Coding-assisted OTFS peak-to-average ratio reduction method

Abstract

The invention discloses a coding-assisted OTFS peak-to-average power ratio reduction method in the technical field of wireless communication, and the method comprises the following steps: determining a forming bit selection mode, determining V bit positions as forming bits in an information bit or a frozen bit of a bit to be coded of a polarization code according to a preset forming scheme, different forming bit combinations are configured for the V forming bits, 2V bit combinations are generated, polarization code coding, symbol modulation and OTFS modulation are sequentially carried out on the bits to be coded containing the different forming bit combinations, and corresponding sending signals are obtained; and selecting the sending signal with the minimum peak-to-average power ratio (PAPR) value to be sent to a channel through an antenna. According to the coding-assisted OTFS peak-to-average power ratio reduction method, the peak-to-average power ratio of the transmitted signal can be remarkably reduced under the condition that the channel coding error correction performance is not reduced.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication, in particular to a coding-aided OTFS peak-to-average ratio reduction method. BACKGROUND

[0002] Orthogonal time frequency space (OTFS) modulation technology is a potential waveform technology for the sixth generation mobile communication (6G), which can effectively combat Doppler shift by placing information symbols in the time-delay-Doppler domain, thereby supporting higher mobility and higher operating frequency bands. However, as a multi-carrier modulation technology, OTFS has a high peak-to-average power ratio (PAPR). High power amplifiers need to work in the linear region, otherwise high adjacent channel interference will occur, resulting in serious reduction of error code performance. Excessive PAPR will cause the high power amplifier to work in the nonlinear characteristic region, resulting in inter-carrier interference, reducing amplification efficiency, increasing hardware cost, causing serious distortion of the transmitted signal, and deteriorating the OTFS transmission performance. Therefore, it is necessary to reduce the PAPR level of OTFS modulation.

[0003] Polar code is a constructive coding scheme that can theoretically prove the capacity of the channel, has a regular coding structure, low decoding complexity, and excellent error correction performance. The theoretical basis of polar code is channel polarization, which recombines and splits a plurality of independent physical channels with the same capacity to generate a plurality of virtual synthetic bit channels with different capacities. On this basis, during encoding, high-reliability bit channels are used to transmit source information, and low-reliability bit channels carry fixed bits known to the transmitter and receiver.

[0004] Existing PAPR reduction techniques include amplitude limiting filtering, companding transformation, selective mapping (SLM), partial transmission sequence (PTS), and neural network-based methods. These methods may introduce distortion, have high computational complexity, or require additional sideband information. SUMMARY

[0005] The present application aims to provide a coding-aided OTFS peak-to-average ratio reduction method, which solves the technical problems of introducing distortion, high computational complexity, or requiring additional sideband information in the prior art. Based on polar codes and OTFS systems, the coding structure characteristics of polar codes are utilized to combine with OTFS modulation, effectively reducing the PAPR of OTFS signals without sacrificing error correction performance, suppressing the peak-to-average ratio level of OTFS transmitted signals, and enabling them to work normally in the linear region of high power amplifiers, thereby improving the performance of information data transmission in wireless channels.

[0006] To achieve the above-mentioned purpose, the technical scheme of the present application is as follows:

[0007] An OTFS peak-to-average ratio reduction method with coding assistance, comprising the following steps:

[0008] S1. Determine the shaping bit selection method, which includes a selection method based on a peak-to-average ratio impact metric criterion and a selection method based on reinforcement learning;

[0009] S2. According to the selected shaping bit selection method, determine V bit positions as shaping bits in the information bits or frozen bits of the polar code to be encoded according to the preset shaping scheme, V being a positive integer;

[0010] S3. Configure different shaping bit combinations for the V shaping bits, generate 2^ V bit combinations, and sequentially encode, modulate, and OTFS modulate the to-be-encoded bits containing different shaping bit combinations to obtain corresponding transmission signals;

[0011] S4. Calculate and select the transmission signal with the minimum peak-to-average power ratio (PAPR) value as the actual transmission signal, which is sent into the channel through the antenna, and the information is recovered from the received signal through demodulation and decoding at the receiving end.

[0012] Further, the selection method based on the peak-to-average ratio impact metric criterion includes the following steps:

[0013] S11. Randomly generate K' bits as source information and place them in the information bits, and assign the remaining frozen bit values to 0, K' being the initial number of information bits;

[0014] S12. According to the selected shaping scheme, determine the range of the shaping bits, generate the time-domain OTFS signal, and sequentially calculate the metric value of the i-th to-be-encoded bit in the range of the shaping bits by calculating the PAPR ratio of the transmission signal when the i-th encoded bit takes different values, i∈{1,2,...,N}, N being the total number of to-be-encoded bits;

[0015] S13. Repeat the process of S11-S12 for T times, T being a positive integer, to obtain the metric value of each to-be-encoded bit in the range of the shaping bits

[0016] S14. Calculate the average value d of the T groups of metric values i of each to-be-encoded bit.

[0017] S15. Sort all d i from large to small, and select the first V d i corresponding bit positions as shaping bits.

[0018] Further, the calculation formula of the metric value of the i-th to-be-encoded bit in S12 is as follows:

[0019]

[0020] wherein, PAPR(i) represents the peak-to-average ratio influence metric value of the i-th (i∈{1,2,...,N}) to-be-encoded bit, The greater the value is, the greater the influence of the to-be-encoded bit on the peak-to-average ratio level of the transmitted signal is;b i b(i) represents the bit value of the i-th to-be-encoded bit; P(b i ) represents the PAPR value of the transmitted signal after encoding and modulation when the value of the i-th to-be-encoded bit is b i .

[0021] Further, the calculation process of the PAPR of the transmitted signal in S12 is as follows:

[0022]

[0023] wherein, x k represents the time-domain discrete symbol after OTFS modulation, U represents the total number of two-dimensional discrete samples in the DD domain, and L represents the oversampling coefficient.

[0024] Further, the selection method based on reinforcement learning specifically includes the following steps:

[0025] S1-1. When V≤3, the selection method based on reinforcement learning is used to determine the shaping bit, a training learning environment is constructed, and the state s is defined as the transmitted signal after polarization encoding and OTFS modulation of the to-be-encoded bit; the action a is the set of V shaping bits selected from the originally information bits or frozen bits, the reward r is the feedback of the PAPR level result of the transmitted signal after the shaping bits are configured with different bits, and the reward value r is calculated according to the formula:

[0026]

[0027] wherein, b(j) represents the bit value of the j-th shaping bit, j∈{1,2,...,V}, s1 to s v represent the shaping bit index, and P(·) is the PAPR calculation function;

[0028] S1-2. The Q-learning algorithm is used for learning training, and the state-action function value calculation formula in the learning training is as follows:

[0029]

[0030] Wherein, Q represents a state-action function value, a represents a learning rate in the Q-learning algorithm; s' represents a future state in the interaction process between the agent and the environment, and a' represents a future state in the interaction process between the agent and the environment.

[0031] S1-3. The optimal strategy is determined by using the greedy criterion after sufficient period of learning training, and a sending signal corresponding to the set of shaping bits when the PAPR level is optimal is output.

[0032] Further, the preset shaping scheme in S2 includes:

[0033] S21. Shaping scheme one: the shaping bits are selected from the information bits of the original polar code, when the shaping scheme one is adopted, the coding rate R is reduced from R=K' / N to R=(K'-V) / N=K / N, at this time, the number of frozen bits F=F' satisfies N=K+V+F.

[0034] S22. Shaping scheme two: the shaping bits are selected from the frozen bits of the original polar code, when the shaping scheme two is adopted, the coding rate R remains R=K' / N, at this time, the number of frozen bits F=F'-V, and the current number of information bits K=K' satisfy N=K+V+F.

[0035] Further, the shaping scheme one is suitable for the resource-restricted sending end uplink in the actual communication scenario; and the shaping scheme two is suitable for the resource-sufficient sending end downlink in the actual communication scenario.

[0036] Further, the symbol modulation in S3 adopts M-QAM modulation or QPSK modulation, M is an integer greater than 1 and commonly takes 4, 16 or 64.

[0037] Further, the OTFS modulation process in S3 includes M-QAM modulation symbol generation, inverse symplectic Fourier transform ISFFT and Heisenberg transform.

[0038] Further, the information recovery from the received signal by demodulation and decoding at the receiving end in S4 includes the following steps:

[0039] S41. After receiving the signal, the receiving end sequentially performs OTFS demodulation, symbol demodulation and polar code decoding.

[0040] S42. When the shaping bits are selected from the frozen bits, the receiving end needs to perform multiple times of decoding for 2^ V shaping bit configuration cases until the source information is recovered through the cyclic redundancy check.

[0041] By adopting the above technical scheme, the present application has the following advantages:

[0042] The application provides an encoding-assisted OTFS peak-to-average ratio reduction method, effectively reduces the PAPR of the OTFS modulated signal, and the PAPR suppression effect is obvious; simulation shows that 1.5dB to 4dB gain can be obtained when CCDF=10^-3. The polar code construction is not changed, that is, the reliability order of the bit channel is not changed, only part of the bit channel position is preferred as the beamforming bit, and the PAPR performance of the transmitted signal is optimized, while the bit channel generation result of the polar code and the decoding algorithm are not changed, the original error correction performance of the polar code is maintained, and the error rate performance of the receiving end is consistent with the conventional scheme. By utilizing the coding structure characteristics of the polar code and the OTFS modulation cascade design, the peak-to-average ratio level of the transmitting end transmitted signal can be significantly reduced, without additional signal pre-distortion module or complex sideband information transmission, the PAPR of the OTFS signal is effectively reduced, the peak-to-average ratio level of the OTFS transmitted signal is suppressed, the OTFS transmitted signal can normally work in the linear region of the high-power amplifier, and the information data transmission performance in the wireless channel is improved. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 The system block diagram corresponding to the encoding-assisted OTFS peak-to-average ratio reduction method of the application is shown in the figure.

[0044] Figure 2 The flowchart of the encoding-assisted OTFS peak-to-average ratio reduction method of the application is shown in the figure.

[0045] Figure 3 The PAPR performance simulation comparison chart of the OTFS signal by using the method of the application and the conventional method is shown in the figure. DETAILED DESCRIPTION

[0046] The technical solutions of the application are described in detail below in combination with the accompanying drawings of the specification. It should be noted that in this document, relationship terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or sequence between the entities or operations. Moreover, the terms "include", "contain" or any other variants thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or equipment.

[0047] The system block diagram corresponding to the encoding-assisted OTFS peak-to-average ratio reduction method of the application is shown in the figure. Figure 1As shown, in high-speed moving, Doppler frequency shift sensitive communication scenarios such as low-orbit satellite Internet satellite-ground communication, Internet of vehicles, a data transmission system adopts OTFS as a modulation method, adopts a polar code as a channel coding scheme, and in specific embodiments, symbol modulation can adopt M-QAM (M-ary quadrature amplitude modulation, M is an integer greater than 1 and is commonly taken as 4, 16 or 64, etc.) modulation data symbols or adopts QPSK modulation, and then maps the symbol to the time-frequency domain. In polar coding, the to-be-coded bits u include three parts: information bits, frozen bits, and configured shaping bits. The bit channel index sets corresponding to them are information bits A, frozen bits F, and shaping bits V, respectively. Different shaping bits are configured, polar code encoding is performed to obtain code word bits c, OTFS modulation is completed after M-QAM modulation symbol, ISFFT, and Heisenberg transformation to obtain a transmission signal x, and the transmission signal with the lowest PAPR is selected as the transmission signal, which is sent into the channel through the transmitting antenna.

[0048] In the polar code encoding process, first, polar code construction is performed to obtain N bit channels sorted by reliability, which carry N to-be-coded bits. In conventional coding, K' information bits are source information to be transmitted, which are carried by the K' bit channels with the highest reliability, and the remaining F' frozen bits are all 0 bits known by the transmitting and receiving ends, which are placed in the remaining bit channels, and N = K' + F'.

[0049] The application also provides an encoding-assisted OTFS PAPR reduction method, specifically as shown Figure 2 The method comprises the following steps:

[0050] S1. Determine the shaping bit selection method, which includes a selection method based on a PAPR impact metric criterion and a selection method based on reinforcement learning;

[0051] The determination of the shaping bit adopts the selection method based on the PAPR impact metric criterion, that is, the PAPR impact metric criterion of the to-be-coded bit is determined, the metric values are sorted from large to small, and the V largest metric values are selected as the shaping bits. The PAPR impact metric of the to-be-coded bit is obtained by calculating . Wherein, represents the PAPR impact metric value of the i-th (i is an element in {1, 2,..., N}) to-be-coded bit, and the greater the value, the greater the impact of the to-be-coded bit on the PAPR level of the transmission signal. i represents the bit value of the i-th to-be-coded bit; Ρ(b i represents the PAPR value of the transmission signal after encoding and modulation when the value of the i-th to-be-coded bit is b i . The calculation process of PAPR is as follows:

[0052]

[0053] Where, x k Let U represent the time-domain discrete symbol after OTFS modulation, U represent the total number of two-dimensional discrete samples in the DD domain, and L represent the oversampling coefficient. Based on the selected shaping scheme, the range of the shaped bits is determined, and the value at each bit position is calculated bit by bit. Then, according to the measurement value Sort the values ​​from largest to smallest, and select the V largest metric values ​​as the shaping bits.

[0054] The selection method for the peak-to-average ratio (PAPR) impact metric in S1 includes the following steps:

[0055] S11. Randomly generate K' bits as source information and place them in the information bits, and assign the remaining frozen bits to 0. K' is the initial number of information bits.

[0056] S12. Based on the selected shaping scheme, determine the range of the shaped bits, generate a time-domain OTFS signal, and sequentially calculate the metric value of the i-th bit to be encoded within the range of the shaped bits. The PAPR ratio of the transmitted signal is obtained by calculating the PAPR ratio of the transmitted signal when the i-th encoded bit takes different values, where i∈{1,2,...,N} and N is the total number of bits to be encoded.

[0057] S13. Repeat steps S11-S12 for a total of T times, where T is a positive integer, to obtain the measurement value of each bit to be encoded within the range of the shaped bit.

[0058] S14. Calculate the T groups of metrics for each bit to be encoded. average value d i ,in,

[0059] S15. For all d i Sort by size from largest to smallest, and select the first V d. i The corresponding bits are used as shaping bits.

[0060] The reinforcement learning-based selection method in S1 specifically includes the following steps:

[0061] S1-1. The reinforcement learning-based selection method is suitable for small V values, such as V≤3. In this case, the reinforcement learning-based selection method can be used to determine the shaping bits, construct a training environment, and define state s as the transmitted signal after polar coding and OTFS modulation of the bit to be encoded; action a is the set of V shaping bits originally selected as information bits or frozen bits; and reward r is the feedback of the PAPR level result of the transmitted signal after configuring different bits in the shaping bits. In a specific embodiment, the reward value r is calculated using the following formula:

[0062]

[0063] wherein, is the bit value of the jth shaping bit, j∈{1,2,...,V}, s1 to s v is the shaping bit index, P(·) is the PAPR calculation function; that is, when the selected action a is the shaping bit set V={s1,s2,...,s V}, the bit value is assigned in its corresponding bit channel After s1 to sV are assigned as 0 or 1 (j∈{1,2,...,V}), the PAPR value of the transmitted signal (represented by the function P(·)) is calculated after encoding and modulation, and the inverse of the minimum value is selected as the reward value r.

[0064] S1-2. As an embodiment of the reinforcement learning algorithm, the Q-learning algorithm is used for learning and training, and the state-action function value calculation formula in the learning and training is:

[0065]

[0066] wherein, Q represents the state-action function value, a represents the learning rate in the Q-learning algorithm; s' represents the future state in the interaction process of the agent and the environment, and a' represents the future state in the interaction process of the agent and the environment.

[0067] S1-3. By using the greedy criterion, the optimal strategy is determined after sufficient learning and training of episodes The output is the shaping bit set corresponding to the optimal PAPR level of the transmitted signal.

[0068] S2. According to the selected shaping bit selection mode, V bit positions are determined as shaping bits in the information bits or frozen bits of the polar code to be encoded according to the preset shaping scheme, and V is a positive integer;

[0069] In specific embodiments, V positions are selected as shaping bits V in the to-be-encoded bit positions (i.e., bit channels), which can carry multiple shaping bits, a total of 2^ V , including {0,0,……0}, {0,0,……1},..., {1,1,……1}.

[0070] The preset shaping scheme in S2 includes:

[0071] S21. Shaping scheme one: when the shaping bits are in the information bits of the original polar code, the shaping bits occupy part of the positions of the information bits, and the coding rate R is lost, that is, R decreases from R=K' / N to R=(K'-V) / N=K / N, the current number of frozen bits F=F', and N=K+V+F is satisfied.

[0072] S22. The second shaping scheme: when the shaping bits are at the frozen bits of the original polar code, the shaping bits occupy part of the positions of the frozen bits, without loss of the coding rate, R = K' / N, the number of remaining frozen bits is F = F' - V, the current number of information bits K = K', and N = K + V + F is satisfied; at this time, since the shaping bits occupying the frozen bits are unknown to the receiving end, the receiving end needs to perform polar code decoding multiple times according to 2^ V possible configuration cases until the information is recovered through CRC (cyclic redundancy check).

[0073] Considering the implementation complexity, resource consumption and the like of the two shaping schemes, the first shaping scheme is suitable for the resource-limited uplink of the sending end in the actual communication scenario, and the second shaping scheme is suitable for the resource-rich downlink of the sending end in the actual communication scenario. That is, in the actual communication scenario, such as a ground cellular network or a satellite-ground mobile phone direct communication, the power, calculation, storage and the like of the user end are limited, and the uplink thereof can adopt the first shaping scheme; the power, calculation, storage and the like of the base station are sufficient, and the downlink thereof can adopt the second shaping scheme.

[0074] S3. After the shaping bits are determined, different shaping bit combinations are configured for the V shaping bits with different bit values, 2^ V bit combinations are generated, including {0, 0, …, 0}, {0, 0, …, 1}, …, {1, 1, …, 1}, and the coded bits containing different shaping bit combinations are sequentially subjected to polar code encoding, symbol modulation and OTFS modulation to obtain corresponding sending signals;

[0075] wherein the symbol modulation adopts M-QAM modulation or QPSK modulation, M is an integer greater than 1 and is commonly 4, 16 or 64. The OTFS modulation process includes M-QAM modulation symbol generation, inverse Sine Fourier transform ISFFT and Heisenberg transform.

[0076] S4. The sending signal with the minimum peak-to-average power ratio PAPR value is calculated and selected as the actual sending signal, which is sent into the channel through an antenna, and the receiving end recovers the information from the received signal through demodulation and decoding.

[0077] The recovery of the information from the received signal by the receiving end through demodulation and decoding in S4 includes the following steps:

[0078] S41. After receiving the signal, the receiving end sequentially performs OTFS demodulation, symbol demodulation and polar code decoding;

[0079] S42. When the shaping bits are selected from the frozen bits, the receiving end needs to perform multiple times of decoding according to 2^ V configuration cases of the shaping bits until the information source is recovered through cyclic redundancy check.

[0080] Since the above method does not change the polar code construction, that is, the reliability order of the N bit channels remains unchanged, only some bit channel positions are selected as shaping bits to optimize the PAPR performance of the transmitted signal, and the bit channel generation result and decoding algorithm of the polar code remain unchanged, the error correction performance of the codeword remains unchanged, and the bit error rate performance of the receiver is consistent with the conventional scheme.

[0081] The technical solution of the coding-assisted OTFS peak-to-average power ratio reduction method according to the present invention is verified by simulation. In a specific embodiment, the simulation conditions are set as follows: the number of information bits K' = 128, the total number of bits to be encoded or the code length after polar code encoding is N = 256, and 16QAM (i.e., M = 16) modulation symbols are used; V = 6 positions are selected from the bits to be encoded as shaping bits V, carrying various shaping bits, with a total of 2^ 6 The DD domain two-dimensional discrete sample points are set to 8×8, with a total of U=64 and an oversampling coefficient L=4. Here, the shaping position is determined by the selection method based on the peak-to-average ratio influence metric. The simulation diagrams of the PAPR performance of the OTFS signal under different schemes (the two shaping schemes of this invention and the conventional scheme) are shown in the figure. Figure 3 As shown, in Figure 3 In the figure, the horizontal axis represents the peak-to-average ratio (PAPR) (dB) after logarithmic processing, that is, the PAPR value is processed by 10lg(.); the vertical axis represents the complementary cumulative density function (CCDF), which is used to measure the probability that the PAPR value exceeds a given threshold value λ, and is defined as CCDF(λ) = Pr(PAPR > λ).

[0082] from Figure 3 Simulation results show that, compared to the conventional polar code-OTFS scheme without PAPR suppression, the proposed coding-assisted OTFS peak-to-average power ratio (PAPR) reduction method significantly suppresses the PAPR level of the transmitted signal. At the same PAPR level, both beamforming schemes reduce CCDF, with beamforming scheme one showing a more significant PAPR suppression effect. (At CCDF = 10...) -3 At that time, the first shaping scheme achieved a gain of about 4dB compared to the conventional scheme, and the second shaping scheme achieved a gain of about 1.5dB compared to the conventional scheme.

[0083] Finally, it should be noted that although the present invention has been described with reference to specific embodiments, those skilled in the art should recognize that the above embodiments are only used to illustrate the present invention and are not intended to limit the present invention. Various equivalent changes or substitutions can be made without departing from the concept of the present invention. Therefore, any changes or modifications to the above embodiments within the essential spirit of the present invention will fall within the scope of the claims of the present invention.

Claims

1. A coding-assisted method for reducing the peak-to-average power ratio (PAPR) of OTFS, characterized in that, Includes the following steps: S1. Determine the shaping position selection method, which includes a selection method based on the peak-to-average ratio influence metric and a selection method based on reinforcement learning; S2. Based on the selected shaping bit selection method and the preset shaping scheme, determine V bit positions as shaping bits in the information bits or frozen bits of the bit to be encoded in the polar code, where V is a positive integer; S3. Configure different combinations of shaped bits for the V shaped bits to generate 2^ V The bits to be encoded, which contain different shaped bit combinations, are sequentially encoded using polar codes, symbol modulation, and OTFS modulation to obtain the corresponding transmitted signals. S4. Calculate and select the transmitted signal with the smallest peak-to-average power ratio (PAPR) as the actual transmitted signal, send it into the channel through the antenna, and the receiver recovers the information from the received signal through demodulation and decoding.

2. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 1, characterized in that, The selection method for the peak-to-average ratio (PAPR) impact metric in S1 includes the following steps: S11. Randomly generate K' bits as source information and place them in the information bits, and assign the remaining frozen bits to 0. K' is the initial number of information bits. S12. Based on the selected shaping scheme, determine the range of the shaped bits, generate a time-domain OTFS signal, and sequentially calculate the metric value of the i-th bit to be encoded within the range of the shaped bits. The PAPR ratio of the transmitted signal is obtained by calculating the PAPR ratio of the transmitted signal when the i-th encoded bit takes different values, where i∈{1,2,...,N} and N is the total number of bits to be encoded. S13. Repeat steps S11-S12 for a total of T times, where T is a positive integer, to obtain the measurement value of each bit to be encoded within the range of the shaped bit. S14. Calculate the T groups of metrics for each bit to be encoded. average value d i ,in, S15. For all d i Sort by size from largest to smallest, and select the first V d. i The corresponding bits are used as shaping bits.

3. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 2, characterized in that, The measurement value of the i-th bit to be encoded in S12 The calculation formula is as follows: in, This represents the peak-to-average power ratio (PAPR) influence metric for the i-th (i∈{1,2,...,N}) bit to be encoded. A larger value indicates a greater impact of the bit to be encoded on the peak-to-average power ratio (PAPR) of the transmitted signal; b i This represents the bit value of the i-th bit to be encoded; P(b i This indicates that, given the values ​​of the other bits to be encoded, the value of the i-th bit to be encoded is calculated as b. i The PAPR value of the transmitted signal after encoding and modulation.

4. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 3, characterized in that, The calculation process of the PAPR of the transmitted signal in S12 is as follows: Where, x k U represents the time-domain discrete symbol after OTFS modulation, U represents the total number of two-dimensional discrete samples in the DD domain, and L represents the oversampling coefficient.

5. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 1, characterized in that, The reinforcement learning-based selection method in S1 specifically includes the following steps: S1-1. When V≤3, a reinforcement learning-based selection method is used to determine the shaping bits, constructing a training environment. State s is defined as the transmitted signal after polar coding and OTFS modulation of the bit to be encoded; action a is the set of V shaping bits originally selected as information bits or freeze bits; reward r is the feedback of the PAPR level result of the transmitted signal after configuring different bits in the shaping bits. The formula for calculating the reward value r is: in, Let s1 be the bit value of the j-th shaped bit, j∈{1,2,...,V}, and s1 to s2 are the values ​​of the shaped bits. v To assign a geometric index, P(·) is the PAPR calculation function; S1-2. Q-learning algorithm is used for learning and training. The formula for calculating the state-action function value during learning and training is as follows: Where Q represents the state-action function value, α represents the learning rate in the Q-learning algorithm; s′ represents the future state of the agent during its interaction with the environment, and a' represents the future state of the agent during its interaction with the environment. S1-3. Using the greedy criterion, after a sufficient number of learning cycles, determine the optimal strategy and output the set of shaped bits corresponding to the optimal PAPR level of the transmitted signal.

6. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 1, characterized in that, The preset shaping scheme in S2 includes: S21. Shaping Scheme 1: The shaping bits are selected from the information bits of the original polar code. When Shaping Scheme 1 is adopted, the coding rate R is reduced from R = K' / N to R = (K'-V) / N = K / N. At this time, the number of frozen bits F = F', which satisfies N = K + V + F. S22. Shaping Scheme 2: The shaping bits are selected from the frozen bits of the original polar code. When Shaping Scheme 2 is adopted, the coding rate R is kept at R = K' / N. At this time, the number of frozen bits is F = F' - V, and the number of current information bits is K = K', satisfying N = K + V + F.

7. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 6, characterized in that, The first beamforming scheme is suitable for uplinks of user terminals with limited power, computing, and storage resources in actual communication scenarios; the second beamforming scheme is suitable for downlinks of base stations with sufficient power, computing, and storage resources in actual communication scenarios.

8. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 6, characterized in that, The symbol modulation in S3 adopts M-QAM modulation or QPSK modulation, where M is an integer greater than 1 and is usually 4, 16 or 64.

9. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 1, characterized in that, The OTFS modulation process in S3 includes M-QAM modulation symbol generation, inverse symmetric Fourier transform (ISFFT), and Heisenberg transform.

10. The coding-assisted OTFS peak-to-average power ratio reduction method according to claim 1, characterized in that, The steps in S4 for the receiver to recover information from the received signal after demodulation and decoding include: S41. After receiving the signal, the receiving end sequentially performs OTFS demodulation, symbol demodulation, and polar code decoding; S42. When the shaping bit is selected from the frozen bit, the receiver needs to target 2^ V The bit configuration is decoded multiple times until the source information is recovered through cyclic redundancy check (CRC).

Images