A pilot-free enhanced Slotted ALOHA large-scale random access and transmission method based on differential phase modulation

Abstract

The present invention proposes a pilot-free enhanced Slotted ALOHA large-scale random access and transmission method based on differential phase modulation, comprising the following steps: Step 1: differential coding modulation and frame structure; Step 2: sending a signal through a wireless channel to a base station; Step 3: receiving end signal detection, including: user preamble detection, user signal detection, and interference elimination using solved information; Step 4: simulation verification and performance evaluation. By strengthening the Slotted ALOHA transmission protocol, the present invention effectively solves the user competition problem in the two-step random access process, improves the success rate of the two-step random access, and thereby reduces the system delay in random access in large-scale communication scenarios. In addition, the present invention saves pilot overhead through differential phase modulation and a differential phase modulation detection algorithm that combines non-coherent detection with iterative detection, and can achieve the same access throughput as the traditional pilot scheme in the absence of a pilot, thereby improving the spectrum efficiency of the system.

Description

Technical field

[0001] To support user access and transmission in large-scale communication scenarios for next-generation mobile communication systems, the present invention provides a pilot-free, enhanced Slotted ALOHA large-scale random access and transmission scheme based on differential phase shift keying (DPSK). This scheme significantly improves user random access and transmission efficiency in large-scale scenarios. This invention belongs to the field of communications and signal processing. [Background Technology]

[0002] The white paper "6G Typical Scenarios and Key Capabilities" points out that ultra-large-scale connections in 6G will expand new capabilities and applications based on the massive Internet of Things in 5G. Figure 1 As shown in Figure 1, ultra-large-scale connections will be used in a variety of fields, including smart cities, smart manufacturing, and satellite internet. The white paper states that 6G will be able to achieve a connection density of 10 to 100 users per square meter [1].

[0003] In traditional communication networks, users need to complete a four-step random access process (Four-step RACH) to access the network. The user and the base station perform four steps: sending a preamble, feeding back a random access response, initiating a random access request, and feeding back contention resolution information. In large-scale connection scenarios, the four-step random access process will result in higher system delays and larger signaling overheads. For example, in satellite Internet systems, due to the large distance between the satellite and the ground, the satellite-to-ground link delay is high, and the two-way link delay of low-orbit satellites is 25ms-75ms. The four-step random access process affects user access efficiency. The 5G-NR protocol proposes a two-step random access process (Two-step RACH). The user and the base station perform two steps: sending a preamble and data, and feeding back a random access response [2]. The two-step random access process can reduce the number of signaling interactions and reduce system delay, but it increases the difficulty of user contention resolution. If the user's two-step random access process fails, the four-step random access process will be initiated, which will cause a larger system delay.

[0004] In order to improve the success rate of two-step random access and reduce the delay of the access system, the present invention adopts the enhanced Slotted ALOHA protocol to solve the problem of user contention in the two-step random access process [3]. In the traditional enhanced Slotted ALOHA protocol, the time axis is divided into several time slots of fixed length. The user copies the data sent by random access and generates several copies. The user randomly selects a time slot to send a data copy. The user inserts the location information pointing to the time slot where the other copies of the user are located in each generated copy. At the receiving end, the time slot with two or more user data copies is called the user contention time slot. The receiving end solves the data information in the non-user contention time slot and recovers the time slot location information where the other copies of the user are located. Based on the recovered time slot location information, the receiving end performs interference elimination in the corresponding time slot. The above operation can convert the user contention time slot into the non-user contention time slot through interference elimination, thereby improving the access throughput of the system. Studies have shown that the enhanced Slotted ALOHA protocol can achieve significant access throughput gains compared to the traditional ALOHA protocol. The traditional enhanced Slotted ALOHA protocol requires the use of channel state information for interference elimination. In the pilot-based channel estimation scheme, due to factors such as the channel coherence time and the number of users, the system cannot allocate orthogonal pilot sequences to users, and the channel estimation algorithm is complex [4], [5]. In addition, the pilot will also occupy a large amount of transmission resources, affecting the system spectrum efficiency. In order to reduce the pilot overhead and improve the spectrum efficiency,

[0005] This invention uses differential phase modulation to achieve signal detection without a pilot signal. It also recovers channel information from user data and performs interference cancellation, enabling enhanced Slotted ALOHA user access and transmission without a pilot signal. The proposed solution addresses user contention in two-step random access, improves the success rate of two-step random access, and reduces access latency in large-scale connection scenarios. Furthermore, the invention reduces pilot signal overhead, improves spectrum efficiency, and enhances user access and transmission efficiency.

[0006] References

[0007] [1] IMT2030 (6G) Promotion Group. 6G Overall Vision and Key Capabilities White Paper [R]. 2021.

[0008] [2] Wang Yingmin, Sun Shaohui. Detailed explanation of 5G mobile communication system design and standards[R]. 2020.

[0009] [3]Casini E, Gaudenzi RD, Herrero O R.Contention Resolution DiversitySlotted ALOHA(CRDSA): An Enhanced Random Access Scheme for Satellite AccessPacket Networks[J]. IEEE Transactions on Wireless Communications, 2007, 6(4): 1408-1419.

[0010] [4]L.Liu and W.Yu.Massive connectivity with massive MIMO-Part I:Device activity detection and channel estimation[J].IEEE Transactions onSignal Processing,2018,66(11):2933-2946.

[0011] [5]Cheng Y, Liu L, Ping L. Orthogonal AMP for Massive Access in Channels with Spatial and Temporal Correlations[J]. IEEE Journal on Selected Areas in Communications, 2021, 39(3):726-740. [Summary of the invention]

[0012] (1) Purpose of the present invention

[0013] For random access and transmission in large-scale connection scenarios, this paper designs a pilot-free enhanced Slotted ALOHA large-scale random access and transmission scheme based on differential phase modulation. This scheme improves the efficiency of resolving user contention in the two-step random access process and reduces access latency. Furthermore, this scheme addresses the high pilot overhead of the traditional enhanced Slotted ALOHA protocol, improving spectrum efficiency. This proposed scheme can improve user random access and transmission efficiency in large-scale communication scenarios.

[0014] (2) Technical solution

[0015] Step 1: Differential Coded Modulation and Frame Structure

[0016] The modulation transmission and reception overall block diagram of the present invention is as follows Figure 2 The modulation and transmission block diagram of the present invention is shown in FIG. Figure 3 shown.

[0017] Consider a large-scale communication system with a single base station and M users. Both the user and the base station use a single antenna for transmission and reception. Consider the number of slots in the enhanced Slotted ALOHA protocol to be T.

[0018] The base station transmits synchronization blocks to users through beam scanning. Users decode the synchronization blocks to achieve downlink synchronization and obtain random access configuration information. When a user initiates a random access request, it first sends a preamble codeword over the physical random access channel, followed by the random access data payload over the physical uplink shared channel.

[0019] For the preamble codeword part sent by the user, the user selects the preamble codeword according to the preamble code configuration information indicated by the base station. The purpose of the preamble codeword is to obtain timing information and complete uplink time synchronization.

[0020] For the data payload sent by the user, consider that the message sequences of all users are of equal length. The message sequence sent by user m (m=1...M) is b m ∈{0,1} k , where k represents the length of the message sequence. m Make a copy and generate d m duplicate message sequence Where d is the index of the replica. User m randomly selects time slots on the time axis, which correspond to The sending time slot of is the number of the randomly selected time slot. Add the time slot position information of the remaining copies to generate The length of the message sequence after adding the time slot position information is k'. The code rates used are R c The encoder is encoded to obtain the encoded sequence The length of the coded sequence is L. After the random interleaver π, the coded sequence after the interleaver is obtained. The user will Mapping to symbol sequence

[0021] N represents the length of the symbol sequence, α q It represents the qth symbol in the constellation point set, q represents the index of the constellation symbol, Q represents the modulation order, and j is the imaginary unit.

[0022] User m pairs of symbol sequences Perform differential phase modulation,

[0023] s m,d [n]=s m,d [n-1]x m,d [n] (1)

[0024] Get user m in the tth m,d The symbol sequence s is sent in the time slot m,d , where n represents discrete time. m,d [n] represents x m,d The nth symbol in s m,d [n] means s m,d The nth symbol in s m,d [0] is determined by the modulation order Q. For example, when the modulation order Q = 4, you can choose

[0025] therefore

[0026] Among them, β q Represents the constellation point of the differential signal.

[0027] User m selects a time slot Send symbol sequence Send, each time slot length is N T symbol duration, N T =N preamble +N payload +N Guard Among them, N preamble Indicates the length of the preamble sequence, N payload =N+1 represents the length of the data payload, N Guard Indicates the length of the guard interval.

[0028] Step 2: Send the signal through the wireless channel to the base station

[0029] The transmitted signal reaches the base station through a wireless channel. Consider a multipath channel between the base station and the user, with P paths. Path 1 has the highest average channel gain, exceeding the remaining paths by more than 20 dB. When considering a channel environment with numerous reflectors, the channel model degenerates to a single-path channel. Doppler shift exists between the base station and the user. For any user, assume that its channel coefficient remains constant within a single time slot and that the Doppler shift remains constant over T time slots.

[0030] The signal received by the base station in the tth time slot is

[0031]

[0032] Among them, p represents the path index, express is the channel coefficient of the pth path between user m and the base station in the tth time slot, represents the normalized delay of the pth path between user m and the base station, represents the normalized Doppler shift between user m and the base station, represents the actual Doppler shift between user m and the base station, f d is the maximum Doppler shift, The arrival angle of the signal follows a uniform distribution of [0,2π]. w is the system bandwidth. t [n] is additive white Gaussian noise that obeys complex Gaussian distribution, is the additive white Gaussian noise power.

[0033] Consider the case where the user signal reaches the receiving end synchronously or asynchronously. In the synchronous case, the delay between different users and the base station is the same, while in the asynchronous case, the delay between different users and the base station is different. The schematic diagram of user synchronization and asynchrony is shown in the figure below. Figure 4 shown.

[0034] Step 3: Receiver signal detection

[0035] Receiver signal detection flow box Figure 5 As shown.

[0036] ①User preamble detection

[0037] The receiver first performs preamble detection to determine the number of competing users within a timeslot and the timing advance. This allows for uplink time synchronization between the user and the base station, providing accurate timing information for subsequent data transmission. The base station performs preamble detection by correlating the preamble signal received in each timeslot with a local preamble word.

[0038] Considering that the channel gain of path 1 is the strongest in a multipath channel, the receiver can accurately obtain the user timing information by detecting the preamble codeword. If only one user is detected, the time slot is a non-user contention time slot. If two or more users are detected, the time slot is a user contention time slot. The present invention assumes that the receiver can perfectly detect the user's timing information. That is, the preamble detection is completely correct and the receiver can know the number of users in each time slot.

[0039] ②User signal detection

[0040] After the receiver obtains the user timing information through preamble detection, it performs signal detection on the user. In each time slot, the receiver takes each timing information as the starting point and performs N payloadHere, consider that user m sends its dth copy in time slot t, i.e., t=t m,d The receiver obtains user timing information through preamble detection Signal detection consists of three steps: differential modulation non-coherent detection, data-assisted Doppler frequency shift and channel coefficient estimation, and differential modulation iterative coherent detection.

[0041] 1. Differential modulation non-coherent detection

[0042] First, using the received signal y t , calculate c′ m,d The log-likelihood ratio of [l],

[0043]

[0044] Among them, c′ m,d [l] represents c′ m,d The lth symbol in , l represents the index of the encoding sequence, Indicates c' m,d [l] = 1 corresponding to the symbol set, represents c′ m,d [l] = 0 corresponding to the symbol set. represents the rounding down operation, and ||·|| represents the modulo operation. The receiver is m,d [l], l = 1, ..., L to obtain its log-likelihood ratio, and get c′ m,d The log-likelihood ratio sequence Λ(c′ m,d ).

[0045] Log-likelihood ratio sequence Λ(c′ m,d ) passes through the deinterleaver to generate sequence c m,d The log-likelihood ratio sequence Λ(c m,d ). m,d ) is input to the decoder to generate the decoded sequence c m,d The log-likelihood ratio Λ DEC (c m,d ). DEC (c m,d ) After passing through the interleaver, the sequence c′ is generated m,d Decoded log-likelihood ratio Λ DEC (c′ m,d ).

[0046] According to c′ m,d [l] Log-likelihood ratio after decoding Λ DEC (c′ m,d [l]), calculate

[0047]

[0048] Among them, Pr(c′ m,d [l] = 1) indicates c′ m,d The probability of [l] = 1, Pr(c′ m,d [l]=0) means c′ m,d The probability that [l] = 0.

[0049] Symbol x m,d [n] is α q ,q=0,…,Q-1, that is, Pr(x m,d [n] = α q ),q=0,…,Q-1,

[0050] c′ m,d [l], l = (n-1)log2Q + 1,…, nlog2Q probability.

[0051] 2. Data-assisted Doppler shift and channel coefficient estimation

[0052] According to s m,d [n-1], s m,d [n],x m,d [n], calculate s m,d [n] is β q ,q=0,…,Q-1 probability Pr(s m,d [n] = β q ),q=0,…,Q-1. For s m,d [n] Select the symbol with the largest probability value and greater than the threshold to make a hard decision, and get s m,d Hard decision result of [n] All hard decision results form a vector According to the probability relationship, The number of elements in is less than or equal to s m,d ,and For s m,t The hard decision results of a continuous range of symbols indexed from 0 to N′ in the vector The hard decision result is used to estimate the Doppler shift and channel gain. Since the channel gain of path 1 is the strongest in the multipath channel, only the channel coefficient and Doppler shift of path 1 are estimated.

[0053] Doppler shift The estimated result is

[0054]

[0055] in,(·) * represents the complex conjugate operation, (·) Hrepresents the matrix conjugate transpose operation, and FFT(·) represents the fast Fourier transform operation.

[0056] Channel coefficient The estimated result is

[0057]

[0058] 3. Differential Phase Modulation Iterative Detection

[0059] Using Doppler shift estimation results and channel coefficient estimation results Perform differential phase modulation iterative detection to further improve detection performance.

[0060] The block diagram of differential phase modulation iterative detection is as follows: Figure 3 As shown, there are four modules: observation node, check node, interleaver / deinterleaver, and decoder.

[0061] The input of the variable node is s m,d Probability information of [n]

[0062]

[0063] Use λ to represent the number of iterations. During the λth iteration, the input of the check node is x fed back to the check node in the λ-1th iteration. m,d [n] Probability information after the decoder q=0,…,Q-1, Represents x m,d [n] Probability information after passing through the decoder. x m,r [n] The initial input of the probability information after the decoder is q=0,…,Q-1。

[0064] The verification node and the observation node obtain x through the Belief Propagation (BP) algorithm. m,d [n] probability information, q=0,…,Q-1。 Indicates x after the BP algorithm in the λth iteration m,d [n] probability information. m,d The probability information of [n] can be obtained as c′ m,d [l],l=(n-1)log2Q+1,…,nlog2Q is the log-likelihood ratio after the BP algorithm in the λth iteration Then we get c′ m,d Log-likelihood ratio sequence after BP algorithm

[0065] Log-likelihood ratio sequence after BP algorithm After the deinterleaver, the sequence c is generated m,d The log-likelihood ratio sequence will sequence Input to the decoder to generate sequence c m,d The log-likelihood ratio sequence after the decoder in the λth iteration is From the sequence Remove the decoder input, that is, The results After the interleaver, the sequence c′ is generated m,d The log-likelihood ratio sequence after decoding in the λth iteration

[0066] use l=(n-1)log2Q+1,…,nlog2Q,

[0067] The x fed back to the check node in the λth iteration can be obtained m,d [n] probability information,

[0068] Right now q=0,…,Q-1. In the last iteration, the decoder outputs the estimated result of the message sequence through the message sequence output module.

[0069] The decoder's built-in verification module can be used to determine the solution Whether the verification relationship is satisfied. If the verification is passed, it is determined that the receiver has decoded the signal b′ m,d The receiver obtains the signal b′ from the decoded signal m,d The message sequence b of user m is obtained from m , and obtain the location information of the time slots where the other copies of user m are located.

[0070] ③ Use the solved information to eliminate interference

[0071] The receiver uses the solved user's message sequence and the position information of the time slots where the remaining copies of the user are located to perform interference cancellation operations.

[0072] Assume that there are M t Users compete, so Indicates M t Assume that M′ has been solved in time slot t t The message sequence of a user, Represents M′ t A collection of users.

[0073] Since the channel gain of path 1 is the strongest in a multipath channel, only the channel coefficient of path 1 is estimated.

[0074] For simplicity of expression, let h′ t express The channel vector is composed of the channel coefficients of path 1 of the user in time slot t. Let h″ t Indicates belonging to a set But not The channel vector is composed of the channel coefficients of path 1 of user in time slot t.

[0075] According to M′ t The message sequences of each user can be used to obtain the signal sequences sent by these users in time slot t. According to the Doppler estimation results, the signal sequence is phase shifted.

[0076]

[0077] in, Represents the rotation vector. Represents the Hadamard product operation. Pad the signal with zeros to expand its length to N. payload +N Guard , the signal after zero padding is s″ m,d .

[0078] For the signal s of user m m,d ,in, Assume that after phase shift and zero filling, the value is s″ m,d (In practice, the above operation cannot be performed because the actual signal and Doppler shift are unknown. This assumption is only for channel estimation.) and Represents the identity matrix, size N T ×N T , E(·) represents the expected operation.

[0079] For h′ t Perform minimum mean square error channel estimation,

[0080]

[0081] in, is the coefficient of the minimum mean square error channel estimation

[0082]

[0083] express The sequence used by the first user for channel estimation in express Mt ' user sequences used for channel estimation.

[0084] y t Interference elimination is performed, and the result of interference elimination is

[0085]

[0086] The receiver uses the result (11) after interference cancellation to detect signals from other users. It should be noted that the receiver uses the signal y′ after interference cancellation in subsequent signal detection. t As the input signal, the receiver's initial received signal y is used when performing interference cancellation. t .

[0087] Step 4: Simulation Verification and Performance Evaluation

[0088] Numerical simulations were conducted on both terrestrial and satellite channels to show how the throughput of the proposed system varies with user load factor. The simulation results reveal the performance improvements in user access and transmission efficiency achieved by the proposed solution in large-scale communication scenarios.

[0089] First, consider a terrestrial channel environment filled with numerous reflectors, thus effectively representing a single-path channel. The channel model assumes a path average channel gain of 1, and the channel coefficients follow a Rayleigh distribution. Different users arrive at the base station synchronously, the Doppler shift between the user and the base station is zero, and the receiver knows both the delay and Doppler shift, meaning that Doppler shift estimation is unnecessary. In the simulation, a system signal-to-noise ratio of 15 dB is used, employing a [7,7,5]8 convolutional code. User preamble detection is assumed to be completely accurate.

[0090] Depend on Figure 7 It can be seen that when the proposed solution uses Q = 2 differential phase modulation, the number of replicas generated by the enhanced Slotted ALOHA is 2,3. If the receiver only solves for non-contention time slots, it can achieve the same throughput as a traditional Q = 2 phase modulation system using ideal channel information. Differential phase modulation reduces pilot overhead while achieving the same throughput, improving the system's spectral efficiency.

[0091] Depend on Figure 8 It can be seen that the proposed detection scheme can achieve similar throughput using Q=2 and Q=4 modulation schemes when using enhanced Slotted ALOHA to generate 3 replicas and the receiver resolves contention time slots. The proposed scheme can be applied not only to low-order differential modulation but also to high-order differential modulation.

[0092] Depend on Figure 9It can be seen that when the number of replicas generated by enhanced Slotted ALOHA is 3, the detection method combining differential phase modulation incoherent detection with iterative detection can effectively improve the system throughput. When the modulation method Q = 2 is used, the system throughput is 70% better than that of differential phase modulation incoherent detection alone.

[0093] In the satellite channel, the Channel Model A satellite channel model in the ITUM.1225 protocol is used for channel modeling. The bit rate is The low-density parity check code (LDPC) is used, assuming that the user preamble detection is completely correct. Consider the two cases where different users arrive at the receiver synchronously and asynchronously. Figure 10-12 It can be seen that the detection scheme proposed in the present invention can estimate and compensate for Doppler frequency shift in low-orbit satellite scenarios and scenarios with high Doppler frequency shift between satellite and ground links, thereby improving system access throughput, reducing user access delay, and reducing the total system access duration. At the same time, considering the situation where users arrive at the asynchronous receiving end, in this case, the receiver can effectively handle the signal delay caused by user asynchrony. At the same time, compared with the case of user synchronization, the asynchronous arrival of users can improve system throughput and reduce access delay. The scheme proposed in the present invention can improve system throughput and reduce the average user access delay compared to the four-step random access of the baseline scheme.

[0094] (3) Advantages and application value

[0095] The present invention provides a pilot-free enhanced Slotted ALOHA access scheme based on differential phase modulation for large-scale random access scenarios. The present invention effectively solves the user competition problem in the two-step random access process by strengthening the Slotted ALOHA transmission protocol, improves the success rate of two-step random access, and thus reduces the system delay in random access in large-scale communication scenarios. In addition, the present invention saves pilot overhead through differential phase modulation and a differential phase modulation detection algorithm that combines non-coherent detection with iterative detection, and can achieve the same access throughput as the traditional pilot scheme in the absence of a pilot, thereby improving the spectrum efficiency of the system. The scheme proposed in the present invention can effectively improve the access throughput in large-scale random access scenarios, and can reduce the average user access delay compared to the four-step random access scheme. This has high application value in supporting low-latency random access and high-spectrum-efficiency signal transmission in ultra-large-scale connection scenarios.

Brief Description of the Drawings

[0096] Figure 1 Schematic diagram of ultra-large-scale connection scenario.

[0097] Figure 2 This is the overall block diagram of the sending and receiving scheme of the present invention.

[0098] Figure 3 This is a block diagram of the sending solution of the sending end of the present invention.

[0099] Figure 4 This is a schematic diagram of signal synchronization and asynchrony at the receiving end of the present invention.

[0100] Figure 5 This is a detection block diagram of the present invention, which combines non-coherent detection with iterative coherent detection at the receiving end.

[0101] Figure 6 This is a block diagram of iterative coherent detection at the receiving end of the present invention.

[0102] Figure 7 In order to compare the system throughput of the scheme proposed in the present invention using Q=2 differential phase modulation and Q=2 traditional phase modulation scheme under the condition of ideal user contention detection and solving only non-contention time slots.

[0103] Figure 8 The system throughput curves of the solution proposed in the present invention using the Q=2 differential phase modulation and Q=4 differential phase modulation schemes.

[0104] Figure 9 This is a comparison chart of the system throughput curves of the solution proposed in the present invention using a combination of incoherent detection and iterative coherent detection under Q=2 differential phase modulation and traditional incoherent detection.

[0105] Figure 10 This is a comparison chart of the throughput of user synchronous and asynchronous systems under the ITU-R M.1225 satellite channel for the solution proposed in the present invention.

[0106] Figure 11 This is a comparison chart of the access delay between user synchronous and asynchronous systems under the ITU-R M.1225 satellite channel for the solution proposed in the present invention.

[0107] Figure 12 This is a comparison chart of the total access time of the synchronous and asynchronous systems under the ITU-R M.1225 satellite channel of the solution proposed in the present invention. [Specific implementation method]

[0108] The present invention will be further described below in conjunction with the accompanying drawings.

[0109] Step 1: Differential Coded Modulation and Frame Structure

[0110] The specific implementation plan of step one is consistent with the content of step one in the technical solution and will not be described in detail here.

[0111] Step 2: Send the signal through the wireless channel to the base station

[0112] The specific implementation plan of step 2 is consistent with the content of step 2 in the technical solution and will not be described in detail here.

[0113] Step 3: Receiver signal detection

[0114] ①User preamble detection

[0115] The receiver detects the preamble codeword within each time slot to determine the number of user contentions and the timing advance. Preamble code detection uses a traditional scheme. The receiver performs a cross-correlation operation on the received signal using the local preamble code sequence. The number and position of peaks in the cross-correlation spectrum determine the number of user contentions and the timing advance.

[0116] The solution proposed in the present invention can directly perform signal detection on the user's contention time slot and non-contention time slot. The main purpose of preamble code detection is to complete the measurement of the timing advance, so as to achieve uplink time synchronization between the user and the base station, and to provide accurate time synchronization information for subsequent data transmission. In the present invention, since the transmitting end adopts differential modulation, the receiving end can perform non-coherent demodulation on the received signal without channel information. In the traditional pilot-based scheme, the receiving end needs to use the transmitted pilot information to perform channel estimation to obtain channel information. The channel information obtained by channel estimation is used to coherently demodulate the received signal. Affected by the different arrival times of different users, when there are many competing users, the orthogonality between the pilots will be broken, affecting the results of the channel estimation.

[0117] ②User signal detection

[0118] 1. Incoherent detection

[0119] According to formula (3) in the non-coherent detection part of the technical solution, c′ can be obtained m,d The log-likelihood ratio sequence Λ(c′ m,d ). Λ(c′ m,d ) After passing through the deinterleaver, the sequence c is generated m,d The log-likelihood ratio sequence Λ(c m,d ). m,d ) is input to the decoder to generate the decoded sequence c m,d The log-likelihood ratio Λ DEC (c m,d ). DEC (c m,d ) After passing through the interleaver, the sequence c′ is generated m,d Decoded log-likelihood ratio Λ DEC (c′ m,d ).

[0120] According to formula (4), we can get c′ m,d [l] = 1 and c′ m,d The probability Pr(c′) of [l]=0m,d [l]=1),

[0121] Pr(c′ m,d [l]=0).

[0122] Symbol x m,d [n] is α q ,q=0,…,Q-1, that is, Pr(x m,d [n] = α q ),q=0,…,Q-1, can be obtained by c′ m,d [l], l = (n-1)log2Q + 1,…, nlog2Q probability.

[0123]

[0124] Among them, b l ∈{0,1}. b l Indicates that when x m,d [n] = α q When c′ m,d The value of [l].

[0125] 2. Data-assisted Doppler shift and channel estimation

[0126] Calculate s m,d [n] is β q ,q=0,…,Q-1 probability Pr(s m,d [n] = β q ),q=0,…,Q-1。

[0127]

[0128] Among them, α (n) Represents x m,d The value of [n], β (n-1) Indicates s m,d The value of [n-1] satisfies α (n) β (n-1) =β q .

[0129] For s m,d [n] Select the symbol with the largest probability value and greater than the probability threshold as its hard decision result, and get the hard decision result According to the probability relationship, The number of elements in is less than or equal to s m,d ,and For s m,d The hard decision results of a continuous range of symbols with indices from 0 to N', where N' is the maximum index value. Doppler shift and channel gain estimation: Since the channel gain of path 1 is the strongest in the multipath channel, only the channel coefficient and Doppler shift of path 1 are estimated.

[0130] Doppler shift The estimated result is

[0131]

[0132] Channel coefficient The estimated result is

[0133]

[0134] 3. Differential Phase Modulation Iterative Detection

[0135] The input of the observation node is s m,d The probability information of [n] is obtained by formula (7).

[0136] In the λth iteration, the input of the check node is x fed back to the check node in the λ-1th iteration. m,d [n] The probability information after the decoder, that is, q=0,…,Q-1. The initial input is, q=0,…,Q-1。

[0137] The check node and the variable node use the BP algorithm to obtain x m,d [n] probability information, q=0,…,Q-1。

[0138] The following describes the process of the BP algorithm.

[0139] In the first step, the observation node and the verification node transmit probability information alternately from top to bottom.

[0140] The observation node passes information downward to the verification node,

[0141]

[0142] Pr ob→c (s m,d [n] = β q ) represents the information transmitted from the observation node to the verification node, Pr c→ob (s m,d [n] = β q ) represents the information transmitted by the verification node to the observation node.

[0143] The verification node passes information downward to the observation node,

[0144]

[0145] In the second step, the observation node and the verification node transmit the probability information alternately from bottom to top.

[0146] The observation node passes information upward to the verification node,

[0147]

[0148] The verification node passes information upward to the observation node,

[0149]

[0150] Among them, β (n+1) Indicates s m,d The value of [n+1], α (n+1) Represents x m,d The value of [n+1].

[0151] Step 3: Verify node output x m,d Probability information of [n]

[0152]

[0153] Among them, β (n) Indicates s m,d The value of [n].

[0154] The above are the main operations of the BP algorithm.

[0155] By x m,d The probability information of [n] can be obtained as c′ m,d [l],l=(n-1)log2Q+1,…,nlog2Q log-likelihood ratio ,

[0156]

[0157] in, Indicates the corresponding c′ m,d The set of symbols where [l] = 1, Indicates the corresponding c′ m,d The set of symbols where [l] = 0 has been defined in the technical solution. Then we get c′ m,d The log-likelihood ratio sequence

[0158] sequence After the deinterleaver, the sequence c is generated m,d The log-likelihood ratio sequence will sequence Input to the decoder to generate the decoded sequence c m,d The log-likelihood ratio sequence From the sequence Remove the decoder input, that is, The results After the interleaver, the sequence c′ is generated m,d The decoded log-likelihood ratio sequence.

[0159] use l=(n-1)log2Q+1,…,nlog2Q,

[0160] The x fed back to the check node in the λth iteration can be obtained m,d [n] probability information,

[0161] Right now q=0,…,Q-1,

[0162]

[0163] Among them, b l ∈{0,1}. b l Indicates that when x m,d [n] = α q When c′ m,d The value of [l].

[0164] In the last iteration, the signal detection result is output through the information bit output module of the decoder The decoder's verification module can determine the solution Is it the message sequence b′ sent by user m? m,d If the check is passed, it is determined that the receiver has decoded the signal b′ m,d The receiver obtains the signal b′ from the decoded signal m,d The message sequence b of user m is obtained from m , and obtain the location information of the time slots where the other copies of user m are located.

[0165] ③Interference elimination

[0166] Interference elimination is the same as that described in step 3 of the technical solution and will not be introduced here.

[0167] Step 4: System simulation and performance analysis

[0168] This section further explains the system simulation parameters and simulation results in step four of the technical solution.

[0169] 1. Ground channel model performance simulation

[0170] Consider a channel environment filled with numerous reflectors. Therefore, the channel model can be simplified to a single-path channel. The average channel gain is 1, and the channel coefficient follows a Rayleigh distribution. Considering that different users arrive at the base station synchronously, the Doppler shift between the user and the base station is zero, and the receiver does not need to estimate the Doppler shift. It is also assumed that the user preamble detection is completely accurate.

[0171] In the simulation, the system signal-to-noise ratio is considered to be 15dB, and the channel coding adopts the [7,7,5]8 convolutional code. The total number of time slots T = 100.

[0172] System load (LOAD) is defined as

[0173]

[0174] The system throughput is defined as

[0175]

[0176] Where Solve(M) is the number of users for which the receiver finally solves successfully.

[0177] Under the above model and system parameters, the access throughput of the enhanced Slotted ALOHA protocol system based on differential phase modulation without pilot is evaluated and compared with the traditional enhanced Slotted ALOHA protocol based on pilot channel estimation. Consider that the receiver only performs signal detection on non-user contention time slots. The number of replicas generated by the enhanced Slotted ALOHA is 2 or 3. The system adopts differential phase modulation with Q=2. For the traditional solution, Q=2 phase modulation and ideal channel estimation are considered. Figure 7 It can be seen that the technical solution proposed in the present invention can approach the access throughput of the traditional solution under ideal user contention detection. It can be seen that the technical solution proposed in the present invention can approach the access performance of the traditional solution under ideal channel estimation in the absence of pilot.

[0178] Under the above channel model, the access throughput of the proposed scheme was evaluated. The system solved for contention slots and enhanced Slotted ALOHA to generate 3 replicas. The system throughput for differential phase modulation orders Q = 2 and Q = 4 was compared. It can be seen that the proposed scheme can achieve similar throughput for modulation orders Q = 2 and Q = 4, with the throughput for Q = 4 being slightly lower than that for Q = 2. The proposed scheme can be applied not only to low-order modulation but also to high-order modulation.

[0179] Under the above channel model, the access throughput of the proposed solution is evaluated, and the system throughput of the conventional non-coherent demodulation and the system throughput of the proposed solution combining non-coherent demodulation and iterative detection are compared. Consider the differential phase modulation method with Q=2. Figure 10 It can be seen that the system throughput of the scheme combining non-coherent demodulation and iterative coherent demodulation proposed in the present invention is better than that of the traditional non-coherent demodulation scheme.

[0180] 2. Performance evaluation under satellite channels

[0181] The access throughput of the solution proposed in the present invention is evaluated under satellite channels.

[0182] The Channel Model A satellite channel model in the ITU-R M.1225 protocol is used for modeling. The channel model parameters are shown in Table 1.

[0183] Table 1 ITU-R M.1225 Channel Model A parameters

[0184]

[0185] Considering the system bandwidth B w =1.08MHz, the maximum Doppler shift is f d = 42kHz. When the satellite elevation angle is 90°, the user signal can be considered to arrive synchronously, with a maximum delay symbol difference of 3. At this time, according to the link budget, the system signal-to-noise ratio is 18.05dB. When the satellite elevation angle is 80°, the user signal can be considered to arrive asynchronously, with a maximum delay symbol difference of 58. At this time, according to the link budget, the system signal-to-noise ratio is 17.9dB. The channel coding uses the code rate LDPC code. Considering that the user preamble detection is completely correct. The proposed scheme is simulated under user synchronization and asynchrony, and its throughput and access delay are analyzed.

[0186] Consider a two-way satellite-to-ground link delay of 8ms, with each time slot length of 2ms. For the two-step random access scheme proposed in this solution, the delay from the time slot in which the user sends the first copy to the time slot in which the receiving end obtains the user's information is considered the solution delay. The sum of the solution delay and the two-way satellite-to-ground link delay is the total access delay. If the user information is still not decoded by the end of the entire random access time slot, the user initiates a four-step random access. Assuming a 100% success rate for the four-step random access, the four-step random access delay is 16ms. The four-step random access delay plus the two-step random access failure delay is the total delay when the user's two-step random access fails.

[0187] Depend on Figure 10-12 It can be seen that the detection scheme proposed in this invention can realize the estimation and compensation of Doppler frequency shift in the scenario where there is high Doppler frequency shift between the satellite and the ground link. When the user arrives at the receiving end synchronously, the system throughput is It can reach 160%. The receiver can effectively compensate for the Doppler frequency shift through Doppler estimation. The average delay of user access can be lower than that of traditional four-step random access when the load rate is lower than 60%. The total time for all users to access can be reduced by 38.04% when the load rate is 160% compared with four-step random access. The scheme proposed in the present invention can effectively reduce the access delay of the satellite-to-ground link. The scheme proposed in the present invention can effectively adapt to the scenario where the Doppler frequency shift of low-orbit satellites is high and users arrive asynchronously. When the user arrives at the receiving end asynchronously, the receiving end can effectively handle the user information delay caused by asynchrony. User asynchrony reduces the degree of competition among users at the receiving end, weakens the competition between users, and improves the system throughput and reduces the system delay compared to when users are synchronized.

Claims

1. A pilot-free enhanced Slotted ALOHA large-scale random access and transmission method based on differential phase modulation, characterized in that: The specific steps are as follows: Step 1: Differential Coded Modulation and Frame Structure Consider a large-scale communication system with a single base station and M users. Both the user and the base station use a single antenna for transmission and reception. The total number of time slots in the enhanced Slotted ALOHA protocol is T. The base station sends synchronization blocks to users through beam scanning. Users complete downlink synchronization and obtain random access configuration information by decoding the synchronization blocks. When a user needs to initiate a random access request, it first sends a preamble codeword over the physical random access channel, followed by the random access data payload over the physical uplink shared channel. Step 2: Send the signal through the wireless channel to the base station The transmitted signal reaches the base station through a wireless channel. Consider a multipath channel between the base station and the user, with P paths. Path 1 has the highest average channel gain, exceeding the remaining paths by more than 20 dB. When considering a channel environment with many reflectors, the channel model degenerates into a single-path channel. There is a Doppler shift between the base station and the user. For any user, assume that its channel coefficient remains constant within a single time slot, and the Doppler shift remains constant over T time slots. Step 3: Receiver signal detection 3.1 User Preamble Detection The receiver first performs preamble detection to detect the number of competing users in the time slot and measure the timing advance to achieve uplink time synchronization between the user and the base station, providing accurate time synchronization information for subsequent data transmission. The base station performs preamble detection by performing correlation operations on the preamble signal received in each time slot using the local preamble word. 3.2 User Signal Detection After the receiver obtains the user timing information through preamble detection, it performs signal detection on the user. In each time slot, the receiver takes each timing information as the starting point and performs signal detection on the subsequent N payload symbols for signal detection; here we consider that user m sends its dth copy in time slot t, that is, t = t m,d ; The receiver obtains user timing information through preamble detection Signal detection consists of three steps: differential modulation non-coherent detection, data-assisted Doppler frequency shift and channel coefficient estimation, and differential modulation iterative coherent detection. 3.3 Using the solved information to eliminate interference The receiver uses the solved user's message sequence and the position information of the time slots where the remaining copies of the user are located to perform interference cancellation operations; Step 4: Simulation Verification and Performance Evaluation Numerical simulations of the system throughput varying with user loading rate are performed under terrestrial and satellite channels respectively.

2. The method of claim 1 , wherein: In step 1, for the preamble code word portion sent by the user, the user selects the preamble code word according to the preamble code configuration information indicated by the base station; The preamble codeword is used to obtain timing information and complete uplink time synchronization.

3. The method of claim 1 , wherein: In step 1, for the data payload sent by the user, the message sequences of all users are considered to be of equal length; the message sequence sent by user m is b m ∈{0,1} k ,m=1…M,where k represents the length of the message sequence; user m has a m Make a copy and generate d m duplicate message sequence Where d is the index of the replica; user m randomly selects time slots on the time axis, which correspond to The sending time slot of is the number of the randomly selected time slot; the user Add the time slot position information of the remaining copies to generate The length of the message sequence after adding the time slot position information is k'; The code rates used are R c The encoder is encoded to obtain the encoded sequence The length of the coded sequence is L; after passing through the random interleaver π, the coded sequence after the interleaver is obtained The user will Mapping to symbol sequence N represents the length of the symbol sequence, α q represents the qth symbol in the constellation point set, q represents the index of the constellation symbol, Q represents the modulation order, and j is the imaginary unit; user m has a certain influence on the symbol sequence. Perform differential phase modulation, s m,d [n]=s m,d [n-1]x m,d [n] (1) Get user m in the tth m,d The symbol sequence s is sent in the time slot m,d , where n represents discrete time; x m,d [n] represents x m,d The nth symbol in s m,d [n] means s m,d The nth symbol in s m,d [0] is determined by the modulation order Q; When the modulation order Q=4, select therefore where β q Represents the constellation points of the differential signal; User m selects a time slot Send symbol sequence Send, each time slot length is N T symbol duration, N T =N preamble +N payload +N Guard ; where N preamble Indicates the length of the preamble sequence, N payload =N+1 represents the length of the data payload, N Guard Indicates the length of the guard interval.

4. The method of claim 3, wherein: In step 2, the signal received by the base station in the tth time slot is Where P represents the path index, is the channel coefficient of the pth path between user m and the base station in the tth time slot, represents the normalized delay of the pth path between user m and the base station, Indicates the user's timing information, represents the normalized Doppler shift between user m and the base station, represents the actual Doppler shift between user m and the base station, f d is the maximum Doppler shift, The arrival angle of the signal obeys the uniform distribution of [0,2π]; B w is the system bandwidth; z t [n] is additive white Gaussian noise that obeys complex Gaussian distribution, is the additive white Gaussian noise power.

5. The method of claim 1 , wherein: In step 3, considering that the channel gain of path 1 is the strongest in the multipath channel, the receiver can accurately obtain the user timing information by detecting the preamble codeword. If only one user is detected, the time slot is a non-user contention time slot. If two or more users are detected, the time slot is a user contention time slot. Assume that the receiver can perfectly detect the user's timing information. That is, the preamble detection is completely correct and the receiver can know the number of users in each time slot.

6. The method of claim 4, wherein: In step three, the differential modulation incoherent detection is: First, using the received signal y t , calculate c′ m,d The log-likelihood ratio of [l], Among them, c′ m,d [l] represents c′ m,d The lth symbol in , l represents the index of the encoding sequence, Indicates c' m,d [l] = 1 corresponding to the symbol set, represents c′ m,d [l] = the set of symbols corresponding to 0; represents the rounding down operation, ||·|| represents the modulo operation; the receiver is for all c′ m,d [l], l = 1, ..., L to obtain its log-likelihood ratio, and get c′ m,d The log-likelihood ratio sequence Λ(c′ m,d ); Log-likelihood ratio sequence Λ(c′ m,d ) passes through the deinterleaver to generate sequence c m,d The log-likelihood ratio sequence Λ(c m,d ); Λ(c m,d ) is input to the decoder to generate the decoded sequence c m,d The log-likelihood ratio Λ DEC (c m,d );Λ DEC (c m,d ) After passing through the interleaver, the sequence c′ is generated m,d Decoded log-likelihood ratio Λ DEC (c′ m,d ); According to c′ m,d [l] Log-likelihood ratio after decoding Λ DEC (c′ m,d [l]), calculate Among them, Pr(c′ m,d [l] = 1) indicates c′ m,d The probability of [l] = 1, Pr(c′ m,d [l]=0) means c′ m,d The probability that [l] = 0; Symbol x m,d [n] is α q ,q=0,…,Q-1, that is, Pr(x m,d [n] = α q ),q=0,…,Q-1, By c′ m,d [l], l = (n-1)log2Q + 1,…, nlog2Q probability.

7. The method of claim 6, wherein: In step 3, the data-aided Doppler shift and channel coefficients are estimated as follows: According to s m,d [n-1], s m,d [n],x m,d [n], calculate s m,d [n] is β q ,q=0,…,Q-1 probability Pr(s m,d [n] = β q ),q=0,…,Q-1; for s m,d [n] Select the symbol with the largest probability value and greater than the threshold to make a hard decision, and get s m,d Hard decision result of [n] All hard decision results form a vector According to the probability relationship, The number of elements in is less than or equal to s m,d ,and For s m,t The hard decision results of a continuous range of symbols indexed from 0 to N′ in the vector The maximum index value of ; Doppler frequency shift and channel gain estimation are performed using the hard decision result; since the channel gain of path 1 in the multipath channel is the strongest, only the channel coefficient and Doppler frequency shift of path 1 are estimated; Doppler shift The estimated results are: in,(·) * represents the complex conjugate operation, (·) H represents the matrix conjugate transpose operation, and FFT(·) represents the fast Fourier transform operation; Channel coefficient The estimated results are:

8. The method of claim 7, wherein: In step 3, the differential phase modulation iterative detection is as follows: using the Doppler frequency shift estimation result and channel coefficient estimation results Perform differential phase modulation iterative detection to improve detection performance; The input of the variable node is s m,d Probability information of [n]: Use λ to represent the number of iterations; in the λth iteration, the input of the check node is x fed back to the check node in the λ-1th iteration. m,d [n] Probability information after the decoder Represents x m,d [n] Probability information after the decoder; x m,r [n] The initial input of the probability information after the decoder is The verification node and the observation node obtain x through the Belief Propagation (BP) algorithm. m,d [n] probability information, Indicates x after the BP algorithm in the λth iteration m,d [n] probability information; through x m,d The probability information of [n] is obtained by c′ m,d [l],l=(n-1)log2Q+1,…,nlog2Q is the log-likelihood ratio after the BP algorithm in the λth iteration Then we get c′ m,d Log-likelihood ratio sequence after BP algorithm Log-likelihood ratio sequence after BP algorithm After the deinterleaver, the sequence c is generated m,d The log-likelihood ratio sequence will sequence Input to the decoder to generate sequence c m,d The log-likelihood ratio sequence after the decoder in the λth iteration is From the sequence Remove the decoder input, that is, The results After the interleaver, the sequence c′ is generated m,d The log-likelihood ratio sequence after decoding in the λth iteration use Get x that is fed back to the check node in the λth iteration m,d [n] probability information, that is, In the last iteration, the decoder outputs the estimated result of the message sequence through the message sequence output module The solution is determined by the decoder's built-in verification module Whether the verification relationship is satisfied; If the check passes, it is determined that the receiver has decoded the signal b′ m,d ; The receiver obtains the decoded signal b′ m,d The message sequence b of user m is obtained from m , and obtain the location information of the time slots where the other copies of user m are located.

9. The method of claim 8, wherein: In step 3, suppose there are M t Users compete, so Indicates M t A set of users; suppose M t ′ user's message sequence, let Indicates M t ′ users; Since the channel gain of path 1 is the strongest in a multipath channel, only the channel coefficient of path 1 is estimated; For simplicity of expression, let h′ t express The channel vector is composed of the channel coefficients of path 1 of the user in time slot t; let h′ t ′ means belonging to the set But not The channel vector composed of the channel coefficients of path 1 of the user in time slot t; According to M t The message sequences of the ′ users are respectively obtained by the signal sequences sent by these users in time slot t, According to the Doppler estimation results, the signal sequence is phase shifted. in, represents the rotation vector; Represents the Hadamard product operation; the signal is padded with zeros to expand its length to N payload +N Guard , the signal after zero padding is s′ m ′,d; For the signal s of user m m,d ,in Suppose the value after phase shift and zero filling is s″ m,d ,and Represents the identity matrix, size N T ×N T , E(·) represents the expected operation; For h′ t Perform minimum mean square error channel estimation, in, is the coefficient of the minimum mean square error channel estimation express The sequence used by the first user for channel estimation in express M t ′ user sequence for channel estimation; y t Interference elimination is performed, and the result of interference elimination is:

10. The method of claim 1, wherein: In step 4, we first consider a terrestrial channel environment filled with many reflectors, so the channel is equivalent to a single-path channel. The average channel gain of the channel model path is 1, and the channel coefficient follows the Rayleigh distribution. We consider that different users arrive at the base station synchronously, the Doppler shift between the user and the base station is 0, and the receiver has known delay and Doppler shift, that is, the receiver does not need to estimate the Doppler shift. In the simulation, the system signal-to-noise ratio is considered to be 15dB, and the convolutional code [7,7,5]8 is used. The user preamble code detection is considered to be completely correct.

Images