A shared codebook multiple access transmission method without user identification approaching the theoretical capacity in finite fields

Abstract

The present invention discloses a method for multiple access transmission with a shared codebook without user identification (URA) that approaches the theoretical capacity under a finite field. The source data of each user terminal is encoded according to the encoding method proposed by the present invention, and the signal is transmitted according to the disclosed sending method. The multi-user signal is dispersed in several subframes. The disclosed encoding and sending methods can construct subframe signals with strong noise resistance. The disclosed decoding method completes the detection and estimation of the active sequence codeword channel state information based on the signal transmission energy and noise statistical information estimation and the pre-stored spread spectrum sequence codebook, and performs soft demodulation and decoding on the state information of the subframe time slot. In addition, the recovered information obtained by parity check is reused for serial interference elimination, and this iterative operation can improve system performance. Under a finite field, the above scheme improves the user capacity of multi-antenna multiple access and approaches the theoretical capacity of multi-antenna without user identification shared codebook multiple access.

Description

Technical Field

[0001] The present invention belongs to the field of wireless / digital communications, and in particular relates to a user-identity-free shared codebook multiple access transmission method approaching theoretical capacity under a finite field. Background Art

[0002] Vertical industry applications, exemplified by the Internet of Things (IoT), have become a primary driver of the development of post-5G and 6G mobile communication systems. Compared to 5G, 6G is expected to offer substantial improvements in performance metrics such as data transmission rate, number of connected devices, and communication latency, to better support diverse application needs. Ultra-large-scale connectivity (also known as mega-connectivity or mega-address access) encompasses a wide range of future industry applications, including low-power, wide-coverage IoT, smart agriculture, and integrated transportation.

[0003] Massive Machine-Type Communications (mMTC) aims to support the dense interaction of machine-type devices in next-generation wireless communication systems, addressing the sporadic traffic flow of many mainstream services, short packet transmission, and the need for low power consumption. In this context, Unsourced Random / Multiple Access (URA) has emerged as a potential supporting technology. Massive MIMO (Massive Multiple Input Multiple Output) antenna technology significantly improves communication system capacity and service quality, and is therefore widely used. The integration of massive antenna systems with URA is imperative.

[0004] Recently, the academic community has demonstrated the performance bounds of user-identifier-free shared codebook multiple access (SHAMA) in massive antenna systems. Currently, no solution has emerged that can approach the theoretically achievable bounds. Based on this reality, the present invention designs a SHAMA multiple access transmission method for user-identifier-free (UID) that approaches the theoretical capacity in a finite field. The system design is simple and easy, with low power consumption, and compared to other transmission schemes, it can approach the theoretical capacity. Summary of the Invention

[0005] The present invention addresses the problem of improving system capacity in a limited domain for a system equipped with a large-scale antenna under mMTC services. The transmission method disclosed in the present invention divides a signal frame of a fixed total length into several subframes, and each subframe is divided into two subframe blocks. The subframe blocks respectively carry spread spectrum sequence signals from multiple users and constellation point modulation signals after multiple transmission modes are arranged. The decoding method first estimates the number of active users contained in each subframe signal based on the signal transmission energy and noise statistical information, and then detects and estimates the channel state information based on the pre-stored spread spectrum sequence codebook. After obtaining the active sequence codeword information and the corresponding channel state information, the subframe time slot is soft-demodulated. The obtained soft demodulation state information is used to simultaneously complete the transmission mode activity detection and the corresponding constellation point soft demodulation. The constellation point soft demodulation information is used to calculate the bit likelihood soft information, and the forward error correction decoding of the channel coding is performed based on the bit likelihood soft information. In addition, the recovered information through the parity check is repeatedly used for serial interference elimination. This iterative operation can improve system performance. Under a finite field, the above scheme improves the user capacity of multi-antenna multiple access and approaches the theoretical capacity achievable by multi-antenna shared codebook multiple access. To achieve the above purpose, the technical solution of the present invention is as follows: a method for transmitting multiple access with no user identification and shared codebook (URA) that approaches the theoretical capacity under a finite field, a method for encoding source bit data, a method for transmitting, and a method for decoding multiple access with no user identification and shared codebook. The disclosed decoding method corresponds to the disclosed encoding method and the transmitting method. At the same time, unlike the multiple access encoding and decoding focused on in the related field under a single-antenna Gaussian address channel, the decoding method disclosed in the present invention is applicable to the uplink communication access process under a large-scale antenna configuration. The single-antenna system does not essentially have the conditions required by the decoding method disclosed in the present invention. Specifically, the decoding method disclosed in the present invention needs to complete the estimation and recovery of the multi-user constellation point symbols and the transmission mode based on the estimated channel state information. However, the single-antenna receiving end cannot provide sufficient signal observations and can only provide a received signal of rank 1, which is not enough to provide information equivalent to the amount to be estimated.

[0006] As an improvement of the present invention, the encoding method includes:

[0007] Source bit data and its division and encoding method;

[0008] Pre-stored spreading sequence codebook and transmission mode codebook;

[0009] Divide and encode the source bit data into a subframe selection bit sequence, a preamble selection bit sequence, a transmission mode selection bit sequence, and a transmission bit sequence, and complete corresponding encoding respectively;

[0010] The user selects the subframe position for sending information through the subframe selection bit sequence;

[0011] Through the preamble selection bit sequence, the user selects the corresponding spreading sequence codeword from the pre-stored spreading sequence codebook;

[0012] The user selects a plurality of transmission modes from a pre-stored transmission mode codebook through the transmission mode selection bit sequence;

[0013] The transmission bit sequence is channel coded and constellation-point modulated, and is arranged in multiple transmission modes selected by the transmission mode selection bit sequence. After the arrangement is completed, channel transmission is completed according to a preset transmission method. In the encoding method, the source bit data is divided into multiple sub-blocks, which are respectively a subframe selection bit sequence, a preamble selection bit sequence, a transmission mode selection bit sequence, and a transmission bit sequence.

[0014] The subframe selection bit sequence determines the subframe position occupied by the user, and the transmission mode selection bit sequence determines the arrangement of subsequent modulation constellation points;

[0015] The subframe selection bit sequence and the transmission mode selection bit sequence do not participate in channel coding and constellation point modulation, thereby reducing the number of modulation constellation points that need to be sent subsequently;

[0016] The transmission mode selection bit sequence selects the multi-terminal transmission mode for the tail of the subframe. Since the size of the transmission mode codebook grows exponentially, the transmission mode selection bit sequence selects multiple transmission modes in segments for subsequent arrangement. Therefore, only a relatively small-sized transmission mode codebook needs to be pre-stored, which helps reduce storage requirements.

[0017] As an improvement of the present invention, the sending method includes:

[0018] The channel with fixed channel usage under finite field is divided into multiple subframe positions;

[0019] The selected spreading sequence codeword is located at the head of the subframe, and together with the constellation point symbol arranged in multiple transmission modes at the tail of the subframe, constitutes a complete subframe;

[0020] completing signal transmission based on the selected subframe position;

[0021] Based on the subframe selection bit sequence, only signals of some users are superimposed at the subframe position in each subframe;

[0022] The receiving end receives all subframe signals in the finite field and recovers the source data for each subframe based on the set decoding method.

[0023] In the transmission method, only the corresponding signals of the encoded leading selection bit sequence and the transmission bit sequence need to be transmitted;

[0024] The collision probability of the preamble signal, i.e., the spreading sequence codeword / pilot signal, is reduced. Because the subframe position is randomly selected by the subframe selection bit sequence, and the preamble selection bit sequence randomly selects the spreading sequence codeword, the probability of multi-user collision of the pilot codeword used to estimate the channel state information can be controlled to be extremely low. (If the pilot signals used to estimate the channel between multiple users collide, only overlapping channel coefficients can be estimated. The overlapping channel coefficients cannot support subsequent symbol detection, resulting in communication failure.)

[0025] The subframe selection bit sequence and the transmission mode selection bit sequence do not participate in channel coding and constellation point modulation, and thus do not generate redundant signal units that need to be sent, thereby reducing transmission redundancy, increasing the unit energy of the modulated constellation points, and improving the system's anti-noise performance;

[0026] The transmission mode sequence selected by the transmission mode selection bit sequence is an independent operation and does not depend on other data source bits. At the same time, the sparse subframe tail structure is conducive to reducing crosstalk between multiple users.

[0027] As an improvement of the present invention, the decoding method for recovering the source data includes (the signal recovery method for each subframe is the same):

[0028] Estimating the number of active users in a subframe based on transmitted signal energy and noise statistics;

[0029] Based on the multi-user spread spectrum sequence codeword signal superimposed on the subframe header, the spread spectrum sequence codeword activity detection and corresponding channel state information estimation are completed;

[0030] Based on the channel state information estimation, soft demodulation of time slot subframe state information is completed;

[0031] Based on the soft demodulation time slot subframe status information, complete the transmission mode active detection and corresponding constellation point soft demodulation;

[0032] Calculate bit likelihood soft information based on the transmission mode activity detection results and constellation point soft demodulation information;

[0033] Performing forward error correction decoding of channel coding based on the bit likelihood soft information;

[0034] Based on the result of the forward error correction decoding, performing a parity check on the decoded data sequence;

[0035] Based on the parity check result, serial interference elimination is performed, and the signal that has undergone the encoding method and transmission method again is removed from the received signal, and the above decoding method is repeated until there is no data sequence that passes the parity check.

[0036] The decoding method for recovering source data,

[0037] The recovery of the subframe selection bit sequence does not require additional computational complexity and energy consumption, and only requires the active estimation of the spread spectrum sequence codeword completed by the subframe header signal at the corresponding subframe position.

[0038] Recover the subframe selection bit sequence of the user at the corresponding subframe position;

[0039] The transmission mode selection bit sequence is recovered without additional signal, but through the time slot status

[0040] Soft information demodulator to estimate;

[0041] The recovery of the transmission mode selection bit sequence and the recovery of the transmission bit sequence are simultaneous.

[0042] The recovery of the bit sequence is accompanied by the estimation of the correct transmission mode;

[0043] The serial interference elimination in the recovery method takes into account the correction of channel information, further enhancing the system source bit data recovery capability

[0044] Specifically:

[0045] For the encoding method:

[0046] The total channel usage in the finite domain is represented as L, the amount of source bit data is B, and the number of active users per antenna is K a , the number of antennas at the receiving end is M.

[0047] The source bit data is divided into B ck ,B pn ,B p ,B c , respectively, selecting a bit sequence for the subframe, selecting a bit sequence for the preamble, selecting a bit sequence for the transmission mode, and sending a bit sequence. ck The bit data value selects the subframe position, so the total channel usage of the finite field needs to be divided into subframes, the subframe header is L p and tail length L c , if the subframe lengths are the same, then the constraint is satisfied

[0048]

[0049] Due to the randomness of subframe position selection, the number of active users at a single subframe position is given by Indicates; through the B pn Bit data selection J p transmission mode codewords, determine the subsequent arrangement; through the B pThe bit data value selects the spread spectrum sequence codeword, so the spread spectrum sequence codebook needs to have Spread spectrum sequence codewords, define the spread spectrum sequence codebook as

[0050]

[0051] The selected spreading sequence codeword is determined by Indicates; said B c Bits of data need to go through channel coding, constellation point symbol modulation and transmission mode arrangement. The channel coding rate is expressed as R c The constellation point symbol modulation order is represented by Q. Therefore, after channel coding and symbol modulation, the number of constellation point symbols that need to be arranged is

[0052]

[0053] According to the above encoding method, the transmission mode codebook can be defined as

[0054]

[0055] The total number of transmission mode codebooks is There are transmission mode codewords, each of which has only S non-zero elements. The position of the non-zero element represents the position of the constellation point symbol, and the position of the zero element represents no transmission signal. p The tail of the subframe after the transmission mode is arranged is Here is an example of how to arrange the constellation points according to the transmission mode codeword. If there is a transmission mode codeword [0,1,0,1,0,0] and a constellation point symbol [s1,s2], the following combinations can be arranged: [0,s1,0,s2,0,0] or

[0056] [0,s1,0,s2,0,0], that is, the order of symbols after arrangement can be flexibly set.

[0057] For the sending method:

[0058] The signal of a single subframe superposition can be expressed as The multi-user spread spectrum sequence codeword signal superposition signal (including additive Gaussian white noise) is: Frame header receiving signal

[0059] The body expression is

[0060]

[0061] in represents a user's spreading sequence codeword vector, is the channel coefficient vector of a certain user, and the additive Gaussian white noise in the frame header is Follow the mean 0 variance σ2 The cyclically symmetric complex Gaussian distribution of multiple users through multiple transmission modes is the superposition signal of the constellation point symbols (including additive Gaussian white noise) The specific expression of the received signal at the end of the frame is:

[0062]

[0063] in is the constellation point signal arranged in multiple transmission modes, and the additive white Gaussian noise at the end of the frame is Follow the mean 0 variance σ 2 The energy constraint of the signal transmission is determined by the bit signal-to-noise ratio (energy-per-bit), which is defined as

[0064]

[0065] The total signal transmission energy constraint of a single user is P, E s The energy usage per unit channel is defined, and the power allocation coefficient α is expressed as the ratio of the power allocated to a single user in the subframe header, that is, the single user subframe header signal satisfies the energy constraint The energy constraint of the subframe tail signal is

[0066] For the decoding method:

[0067] The processing method for each subframe is the same, and the following is a detailed description of the received signal of a single subframe.

[0068] Due to the randomness of subframe position selection, the receiver needs to estimate the number of active users in each subframe.

[0069] The number of active users is calculated based on the energy constraint of the transmitted signal and the noise statistics. Make rounding estimates

[0070]

[0071] in is the average received signal energy of a single antenna. Number of active users That is equivalent to the number of active spreading sequence codeword sequences.

[0072] After estimating the signal reception energy, the receiver needs to complete the estimation of the channel state information:

[0073] The subframe header receiving signal can be equivalent to

[0074] Y p =AH+Z p

[0075] The channel matrix H has only The rows are non-zero vectors, and the rest are zero vectors; the non-zero vectors of the channel matrix represent the channel coefficients of the active spreading sequence codewords; solving the channel coefficients can be equivalent to solving the compressed sensing problem. Many common methods can solve this type of problem. Here, according to the idea of ​​multi-antenna orthogonal matching pursuit (SOMP), the active detection of the spreading sequence codewords and the corresponding channel state information estimation are performed based on the pre-stored common spreading sequence codebook; in each loop, SOMP selects the measurement matrix column with the largest correlation with the current residual and adds its index to the output list, thus completing the active detection of the spreading sequence codewords; the detected active spreading sequence codewords are used to estimate the channel coefficients and update the residuals; the number of loops can be set to The residual R can be initialized to the subframe header received signal, and the spread spectrum sequence codeword correlation / activity detection is calculated as

[0076]

[0077] The sequence number of the active codeword is stored in In the middle; the orthogonal projection is constructed by determining the active codeword sequence, let Then the residual is updated as

[0078]

[0079] in is the Moore-Penrose generalized inverse, and the sequence numbers of all active codewords are obtained through repeated iterations. The channel estimation corresponding to the active spreading sequence codeword can be completed in different ways, such as the error criterion based on the minimum mean square error (MMSE) channel coefficient estimation is

[0080]

[0081] After obtaining the estimated channel coefficients, the receiving end performs soft demodulation on the state information of the subframe tail time slot: First, define the subframe time slot state for

[0082]

[0083] Where s0=0 is idle state, {s1,…,s Q} is the active state, i.e. the constellation point symbol after arrangement.

[0084] is the energy constraint for a single constellation point symbol.

[0085] No. 1 c The soft demodulation signal model of a time slot is:

[0086]

[0087] in Y c A column represents the received signal of a single time slot, and z is the corresponding additive Gaussian white noise. The soft demodulation method of the time slot status information is the same, and different time slots can be performed independently. The subsequent detailed description is for a single

[0088] Time slot signal processing. The time slot signal model after channel equalization can be obtained:

[0089]

[0090] in z can be equivalently represented as a matrix with zero mean and covariance Gaussian noise. State soft information demodulation aims to calculate the time slot state corresponding to each element in x The posterior probability of the state element in , Used to represent the element x to be estimated i ∈x and the posterior probability of each time slot state. To solve the element x i The state posterior probability, introducing the guidance vector The guidance vector has only one non-zero element 1 at the i-th position. After introducing the guidance vector, we can get

[0091]

[0092] The signal except the target element (Target) is regarded as pseudo noise Pseudo noise has a statistical mean matrix (Mean Matrix) E i , Covariance Matrix Ω i , Pseudo Covariance Matrix i , considering the interference as noise, the above statistics can be obtained by the following calculation

[0093]

[0094] Process variables ω j ,ξ j It can be obtained by the following calculation

[0095]

[0096] At the same time, the prior probability of each time slot state can be obtained through the encoding method and the sending method.

[0097]

[0098] After obtaining the above statistics, the fixed estimation element is a certain time slot state And calculate the signal residual in this state

[0099]

[0100] Based on this residual calculation, the likelihood value of the corresponding state can be calculated

[0101]

[0102] in Represent the real and imaginary parts of the complex number, and the process variable Λ i Noise statistics can be obtained

[0103]

[0104] After the above solution, the element x i The posterior probability of the time slot state can be calculated by the likelihood value

[0105]

[0106] The above time slot state soft information demodulation solution process is iterated repeatedly, divided into an inner loop and an outer loop. The inner loop calculates the element x i The posterior probability of each time slot state, the outer loop completes the update of the posterior probability of the time slot state of all elements of x through the update of pseudo noise.

[0107] Based on the soft information of all subframe time slot states, the receiver needs to complete the soft demodulation of the transmission mode codeword activity detection and the arrangement of constellation symbols:

[0108] Based on the encoding method and the sending method, the subframe tail has J p Active transmission modes need to be detected, and the single-slot state soft information Represents the posterior probability of the idle state, combined with the estimated soft information of all time slot states, to define the active probability matrix

[0109]

[0110] in Represents an element with r rows and c columns, Represents the idle state posterior probability of the rth element in the cth time slot. Active detection is performed on different segments of each row of the active probability matrix. The transmission mode sequence codeword with the maximum sum of active posterior probabilities will be determined as the active transmission mode sequence. The determination process can be equivalent to the inner product of the active probability matrix and the pre-stored transmission mode sequence codebook.

[0111]

[0112] Based on the transmission mode detection result and the estimated posterior probability of the constellation point symbol of the corresponding arrangement position, the bit likelihood soft information (LLR, Log-Likelihood Ratio) can be obtained. Taking Quadrature Phase Shift Keying (QPSK, Quadrature PhaseShift Keying) as an example, the modulation order is Q=4 and the time slot state is One constellation point symbol carries two bits of information, namely Based on this example, the bit likelihood soft information can be calculated

[0113]

[0114] Based on the calculated bit likelihood soft information, the receiver performs forward error correction decoding of the channel coding. Based on the results of the forward error correction channel coding soft decoding, the receiver performs successive interference cancellation (SIC):

[0115] The receiving end performs bit judgment and parity check on the forward error correction decoding result of the channel coding. The bit data that passes the parity check is re-encoded and modulated to form an elimination signal matrix, which is recorded as X SIC , before serial interference cancellation, the channel estimate is corrected using the cancellation signal matrix

[0116]

[0117] Where Y is the subframe received signal, H SIC is the corrected channel estimate, since X SIC The channel length is longer than the subframe header signal, and more observations will lead to relatively accurate channel estimation.

[0118] Y=YH SIC X SIC

[0119] The receiving end repeatedly performs the above solution operation until no bit data that can pass the parity check can be generated.

[0120] An electronic device comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the method for multiple access transmission with a shared codebook without user identification and approaching the theoretical capacity under a finite field is implemented.

[0121] A computer-readable storage medium stores computer instructions, wherein when the computer instructions are executed by a processor, the method for multiple access transmission without user identification and shared codebook approaching the theoretical capacity under a finite field is realized.

[0122] Compared with the existing technology, the main feature of the present invention is that the transmission energy efficiency is higher. Only a small part of the encoded signal consumes energy, and the recovery of most data bits does not require additional energy consumption. At the same time, the complexity of the present invention is lower than that of similar schemes. In addition, since there are fewer transmission symbols entering the channel, the higher level of unit transmission energy gives the present invention extremely strong anti-interference performance. Therefore, the random multiple access of the present invention under a finite field can achieve system performance close to the theoretical capacity.

[0123] The present invention has the beneficial effects:

[0124] 1. The transmission method of the present invention is simple and easy to construct, has high transmission efficiency, and has a small amount of signal transmission;

[0125] 2. The transmission method of the present invention has the characteristics of strong anti-interference ability. Since the amount of transmitted signal is small, the unit energy of the transmitted signal is high under the same transmission energy, and it has excellent ability to resist background noise and crosstalk between users;

[0126] 3. The transmission method of the present invention has the characteristics of low pilot signal collision;

[0127] 4. The transmission method of the present invention has the characteristics of relatively low complexity;

[0128] 5. The transmission scheme of the present invention has the characteristics of low power consumption and system capacity far exceeding that of similar schemes. It has the potential to reach the theoretical capacity under finite domain when equipped with a large-scale antenna system.

[0129] The transmission scheme of the present invention has the characteristics of low power consumption, and the system capacity is much higher than that of similar schemes. It has the potential to reach the theoretical capacity under finite domain when equipped with a large-scale antenna system. Under a fixed number of users, the present invention requires a lower bit signal-to-noise ratio to achieve a certain system error probability, such as Figure 6 As shown, the vertical axis represents the user bit signal-to-noise ratio required for the system error rate to not exceed 0.05 under a certain number of active users. When the number of antennas is 50, the energy consumption of the present invention is the lowest within 600 users, and when the number of antennas is 200

[0130] At the same time, the present invention is close to the theoretical limit, demonstrating its characteristics of low energy consumption and high capacity. BRIEF DESCRIPTION OF THE DRAWINGS

[0131] Figure 1 Schematic diagram of the encoding method and sending method of the present invention;

[0132] Figure 2 Schematic diagram of a receiving end of the decoding method of the present invention;

[0133] Figure 3 This is a comparison chart of the frame error rate caused by pilot signal collision in multi-subframe transmission of the present invention and in normal non-subframe transmission;

[0134] Figure 4 This is a comparison chart of the computational complexity of the present invention and similar solutions;

[0135] Figure 5 Comparison chart of the unit energy of the transmitted signal between the present invention and similar solutions;

[0136] Figure 6 This is a comparison chart of the system capacity of the present invention and similar solutions. DETAILED DESCRIPTION

[0137] The present invention will be further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, modifications of various equivalent forms of the present invention made by those skilled in the art all fall within the scope defined by the claims attached to the present invention.

[0138] Example 1: See Figure 1 、 Figure 2 , a user-identified shared codebook multiple access transmission method (URA, Unsourced Random Access) approaching the theoretical capacity under a finite field, a source bit data encoding method, a sending method, and a decoding method for user-identified shared codebook multiple access transmission;

[0139] 1) The encoding method includes:

[0140] Source bit data and its division and encoding method;

[0141] Pre-stored spreading sequence codebook and transmission mode codebook;

[0142] Divide and encode the source bit data into a subframe selection bit sequence, a preamble selection bit sequence, a transmission mode selection bit sequence, and a transmission bit sequence, and complete corresponding encoding respectively;

[0143] The user selects the subframe position for sending information through the subframe selection bit sequence;

[0144] Through the preamble selection bit sequence, the user selects the corresponding spreading sequence codeword from the pre-stored spreading sequence codebook;

[0145] The user selects a plurality of transmission modes from a pre-stored transmission mode codebook through the transmission mode selection bit sequence;

[0146] The transmission bit sequence is channel coded and constellation modulated, and is arranged in a plurality of transmission modes selected by the transmission mode selection bit sequence. After the arrangement is completed, channel transmission is completed according to a preset transmission method;

[0147] 2) The sending method includes:

[0148] The channel with fixed channel usage under finite field is divided into multiple subframe positions;

[0149] The selected spreading sequence codeword is located at the head of the subframe, and together with the constellation point symbol arranged in multiple transmission modes at the tail of the subframe, constitutes a complete subframe;

[0150] completing signal transmission based on the selected subframe position;

[0151] Based on the subframe selection bit sequence, only signals of some users are superimposed at the subframe position in each subframe;

[0152] The receiving end receives all subframe signals in the finite field and recovers the source data for each subframe based on the set decoding method.

[0153] 3) The decoding method for recovering the source data includes (the signal recovery method for each subframe is the same):

[0154] Estimating the number of active users in a subframe based on transmitted signal energy and noise statistics;

[0155] Based on the multi-user spread spectrum sequence codeword signal superimposed on the subframe header, the spread spectrum sequence codeword activity detection and corresponding channel state information estimation are completed;

[0156] Based on the channel state information estimation, soft demodulation of time slot subframe state information is completed;

[0157] Based on the soft demodulation time slot subframe status information, complete the transmission mode active detection and corresponding constellation point soft demodulation;

[0158] Calculate bit likelihood soft information based on the transmission mode activity detection results and constellation point soft demodulation information;

[0159] Performing forward error correction decoding of channel coding based on the bit likelihood soft information;

[0160] Based on the result of the forward error correction decoding, performing a parity check on the decoded data sequence;

[0161] Based on the parity check result, serial interference elimination is performed, and the signal that has undergone the encoding method and transmission method again is removed from the received signal, and the above decoding method is repeated until there is no data sequence that passes the parity check.

[0162] Specifically:

[0163] For the encoding method:

[0164] The total channel usage in the finite domain is represented as L, the amount of source bit data is B, and the number of active users per antenna is K a , the number of antennas at the receiving end is M.

[0165] The source bit data is divided into B ck ,B pn ,B p ,B c , respectively, selecting a bit sequence for the subframe, selecting a bit sequence for the preamble, selecting a bit sequence for the transmission mode, and sending a bit sequence. ck The bit data value selects the subframe position, so the total channel usage of the finite field needs to be divided into subframes, the subframe header is L p and tail length L c , if the subframe lengths are the same, then the constraint is satisfied

[0166]

[0167] Due to the randomness of subframe position selection, the number of active users at a single subframe position is given by Indicates; through the B pn Bit data selection J p transmission mode codewords, determine the subsequent arrangement; through the B p The bit data value selects the spread spectrum sequence codeword, so the spread spectrum sequence codebook needs to have Spread spectrum sequence codewords, define the spread spectrum sequence codebook as

[0168]

[0169] The selected spreading sequence codeword is determined by Indicates; said B c Bits of data need to go through channel coding, constellation point symbol modulation and transmission mode arrangement. The channel coding rate is expressed as R c The constellation point symbol modulation order is represented by Q. Therefore, after channel coding and symbol modulation, the number of constellation point symbols that need to be arranged is

[0170]

[0171] According to the above encoding method, the transmission mode codebook can be defined as

[0172]

[0173] The total number of transmission mode codebooks is There are transmission mode codewords, each of which has only S non-zero elements. The position of the non-zero element represents the position of the constellation point symbol, and the position of the zero element represents no transmission signal. p The tail of the subframe after the transmission mode is arranged is Here is an example of how to arrange the constellation points according to the transmission mode codeword. If there is a transmission mode codeword [0,1,0,1,0,0] and a constellation point symbol [s1,s2], the arrangement can be the following combinations [0,s1,0,s2,0,0] or [0,s1,0,s2,0,0], that is, the order of the symbols after arrangement can be flexibly set.

[0174] For the sending method:

[0175] The signal of a single subframe superposition can be expressed as The multi-user spread spectrum sequence codeword signal superposition signal (including additive Gaussian white noise) is: The specific expression of the frame header receiving signal is:

[0176]

[0177] in represents a user's spreading sequence codeword vector, is the channel coefficient vector of a certain user, and the additive Gaussian white noise in the frame header is Follow the mean 0 variance σ 2 Cyclic symmetric complex Gaussian distribution; Multiple users through multiple transmission modes arranged constellation point symbol superposition signal

[0178] (including additive white Gaussian noise) is The specific expression of the received signal at the end of the frame is:

[0179]

[0180] in is the constellation point signal arranged in multiple transmission modes, and the additive white Gaussian noise at the end of the frame is Follow the mean 0 variance σ 2 The energy constraint of the signal transmission is determined by the bit signal-to-noise ratio (energy-per-bit), which is defined as

[0181]

[0182] The total signal transmission energy constraint of a single user is P, E s The energy usage per unit channel is defined, and the power allocation coefficient α is expressed as the ratio of the power allocated to a single user in the subframe header, that is, the single user subframe header signal satisfies the energy constraint The energy constraint of the subframe tail signal is

[0183] For the decoding method:

[0184] The processing method for each subframe is the same, and the following is a detailed description of the received signal of a single subframe.

[0185] Due to the randomness of subframe position selection, the receiver needs to estimate the number of active users in each subframe.

[0186] The number of active users is calculated based on the energy constraint of the transmitted signal and the noise statistics. Make rounding estimates

[0187]

[0188] in is the average received signal energy of a single antenna. Number of active users That is equivalent to the number of active spreading sequence codeword sequences.

[0189] After estimating the signal reception energy, the receiver needs to complete the estimation of the channel state information:

[0190] The subframe header receiving signal can be equivalent to

[0191] Y p =AH+Z p

[0192] The channel matrix H has only The rows are non-zero vectors, and the rest are zero vectors. The non-zero vectors of the channel matrix represent the channel coefficients of the active spreading sequence codewords. Solving the channel coefficients is equivalent to solving the compressed sensing problem, and many common methods can solve this type of problem. Here, according to the idea of ​​Simultaneous Orthogonal Matching Pursuit (SOMP), active detection of spreading sequence codewords and estimation of their corresponding channel state information are performed based on the pre-stored common spreading sequence codebook. In each loop, SOMP selects the measurement matrix column with the greatest correlation with the current residual and adds its index to the output list, thus completing the active detection of the spreading sequence codeword.

[0193] The detected active spreading sequence codeword is used to estimate the channel coefficient and update the residual; the number of cycles can be set to The residual R can be initialized to the subframe header received signal, and the spread spectrum sequence codeword correlation / activity detection is calculated as

[0194]

[0195] The sequence number of the active codeword is stored in In the middle; the orthogonal projection is constructed by determining the active codeword sequence, let Then the residual is updated as

[0196]

[0197] in is the Moore-Penrose generalized inverse, and the sequence numbers of all active codewords are obtained through repeated iterations. The channel estimation corresponding to the active spreading sequence codeword can be completed in different ways, such as the error criterion based on the minimum mean square error (MMSE) channel coefficient estimation is

[0198]

[0199] After obtaining the estimated channel coefficients, the receiver performs soft demodulation of the state information of the tail time slot of the subframe:

[0200] First, define the subframe time slot status for

[0201]

[0202] Where s0=0 is idle state, {s1,…,s Q} is the active state, i.e. the constellation point symbol after arrangement.

[0203] is the energy constraint for a single constellation point symbol.

[0204] No. 1 c The soft demodulation signal model of a time slot is:

[0205]

[0206] in Y c A column represents the received signal of a single time slot, and z is the corresponding additive Gaussian white noise. The soft demodulation method of the time slot status information is the same, and different time slots can be performed independently. The subsequent detailed description is for a single

[0207] Time slot signal processing. The time slot signal model after channel equalization can be obtained:

[0208]

[0209] in z can be equivalently represented as a matrix with zero mean and covariance Gaussian noise. State soft information demodulation aims to calculate the time slot state corresponding to each element in x The posterior probability of the state element in , Used to represent the element x to be estimated i ∈x and the posterior probability of each time slot state. To solve the element x i The state posterior probability, introducing the guidance vector The guidance vector has only one non-zero element 1 at the i-th position. After introducing the guidance vector, we can get

[0210]

[0211] The signal except the target element (Target) is regarded as pseudo noise Pseudo noise has a statistical mean matrix (Mean Matrix) E i , Covariance Matrix Ω i , Pseudo Covariance Matrix i , considering the interference as noise, the above statistics can be obtained by the following calculation

[0212]

[0213] Process variables ω i ,ξ j It can be obtained by the following calculation

[0214]

[0215] At the same time, the prior probability of each time slot state can be obtained through the encoding method and the sending method.

[0216]

[0217] After obtaining the above statistics, the fixed estimation element is a certain time slot state And calculate the signal residual in this state

[0218]

[0219] Based on this residual calculation, the likelihood value of the corresponding state can be calculated

[0220]

[0221] in Represent the real and imaginary parts of the complex number, and the process variable Λ i Noise statistics can be obtained

[0222]

[0223] After the above solution, the element x i The posterior probability of the time slot state can be calculated by the likelihood value

[0224]

[0225] The above time slot state soft information demodulation solution process is iterated repeatedly, divided into an inner loop and an outer loop. The inner loop calculates the element x i The posterior probability of each time slot state, the outer loop completes the update of the posterior probability of the time slot state of all elements of x through the update of pseudo noise.

[0226] Based on the soft information of all subframe time slot states, the receiver needs to complete the soft demodulation of the transmission mode codeword activity detection and the arrangement of constellation symbols:

[0227] Based on the encoding method and the sending method, the subframe tail has J p Active transmission modes need to be detected, and the single-slot state soft information Represents the posterior probability of the idle state, combined with the estimated soft information of all time slot states, to define the active probability matrix

[0228]

[0229]

[0230] in Represents an element with r rows and c columns, Represents the idle state posterior probability of the rth element in the cth time slot. Active detection is performed on different segments of each row of the active probability matrix. The transmission mode sequence codeword with the maximum sum of active posterior probabilities will be determined as the active transmission mode sequence. The determination process can be equivalent to the inner product of the active probability matrix and the pre-stored transmission mode sequence codebook.

[0231]

[0232] Based on the transmission mode detection result and the estimated posterior probability of the constellation point symbol of the corresponding arrangement position, the bit likelihood soft information (LLR, Log-Likelihood Ratio) can be obtained. Taking Quadrature Phase Shift Keying (QPSK, Quadrature PhaseShift Keying) as an example, the modulation order is Q=4 and the time slot state is One constellation point symbol carries two bits of information, namely Based on this example, the bit likelihood soft information can be calculated

[0233]

[0234] Based on the calculated bit likelihood soft information, the receiver performs forward error correction decoding of the channel coding. Based on the results of the forward error correction channel coding soft decoding, the receiver performs successive interference cancellation (SIC):

[0235] The receiving end performs bit judgment and parity check on the forward error correction decoding result of the channel coding. The bit data that passes the parity check is re-encoded and modulated to form an elimination signal matrix, which is recorded as X SIC , before serial interference cancellation, the channel estimate is corrected using the cancellation signal matrix

[0236]

[0237] Where Y is the subframe received signal, H SIC is the corrected channel estimate, since X SIC The channel length is longer than the subframe header signal, and more observations will lead to relatively accurate channel estimation.

[0238] Y=YH SIC X SIC

[0239] The receiving end repeatedly performs the above solution operation until no bit data that can pass the parity check can be generated.

[0240] Example 2: The channel usage of the finite domain can be flexibly adjusted. Taking a single carrier of 2GHZ as an example, the coherence time range is 1ms to 45ms (depending on the moving speed of the device transmitter 3km / h-120km / h). At the same time, in an outdoor environment, when matching the coherence bandwidth, the sampling frequency of the receiving device is usually between 100kHZ-500kHZ. Therefore, the channel usage of the finite domain can be flexibly set in the range of 100-20,000. The consistency of spectrum efficiency must be maintained when comparing solutions. The frame error rate (including the false alarm rate and the missed detection rate) is usually used as a performance evaluation indicator for a shared codebook multiple access system without user identification.

[0241] Number of active users: 100, 500, 1000;

[0242] Multi-subframe transmission B ck =4, with no subframe transmission;

[0243] Figure 3This is a comparison chart of the frame error rate caused by pilot signal collision in multi-subframe transmission of the present invention and in normal subframe-free transmission; the comparison results show that multi-subframe transmission can significantly reduce the frame error rate caused by pilot signal collision, reducing the collision frame error rate by about 100 to 1000 times.

[0244] Example 3:

[0245] The amount of source bit data is B = 470;

[0246] The channel usage is L=15040, and the spectrum efficiency is consistent with the theoretical boundary of the comparison scheme, both of which are 0.0312;

[0247]

[0248] The source bit data is divided into B p =14,B c =36,B ck =4,B pn =416;

[0249] The number of subframes is The subframe length is L / J=940, and the subframe header length and tail length are L p =172,L c =768;

[0250] The energy distribution coefficient is set to α = 0.1;

[0251] The spreading sequence codebook size is Generated by an undersampled discrete Fourier transform matrix (DFT), which complies with energy constraints;

[0252] The channel coding uses the 5GNR Polar code with a length of 256 and a 24-bit Cyclic Redundancy Check (CRC). The 24-CRC generator polynomial is x 24 +x 23 +x 21 +x 20 +x 17 +x 15 +x 13 +x 12 +x 8 +x 4 +x 2 +x+1;

[0253] The constellation point modulation adopts quadrature phase shift keying QPSK;

[0254] The number of transmission modes is J p =32, that is, every Bpn / J p = 13 bits to select a transmission mode, a total of 32, the transmission mode codebook size is The transmission mode codeword length is L c / J p =24, the number of non-zero elements in the transmitted codeword is 4;

[0255] Figure 4 This is a comparison chart of the computational complexity of the present invention and similar solutions; the results show that the present invention has a relatively low level of system computational complexity among similar solutions;

[0256] Figure 5 This is a comparison chart of the unit energy of transmitted signals between the present invention and similar solutions. The unit energy of transmitted signals is defined as the single-user transmission energy constraint divided by the number of transmitted units, i.e., (transmitted energy) / (number of preamble selection bits + number of transmitted constellation point symbols). This result shows that the present invention has higher unit energy of transmitted signals than similar solutions, and therefore has higher noise and interference immunity under the same energy constraint.

[0257] Figure 6 This is a comparison chart of the system capacity of the present invention and similar solutions, comparing the bit signal-to-noise ratio (in decibels dB) required to achieve a system frame error rate of 0.05 under different total active user numbers. The results show that when the receiving end is equipped with 50 antennas, the solution of the present invention can obtain the theoretical achievable capacity under a total number of active users of 100 to 400; when the receiving end is equipped with 200 antennas, the solution of the present invention can basically maintain the theoretical achievable capacity under a total number of active users of 100 to 1000. It proves that the present invention has low power consumption and high capacity system performance compared with similar solutions, and has the potential to approach the theoretical capacity under a finite domain. It should be noted that the above embodiments are not used to limit the scope of protection of the present invention, and equivalent transformations or substitutions made on the basis of the above technical solutions fall within the scope of protection of the claims of the present invention.

Claims

1. A method for multiple access transmission with a shared codebook without user identification that approaches the theoretical capacity under a finite field, characterized in that: Source bit data encoding method, sending method, and decoding method for multiple access transmission without user identification and shared code book; The encoding method comprises: Source bit data and its division and encoding method; Pre-stored spreading sequence codebook and transmission mode codebook; Divide and encode the source bit data into a subframe selection bit sequence, a preamble selection bit sequence, a transmission mode selection bit sequence, and a transmission bit sequence, and encode each bit sequence separately; The user selects the subframe position for sending information through the subframe selection bit sequence; Through the preamble selection bit sequence, the user selects the corresponding spreading sequence codeword from the pre-stored spreading sequence codebook; The user selects a plurality of transmission modes from a pre-stored transmission mode codebook through the transmission mode selection bit sequence; The transmission bit sequence is channel coded and constellation modulated, and is arranged in a plurality of transmission modes selected by the transmission mode selection bit sequence. After the arrangement is completed, channel transmission is completed according to a preset transmission method; The sending method includes: The channel with fixed channel usage under finite field is divided into multiple subframe positions; The selected spreading sequence codeword is located at the head of the subframe, and together with the constellation point symbol arranged in multiple transmission modes at the tail of the subframe, constitutes a complete subframe; completing signal transmission based on the selected subframe position; Based on the subframe selection bit sequence, only signals of some users are superimposed at the subframe position in each subframe; The receiving end receives all subframe signals in the finite field and recovers the source data for each subframe based on the set decoding method. The decoding methods for source data recovery include: Estimating the number of active users in a subframe based on transmitted signal energy and noise statistics; Based on the multi-user spread spectrum sequence codeword signal superimposed on the subframe header, the spread spectrum sequence codeword activity detection and corresponding channel state information estimation are completed; Based on the channel state information estimation, soft demodulation of time slot subframe state information is completed; Based on the soft demodulation time slot subframe status information, complete the transmission mode active detection and corresponding constellation point soft demodulation; Calculate bit likelihood soft information based on the transmission mode activity detection results and constellation point soft demodulation information; Performing forward error correction decoding of channel coding based on the bit likelihood soft information; Based on the result of the forward error correction decoding, performing a parity check on the decoded data sequence; Based on the result of the parity check, serial interference cancellation is performed, and the signal that has undergone the encoding method and the sending method again is removed from the received signal, and the above decoding method is repeated until there is no data sequence that passes the parity check.

2. The method for multiple access transmission without user identification and shared codebook approaching theoretical capacity in a finite field according to claim 1, characterized in that: In the encoding method, The total channel usage in the finite domain is represented as L, the amount of source bit data is B, and the number of active users per antenna is K a , the number of antennas at the receiving end is M, The source bit data is divided into B ck ,B p ,B pn ,B c , respectively, the subframe selection bit sequence, the preamble selection bit sequence, the transmission mode selection bit sequence, the sending bit sequence, through B ck The bit data value selects the subframe position, so the total channel usage of the finite field needs to be divided into subframes, the subframe header is L p and tail length L c , if the subframe lengths are the same, then the constraint is satisfied Due to the randomness of subframe position selection, the number of active users at a single subframe position is given by Indicates; through the B pn Bit data selection J p transmission mode codewords, determine the subsequent arrangement; through the B p The bit data value selects the spread spectrum sequence codeword, so the spread spectrum sequence codebook needs to have Spread spectrum sequence codewords, define the spread spectrum sequence codebook as The selected spreading sequence codeword is determined by Indicates; said B c Bits of data need to go through channel coding, constellation point symbol modulation and transmission mode arrangement. The channel coding rate is expressed as R c The constellation point symbol modulation order is represented by Q. Therefore, after channel coding and symbol modulation, the number of constellation point symbols that need to be arranged is According to the above encoding method, the transmission mode codebook is defined as The transmission mode codebook has a total of There are transmission mode codewords, each of which has only S non-zero elements. The position of the non-zero element represents the position of the constellation point symbol, and the position of the zero element represents no transmission signal. p The tail of the subframe after the transmission mode is arranged is express.

3. The method for multiple access transmission with shared codebook without user identification and approaching theoretical capacity in a finite field according to claim 2, characterized in that: In the sending method, The signal of a single subframe superposition is represented as The multi-user spread spectrum sequence codeword signal superposition signal is The specific expression is in represents a user's spreading sequence codeword vector, is the channel coefficient vector of a certain user, and the additive Gaussian white noise in the frame header is Follow the mean 0 variance σ 2 The cyclically symmetric complex Gaussian distribution of multiple users is: The specific expression is in is the constellation point signal arranged in multiple transmission modes, and the additive white Gaussian noise at the end of the frame is Follow the mean 0 variance σ 2 The cyclically symmetric complex Gaussian distribution of the signal is determined by the bit signal-to-noise ratio (energy-per-bit), which is defined as The total signal transmission energy constraint of a single user is P, E s The energy usage per unit channel is defined, and the power allocation coefficient α is expressed as the ratio of the power allocated to a single user in the subframe header, that is, the single user subframe header signal satisfies the energy constraint The energy constraint of the subframe tail signal is 4. The method for multiple access transmission with shared codebook without user identification and approaching theoretical capacity in a finite field according to claim 3, characterized in that: In the decoding method, Due to the randomness of subframe position selection, the receiving end needs to estimate the number of active users in each subframe, and estimate the number of active users based on the energy constraint of the transmitted signal and the noise statistics. Make rounding estimates in is the average received signal energy of a single antenna, the number of active users That is equivalent to the number of active spreading sequence codeword sequences, After estimating the signal reception energy, the receiver needs to complete the estimation of the channel state information: The subframe header received signal is equivalent to Y p =AH+Z p The channel matrix H has only The rows are non-zero vectors, and the rest are zero vectors; the non-zero vectors of the channel matrix represent the channel coefficients of the active spreading sequence codewords; In each loop, the synchronous orthogonal matching pursuit (SOMP) selects the measurement matrix column with the largest correlation with the current residual and adds its index to the output list, thus completing the active detection of the spreading sequence codeword. The detected active spreading sequence codeword is used to estimate the channel coefficient and update the residual. The number of loops is set to The residual R is initialized to the subframe header received signal, and the spread spectrum sequence codeword correlation / activity detection is calculated as The sequence number of the active codeword is stored in In the middle; the orthogonal projection is constructed by determining the active codeword sequence, let Then the residual is updated as in is the Moore-Penrose generalized inverse, and the sequence numbers of all active codewords are obtained through repeated iterations. The channel coefficient is estimated based on the minimum mean square error (MMSE) error criterion as follows: After obtaining the estimated channel coefficients, the receiver performs soft demodulation of the state information of the tail time slot of the subframe: First, define the subframe time slot status for Where s0=0 is idle state, {s1,…,s Q } is the active state, i.e. the constellation point symbol after arrangement. is the energy constraint of the single constellation point symbol, No. 1 c The soft demodulation signal model of a time slot is: in Y c A column represents the received signal of a single time slot, z is the corresponding additive white Gaussian noise, the time slot state information soft demodulation method is the same, different time slots can be performed independently, the time slot signal model after channel equalization is: in z is equivalent to a matrix with zero mean and covariance Gaussian noise, state soft information demodulation aims to calculate the time slot state corresponding to each element in x The posterior probability of the state element in , Used to represent the element x to be estimated i ∈x and the posterior probability of each time slot state, to solve the element x i The state posterior probability, introducing the guidance vector The guidance vector has only one non-zero element 1 at the i-th position. After introducing the guidance vector, we get The signal except the target element (Target) is regarded as pseudo noise Pseudo noise has a statistical mean matrix (Mean Matrix) E i , Covariance Matrix Ω i , Pseudo Covariance Matrix i , considering the interference as noise, the above statistics can be obtained by the following calculation Process variables ω j ,ξ j Obtained by the following calculation At the same time, the prior probability of each time slot state is obtained through the encoding method and the sending method. After obtaining the above statistics, the fixed estimation element is a certain time slot state And calculate the signal residual in this state Calculate the likelihood value of the corresponding state based on the residual calculation in Represent the real and imaginary parts of the complex number, and the process variable Λ i Obtained from noise statistics After the above solution, the element x i The posterior probability of the time slot state is calculated by the likelihood value The above time slot state soft information demodulation solution process is iterated repeatedly, divided into an inner loop and an outer loop. The inner loop calculates the element x i The posterior probability of each time slot state, the outer loop completes the update of the posterior probability of the time slot state of all elements of x through the update of pseudo noise, Based on the soft information of all subframe time slot states, the receiver needs to complete the soft demodulation of the transmission mode codeword activity detection and the arrangement of constellation symbols: Based on the encoding method and the transmission method, the subframe tail has J p Active transmission modes need to be detected, and the single-slot state soft information Represents the posterior probability of the idle state, combined with the estimated soft information of all time slot states, to define the active probability matrix in Represents an element with r rows and c columns, Represents the idle state posterior probability of the rth element in the cth time slot. Active detection is performed on different segments of each row of the active probability matrix. The transmission mode sequence codeword with the maximum sum of active posterior probabilities will be determined as the active transmission mode sequence. The determination process is equivalent to the inner product of the active probability matrix and the pre-stored transmission mode sequence codebook. Based on the transmission mode detection result and the posterior probability of the constellation point symbol estimated at the corresponding arrangement position, the bit likelihood soft information (LLR, Log-Likelihood Ratio) is obtained, the quadrature phase shift keying (QPSK, Quadrature Phase ShiftKeying), the modulation order is Q=4, and the time slot state is One constellation point symbol carries two bits of information, namely Calculate bit likelihood soft information Based on the calculated bit likelihood soft information, the receiver performs forward error correction decoding of the channel coding. Based on the results of the forward error correction channel coding soft decoding, the receiver performs successive interference cancellation (SIC): The receiving end performs bit judgment and parity check on the forward error correction decoding result of the channel coding. The bit data that passes the parity check is re-encoded and modulated to form an elimination signal matrix, which is recorded as X SIC , before serial interference cancellation, the channel estimate is corrected using the cancellation signal matrix Where Y is the subframe received signal, H SIC is the corrected channel estimate, since X SIC The channel length is longer than the subframe header signal. More observations will lead to relatively accurate channel estimation. The receiving end performs serial interference. Y=Y-H SIC X SIC The receiving end repeats the above solution operation until no more bits can be generated that pass the parity check. according to.

5. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the shared codebook multiple access transmission method without user identification and approaching the theoretical capacity under a finite field as described in any one of claims 1 to 4 is implemented.

6. A computer-readable storage medium having computer instructions stored thereon, characterized in that: When the computer instructions are executed by a processor, the method for multiple access transmission with a shared codebook and no user identification approaching the theoretical capacity in a finite field according to any one of claims 1 to 4 is implemented.

Images