Sky wave large-scale MIMO DFTS-OFDM uplink Turbo receiving method and system

Abstract

The invention discloses a sky wave large-scale MIMO DFTS-OFDM uplink Turbo receiving method and a sky wave large-scale MIMO DFTS-OFDM uplink Turbo receiving system. The method comprises the following steps: firstly, processing a frequency domain signal received by a base station according to subcarriers one by one to obtain a dimension reduction vector of each user group for sparse detection of packet interference; then, carrying out packet interference sparse signal detection by utilizing the dimension reduction vector of each user group and a pre-designed packet interference sparse detector selected according to the precedence variance; and finally, carrying out user-by-user combined all subcarrier processing on the signal detection result, and carrying out IDFT, extrinsic information calculation, de-interleaving, soft input and soft output decoding, interleaving, prior information updating and DFT on the frequency domain signal detection result of each user to obtain a prior mean value and a prior variance of the frequency domain signal of each user for the next round of iterative detection decoding. According to the method, Turbo receiving processing under DFTS-OFDM modulation is carried out by utilizing the interference sparse characteristic of a sky wave large-scale MIMO channel, and the power efficiency and the receiving performance can be effectively improved while the design and implementation complexity is reduced.

Description

Technical Field

[0001] This invention belongs to the field of communication technology, and in particular relates to a method and system for uplink Turbo reception of skywave massive MIMO DFTS-OFDM. Background Technology

[0002] Skywave communication, a key method for long-distance transmission, primarily utilizes the 3-30 MHz shortwave band, achieving beyond-line-of-sight signal propagation through the alternating reflection of electromagnetic waves in the ionosphere and on the ground. However, due to the limited spectrum resources and the complexity of the ionospheric environment, the transmission rate of traditional skywave communication faces significant bottlenecks. Massive MIMO technology, with its large number of antenna elements configured at the base station, can effectively support parallel data transmission from multiple users, thereby significantly improving spatial multiplexing capabilities and spectral efficiency. Therefore, research on skywave massive MIMO communication is of great significance.

[0003] Skywave MIMO uplink reception is a critical issue for skywave MIMO communication. However, existing skywave MIMO uplink receivers all operate under OFDM modulation. With increasing system bandwidth, the peak-to-average power ratio (PAPR) of the transmitted signal under OFDM waveforms will be very high. Furthermore, considering that skywave communication transmission distances are typically thousands of kilometers, skywave MIMO-OFDM uplink transmission places high demands on the power amplifiers of user terminals. DFTS-OFDM waveforms, benefiting from their low PAPR characteristics, have been widely used in terrestrial cellular communication and underwater acoustic communication, and are expected to serve as the uplink transmission waveform for skywave MIMO to improve the power efficiency of user terminals.

[0004] By iteratively transmitting soft information about information bits between a soft-input soft-output detector and a soft-input soft-output decoder, Turbo receivers can effectively improve uplink reception performance. However, for massive MIMO scenarios with significantly increased base station antenna counts and service user numbers, the signal detection process in traditional MMSE Turbo receivers faces unacceptable computational complexity. Furthermore, unlike linear reception, the detectors in Turbo receivers need to be recalculated based on updated prior variances in each iteration, making multiplexing impossible within the channel coherence time and bandwidth. This limitation restricts the practical application of Turbo receivers. Therefore, to improve the power efficiency of user terminals and reduce uplink reception complexity, researching efficient uplink Turbo reception methods for skywave massive MIMO DFTS-OFDM, considering the channel characteristics of skywave massive MIMO, is of significant value. Summary of the Invention

[0005] Purpose of the invention: To address the shortcomings of existing technologies, this invention proposes a skywave massive MIMO DFTS-OFDM uplink Turbo reception method and system. By utilizing the interference sparsity of the channel and combining and improving the Turbo reception framework, the power efficiency and detection decoding performance of the user terminal are effectively improved while reducing the design and implementation complexity.

[0006] Technical Solution: To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] Skywave massive MIMO DFTS-OFDM uplink Turbo reception method includes:

[0008] The frequency domain signal received by the base station is processed on a per-subcarrier basis to obtain the dimension-reduced vector for each user group for sparse detection of packet interference;

[0009] Packet interference sparse signal detection is implemented by using the dimension-reduced vectors of each user group and a pre-designed packet interference sparse detector selected based on the prior variance. The pre-designed packet interference sparse detector is calculated using the channel Gram matrix and a set of quantization variances. Each quantization variance corresponds to a candidate detector. The detector corresponding to the quantization variance closest to the prior variance is selected for packet interference sparse signal detection.

[0010] The signal detection results are processed on a user-by-user basis, involving all subcarriers. For each user's frequency domain signal detection results, IDFT, extrinsic information calculation, deinterleaving, soft-input soft-output decoding, interleaving, prior information update, and DFT are performed to obtain the prior mean and prior variance of each user's frequency domain signal. These are fed back to signal reconstruction and the selection of a pre-designed sparse detector for group interference, and then used for the next round of iterative detection and decoding. After several rounds of iteration, the information bits of each user are finally decoded and output through a soft-input soft-output decoder.

[0011] Further, the step of processing the base station received frequency domain signal on a per-subcarrier basis to obtain the dimension-reduced vector for each user group used for sparse detection of packet interference includes:

[0012] Matched filtering is applied to the spatial received signals of each subcarrier. The mean-removed matched filtered signal is obtained by subtracting the signal reconstructed using the a priori mean of the frequency domain signal from the matched filtered signal.

[0013] Interference selection is performed on the mean-removed matched filter signal using the interference groups of each user group, and the dimension-reduced vectors of each subcarrier and each user group are extracted for sparse detection of grouped interference.

[0014] Furthermore, in the grouped interference sparsity detection, user groups and interference groups of each user group can be constructed using instantaneous channel state information or statistical channel state information. Specifically, the channel correlation or statistical channel correlation between each user is calculated using instantaneous channel state information or statistical channel state information, and user groups are constructed using the channel correlation or statistical channel correlation through clustering methods, and corresponding interference groups are constructed through user groups.

[0015] Furthermore, the prior variance used for selecting the pre-designed packet interference sparse detector is obtained by averaging the prior variances of the data symbols transmitted on each subcarrier by each user.

[0016] Furthermore, the pre-designed packet interference sparse detector is only related to the channel and variance quantization method, and can be reused within the channel coherence time and coherence bandwidth, which can significantly reduce the design complexity of the detector in Turbo reception.

[0017] Furthermore, the detection of the grouped interference sparse signal is achieved by multiplying the conjugate transpose of the pre-designed grouped interference sparse detector selected by each group according to the prior variance with the dimension-reduced vector of each group obtained by interference selection, and then adding it to the prior mean of the frequency domain symbol of each group.

[0018] Furthermore, the process of performing matched filtering on the spatial received signals of each subcarrier is achieved by utilizing IFFT of several base station antenna dimensions and the sparsity of the beam domain channel.

[0019] A skywave massive MIMO DFTS-OFDM uplink Turbo receiver system is used to implement the aforementioned skywave massive MIMO DFTS-OFDM uplink Turbo receiver method, including:

[0020] The matched filtering and interference selection module is used to process the frequency domain signal received by the base station on a per-subcarrier basis to obtain the dimension-reduced vector of each user group for sparse detection of group interference.

[0021] The detector selection module is used to select a pre-designed sparse detector for grouped interference based on the prior variance.

[0022] The signal detection module is used to perform packet interference sparse signal detection using the dimensionality reduction vectors of each user group and the selected packet interference sparse detector;

[0023] The iterative update module is used to process the signal detection results on a user-by-user basis, jointly processing all subcarriers. It performs IDFT, extrinsic information calculation, deinterleaving, soft-input soft-output decoding, interleaving, prior information update, and DFT on the frequency domain signal detection results of each user to obtain the prior mean and prior variance of the frequency domain signal of each user. These are fed back to signal reconstruction and the selection of pre-designed sparse detectors for group interference, and then proceed to the next round of iterative detection and decoding. After several rounds of iteration, the information bits of each user are finally decoded and output through the soft-input soft-output decoder.

[0024] A computer system includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the aforementioned Skywave Massive MIMO DFTS-OFDM uplink Turbo reception method.

[0025] A computer program product includes a computer program that, when executed by a processor, implements the steps of the aforementioned Skywave Massive MIMO DFTS-OFDM uplink Turbo reception method.

[0026] Beneficial Effects: Compared with existing technologies, this invention has the following advantages: 1. This invention utilizes the interference sparsity characteristics of skywave massive MIMO channels and the Turbo receiver framework to design the DFTS-OFDM uplink receiver. It implements low-dimensional packet interference sparse frequency domain subcarrier signal detection for each user group, achieving significant performance improvement compared to linear reception while effectively reducing design and implementation complexity and improving the overall transmission efficiency of the system. 2. This invention improves the detector, which requires recalculation in each iteration of traditional Turbo reception, into a pre-designed packet interference sparse detector that is only related to the channel Gram matrix and a set of quantization variances. In actual iterations, the receiver only needs to select the detector based on the current prior variance, greatly improving the engineering feasibility of the solution in high-speed data transmission scenarios. Attached Figure Description

[0027] Figure 1 This is a block diagram of a skywave massive MIMO DFTS-OFDM uplink Turbo receiver system according to an embodiment of the present invention;

[0028] Figure 2 The figure shows the bit error rate (BER) performance of the uplink Turbo receiver of the skywave massive MIMO DFTS-OFDM according to an embodiment of the present invention under different user packet numbers;

[0029] Figure 3 The graph shows the BER performance of the skywave massive MIMO DFTS-OFDM uplink Turbo receiver in this embodiment of the invention at different variance quantization orders. Detailed Implementation

[0030] To better understand the purpose, structure, and function of this invention, the following detailed description of the uplink Turbo reception method of the Skywave Massive MIMO DFTS-OFDM is provided in conjunction with the accompanying drawings.

[0031] The skywave massive MIMO DFTS-OFDM uplink Turbo reception method disclosed in this invention embodiment is shown in the system block diagram of the receiver. Figure 1 The base station receives the signal and passes it through a packet interference sparse Turbo receiver to obtain the decoding results of each user's transmitted code bits. For example... Figure 1 As shown in this embodiment, the uplink Turbo reception method for skywave massive MIMO DFTS-OFDM mainly includes:

[0032] Step S1: Process the frequency domain signal received by the base station on a per-subcarrier basis to obtain the dimension-reduced vector for each user group for sparse detection of packet interference;

[0033] Step S2: Detect sparse packet interference signals by using the dimension-reduced vectors of each user group and a pre-designed packet interference sparse detector selected based on the prior variance; the pre-designed packet interference sparse detector is calculated using the channel Gram matrix and a set of quantization variances, each quantization variance corresponds to a candidate detector, and the detector corresponding to the quantization variance closest to the prior variance is selected for packet interference sparse signal detection.

[0034] Step S3: Perform user-by-user joint processing of all subcarriers on the signal detection results. For each user's frequency domain signal detection results, perform IDFT, extrinsic information calculation, deinterleaving, soft-input soft-output decoding, interleaving, prior information update, and DFT to obtain the prior mean and prior variance of each user's frequency domain signal. Feedback is used for signal reconstruction and selection of pre-designed sparse detectors for group interference, and the next round of iterative detection and decoding is performed. After several rounds of iteration, the information bits of each user are finally decoded and output through the soft-input soft-output decoder.

[0035] The method of this invention is mainly applicable to skywave massive MIMO DFTS-OFDM systems equipped with large-scale antenna arrays to simultaneously serve multiple users. The specific implementation process of the uplink receiving method involved in this invention will be described in detail below with reference to a specific communication system example. It should be noted that this invention is not only applicable to the specific system model described in this embodiment, but also to other configured system models.

[0036] I. System Model

[0037] Considering base station configuration A uniform linear array of antennas serves simultaneously A skywave massive MIMO DFTS-OFDM system for a single antenna user. Assume the channel remains invariant within each DFTS-OFDM symbol, and utilize traditional OFDM symbols. In the subcarriers Each subcarrier is used for uplink data transmission. In a bit-interleaved coded modulation system, after coding interleaving, the original transmitted bitstreams of each user are mapped to a constellation. The symbols on, among which From bit vector The modulation order is obtained by mapping. N is the number of bits mapped to a single symbol. Define respectively... and users respectively The Each bit vector and symbol is transmitted, thus transmitting the signal in the frequency domain. pass of Point DFT generation. Let , ,but ,in for 3D normalized DFT matrix. Using ,user Time-domain transmitted signal on a DFTS-OFDM symbol Generated after subcarrier mapping and OFDM modulation:

[0038] (1)

[0039] in For subcarrier mapping matrix, It is a zero matrix. It is a K-dimensional identity matrix. After adding the cyclic prefix, it comes from... The time-domain signal of each user is transmitted in the skywave channel.

[0040] After removing the cyclic prefix at the base station side, the base station's... The time-domain received signal of the antenna is Therefore, the signal is received in the space-time domain. It can be represented as

[0041] (2)

[0042] Among users space-time domain channel Let be a one-vector block cyclic matrix, and let the space-time noise vector satisfy... , This represents the noise power. Therefore, the received signal in the spatial frequency domain... After performing OFDM demodulation and subcarrier demapping on each antenna of the base station, the following was obtained:

[0043] (3)

[0044] in , , ,and For users The spatial frequency domain channel, which has a vector block diagonal structure, can be represented as:

[0045] (4)

[0046] formula In the middle, users In subcarrier The channel vector at that location can be written as

[0047] (5)

[0048] in For users The diameter number, For users No. Complex gain of the stripe diameter, and Each with users No. Parameters related to the time delay and direction of the stripe. Pointing. Direction vector satisfy According to the beam-based channel model, the user In the The channel on each subcarrier can be approximated as:

[0049] (6)

[0050] in For users In subcarrier The beam domain channel vectors exhibit significant sparsity. The number of sampling points in the beam domain and This is a refinement factor. In In the middle, beam matrix Depend on The direction vectors in each direction constitute:

[0051] (7)

[0052] in , It is a matrix consisting of the first S rows of a 2S-dimensional identity matrix.

[0053] To demonstrate channel statistical characteristics, for users , No. The channel covariance matrix on each subcarrier is defined as follows:

[0054] (8)

[0055] in This is the beam-domain channel coupling vector, also defined as beam-domain statistical channel state information. In skywave massive MIMO systems, It exhibits significant sparsity, is independent of subcarriers, and remains unchanged over long periods. For simplicity, let... and . use ,user The set of non-zero beams can be defined as

[0056] (9)

[0057] This set corresponds to or The non-zero element subscript. In this embodiment, it is assumed that channel estimation has been performed during uplink data transmission, therefore the accurate beam-domain instantaneous channel state information is... and statistical channel state information Both are known at the base station.

[0058] II. Skywave Massive MIMO DFTS-OFDM Packet Interference Sparse Turbo Reception Method

[0059] For ease of discussion, the spatial frequency domain received signal in (3) is rewritten as

[0060] (10)

[0061] in It consists of frequency-domain transmitted signals from all users, and It contains all the message symbols sent by users. It can be verified. .also, ,in For subcarriers The spatial channel matrix on the space, and the equivalent channel Clearly, the signal model in (10) can also be represented in a subcarrier-by-subcarrier form:

[0062] (11)

[0063] in , and Subcarriers The frequency domain received signal, transmitted signal, and noise vector are all considered. According to (6), the subcarrier... The spatial channel matrix on can be modeled as ,in For subcarriers The sparse beam domain channel matrix.

[0064] Before designing the Turbo receiver, the first focus is on the transmitted symbols. The detection problem. Symbol The prior mean and variance are each a complex-valued vector. and a real-valued diagonal matrix In addition, symbols The average energy of each element is normalized to 1. To reduce the complexity of detection, the average energy of each element is normalized to 1. Approximately ,in , At this time, regarding MMSE detection can be converted into subcarrier-by-subcarrier frequency domain MMSE detection.

[0065] (12)

[0066] in For the MMSE detection results of the transmitted symbols on all subcarriers, and

[0067] (13)

[0068] (14)

[0069] in , , and This is the channel Gram matrix. and They can be considered as being based on subcarriers. MMSE detection and detector for received signals, where subcarriers The received signal is

[0070] (15)

[0071] Therefore, the discussion on signal detection can be conducted only on the received signal model (15) on each subcarrier.

[0072] Although MMSE detection can be performed in the frequency domain on a subcarrier-by-subcarrier basis, the computational complexity of (15) remains high for skywave massive MIMO scenarios with a large number of base station users and service users. Therefore, this paper explores a low-complexity signal detection method by utilizing the channel characteristics of skywave massive MIMO. The MMSE detection equation (13) can be rewritten as follows:

[0073] (16)

[0074] in Equation (16) shows that the optimal MMSE detection is equivalent to the signal after matched filtering. MMSE detection, among which

[0075] (17)

[0076] and This is the noise after matched filtering. Therefore, the observation... For optimal MMSE detection, this is sufficient. For skywave massive MIMO channels, the channel Gram matrix... Typically, interference sparsity is exhibited; this sparsity is defined as interference sparsity, indicating that each user experiences non-negligible interference only with a small number of other users. Utilizing interference sparsity is expected to reduce the computational complexity of signal detection. To achieve this, all users are first divided into... User groups ,and From user group The frequency domain transmitted signal is ,in Extract the matrix for the user, from the identity matrix. From It is constructed from columns. It can be proven that when the user channels of different user groups are mutually orthogonal, each user group only uses low-dimensional observations. The observations for MMSE detection are sufficient. In practice, the condition that user channels of different user groups are mutually orthogonal is difficult to strictly satisfy, while interference sparsity can approximate this condition. In this case, the interference groups for each user group are defined. And construct a set , Using this set as a low-dimensional observation extraction set can reduce detection complexity while maintaining detection performance. This is then used for detection. The observation vector is constructed as follows

[0077] (18)

[0078] in , ,and The observation extraction matrix is ​​derived from the identity matrix. From It is constructed from columns.

[0079] By minimizing MSE, Sparse detection of grouped interference is

[0080] (19)

[0081] in Grouped interference sparse detector

[0082] (20)

[0083] and During Turbo reception, the prior variance of the transmitted symbols... Since the detector is updated in each iteration, it also needs to be recalculated in each iteration, making it impossible to reuse within the coherence time and bandwidth. To address this issue, a pre-design method for grouped interference sparse detectors is proposed using prior variance quantization.

[0084] First, by utilizing the sparsity of interference, the observation vector It can be approximated as

[0085] (twenty one)

[0086] To further simplify the representation of the grouped interference sparse detector, the prior variance is approximated as... The mean variance That is, the prior variance is obtained by averaging the prior variances of the data symbols transmitted on each subcarrier by each user. Therefore, the sparse detector for packet interference based on the observation vector (21) can be written as

[0087] (twenty two)

[0088] in In (22), Still need to follow The update requires recalculation. To avoid this problem and allow... Pre-design before iteration, including variance Quantified as Value and for user groups Generate the corresponding Candidate detectors , of which The candidate detectors are

[0089] (twenty three)

[0090] and During the Turbo iteration process, variance The value of typically decreases rapidly, therefore exponential quantization is used.

[0091] (twenty four)

[0092] Therefore, the detection of sparse signals with grouped interference can be expressed as:

[0093] (25)

[0094] in For user groups Corrected observations , To reconstruct the signal, the detector Select from candidate detectors

[0095] (26)

[0096] in To make the quantized value closest The subscript.

[0097] As described above, according to step S1, the base station received frequency domain signal is processed on a per-subcarrier basis to obtain the dimension-reduced vector for each user group used for sparse detection of packet interference. The steps for obtaining the dimension-reduced vector are as follows: First, matched filtering is applied to the spatial domain received signal of each subcarrier, and then the matched-filtered received signal is... The signal reconstructed using the prior means of the frequency domain signal The mean-reduced matched-filtered signal is obtained by subtraction; then, interference selection is performed on the mean-reduced matched-filtered signal using the interference groups of each user group, and the dimension-reduced vector for each subcarrier and each user group for sparse detection of group interference is extracted. Based on step S2, using the dimensionality-reduced vectors of each user group and a pre-designed sparse detector for group interference selected according to the prior variance, the pre-designed sparse detector for group interference is implemented in the detection of sparse signals. Using the channel Gram matrix and a set of quantized variances According to formula (23), each quantization variance corresponds to a candidate detector. The detector corresponding to the quantization variance closest to the prior variance is selected for group interference sparse signal detection. Group interference sparse signal detection is achieved by using the conjugate transpose of the pre-designed group interference sparse detectors selected by each group based on the prior variance. The dimensionality reduction vectors obtained by interference selection Multiply by, and then by the frequency domain sign prior mean of each group. This is achieved through addition.

[0098] The aforementioned pre-designed sparse detector for packet interference is only related to the channel and the variance quantization method. It can be reused within the channel coherence time and coherence bandwidth, which can significantly reduce the design complexity of the detector in Turbo receivers.

[0099] The efficient computation of signal detection (25) will be discussed next, focusing primarily on the following: Relevant matched filter signals The computation is computationally efficient. In this embodiment, the process of performing matched filtering on the spatial received signals of each subcarrier is achieved by utilizing IFFTs of several base station antenna dimensions and the sparsity of the beam-domain channel. Specifically, according to the beam-based channel model, the matched-filtered signal can be expressed as...

[0100] (27)

[0101] To facilitate the calculation of (27), some DFT matrices have the following structure:

[0102] (28)

[0103] in Permutation matrix The Listed as Substituting (28) into (27) yields...

[0104] (29)

[0105] then, The calculation can be performed through several Point IFFT and The sparsity is effectively achieved and can be computed before iteration.

[0106] Calculated after, The detection can be obtained using the following formula.

[0107] (30)

[0108] in . The detected MSE is

[0109] (31)

[0110] in , In the Turbo reception process, in order to calculate the soft information output by the detector, let... Therefore, bits The outer log-likelihood ratio (LLR) is

[0111] (32)

[0112] in , for The prior LLR, initialized to 0 and updated via the decoder, is also included. The external mean and external variance are respectively and ,in After deinterleaving, decoding, and interleaving, the prior LLR of each bit is updated accordingly, and The prior mean and prior variance can be updated using the following formulas.

[0113] (33)

[0114] (34)

[0115] The prior probabilities of each symbol and bit are as follows:

[0116] (35)

[0117] (36)

[0118] As prior information for feedback, and Through respectively and Update, and via The signal is converted to the frequency domain for frequency-domain subcarrier signal detection in the next iteration. After several iterations, all the information bits sent by the users are finally recovered by the decoder.

[0119] Based on the above discussion, the large-scale MIMO DFTS-OFDM packet interference sparse Turbo reception includes the following steps: Step 1: Initialization , ;

[0120] Step 2: Calculate using (29) ;

[0121] Step 3: Calculate using (36) And updated via (33) and (34) and ;

[0122] Step 4: Calculation and utilize Sparsity computation ;

[0123] Step 5: Select via (26) ;

[0124] Step 6: Calculate using (25) ;

[0125] Step 7: Calculate using (30) and (31) respectively. and ;

[0126] Step 8: Calculate using (32) ;

[0127] Step 9: If Update after deinterleaving, decoding, and interleaving. ,set up Then proceed to step 3;

[0128] Step 10: Output the deinterleaved and decoded information bits.

[0129] In packet-sparse Turbo receivers, the construction of user groups and interference groups is crucial. To construct user groups, user groups are defined. and users The channel correlation is

[0130] (37)

[0131] then, It can roughly represent the user and users The degree of interference between them. Therefore, user groups can be constructed according to the following clustering method:

[0132] Step 1: Divide each user into a separate user group, with each user group containing only one user;

[0133] Step 2: If the current number of user groups is less than the preset number of user groups, merge the two user groups with the highest minimum channel correlation into a new user group;

[0134] Step 3: If the current number of user groups reaches the preset number of user groups, output the current user grouping result; otherwise, go to step 2.

[0135] After the user groups are created, in order to construct the corresponding interference groups for each user group, users are defined. and user groups The average channel correlation between them is

[0136] (38)

[0137] Therefore, the user group Corresponding interference group It can be constructed using the following method:

[0138] Step 1: Interference group Initialize to an empty set;

[0139] Step 2: If set The number of elements is not greater than a preset value. For those that do not belong users Calculate the average user relevance ,like If it exceeds the preset threshold, then Add to collection ;

[0140] Step 3: If set The number of elements has reached the preset value Or for those who do not belong users , If all values ​​are less than the preset threshold, then output the interference group. Otherwise, proceed to step 2.

[0141] Based on the above discussion, the construction of user groups and interference groups depends on the channel correlation between users, a parameter that needs to be updated as instantaneous channel state information changes. When user mobility is high or the ionospheric environment changes rapidly, the time-varying nature of the channel becomes very significant, leading to substantial update overhead for both user and interference groups. To overcome this problem, statistical channel state information is further used for user grouping. Similar to the definition of channel correlation, user groups are further defined... and users Statistical channel correlation between

[0142] (39)

[0143] and users and user groups Average statistical channel correlation between

[0144] (40)

[0145] Therefore, the method for constructing user groups and interference groups using statistical channel state information can be obtained by replacing the channel correlation and average channel correlation in the above methods with statistical channel correlation and average statistical channel correlation, respectively. Thanks to the slow-changing nature of statistical channel state information, user groups and interference groups constructed using statistical channel state information can be reused across multiple subcarriers and over a relatively long period.

[0146] VI. Implementation Results

[0147] To enable those skilled in the art to better understand the present invention, the performance results of the uplink receiving method in this embodiment under a specific configuration are given below.

[0148] Considering a skywave massive MIMO DFTS-OFDM communication system, the system parameters are configured as follows: carrier frequency Subcarrier spacing Number of subcarriers Number of effective subcarriers Number of base station antennas Base station antenna spacing Number of users Beam domain refinement factor The modulation scheme is 16th-order orthogonal amplitude modulation, and the channel coding uses low-density parity-check code with a code length of 2^112 and a code rate of 3 / 4. A row-column interleaver of length 528 is used. To measure uplink reception performance, the BER performance of the packet interference sparse Turbo receiver and the MMSE Turbo receiver under different signal-to-noise ratios (SNR) is compared. Here, SNR refers to the received SNR, measured in dB.

[0149] Figure 2 The BER performance graphs of the method in this embodiment under different numbers of user groups are given, where Represents the number of user groups. This represents the number of iterations, and MMSE TR represents the MMSE turbo receiver. The prior variance quantization order is set to... User groups and interference groups are constructed using statistical channel state information, where the statistical average channel correlation threshold is set to... As shown in the figure, the result of the grouped interference sparse Turbo receiver after 5 iterations is close to the result of the MMSE iterative receiver after 3 iterations, and the BER is [missing value]. Compared to the MMSE linear receiver (in the case of a single iteration of the MMSE Turbo receiver), it has a 2 dB gain. Furthermore, the performance of the packet interference sparse Turbo receiver is similar after 5 iterations for different numbers of packets, demonstrating the robustness of the proposed scheme.

[0150] Figure 3 The BER performance graphs of the method in this embodiment are shown at different variance quantization orders, where the number of user groups is... User groups and interference groups are constructed using statistical channel state information, where the statistical average channel correlation threshold is set to... As can be seen from the graph, when the number of iterations... At this point, a higher order of variance quantization will result in better performance. At that time, the packet interference sparse Turbo receiver is in The time-average performance is better than that of the MMSE linear receiver. It achieves performance close to that of a 3-iteration MMSE Turbo receiver. Even at a quantization order of... At that time, the packet interference sparse Turbo receiver in the number of iterations The performance of the receiver is basically the same as that of the MMSE linear receiver, and the prior variance of the packet interference sparse Turbo receiver is fixed at this time. And only the prior mean is updated in each iteration.

[0151] The present invention also discloses a skywave massive MIMO DFTS-OFDM uplink Turbo receiver system, used to implement the skywave massive MIMO described in any of the foregoing embodiments. The DFTS-OFDM uplink Turbo reception method includes: a matched filtering and interference selection module, used to process the frequency domain signal received by the base station on a per-subcarrier basis to obtain the dimensionality-reduced vector for each user group for sparse packet interference detection; a detector selection module, used to select a pre-designed sparse packet interference detector based on the prior variance; a signal detection module, used to perform sparse packet interference signal detection using the dimensionality-reduced vector of each user group and the selected sparse packet interference detector; and an iterative update module, used to process the signal detection results on a per-user basis along all subcarriers, performing IDFT, extrinsic information calculation, deinterleaving, soft-input soft-output decoding, interleaving, prior information update, and DFT on the frequency domain signal detection results of each user to obtain the prior mean and prior variance of the frequency domain signal of each user, which are fed back for signal reconstruction and selection of the pre-designed sparse packet interference detector, respectively, for the next round of iterative detection and decoding; after several rounds of iteration, the information bits of each user are finally decoded and output through a soft-input soft-output decoder.

[0152] This invention also discloses a computer system, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, it implements the steps of the uplink Turbo reception method of any of the foregoing embodiments.

[0153] This invention also discloses a computer program product, including a computer program that, when executed by a processor, implements the steps of the skywave massive MIMO DFTS-OFDM uplink Turbo reception method described in any of the foregoing embodiments.

[0154] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A method for uplink Turbo reception of skywave massive MIMO DFTS-OFDM, characterized in that, include: The frequency domain signal received by the base station is processed on a per-subcarrier basis to obtain the dimension-reduced vector for each user group for sparse detection of packet interference; Packet interference sparse signal detection is implemented by using the dimension-reduced vectors of each user group and a pre-designed packet interference sparse detector selected based on prior variance; The pre-designed packet interference sparse detector is calculated using the channel Gram matrix and a set of quantization variances. Each quantization variance corresponds to a candidate detector. The detector corresponding to the quantization variance that is closest to the prior variance is selected for packet interference sparse signal detection. The signal detection results are processed by combining all subcarriers on a user-by-user basis. For each user's frequency domain signal detection results, IDFT, extrinsic information calculation, deinterleaving, soft-input soft-output decoding, interleaving, prior information update, and DFT are performed to obtain the prior mean and prior variance of each user's frequency domain signal. These are fed back to signal reconstruction and the selection of pre-designed group interference sparse detectors for the next round of iterative detection and decoding. After several rounds of iteration, the user information bits are finally decoded and output through a soft-input soft-output decoder.

2. The skywave massive MIMO DFTS-OFDM uplink Turbo reception method according to claim 1, characterized in that, The step of processing the base station received frequency domain signal on a per-subcarrier basis to obtain the dimension-reduced vector for each user group used for sparse detection of packet interference includes: Matched filtering is applied to the spatial received signals of each subcarrier. The mean-removed matched filtered signal is obtained by subtracting the signal reconstructed using the a priori mean of the frequency domain signal from the matched filtered signal. Interference selection is performed on the mean-removed matched filter signal using the interference groups of each user group, and the dimension-reduced vectors of each subcarrier and each user group are extracted for sparse detection of grouped interference.

3. The skywave massive MIMO DFTS-OFDM uplink Turbo reception method according to claim 1, characterized in that, In the grouped interference sparsity detection, user groups and interference groups for each user group are constructed using instantaneous channel state information or statistical channel state information. This includes: calculating the channel correlation or statistical channel correlation between each user using instantaneous channel state information or statistical channel state information, constructing user groups using the channel correlation or statistical channel correlation through a clustering method, and constructing corresponding interference groups for each user group.

4. The skywave massive MIMO DFTS-OFDM uplink Turbo reception method according to claim 1, characterized in that, The prior variance used for selecting the pre-designed packet interference sparse detector is obtained by averaging the prior variances of the data symbols transmitted on each subcarrier for each user.

5. The skywave massive MIMO DFTS-OFDM uplink Turbo reception method according to claim 1, characterized in that, The pre-designed group interference sparse detector is only related to the channel and variance quantization method, and can be reused within the channel coherence time and coherence bandwidth.

6. The skywave massive MIMO DFTS-OFDM uplink Turbo reception method according to claim 1, characterized in that, The detection of the grouped interference sparse signal is achieved by multiplying the conjugate transpose of the pre-designed grouped interference sparse detector selected by each group according to the prior variance with the dimension-reduced vector of each group obtained by interference selection, and then adding it to the prior mean of the frequency domain symbol of each group.

7. The skywave massive MIMO DFTS-OFDM uplink Turbo reception method according to claim 1, characterized in that, The process of performing matched filtering on the spatial received signals of each subcarrier is achieved by utilizing the IFFT of several base station antenna dimensions and the sparsity of the beam domain channel.

8. A skywave massive MIMO DFTS-OFDM uplink Turbo receiving system, used to implement the skywave massive MIMO DFTS-OFDM uplink Turbo receiving method according to any one of claims 1-7, characterized in that, include: The matched filtering and interference selection module is used to process the frequency domain signal received by the base station on a per-subcarrier basis to obtain the dimension-reduced vector of each user group for sparse detection of group interference. The detector selection module is used to select a pre-designed sparse detector for grouped interference based on the prior variance. The signal detection module is used to perform packet interference sparse signal detection using the dimensionality reduction vectors of each user group and the selected packet interference sparse detector; The iterative update module is used to process the signal detection results by user-to-user joint processing of all subcarriers. It performs IDFT, external information calculation, deinterleaving, soft-input soft-output decoding, interleaving, prior information update and DFT on the frequency domain signal detection results of each user to obtain the prior mean and prior variance of the frequency domain signal of each user. These are fed back to signal reconstruction and selection of pre-designed group interference sparse detectors for the next round of iterative detection and decoding. After several rounds of iteration, the user information bits are finally decoded and output through a soft-input soft-output decoder.

9. A computer system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the computer program is executed by the processor, it implements the steps of the uplink Turbo reception method of the skywave massive MIMO DFTS-OFDM according to any one of claims 1-7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the uplink Turbo reception method of the skywave massive MIMO DFTS-OFDM according to any one of claims 1-7.

Images