaped) systematic bits, and a subset of c bits of these bits are placed at puncturing positions by a first interleaver π′1000. In this example, the code rate is R=((m−1)k′) / (mk′+c), and the parity encoder 100 generates k′+c bits b4. A second interleaver π1400 is added before the demultiplexer and after the RM. This second interleaver places the extra c bits b4 in the positions of the missing punctured bits b1. In FIG. 9, B2 is the original bit stream, containing k′ shaped bits b2 among the k=3k′ bits of the input bit sequence. B3 is also the original bit stream, containing k′ shaped bits b3 among the k=3k′ bits of the input bit sequence. In contrast, B1 contains only k'-c shaped bits b1 as a result of puncturing. B1 is not the original bit stream because it does not contain k' shaped bits b2 among the k=3k' bits of the input bit sequence. B4 contains only parity bits that are not part of the input bit sequence. Therefore, B4 is not the original bit stream.

[0059] Figure 10 shows a second example adaptation of the 5G NR LDPC scheme. In this example, only c unshaped systematic bits are used, where c is the number of systematic bits punctured. As a result, the code rate is reduced to R = ((m-2)k' + c) / (mk' + c). In Figure 10, B2 and B3 are original bit streams that remain unchanged throughout the illustrated workflow. In contrast, B1 is empty and therefore not an original bit stream. Furthermore, B1' and B4 contain only parity bits and are not original bit streams.

[0060] Note that c' unshaped systematic bits, where c'=c+d and d≧0, can also be used (specifically instead of k' or c as done in the above example).

[0061] Next, we propose a technique that focuses on puncturing systematic bits, including shaped systematic bits.

[0062] A note on cyclic redundancy checks (CRCs): The CRC used by the receiver to verify whether a message has been decoded correctly is calculated from the information bits before LDPC coding. The generated CRC bits are appended to the information bit sequence. The resulting sequence is used as the systematic bits of the channel code. The CRC bits can be assumed to be equiprobable. Therefore, the CRC bits can be used either as unstructured systematic bits or as input to the DM along with the other information bits.

[0063] When systematically puncturing shaped bits, to maintain shaping, each symbol must be labeled with a shaped bit. For example, in Figure 5, all symbols must be labeled with the set b1b2b3 (and one non-shaped bit b4). Therefore, when systematic bit b1 is punctured, the corresponding bit b2b3 and one parity bit b4 must also be punctured.

[0064] Consider the coding scheme of Figure 5, with a code rate R = (m-1) / m (R = 3 / 4 in the example provided). At the output of the error-correcting code, the bits are ordered as B1B2B3B4, which is the concatenation of three labeling bit streams B1B2B3, each formed from the systematic bits output by the shaping scheme, and bit stream B4, formed from the parity bits output by the parity encoder.

[0065] Suppose we achieve a rate R=6 / 7 by transmitting only a subset of the coded bits (and puncturing some parity bits), i.e., by transmitting 7k' / 2 bits instead of 4k'=8k' / 2 bits. In that case, we must puncture a total of k' / 2 bits, as shown in the proposed puncturing scheme 600 in Figure 6. That is, for each original labeling bit stream, we must puncture k' / (2*m) bits 602.

[0066] This can be achieved by the system shown in Figure 11, which corresponds to a modification of the system of Figure 5 by placing π as an additional interleaver 700 after the parity encoder and before the rate matcher, so that the appropriate bits are punctured. The bit ordering 702 after this new interleaver 700 is shown in Figure 12. A RM (similar to that in the 5G NR LDPC coding scheme) that punctures the last coded bit then discards the punctured bit 800, represented here as the rightmost part.

[0067] Note that the same result can be achieved without adding an interleaver by simply replacing the RM 108 with the demultiplexer 110, as shown in Figure 13. In this case, the RM simply discards the k' / (2*m) last symbols.

[0068] Alternatively, the labeling bit streams and puncturing positions can have different forms, as long as k' / (2*m) bits of each original labeling bit stream are punctured. This alternative can be obtained by a (first) interleaver π' 1000 placed after the multiplexer and before the parity encoder, as shown in Figure 14. In this case, the (second) interleaver π 700 rearranges the bits so that the ordering of Figure 12 is obtained at the output. Figure 15 shows an example 1002 of a different ordering after π'. This ordering represents the systematic bits provided to the parity encoder and the parity bits output by the parity encoder, as well as alternative puncturing positions 1100 corresponding to the punctured bits. It should also be noted that any parity bit (not necessarily the last bit) can be punctured.

[0069] In summary, the proposed technique can implement a rate R=((m−1)k′) / (mk′−x), under the constraint that x / m bits of each original labeling bit stream are punctured, where x is the number of punctured bits. The code used has a rate of ((m−1)k′) / mk′. Note that this latter rate can be obtained by using parity coding to generate more parity bits and puncturing some of these parity bits directly (i.e., within block P in the figure).

[0070] It should be noted that the original labeling bit stream defined above with respect to the puncturing of systematic shaped bits will be established after the puncturing of systematic non-shaped bits (i.e., when saying "a bit stream is original if it would have been obtained without puncturing", the puncturing of systematic shaped bits is not taken into account).

[0071] In summary, for both systematic shaped and unsystematic puncturing of bits, the combined proposed technique can implement rate ((mq)k'+c') / ((mq)k'+qk'-d)=((mq)k'+c') / (mk'-d), where: The code rate is ((mq)k'+c') / (mk'+cd), c is the number of systematic non-shaped bits punctured, c'=c+d(d≧0), x is the number of punctured bits under the constraint that x / m bits of each original labeling bit stream are punctured.

[0072] An example of a system combining both of the proposed techniques is shown in Figure 16. An example of a possible coded sequence 1604 at the output of the parity coding before puncturing and interleaving is shown in Figure 17. After puncturing, the output coded sequence 1602 is demultiplexed according to a mapping scheme and used to label the k' symbols to be transmitted.

[0073] Next, we propose applying the system in FIG. 16 to HARQ.

[0074] Consider the scheme of Figure 16 for c = 0 and m = 4 bits per symbol. After puncturing the systematic bits, we obtain 2k' systematic bits and 2k' parity bits. Assume that we design two rounds of HARQ. The target rate of the first round is R0 = 2 / 3. The remaining bits are transmitted in the second round so that a rate R1 = 2 / 4 is achieved. At the output 1604 of the error correcting code, the bits are in the order shown in Figure 18.

[0075] If only 3k' bits out of 4k' bits have to be transmitted in the first round, then to achieve rate R0=2 / 3, x=k' bits have to be punctured. The bit streams are punctured as follows: k' / 4 bits of each bit stream B1B2B3B4 are punctured. In this case, the overall system is as shown in Figure 19, where π'1900 is the added interleaver. In the second round, the RM selects the rightmost part of the input stream (see Figure 12) and punctures the leftmost part.

[0076] References list [BSS2015] G. Boecherer, F. Steiner, and P. Schulte, “Bandwidth Efficient and Rate-Matched Low-Density Parity-Check Coded Modulation,” IEEE Transactions on Communications, vol. 63, no. 12, Dec. 2015. [DPS2020] Erik Dahlman, Stefan Parkvall, and Johan Skoeld, “5G NR, The next generation wireless access technology,” 2018. [CG2022] Vincent Corlay and Nicolas Gresset, “A Simple Sign-Bit Probabilistic Shaping Scheme” IEEE Communications Letters, vol. 26, no. 4, Apr. 2022.

Claims

1. 1. A method for transmitting k″ symbols in a data transmitter, where k″ is a positive integer, the method comprising: obtaining an input bit sequence representing a first data flow and a second data flow; the first data flow includes k' x (m-q) shaped bits representing at least m-q input labeling bit streams, where k' is a positive integer greater than or equal to k", m is the number of bit streams output by the demultiplexer, and q is a positive integer; the second data flow includes c' unshaped bits, c' being a positive integer less than or equal to (q-1)k', and the c' unshaped bits including at most q-1 quantification bits and / or sign bits; puncturing at least a portion of the unshaped bits of the second data flow to obtain at most q-1 incomplete bit streams; obtaining the k" symbols based on the at most q-1 incomplete bit streams and the m-q input labeling bit streams of the first data flow; A method comprising:

2. ordering the input bit sequence by placing c of the c' unshaped bits at systematic puncturing positions; c is a positive integer equal to or less than c', 2. The method of claim 1, wherein puncturing at least a portion of the unshaped bits of the second data flow comprises puncturing at least c of the unshaped bits at the systematic puncturing positions.

3. designating at least c positions of the c' unshaped bits in the input bit sequence as systematic puncturing positions; 2. The method of claim 1, wherein puncturing at least a portion of the unshaped bits of the second data flow comprises puncturing at least c of the unshaped bits at the systematic puncturing positions.

4. obtaining an output of an encoder receiving as input bits associated with said input bit sequence using a systematic error correction coding scheme, the output of the encoder comprises qk'-d parity bits, where d=c'-c, and d is a positive integer or 0; completing the at most q-1 incomplete bit streams using (qk'-d)-k' of the qk'-d parity bits to obtain at most q-1 complete bit streams; Further comprising: Obtaining the k″ symbols includes: the mq input labeling bit streams of the first data flow; said at most q-1 completed bitstreams; parity bits output by the encoder that are not used to complete the at most q-1 incomplete bit streams; The method according to claim 2 or 3, wherein the method is based on

5. Obtaining the k″ symbols based on m original labeling bit streams includes: ordering an output bit sequence comprising the m original labeling bit streams by placing x bits at additional puncturing positions, where x is a positive integer and the x bits comprise x / m bits of each of the m original labeling bit streams; puncturing the x bits located at the additional puncturing positions to obtain m punctured labeling bit streams; obtaining the k″ symbols by labeling a corresponding input symbol using one bit of each of the punctured labeling bit streams; The method of claim 4, comprising:

6. The method of any one of claims 2 to 5, wherein c'=k'.

7. The method according to any one of claims 2 to 5, wherein c = c'.

8. The method of claim 7 , wherein the additional puncturing positions correspond to the end of the output bit sequence after ordering the output bit sequence.

9. Obtaining the k″ symbols includes: labeling a corresponding input symbol using one bit of each original labeling bit stream to obtain the k″ symbols and an additional symbol; discarding said additional symbols using a rate matcher; The method according to any one of claims 1 to 4, comprising:

10. 10. The method according to claim 1, further comprising, after transmitting the k″ symbols, transmitting another symbol having the c punctured unshaped bits and the x punctured bits as labeling bits.

11. 1. A data transmitter configured to transmit k″ symbols, where k″ is a positive integer, said transmission comprising at least: obtaining an input bit sequence representing a first data flow and a second data flow; the first data flow includes k' x (mq) shaped bits; the first data flow represents at least m input labeling bit streams; the second data flow includes c' unshaped bits associated with the shaped bits; m is the number of bitstreams output by the demultiplexer; q is the number of unshaped bits of the second data flow associated with at least a portion of the shaped bits of the first data flow; the unshaped bits include q' quantification bits and a sign bit; c' is a positive integer less than or equal to k'; puncturing at least a portion of the unshaped bits of the second data flow, thereby obtaining at most q-1 incomplete bit streams; obtaining the k″ symbols based at least on the at most q−1 incomplete bit streams and the m−q input labeling bit streams of the first data flow; The data transmitter is performed by

12. A computer program comprising instructions that, when said computer program is run by a computer, cause said computer to carry out the method of any one of claims 1 to 10.

Images