Source-channel joint coding method for short packet communication

Abstract

This invention discloses a source-channel joint coding scheme for short packet communication. The scheme includes: at the transmitting end, firstly, performing two-stage classified enumerative (CE) coding on the message sequence u to be transmitted to obtain a variable-length description sequence (w). u i u ), where w u Indicates the type of the message, i u This indicates the order of message u within this type class; subsequently, the variable-length sequence (w) will be... u i u After multi-rate matching channel coding, a fixed-length transmit codeword c is obtained. During modulation, transmission energy is allocated according to the channel code rate, with low-rate transmit codewords receiving low transmission energy. At the receiving end, type-specific test decoding is used to recover the transmitted message. For a given type, the corresponding channel code decoding is performed first, followed by CE decoding, ultimately yielding an estimate of the message sequence. This invention has a significant coding gain compared to existing schemes, and its performance can approach the source-channel joint coding bound under finite length conditions.

Description

Technical Field

[0001] This invention belongs to the technical field of digital communication, and specifically relates to a source-channel joint coding method for short packet communication. Background Technology

[0002] Short packet communication or short code length data communication is a typical communication scenario in the Internet of Things (IoT), wireless sensor networks, and 5G mobile communication networks. Reliability and energy efficiency are two key indicators for such scenarios. In short packet or short code length communication scenarios, the Shannon separation coding theorem, which assumes unlimited latency, no longer applies. In fact, in short packet or short code length communication scenarios, joint source-channel coding (JSCC) can achieve better performance than separate source-channel coding (SSCC).

[0003] Traditional JSCC mainly includes two types of designs: one type of JSCC is based on the existing source coding scheme, which is combined with the channel coding scheme and the source and channel joint decoding is performed; the other type of JSCC is based on the joint factor graph design of the source code and the channel code, including typical dual LDPC design and dual polar design.

[0004] On the one hand, existing JSCC schemes mainly focus on medium-length or medium-long code designs, lacking effective designs for short packets or short code lengths. On the other hand, the gain of existing JSCCs often depends on known or estimated source statistical distributions, while obtaining or estimating source statistical distributions is difficult in typical short packet or short code length regions, requiring a universal JSCC design that does not rely on source statistical distributions. Summary of the Invention

[0005] The main objective of this invention is to overcome the shortcomings and deficiencies of the prior art and to provide a source-channel joint coding method for short packet communication in wireless communication scenarios. This method has a significant coding gain compared to existing schemes and its performance can approach the source-channel joint coding bound under finite length.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] This invention provides a source-channel joint coding method for short packet communication, comprising the following steps:

[0008] (1) At the sending end, assume a binary sequence The message space is defined as a sequence of messages to be transmitted, with a total number of type m and a partitioning criterion. Divide into m subsets A i ,Right now Furthermore, any two distinct subsets are disjoint, and 0i≤m-1. Encoding and energy allocation are performed according to the following steps:

[0009] (1.1) Based on the above classification criteria, the message sequence u to be transmitted is classified and enumerated in two stages, i.e., classified enumeration, or CE encoding for short, to obtain the description sequence v = (w u i u ),in It is the binary representation of the U-shaped message to be transmitted. express Two-dimensional space, The message sequence u to be transmitted is in this type subset A i The binary representation of sorted data. express Two-dimensional space;

[0010] (1.2) The description sequence v is encoded into a transmission codeword c of length n by multi-rate channel coding;

[0011] (1.3) During modulation and transmission, the transmission energy a is allocated to the i-th type sequence according to the channel code rate. i The basic principle of energy allocation is to allocate low energy to low bitrates and high energy to high bitrates.

[0012] (2) At the receiving end, targeted type test decoding is used to recover the message sequence u. The specific method is as follows:

[0013] (2.1) Sequential or parallel pairing of type i = i0, i1, ..., i m-1 Decoding is performed, and the decoding process includes the following steps:

[0014] (2.1.1) Based on the received sequence y, perform the corresponding list decoding on type i to obtain a series of candidate codewords.

[0015] (2.1.2) Select the most likely and valid candidate codeword, perform CE decoding, and obtain an estimate of the message sequence.

[0016] As a preferred technical solution, in step (1.1), the CE encoding used describes the binary sequence by the type of the binary sequence and its order in the corresponding type. The type division is arbitrary, including but not limited to direct division based on sequence Hamming weight and symmetric Hamming weight division; the order is also arbitrary, including but not limited to dictionary order.

[0017] As a preferred technical solution, the multi-rate code used in step (1.2) is arbitrary, including but not limited to random code, polar code, low-density consistency check code, turbo code, and polarization-adjusted convolutional code.

[0018] As a preferred technical solution, the multi-rate code used is an additional Cyclic Redundancy Check (CRC) code. The CRC encoding and the aforementioned channel encoding are concatenated to form a multi-rate channel encoding module.

[0019] As a preferred technical solution, in step (1.3), the power allocation is based on arbitrary type coding bounds or performance estimation techniques, including but not limited to random code joint bounds, test dependency bounds, normal approximations, etc., or based on the simulation performance of a specific type, or arbitrarily allocated.

[0020] As a preferred technical solution, in step (2.1), the step of sequentially sorting the type i = i0, i1, ... i m-1 Decoding is performed in the following order: i = 0, 1, ..., m-1, or i = m-1, m-2, ..., 0, or a specific order based on the transmission energy estimate, or any order, or in parallel for all types.

[0021] As a preferred technical solution, in step (2.1), the step of sequentially sorting the type i = i0, i1, ... i m-1 In the decoding process, the decoding can be terminated early based on judgments such as valid codewords, thresholds, or list decoding early exit conditions.

[0022] As a preferred technical solution, in step (2.1.1), the corresponding list decoding is performed based on the received sequence y and type i. Here, list decoding can be any decoding method, including but not limited to sequential statistical decoding, locally constrained sequential statistical decoding, Chase decoding, bit flipping decoding, and serial interference cancellation list decoding; in addition, it also includes but is not limited to special list sizes of l. max =1 serial interference cancellation decoding, belief propagation (belief decoding and other traditional decoding schemes).

[0023] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0024] 1. The coding design of this invention is simple and easy to implement, avoiding the complex optimization design in traditional JSCC;

[0025] 2. The transceiver scheme of this invention does not rely on the statistical information of the information source, has universality, and is suitable for short packet or short code long communication scenarios;

[0026] 3. This invention proposes an energy allocation strategy that allocates low transmission energy to low code rate channel codes and high transmission energy to high code rate channel codes, which can effectively improve energy efficiency.

[0027] 4. The source-channel joint coding scheme proposed in this invention (with unknown source statistical characteristics) has a significant coding gain compared to the existing dual-polar scheme (with perfectly known source statistical characteristics) (at a frame error rate of 10). -3 It has a gain of about 0.3dB and can approach the finite code length JSCC boundary. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a schematic diagram of the encoding and transmission of the source-channel joint coding method for short packet communication according to the present invention;

[0030] Figure 2 This is a flowchart of the serial decoding process of the source-channel joint coding method for short packet communication according to the present invention;

[0031] Figure 3 This is a schematic diagram illustrating the performance of serial decoding of the Bernoulli source in Embodiment 1 of the present invention;

[0032] Figure 4 This is a schematic diagram illustrating the performance of parallel decoding of the Bernoulli source in Embodiment 2 of the present invention. Detailed Implementation

[0033] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present application without creative effort are within the scope of protection of the present application.

[0034] In this application, the reference to "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a mutually exclusive, independent, or alternative embodiment. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described in this application can be combined with other embodiments.

[0035] Example 1

[0036] This embodiment 1 provides a source-channel joint coding method for short packet communication. Considering the message sequence as probability Pr{u i A Bernoulli source with a length of k=64 and a class of m=8 has a transmitted codeword length of n=128. First, u is encoded using CE to generate a variable-length compressed sequence (w). u i u ), secondly, regarding (w u i u The generator polynomial is x. 12 x 11 +x 3 +x 2 +x+1 Cyclic Redundancy Check (CRC) encoding, adding 12 bits of CRC information p. u , to obtain (w u i u ,p u Finally, (w) u i u ,p u ) Perform random encoding, i.e. Obtain the transmitted codeword c of length n. Achieve the target frame error rate of 10% according to the RCU boundary. -3 The required signal-to-noise ratio (SNR) is used to set the transmission energy a for each type i. i For simplicity, this embodiment 1 linearly scales the transmission amplitude to change the SNR. The signal is transmitted through an additive white Gaussian noise (AWGN) channel after binary phase-shift keying (BPSK) modulation. At the receiver, the source distribution is unknown, and serial test-type decoding is used for decoding. The list decoding uses the LC-OSD decoding algorithm with parameters l. max =2 14 And δ = 8. For comparison, this Example 1 assumes w... uWith the decoder known, the lower bound of the GA is plotted, and the simulation results are as follows: Figure 3 As shown, in the low to medium SNR region, the performance of the proposed scheme matches well with the weighted frame error rate (FER) estimation bound and the weighted RCU bound, and is within 0.5 dB of Gallager's source-channel joint coding bound.

[0037] Example 2

[0038] This embodiment 2 provides a source-channel joint coding method for short packet communication. Considering the message sequence as probability Pr{u i A Bernoulli source with a length of k=64 and a transmission codeword length of n=128 is used, divided into m=8 classes. First, u is encoded using CE to generate a variable-length compressed sequence (w). u i u ), secondly, regarding (w u i u The generator polynomial is x. 16 x 12 +x 5 +1 CRC encoding, adding 16 bits of cyclic redundancy check information p u , to obtain (w u i u ,p u Finally, (w) u i u ,p u Perform CRC-assisted Reed-Muller encoding to obtain a transmit codeword c of length n. Based on the RCU bounds, achieve the target FER = 10. -3 The required SNR is used to set the transmission energy a for each type i. i The signal is modulated using BPSK and transmitted through an AWGN channel. For simplicity, we linearly scale the transmitted amplitude to change the SNR. At the receiver, the source distribution is unknown, and parallel test-class decoding is used for decoding. The list decoding employs the LC-OSD decoding algorithm with parameters l. max =2 14 And δ = 8. This Example 2 uses the separation and joint decoding performance of existing dual-polarization JSCC schemes as a comparison object, and also lists the SSCC boundary and JSCC boundary. Simulation results are as follows... Figure 4 As shown, compared to the dual-polarization JSCC scheme, the JSCC scheme proposed in this embodiment 2 has a significant coding gain, especially at FER=10. -3 It has a gain of about 0.3dB and can approach the finite code length JSCC limit.

[0039] Example 3

[0040] This embodiment 3 provides a source-channel joint coding method for short packet communication. It considers a time-varying Bernoulli source, i.e., the source distribution rate Pr{u i =1} is fixed in one frame, and samples are taken independently, uniformly, and randomly from the interval (0,1) between frames. The transmission sequence length is k=64, divided into m=8 classes, and the transmission codeword length is n=128. First, u is encoded using CE to generate a variable-length compressed sequence (w u i u ), secondly, regarding (w u i u The generator polynomial is x. 16 +x 12 +x 5 +1 CRC encoding, adding 16 bits of cyclic redundancy check information p u , to obtain (w u i u ,p u Finally, (w) u i u ,p u Perform CRC-assisted Reed-Muller encoding to obtain a transmit codeword c of length n. Based on the RCU boundaries, achieve the target FER = 10. -1 10 -2 10 -3 and 10 -4 The required SNR is used to set the transmission energy a for each type i. i The signal is transmitted through an additive white Gaussian noise (AWGN) channel after binary phase-shift keying (BPSK) modulation. At the receiver, the source distribution is unknown, and parallel test-type decoding is used. The list decoding employs the LC-OSD decoding algorithm with parameters l. max =2 14 And δ = 8. The simulation results are shown in Table 1. It can be seen that the performance of the parallel decoder is close to the expected target FER, and the difference with the performance of the GA decoder is very small.

[0041] Table 1. Performance comparison of parallel decoding and sprite-assisted decoding under different target FERs.

[0042] Target FER <![CDATA[10 -1 ]]> <![CDATA[10 -2 ]]> <![CDATA[10 -3 ]]> <![CDATA[10 -4 ]]> Parallel Decoding 2.44e-2 2.15e-3 2.81e-4 8.38e-5 Elf-assisted decoding 2.43e-2 2.03e-3 2.72e-4 6.83e-5

[0043] It should be noted that, for the sake of simplicity, the aforementioned method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously.

[0044] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0045] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0046] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any changes, modifications, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A source-channel joint coding method for short packet communication, characterized in that, Includes the following steps: (1) At the sending end, assume a binary sequence The message space is defined as a sequence of messages to be transmitted, with a total number of type m and a partitioning criterion. Divide into m subsets A i ,Right now And any two distinct subsets are disjoint, 0≤i≤m-1, and the encoding and energy allocation are performed according to the following steps: (1.1) Based on the above classification criteria, the message sequence u to be transmitted is classified and enumerated in two stages, i.e., classified enumeration, or CE encoding for short, to obtain the description sequence v = (w u i u ),in It is the binary representation of the U-shaped message to be transmitted. express Two-dimensional space, The message sequence u to be transmitted is in this type subset A i The binary representation of sorting in the middle. express Two-dimensional space; (1.2) The description sequence v is encoded into a transmission codeword c of length n by multi-rate channel coding; (1.3) During modulation and transmission, the transmission energy a is allocated to the i-th type sequence according to the channel code rate. i The basic principle of energy allocation is to allocate low energy to low bitrates and high energy to high bitrates. (2) At the receiving end, targeted type test decoding is used to recover the message sequence u. The specific method is as follows: (2.1) Sequential or parallel pairing of type i = i0, i1, ..., i m-1 Decoding is performed, and the decoding process includes the following steps: (2.1.1) Based on the received sequence y, perform the corresponding list decoding on type i to obtain a series of candidate codewords. (2.1.2) Select the most likely and valid candidate codeword, perform CE decoding, and obtain an estimate of the message sequence.

2. The source-channel joint coding method for short packet communication according to claim 1, characterized in that, In step (1.1), the CE encoding used describes the binary sequence by the type class of the binary sequence and its order in the corresponding type class. The type class division is arbitrary, including but not limited to direct division based on sequence Hamming weight and symmetric Hamming weight division; the order is also arbitrary, including but not limited to dictionary order.

3. The source-channel joint coding method for short packet communication according to claim 1, characterized in that, In step (1.2), the multi-rate code used can be arbitrary, including but not limited to random code, polar code, low-density consistency check code, turbo code, and polarization-adjusted convolutional code.

4. The source-channel joint coding method for short packet communication according to claim 3, characterized in that, The multi-rate code used can be supplemented with a cyclic redundancy check (CRC) code. The CRC code and the channel code mentioned above are concatenated to form a multi-rate channel coding module.

5. The source-channel joint coding method for short packet communication according to claim 1, characterized in that, In step (1.3), the power allocation is based on arbitrary type coding bounds or performance estimation techniques, including but not limited to random code joint bounds, test dependency bounds, normal approximations, or simulation performance based on specific types, or arbitrary allocation.

6. The source-channel joint coding method for short packet communication according to claim 1, characterized in that, In step (2.1), the sequential sorting of type i = i0, i1, ... i m-1 Decoding is performed in the following order: i = 0, 1, ..., m-1, or i = m-1, m-2, ..., 0, or a specific order based on the transmission energy estimate, or any order, or in parallel for all types.

7. The source-channel joint coding method for short packet communication according to claim 1, characterized in that, In step (2.1), the sequential sorting of type i = i0, i1, ... i m-1 In the decoding process, the decoding can be terminated early by judging the conditions for early termination based on the legal codeword, threshold, or list decoding.

8. The source-channel joint coding method for short packet communication according to claim 1, characterized in that, In step (2.1.1), the corresponding list decoding is performed based on the received sequence y and type i. Here, list decoding can be any decoding method, including but not limited to sequential statistical decoding, locally constrained sequential statistical decoding, Chase decoding, bit-flipping decoding, and serial interference cancellation list decoding; in addition, it also includes but is not limited to special list sizes. Serial interference cancellation decoding, belief propagation decoding, and other traditional decoding schemes.

Images