Low-complexity multi-user detection method and device in satellite Internet of Things scene

Abstract

The embodiment of the invention relates to the field of data processing, and provides a low-complexity multi-user detection method and device in a satellite Internet of Things scene, and the method comprises the steps: building a satellite communication channel model of an uplink low earth orbit (LEO) satellite Internet of Things system based on a large-scale multiple-input multiple-output (mMIMO) model; determining a second satellite communication pilot signal received by the satellite side based on a satellite communication channel model of the uplink LEO satellite Internet of Things system and a first satellite communication pilot signal sent by the user side; carrying out sparse characteristic mining processing on the second satellite communication pilot signal to obtain a sparse characteristic and a quantized receiving signal; carrying out sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix; performing multi-user detection according to the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relation data; and under the condition that the speed reliability relation data meets a first preset condition, determining the reference channel estimation matrix as a target channel estimation matrix.

Description

Technical Field

[0001] The present application relates to the fields of data processing and wireless communication technology, and in particular to a low-complexity multi-user detection method and device in a satellite Internet of Things scenario. Background Art

[0002] In recent years, low-Earth orbit (LEO) satellite communication systems have developed rapidly. With their wider field of view, they can complement the coverage capabilities of terrestrial IoT in deserts, mountainous areas, and oceans, achieving seamless global coverage and providing access services for a large number of IoT devices. Massive Multiple Input Multiple Output (mMIMO) technology is a promising research direction for sixth-generation (6G) mobile communication systems, offering significant advantages in energy efficiency, reliability, and peak rate improvements. By equipping LEO satellites with massive antenna arrays, constructing an mMIMO LEO satellite communication system and adjusting channel gain and direction in real time based on user channel conditions, the communication capabilities of LEO satellite systems can be significantly improved. Currently, researchers have proposed various transmission schemes based on mMIMO technology. However, while improving the performance of LEO satellite communication systems, mMIMO faces challenges such as the high speed of LEO satellites, complex transmission environments, and limited onboard data processing capabilities. Further exploration is needed to ensure performance while reducing complexity, balancing the inherent contradictions between the large number of mMIMO connections and the limited computing resources of LEO satellites. Summary of the Invention

[0003] The embodiments of the present application provide a low-complexity multi-user detection method and device in a satellite Internet of Things scenario, which can automatically determine the corresponding target business products based on the user's product demand information, thereby improving the efficiency of low-complexity multi-user detection in the target satellite Internet of Things scenario.

[0004] A first aspect of an embodiment of the present application provides a low-complexity multi-user detection method in a satellite Internet of Things scenario, the method comprising:

[0005] Based on the massive multiple-input multiple-output (mMIMO) model, a satellite communication channel model for the uplink low-Earth orbit (LEO) satellite IoT system is established.

[0006] Determine a second satellite communication pilot signal received by the satellite side based on a satellite communication channel model of the uplink LEO satellite Internet of Things system and a first satellite communication pilot signal sent by the user side;

[0007] performing sparse feature mining processing on the second satellite communication pilot signal to obtain a sparse feature and a quantized received signal;

[0008] Perform sparse signal recovery processing based on the sparse characteristics and quantized received signals to obtain a reference channel estimation matrix;

[0009] Perform multi-user detection based on channel uncertainty and reference channel estimation matrix to obtain speed reliability relationship data;

[0010] When the speed reliability relationship data satisfies a first preset condition, the reference channel estimation matrix is determined as the target channel estimation matrix.

[0011] In this example, a satellite communication channel model of an uplink low-Earth orbit (LEO) satellite IoT system is established based on a massive multiple-input multiple-output (mMIMO) model. A second satellite communication pilot signal received by the satellite side is determined based on the satellite communication channel model of the uplink LEO satellite IoT system and a first satellite communication pilot signal sent by the user side. Sparse feature mining is performed on the second satellite communication pilot signal to obtain sparse features and a quantized received signal. Sparse signal recovery is performed based on the sparse features and the quantized received signal to obtain a reference channel estimation matrix. Multi-user detection is performed based on the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data. When the speed reliability relationship data meets a first preset condition, the reference channel estimation matrix is determined to be the target channel estimation matrix. This is beneficial for simplifying the complexity of calculations in satellite communication scenarios with limited computing resources and obtaining a more accurate target channel estimation matrix.

[0012] A second aspect of an embodiment of the present application provides a low-complexity multi-user detection device in a satellite Internet of Things scenario, the device comprising:

[0013] The first processing unit is configured to establish a satellite communication channel model for an uplink low earth orbit (LEO) satellite IoT system based on a massive multiple-input multiple-output (mMIMO) model.

[0014] A first determining unit is configured to determine a second satellite communication pilot signal received by a satellite side based on a satellite communication channel model of the uplink LEO satellite Internet of Things system and a first satellite communication pilot signal sent by a user side;

[0015] A second processing unit is configured to perform sparse feature mining processing on the second satellite communication pilot signal to obtain a sparse feature and a quantized received signal;

[0016] a third processing unit, configured to perform sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix;

[0017] a fourth processing unit, configured to perform multi-user detection based on the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data;

[0018] The second determining unit is configured to determine, when the speed reliability relationship data satisfies a first preset condition, that the reference channel estimation matrix is a target channel estimation matrix.

[0019] A third aspect of an embodiment of the present application provides a terminal, comprising a processor, an input device, an output device, and a memory, wherein the processor, input device, output device, and memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions and execute the step instructions in the first aspect of the embodiment of the present application.

[0020] The fourth aspect of the embodiments of the present application provides a computer-readable storage medium, wherein the above-mentioned computer-readable storage medium stores a computer program for electronic data exchange, wherein the above-mentioned computer program enables a computer to execute some or all of the steps described in the first aspect of the embodiments of the present application.

[0021] A fifth aspect of the embodiments of the present application provides a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program, wherein the computer program is operable to cause a computer to perform some or all of the steps described in the first aspect of the embodiments of the present application. The computer program product may be a software installation package. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0023] Figure 1 The present invention provides a flowchart of a low-complexity multi-user detection method in a satellite Internet of Things scenario;

[0024] Figure 2 An architectural diagram of a satellite communication channel model for an uplink LEO satellite IoT system is provided for an embodiment of the present application;

[0025] Figure 3 A schematic diagram of the structure of a terminal provided in an embodiment of the present application;

[0026] Figure 4 The present invention provides a schematic structural diagram of a low-complexity multi-user detection device in a satellite Internet of Things scenario. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0028] The terms "first," "second," and the like in the specification and claims of this application and the accompanying drawings are used to distinguish between different objects, not to describe a particular order. Furthermore, the terms "including," "having," and any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements but may optionally include steps or elements not listed, or may optionally include other steps or elements inherent to the process, method, product, or apparatus.

[0029] References to "embodiments" in this application mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described in this application may be combined with other embodiments.

[0030] In order to better understand the low-complexity multi-user detection method in a satellite Internet of Things scenario provided by an embodiment of the present application, the scenario of applying the low-complexity multi-user detection method in the satellite Internet of Things scenario is first briefly introduced below. In order to build a three-dimensional network integrating air, space, and land with global wide coverage and realize seamless three-dimensional super connection, the future satellite Internet of Things communication system faces a key challenge: in the actual mMIMO LEO satellite communication scenario, the on-board computing resources are limited, and it is difficult to meet the ultra-high overhead required for the communication system calculation. How to achieve massive access while meeting the low-overhead requirements is a key issue to be solved in the development of the mMIMO LEO satellite Internet of Things industry. Therefore, in the LEO satellite communication scenario, when a large number of users access at the same time, there is an urgent need for an optimization solution for computing resources to realize the satellite Internet of Things.

[0031] Due to the limited onboard data processing capabilities, large connection scale, and sparse user activity in satellite IoT, classic signal detection algorithms are no longer suitable for the complex and diverse wireless transmission scenarios of the future. Compressed sensing theory can fully exploit the sparse nature of signals, enabling signal recovery from a relatively small number of measurements. Therefore, incorporating compressed sensing theory to address the sparsity of the user activity matrix and the sparse channel impulse response (CIR) in LEO satellite IoT, the channel estimation problem is formulated and solved as a sparse signal recovery problem, avoiding the high computational overhead caused by the excessive number of connected users. Furthermore, the limited computing power of LEO satellites cannot meet the high computational complexity requirements of traditional linear multi-user detection algorithms. By incorporating a channel state information (CSI) error estimation model, the detection algorithm's reliance on accurate CSI can be reduced, improving its robustness. Finally, the multi-user detection problem is formulated as a strict convex quadratic optimization problem and solved iteratively using the BFGS quasi-Newton method. This overcomes the high computational complexity of traditional linear detection schemes due to matrix inversion, making it applicable to scenarios with limited LEO satellite payloads.

[0032] The purpose of the present invention is to propose a multi-user detection scheme for satellite Internet of Things communication for massive user access under the condition of limited satellite computing resources. This method combines LEO satellite communications with the relatively mature massively multiple-in-one (MMIMO) technology in recent years. Taking advantage of the large propagation distance between the satellite and the ground and the high-speed mobility of LEO satellites, a satellite communication channel model is established in the uplink based on key channel analysis techniques such as spatial geometry analysis. The sparse characteristics of the active user matrix in the satellite IoT scenario and the sparse characteristics of its uplink transmission channel in the angle domain are utilized to formulate the joint user activity detection and channel estimation problem as a sparse signal recovery problem, which is solved using the Simultaneous Orthogonal Matching Pursuit (SOMP) algorithm with low computational complexity. A low-computational-complexity multi-user detection scheme is designed based on the Broyden-Fletcher-Goldfarb-Shanno (BFGS) quasi-Newton algorithm. Combined with a CSI error estimation model, the multi-user detection algorithm's reliance on accurate CSI is reduced. Finite Block Length (FBL) information theory is used to derive the maximum achievable user data rate and post-processing signal-to-noise ratio (PSNR). Ratio, PPSNR) to explore the data rate achievable boundaries under actual constraints in satellite IoT communication scenarios.

[0033] See also Figure 1 , Figure 1 The present invention provides a flowchart of a low-complexity multi-user detection method in a satellite Internet of Things scenario, which includes:

[0034] S10: Based on the massive multiple-input multiple-output (mMIMO) model, a satellite communication channel model for the uplink low-Earth orbit (LEO) satellite IoT system is established.

[0035] The mMIMO model is an advanced wireless communication technology that enables simultaneous communication with numerous user devices by equipping base stations (LEO satellites in this case) with a large number of antennas. The mMIMO model can significantly improve the spectrum efficiency, energy efficiency, and reliability of communication systems. Through spatial multiplexing and diversity technologies, it allows multiple users to transmit data simultaneously on the same time-frequency resources, effectively increasing system capacity. In this application, the mMIMO model can be used to provide an infrastructure for establishing a satellite communication channel model for a satellite IoT system to support access by a large number of users.

[0036] The satellite communication channel model for an uplink LEO satellite IoT system can be understood as a mathematical model used to describe the transmission characteristics of signals in the channel when transmitting signals from a ground-based IoT user device to a LEO satellite. When establishing the satellite communication channel model for an uplink LEO satellite IoT system, factors such as line-of-sight (LoS) and non-line-of-sight (non-LoS) propagation paths, Doppler shift, and propagation delay can be considered, but this application does not impose any restrictions on these factors.

[0037] See also Figure 2 , Figure 2 The schematic diagram of the satellite communication channel model for the uplink LEO satellite IoT system is shown in Figure 2. For example, assume that in a given time slot, a LEO satellite serves K single-antenna users simultaneously, and each satellite is equipped with N a =N x N y A uniform planar array consisting of N receiving antennas, x and N y are the number of antennas along the x-axis and y-axis respectively. In addition, each user transmits an orthogonal frequency division multiplexing (OFDM) symbol containing D bits of information through N subcarriers, occupying a bandwidth of B and a delay of t s , the length of the satellite communication pilot signal is m, which is not limited in this application.

[0038] S20: Determine a second satellite communication pilot signal received by the satellite side based on the satellite communication channel model of the uplink LEO satellite Internet of Things system and the first satellite communication pilot signal sent by the user side.

[0039] The first satellite communication pilot signal can be understood as a satellite communication pilot signal sent by the user side to assist in channel estimation and synchronization. In an OFDM system, the satellite communication pilot signal has a specific distribution in the frequency domain, and its time domain representation can be obtained by inverse Fourier transforming the frequency domain representation, which is not limited in this application. The second satellite communication pilot signal can be understood as a satellite communication pilot signal received by the satellite side after channel transmission and various processing. This second satellite communication pilot signal can be a signal that has been affected by the first satellite communication pilot signal after channel transmission and noise interference.

[0040] It should be understood that determining the second satellite communication pilot signal received by the satellite side based on the satellite communication channel model of the uplink LEO satellite IoT system and the first satellite communication pilot signal sent by the user side refers to the process of converting the first satellite communication pilot signal sent by the user side into the second satellite communication pilot signal received by the satellite side after the first satellite communication pilot signal is transmitted through the channel and is affected by noise interference. Step S20, i.e., determining the second satellite communication pilot signal received by the satellite side based on the satellite communication channel model of the uplink LEO satellite IoT system and the first satellite communication pilot signal sent by the user side, may include the following steps:

[0041] S21: Determine a satellite uplink transmission spatial domain channel impulse response (CIR) based on a satellite communication channel model of the uplink LEO satellite IoT system;

[0042] S22: Determine a time domain representation of the first satellite communication pilot signal according to the frequency domain representation of the first satellite communication pilot signal sent by the user side;

[0043] S23: performing compensation processing on the time domain representation of the first satellite communication pilot signal to obtain a first compensated satellite communication pilot signal;

[0044] S24: Determine a second satellite communication pilot signal received by the satellite side according to the first compensated satellite communication pilot signal and the satellite uplink transmission spatial domain CIR.

[0045] Among them, the satellite uplink transmission spatial domain CIR can be a function that describes the impact of the satellite uplink transmission channel on the input signal. It can reflect the signal changes caused by factors such as multipath propagation, Doppler frequency shift, and signal attenuation during the transmission process from the user side to the satellite side.

[0046] Considering that LEO satellite communication systems are usually line-of-sight propagation, but user terminals are located in areas blocked by obstacles such as mountains and forests, when the direct component is blocked, it will lead to non-LoS propagation caused by multipath effects. Therefore, it is reasonable to assume that there are LoS paths and a limited number of non-LoS paths in the satellite Internet of Things scenario. Therefore, the spatial domain CIR of the satellite uplink transmission between user k and LEO at time t and frequency f can be expressed as:

[0047]

[0048] in, represents the satellite uplink transmission spatial domain CIR between user k and LEO satellite at time t and frequency f, where t represents time and f represents frequency; κ k is the Rician coefficient; represents the LoS path channel gain of user k; e represents a natural constant, j represents an imaginary unit, and π represents pi; represents the LoS path Doppler shift of user k; The LoS path propagation delay of user k; L k represents the number of non-LoS paths for user k; l represents the lth non-LoS path for user k; a k,l is the channel gain of the lth non-LoS path of user k; v k,l The Doppler shift of the lth non-LoS path of user k; τ k,l represents the propagation delay of the lth non-LoS path of user k; is the array response vector on the satellite side.

[0049] It should be noted that the first term on the right side of the equal sign represents the LoS path, and the second term represents the non-LoS path. Optionally, for the sake of brevity, the superscript of the LoS path is omitted in the following text.

[0050] In communication systems, frequency domain representation can describe the different frequency components of a signal, as well as their amplitude and phase information. The mathematical description of the first satellite communication pilot signal sent by the user in the frequency domain can be used to analyze the signal's frequency characteristics, such as the energy distribution of different frequency components. Corresponding to the frequency domain representation, the time domain representation can describe how the signal changes over time. Specifically, the frequency domain representation of the first satellite communication pilot signal can be derived from its corresponding time domain representation through an inverse Fourier transform, i.e., the time domain representation of the first satellite communication pilot signal.

[0051] Optionally, the satellite communication pilot signal (i.e., the first satellite communication pilot signal) sent by user k is Take θ as an example, therefore, the signal can be expressed in the time domain as:

[0052]

[0053] Among them, x k (t) is the time domain representation of the first satellite communication pilot signal; t represents time; x k,r is the frequency domain representation of the first satellite communication pilot signal; e represents a natural constant, j represents an imaginary unit, and π represents the ratio of pi; Δf is the subcarrier spacing; m is the length of the satellite communication pilot signal, and r is the index of the satellite communication pilot signal.

[0054] Furthermore, in LEO satellite communication systems, due to the high-speed satellite movement and long-distance transmission, signals experience Doppler shift and propagation delay, which can lead to signal distortion and synchronization difficulties. Compensation processing can correct for these effects. By performing compensation processing on the time-domain representation of the first satellite communication pilot signal, the negative effects of Doppler shift and propagation delay can be eliminated or reduced. The first compensated satellite communication pilot signal can be understood as the satellite communication pilot signal obtained after compensation processing.

[0055] In LEO satellite communication systems, since the orbit of the satellite is generally known, the Doppler frequency shift caused by the side movement of the LEO satellite can be and propagation delay due to long-distance transmission After pre-compensation, the satellite communication pilot signal after compensation can be expressed as:

[0056]

[0057] in, is the compensated satellite communication pilot signal, i.e., the first compensated satellite communication pilot signal; x k ( ) is the time domain representation of the first satellite communication pilot signal; t represents time; is the propagation delay caused by long-distance transmission; e represents a natural constant, j represents an imaginary unit, and π represents the ratio of circumference to circumference; It is the Doppler shift caused by the lateral movement of the LEO satellite.

[0058] Therefore, the second satellite communication pilot signal received by the satellite side can be expressed as:

[0059]

[0060] in, is a satellite communication pilot signal received by the satellite side, i.e., a second satellite communication pilot signal; It is h k The inverse Fourier transform of (t,f) with respect to τ, Its elements are independent and identically distributed Additive complex Gaussian noise.

[0061] It is understandable that the second satellite communication pilot signal may include information such as channel characteristics, transmission signals, and noise, which is an important basis for subsequent channel estimation and multi-user detection. This application does not impose any restrictions on this.

[0062] S30: Performing sparse feature mining processing on the second satellite communication pilot signal to obtain sparse features and a quantized received signal.

[0063] This application can use the sparse characteristics of the active user matrix and channel in the angle domain in the satellite Internet of Things scenario for analysis and processing, providing a basis for subsequently converting the channel estimation problem into a sparse signal recovery problem, which helps to reduce computational complexity. Since there may be only a few active users sending signals at the same time, and the satellite uplink transmission channel exhibits good sparsity in the angle domain (the number of non-LoS path signals is limited), by converting the received signal to the angle domain and processing the angle domain received signal, the angle domain channel matrix can be determined, thereby mining the sparse characteristics of the user and channel matrices in the angle domain. This application does not impose any restrictions on this.

[0064] The quantized received signal can be understood as the result of quantizing the signal received from the satellite. Quantization is the process of converting a continuous analog signal into a discrete digital signal. Because the quantized received signal retains key information while facilitating digital signal processing and storage, it can be used for subsequent sparse signal recovery.

[0065] It should be understood that performing sparse feature mining on the second satellite communication pilot signal to obtain sparse features and a quantized received signal refers to the process of mining the sparse features of the user and channel matrices in the angle domain and quantizing the signal received on the satellite side to obtain a quantized received signal. Step S30, i.e., performing sparse feature mining on the second satellite communication pilot signal to obtain sparse features and a quantized received signal, may include the following steps:

[0066] S31: Determine satellite-side received signals of K users based on the second satellite communication pilot signal;

[0067] S32: performing angle domain conversion processing on the satellite side received signal to obtain an angle domain received signal;

[0068] S33: Determine an angle domain channel matrix according to the angle domain received signal;

[0069] S34: Determine, based on the angle-domain channel matrix, a sparse characteristic of the K users and the channel matrix in the angle domain;

[0070] S35: Determine a quantized received signal according to the angle domain channel matrix.

[0071] In the satellite IoT scenario, only a few active users send signals at the same time, and the user activity status is unknown a priori. Therefore, user activity detection is required before channel estimation and multi-user detection. Therefore, the signal received by the satellite at the same time, that is, the satellite-side received signal of K users, can be expressed as:

[0072]

[0073] Where Y represents the satellite side receiving signal of K users, K represents the number of users, k represents user k, and b k ∈{0,1} represents the active state of user k, x k represents the second satellite communication pilot signal sent by user k, h k represents the spatial domain CIR of user k and satellite, Z represents additive complex Gaussian noise, X=[x1… x k ] represents the second satellite communication pilot signal sent by all users at the same time; represents the spatial domain channel between K users and the LEO satellite.

[0074] It should be noted that b k ∈{0,1} represents the active state of user k, when b k = 1 indicates that user k is active; otherwise, it is inactive. represents the set of active users and K a Indicates the number of currently active users.

[0075] Because LEO satellite uplink transmission channels exhibit significant sparseness in the angular domain, converting satellite-side received signals from their original time or frequency domains to the angular domain can more effectively exploit signal characteristics. This conversion is typically achieved through mathematical transformations such as the discrete Fourier transform (DFT), converting the signal from one domain representation to the angular domain for easier analysis.

[0076] The angle-domain received signal can be understood as the signal obtained after angle-domain conversion processing. In the angle domain, the different angular components of the signal can reflect the directional information of different user signals and the angular correlation characteristics of the channel. For example, different user signals may have different energy distributions in the angle domain. By analyzing the angle-domain received signal, different user signals and their relationship with the channel can be more clearly distinguished, and this application does not impose any restrictions on this.

[0077] Optionally, formula (5) is converted to the angle domain to fully utilize its sparse characteristics in the angle domain. At this time, the received signal, that is, the angle domain received signal, can be expressed as:

[0078] R=YW DFT (:,i)=XH T *W DFT (:,i)+Z*W DFT (:,i),(6)

[0079] Where R represents the angle domain received signal, Y represents the satellite communication pilot signal received by the satellite side, and W DFT (:,i) is the discrete Fourier transform matrix, X represents the second satellite communication pilot signal sent by all users at the same time, H T represents the spatial domain channel between the user and the LEO satellite, and Z represents the additive complex Gaussian noise.

[0080] The angle-domain channel matrix can be understood as a matrix determined based on the angle-domain received signals. It describes the characteristics of the channels between K users and the satellite in the angle domain. The matrix elements contain information such as the channel gain and phase change for different user signals in the angle domain. The angle-domain channel matrix can be used to determine the sparse nature of the channel and perform signal detection and estimation. For example, in multi-user detection, the angle-domain channel matrix can help determine which user signals have strong energy in the current angle domain and which signals at angles can be ignored, thereby simplifying signal processing.

[0081] In satellite IoT scenarios, only a few users are active at any given time, and the number of signals along non-line-of-sight (LoS) paths is limited. This results in a sparse representation of the K users and the channel matrix in the angular domain. This sparsity manifests itself as signal components at only a few angles having significant energy, while those at most angles have very little or even negligible energy. Identifying this sparsity can facilitate signal processing using compressed sensing theory, recovering high-dimensional channel information from a small amount of measurement data and reducing computational complexity. For example, in channel estimation, the sparsity property can be used to focus on angular components with significant energy, thereby reducing unnecessary computational effort.

[0082] Since the path loss of LEO satellite communication is very high, the number of signals received by the satellite that belong to non-LoS paths is relatively limited. The sparse characteristics of CIR in the angle domain can be expressed as where supp{·} is the support set, |A| c is the cardinality of set A. And since there are only a few active users transmitting signals to the satellite in the same time slot, that is, Therefore, the sparse characteristics of the channel matrix between K users and LEO satellites in the angle domain can be expressed as:

[0083]

[0084] Among them, H a represents the angle domain channel matrix, K represents the active user set, L k represents the number of non-LoS paths of user k, K represents the number of users, and m represents the length of the satellite communication pilot signal.

[0085] The quantized received signal can be understood as the result of quantizing the angle domain received signal according to the angle domain channel matrix. The quantization process can be a process of converting a continuous analog signal into a discrete digital signal. For example, by setting certain quantization rules, the continuous values of the amplitude and phase of the angle domain received signal can be mapped to a finite number of discrete values. The quantized received signal can facilitate digital signal processing and storage, and at the same time can reduce redundant information in the signal transmission and processing process to a certain extent, thereby improving system efficiency. In actual communication systems, the quantized received signal can be used as input data for subsequent signal processing steps (such as sparse signal recovery, multi-user detection, etc.), and this application does not impose any restrictions on this.

[0086] Optionally, formula (6) can be vectorized row by row as follows:

[0087]

[0088] Among them, r represents the received signal after vectorization, represents noise, represents the angle domain channel matrix after vectorization, Indicates size N a The identity matrix, It represents the result of multiplying the satellite communication pilot signal sent by the user by the identity matrix (no specific name), which can be used as the measurement matrix in sparse signal recovery.

[0089] It should be noted that the angle domain matrix h is obtained from formula (7): a It has a sparse structural characteristic. Therefore, the sparse characteristics of the matrix can be used to recover the high-dimensional channel matrix.

[0090] S40: Perform sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix.

[0091] The reference channel estimation matrix can be understood as the channel estimation result obtained after sparse signal recovery processing. This reference channel estimation matrix can be an approximate estimate of the actual channel and can reflect the channel characteristics between the user and the satellite, including but not limited to information such as channel gain and phase. In subsequent multi-user detection, the reference channel estimation matrix can serve as an important parameter for detecting user signals, which is not limited in this application.

[0092] This application is based on the theory of compressed sensing and uses the sparse characteristics of the signal to recover the original high-dimensional signal from a small number of measurements. In this application, the joint user activity detection and channel estimation problem is formulated as a sparse signal recovery problem and can be solved by the synchronous orthogonal matching pursuit (SOMP) algorithm. Among them, the SOMP algorithm uses the correlation maximization criterion to select multiple users, and uses the least squares method to solve the channel estimation matrix corresponding to the set of active users, and continuously updates the residual signal until the preset conditions are met, and finally the above-mentioned reference channel estimation matrix can be obtained.

[0093] It should be understood that performing sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix refers to the process of formulating the channel estimation problem as a sparse signal recovery problem and solving it using the SOMP algorithm. In step S40, that is, performing sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix, may include the following steps:

[0094] S41: Initializing the quantized received signal as an initial residual signal;

[0095] S42: Selecting a current active user based on the initial residual signal and the measurement matrix using a maximum correlation criterion;

[0096] S43: The current active user and the historical active users are combined into an active user set;

[0097] S44: Based on the active user set, select corresponding column vectors from the measurement matrix to obtain a first matrix;

[0098] S45: Determine a first channel estimation matrix corresponding to the active user set according to the first matrix using a least squares method;

[0099] S46: performing update processing on the initial residual signal according to the first channel estimation matrix to obtain a first residual signal;

[0100] S47: When the first residual signal satisfies a second preset condition, determine the first channel estimation matrix as a reference channel estimation matrix.

[0101] Currently, compressed sensing theory has proven that if the objective function is sparse or approximately sparse, high-dimensional signals can be accurately recovered from low-dimensional observation signals. Because the SOMP algorithm can fully utilize the sparse characteristics of active users in satellite IoT scenarios and the sparse characteristics of uplink transmission channels in the angular domain, we choose to use the SOMP algorithm to solve the sparse signal recovery problem.

[0102] The specific steps of the SOMP algorithm are as follows:

[0103] ① Initialization: The residual signal is the received signal on the satellite side, that is, R (0) =r, active user set

[0104] The initial residual signal can be the quantized received signal. In the sparse signal recovery process, the residual signal represents the difference between the current estimate and the true signal. Initially, there are no other more accurate estimates, so the quantized received signal is used as the initial difference signal. This is then iterated and continuously updated to bring it closer to the true signal.

[0105] ②Use the relevance maximization criterion to select multiple users:

[0106]

[0107] Among them, η i Indicates the multiple users selected in this iteration, that is, the current active users; C i =Φ H R (i-1) , R (i-1) Represents the residual signal of the last iteration result, Γ i-1 =I KNa / Λ i-1 . And order:

[0108] Λ i ={η i}∪Λ i-1 (10)

[0109] Among them, Λ i represents the set of active users in this iteration, η i Represents multiple users selected in this iteration, i.e., currently active users; Λ i-1 Represents the set of active users in the previous iteration, that is, historical active users.

[0110] In compressed sensing theory, the measurement matrix is the key matrix that connects the high-dimensional original signal and the low-dimensional observation signal. The measurement matrix performs a linear projection of the original signal to obtain low-dimensional observations. In this scenario, the measurement matrix is related to the channel characteristics and signal transmission method of the satellite IoT system and is used to extract information related to the user signal from the quantized received signal.

[0111] The maximum correlation criterion is a strategy for selecting currently active users. Active users in satellite IoT scenarios are sparse, meaning only a few transmit signals at a time. By calculating the correlation between the initial residual signal and each column of the measurement matrix, the users corresponding to the columns with the highest correlation are selected as currently active users. The greater the correlation, the greater the contribution of the user's signal to the current residual signal, making them more likely to be active users.

[0112] ③Through Λ i Select the corresponding i-th column in the measurement matrix Φ and use Then the least squares method is used to solve the channel estimation matrix corresponding to the active user set selected in this iteration, that is,

[0113]

[0114] in, represents the channel estimation matrix of this iteration, that is, the first channel estimation matrix; Indicates that through Λ i Select the first matrix composed of the corresponding i-th column in the measurement matrix Φ, r is the received signal, h a Represents the angle domain channel.

[0115] The first matrix can be understood as a matrix consisting of column vectors corresponding to users in the active user set, selected from the measurement matrix. This first matrix concentrates channel information related to active users and can be used to subsequently calculate the channel estimation matrix corresponding to the active user set. It is an important intermediate data in the channel estimation process.

[0116] The least squares method can be used to find the optimal fitting model parameters in measurement data with errors. In this step, the least squares method can determine the channel estimation matrix corresponding to the active user set by minimizing the sum of squared errors between the estimated signal and the received signal. The first channel estimation matrix can be understood as the result obtained by processing the first matrix using the least squares method. The first channel estimation matrix can be an estimate of the channel between the active user and the satellite, and can reflect the characteristics of the user signal during channel transmission, such as channel gain, phase change, etc. This first channel estimation matrix can be continuously optimized with iterations, gradually approaching the actual channel conditions.

[0117] ④ Signal estimation matrix based on this iteration Update the residual signal, that is, get the first residual signal R (i) .

[0118] The estimated signal calculated using the first channel estimation matrix is then subjected to an operation (typically a subtraction) on the initial residual signal to obtain a first residual signal. The first residual signal can reflect the remaining difference considering the currently estimated channel, which helps provide a more accurate basis for the next iteration.

[0119] ⑤ Determine the residual signal (i.e. the first residual signal) and noise power σ 2 If the size relationship Then stop the iteration; if Repeat from step ② and continue to iterate; the final channel estimation matrix, that is, the reference channel estimation matrix, can be

[0120] The second preset condition can be understood as a manually set judgment standard, which can be used to decide whether to stop the iteration. The common second preset condition can be as above, such as judging that the energy (such as power) of the first residual signal is less than a certain threshold (such as noise power σ 2 ); Optionally, the second preset condition may also be determining whether the variation of the residual signal is less than a certain threshold, etc., which is not limited in this application. When the first residual signal satisfies the second preset condition, it is considered that the current channel estimation is sufficiently accurate.

[0121] From formula (8), we can see that By further Performing discrete inverse Fourier transform can obtain the channel estimation matrix in the spatial domain

[0122] In addition, the active user set K can be determined by formula (12).

[0123]

[0124] in, represents the user's active state, η(x,ω) is the threshold function, Represents the channel estimation matrix, that is, when When user k’s active state is b k =1, and its decision threshold ω is generally set to 0.1 times or 0.2 times the energy of the transmitted signal.

[0125] S50: Perform multi-user detection according to the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data.

[0126] Channel uncertainty can be used to describe the degree of channel estimation inaccuracy. Since satellites cannot obtain perfect channel state information (such as CSI), channel estimation errors exist. Channel uncertainty can be understood as a quantified representation of the degree of this error. In multi-user detection, by comprehensively considering channel uncertainty and the reference channel estimation matrix, the detection algorithm's reliance on accurate CSI can be reduced, thereby improving the algorithm's robustness and making the detection results more reliable.

[0127] Multi-user detection can be understood as the process of separating the signals sent by individual users from the mixed signals received by the satellite and detecting the active users. Traditional linear multi-user detection algorithms have high computational complexity and are difficult to apply when satellite computing resources are limited. This application utilizes a multi-user detection algorithm based on the BFGS quasi-Newton method and incorporates channel uncertainty to avoid matrix inversion operations. This not only reduces computational complexity but also improves the robustness of the algorithm, resulting in speed-reliability relationship data.

[0128] Speed reliability relationship data can be understood as data that can reflect the relationship between signal transmission speed and signal quality (such as PPSNR) in multi-user detection results. In this application, the received signal is processed by a multi-user detection algorithm to obtain detection results of different users. The detection result may include relevant information about the signal transmission speed and quality, such as the maximum achievable data rate of the user (i.e., speed-related) and PPSNR (i.e., quality-related). This relevant information can be used to evaluate the performance of the multi-user detection algorithm and determine whether the reference channel estimation matrix meets the requirements, and this application does not impose any restrictions on this.

[0129] It should be understood that performing multi-user detection based on the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data refers to the process of how to perform multi-user detection and how to determine the speed reliability relationship data based on the multi-user detection results. In step S50, performing multi-user detection based on the channel uncertainty and the reference channel estimation matrix to obtain the speed reliability relationship data may include the following steps:

[0130] S51: determining initial gradient data according to the reference channel estimation matrix, channel uncertainty and the satellite side received signal;

[0131] S52: Determine initial search direction data according to the initial gradient data;

[0132] S53: Determine a minimum mean square error filter matrix according to the noise power, the channel hardening characteristic, the reference channel estimation matrix and the channel uncertainty;

[0133] S54: updating the iteration step size according to the minimum mean square error filter matrix, the initial search direction data, and the initial gradient data to obtain a first iteration step size;

[0134] S55: updating the iteration point according to the first iteration step size, the initial search direction data, and the initial iteration point to obtain a first iteration point;

[0135] S56: If a third preset condition is met, determining that the first iteration point is a transmitted signal after multi-user detection;

[0136] S57: Determine speed reliability relationship data based on the transmitted signal after multi-user detection, the minimum mean square error filter matrix and the transmission power of the user data signal.

[0137] Linear detection schemes have been proven to be near-optimal in mMIMO systems, but require high-complexity matrix inversion operations. The computing power of satellites is limited and cannot meet such high computational complexity requirements. Therefore, this application adopts the recognized and effective BFGS quasi-Newton method for multi-user detection.

[0138] In addition, since the satellite side cannot obtain perfect CSI, the channel estimation error will lead to the degradation of multi-user detection performance and increase the probability of detection error. Therefore, the CSI error estimation model is comprehensively considered. Where H represents the real channel matrix, is the channel matrix estimated by the SOMP algorithm, and ΔH represents the channel uncertainty. By combining this with the CSI error estimation model, the multi-user detection algorithm can reduce its reliance on accurate CSI and improve its robustness, making it suitable for scenarios with limited satellite payload.

[0139] The specific steps of the BFGS quasi-Newton multi-user detection algorithm are as follows:

[0140] ① Initialization: Assume S represents the number of iterations, let s = 0. Iteration point p (0) =0, gradient value (i.e. initial gradient data) Search direction (i.e. initial search direction data) in

[0141] Initial gradient data can be understood as the gradient value determined by combining the reference channel estimation matrix, channel uncertainty, and the satellite-side received signal. The reference channel estimation matrix reflects the channel estimation, while the channel uncertainty reflects the degree of error in the channel estimation. The satellite-side received signal contains information about the user's transmitted signal after it has been transmitted through the channel. By performing calculations on this data (specifically, correlation calculations based on the signal model and estimation error), initial gradient data can be obtained. This represents the change trend of the objective function in its initial state and provides a basis for determining the subsequent search direction.

[0142] The initial search direction can be determined based on the initial gradient data. Typically, the initial search direction is associated with the negative direction of the initial gradient (e.g., the negative gradient direction). In optimization algorithms, the negative gradient direction indicates the direction in which the function value decreases most rapidly. Using this as the initial search direction is expected to quickly find the optimal solution to the objective function, helping to quickly approximate the true user signal in multi-user detection.

[0143] ② Update the iteration step δ (s) :

[0144]

[0145] Among them, δ (s) represents the updated iteration step, i.e. the first iteration step; p (s) Indicates the iteration point; d (s) Indicates the search direction; M = H T H+σ 2 υ represents the minimum mean square error filter matrix; υ represents the channel hardening characteristic, that is, in the satellite IoT scenario, the number of transmitting antennas on the user side is always 1, but as the number of receiving antennas N on the satellite side increases a As θ increases, we can see that the column vectors of the channel matrix H gradually become orthogonal, that is:

[0146]

[0147] Among them, N a The number of receiving antennas on the satellite side. H represents the channel matrix.

[0148] The minimum mean square error (MMSE) filter matrix can be understood as a mean square error (MSE) filter matrix determined by combining noise power, channel hardening characteristics, a reference channel estimation matrix, and channel uncertainty. Noise power affects signal quality, while channel hardening indicates that the orthogonality of the channel matrix column vectors increases with the number of satellite-side receiving antennas. The reference channel estimation matrix and channel uncertainty are used to modify the filter matrix. This minimum mean square error (MMSE) filter matrix can be used to minimize the mean square error (MSE) in multiuser detection, bringing the received signal estimate closer to the true signal. It is a crucial parameter for subsequent updates to the iteration step size and signal processing.

[0149] The first iteration step size can be updated using the minimum mean square error filter matrix, initial search direction data, and initial gradient data. A suitable iteration step size can be determined by combining these data using a formula (such as the iteration step size calculation formula in related algorithms). The iteration step size determines the distance advanced in the search direction during each iteration. A suitable step size ensures rapid algorithm convergence while avoiding missing the optimal solution.

[0150] ③Based on iterative step size δ (s) , update the iteration point, that is, get the first iteration point:

[0151] p (s+1) =p (s) +δ (s) d (s) (15)

[0152] Among them, p (s+1) represents the updated iteration point, i.e. the first iteration point; p (s) Indicates the iteration point of this iteration; δ (s) It represents the iteration step, that is, the first iteration step; d (s) Indicates the search direction.

[0153] The first iteration point can be obtained based on the first iteration step size, the initial search direction data, and the initial iteration point update. The initial iteration point is generally set to a certain initial value (e.g., a zero vector), and the first iteration point can be obtained by moving along the initial search direction according to the first iteration step size. The iteration point is continuously updated in each iteration, gradually approaching the optimal solution, and this application does not impose any restrictions on this.

[0154] It can be understood that after obtaining the first iteration point, the gradient data can be updated according to the first iteration point, the minimum mean square error filter matrix and the initial gradient data to obtain the first gradient data; the unit matrix can be corrected according to the initial iteration point, the first iteration point, the initial gradient data and the first gradient data to obtain the first approximate matrix; and the search direction can be updated according to the first approximate matrix and the first gradient data to obtain the first search direction data.

[0155] That is to say, the gradient value can be updated according to formula (16), that is, the first gradient data is obtained:

[0156]

[0157] in, represents the updated gradient value, M represents the minimum mean square error filter matrix, and p (s+1) represents the updated iteration point,

[0158] The first gradient data can be updated based on the first iteration point, the minimum mean square error filter matrix, and the initial gradient data. As the iteration point is updated, the gradient of the objective function at the new iteration point may change. The initial gradient data can be modified by combining it with the minimum mean square error filter matrix to obtain the first gradient data. This first gradient data can reflect the changing trend of the objective function at the first iteration point, providing information for subsequent iterations.

[0159] ④Through vector m (s) =p (s+1) -p (s) and Based on formula (17) the approximate matrix A (s) Make corrections (approximate matrix A (0) is the identity matrix), we get the first approximate matrix:

[0160]

[0161] Among them, A (s+1) represents the modified approximate matrix, that is, the first approximate matrix; A (s) Represents the approximate matrix before correction, initially the unit matrix A (0) ;m (s) =p (s+1) -p (s) and is a vector pair.

[0162] The first approximate matrix can be obtained by modifying the identity matrix based on the initial iteration point, the first iteration point, the initial gradient data, and the first gradient data. During the iteration process, the identity matrix is adjusted by calculating the above data (e.g., based on the matrix correction formula in the BFGS quasi-Newton method) to obtain the first approximate matrix. This approximate matrix can be used to approximate the inverse of the Hessian matrix of the objective function, playing a key role in subsequently updating the search direction, helping the algorithm converge to the optimal solution more quickly.

[0163] Furthermore, according to the modified approximate matrix A (s) Calculate the new search direction, that is, get the first search direction data:

[0164]

[0165] Among them, d (s+1) Indicates the new search direction, i.e. the first search direction data; A (s+1) represents the modified approximate matrix, that is, the first approximate matrix; represents the updated gradient value, that is, the first gradient data.

[0166] The first search direction data can be updated based on the first approximation matrix and the first gradient data. Using the first approximation matrix and the first gradient data, a new search direction, that is, the first search direction data, is calculated according to a formula (such as the search direction update formula). The new search direction is determined after considering the gradient information of the current iteration point and the approximation matrix, and it is expected to more effectively guide the iteration towards the optimal solution.

[0167] And let s = s + 1.

[0168] ⑤ If s < S, repeat from step ② to continue the iteration, otherwise stop the iteration, and the final detection result is We solve the multi-user detection algorithm by using the BFGS quasi-Newton algorithm, avoiding the matrix inversion step in the traditional linear detection algorithm, thereby reducing its computational complexity while ensuring the multi-user detection performance.

[0169] The third preset condition can be understood as a preset judgment criterion, which can be used to determine whether the iteration process can stop. Common third preset conditions can be as shown in step ⑤ above, or the norm of the search direction is less than a certain extremely small threshold, or the change in the objective function value is less than a certain set value, etc. When the first search direction data meets the third preset condition, it means that the algorithm has converged to a relatively stable state, and the iteration result at this time can be considered a good approximate solution.

[0170] Furthermore, assuming that the transmit power of the user data signal is ρ, the received signal after multi-user detection can be expressed as:

[0171]

[0172] where, M represents the minimum mean square error filtering matrix, Y represents the received signal, ρ represents the transmit power, M is the minimum mean square error filtering matrix, represents the transmitted signal after multi-user detection, represents the channel estimation matrix, ΔH represents the channel estimation error matrix, and Z represents the noise.

[0173] Furthermore, the PPSNR expression of the user can be deduced as:

[0174]

[0175] where, γ k represents the PPSNR of the user, ρ represents the transmit power, K a represents the number of active users at the current moment, M is the minimum mean square error filtering matrix, represents the transmitted signal after multi-user detection, represents the channel estimation matrix, ΔH represents the channel estimation error matrix, and Z represents the noise.

[0176] In the satellite IoT communication scenario, the channel coding block length M is equal to the number of subcarriers occupied by the transmitted signal data, that is, M = Nm = Bt s -m. Therefore, according to FBL information theory, the relationship between the user's maximum achievable data rate and PPSNR can be expressed as:

[0177]

[0178] Among them, R k represents the maximum achievable data rate of the user, γ k represents the user’s PPSNR, C(γ k )=log2(1+γ k ) indicates that PPSNR is γ k The Shannon capacity when It is a correction for the lower PPSNR. The Q function expression is ε k =ε k (m,ρ′,ρ) represents the user’s error probability, where ρ′ is the transmission power of the satellite communication pilot signal, B represents the carrier bandwidth, and t s represents the time delay, and m represents the length of the satellite communication pilot signal. Therefore, according to formula (21), the relationship between the maximum achievable data rate and PPSNR can be expressed as

[0179]

[0180] Where R represents the maximum achievable data rate for all users, K a represents the number of active users at the current moment, C(γ k ) indicates that PPSNR is γ k Shannon capacity, V(γ k ) is a correction for lower PPSNR, B represents the carrier bandwidth, t s represents the time delay, m represents the length of the satellite communication pilot signal, and the Q function expression is ε k =ε k (m,ρ′,ρ) represents the user’s error probability.

[0181] S60: When the speed reliability relationship data satisfies a first preset condition, determining the reference channel estimation matrix as a target channel estimation matrix.

[0182] Among them, the first preset condition can be understood as a preset condition set in advance for determining whether the speed reliability relationship data meets the requirements. The first preset condition can be set manually or by the system default setting, and this application does not impose any restrictions on this. Optionally, the first preset condition can also be used to evaluate the rationality and accuracy of the multi-user detection results. Specifically, the maximum achievable data rate can be set to reach a certain threshold, or the PPSNR can be set to be higher than a certain threshold, etc., so as to obtain the threshold of the speed reliability relationship data as the first preset condition, and this application does not impose any restrictions on this.

[0183] When the speed-reliability relationship data meets the first preset condition, it can be considered that the accuracy and reliability of the reference channel estimation matrix meet the requirements, and the reference channel estimation matrix can be used as the target channel estimation matrix; otherwise, when the speed-reliability relationship data does not meet the first preset condition, it can be considered that the accuracy and reliability of the reference channel estimation matrix cannot meet the requirements, and it may be necessary to adjust the parameters of the SOMP algorithm and / or re-perform multi-user detection.

[0184] The target channel estimation matrix can be understood as the channel estimation matrix determined after multi-user detection and condition judgment, used for channel estimation and signal processing in actual communication systems. This target channel estimation matrix is the output of the entire multi-user detection method and can be used for subsequent data transmission, demodulation, decoding, and other operations, directly affecting the performance of the communication system.

[0185] To address the massive user access and limited satellite computing resources, a multi-user detection scheme for satellite IoT communications is proposed. This scheme establishes a satellite communication channel model for LEO satellite IoT uplink transmission: Based on the massive multi-input multiple-input multiple-output (MMIMO) system, the uplink transmission CIR is modeled and pre-compensated at the transmitter for the severe Doppler shift and transmission delay issues in the satellite-to-ground link, simplifying the computational complexity of subsequent signal processing. A low-overhead joint user activity detection and channel estimation algorithm is designed: Based on compressed sensing theory, the joint user activity detection and channel estimation problem is formulated as a sparse signal recovery problem and solved using the SOMP algorithm. This ensures algorithm performance while avoiding the high computational overhead caused by too many connected users. A low-computational-complexity multi-user detection scheme is implemented: Incorporating a CSI error estimation model improves algorithm robustness; a low-computational-complexity multi-user detection scheme based on the BFGS quasi-Newton algorithm addresses the practical challenges of limited satellite payloads.

[0186] It can be seen that in the above scheme, a satellite communication channel model of an uplink low earth orbit (LEO) satellite Internet of Things system is established based on a large-scale multiple-input multiple-output (mMIMO) model; based on the satellite communication channel model of the uplink LEO satellite Internet of Things system and the first satellite communication pilot signal sent by the user side, the second satellite communication pilot signal received by the satellite side is determined; sparse feature mining processing is performed on the second satellite communication pilot signal to obtain sparse features and a quantized received signal; sparse signal recovery processing is performed based on the sparse features and the quantized received signal to obtain a reference channel estimation matrix; multi-user detection is performed based on the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data; when the speed reliability relationship data meets the first preset condition, the reference channel estimation matrix is determined as the target channel estimation matrix, which is beneficial to simplify the calculation complexity in satellite communication scenarios with limited computing resources and obtain a more accurate target channel estimation matrix.

[0187] For the same example as above, please refer to Figure 3 , Figure 3 A schematic diagram of the structure of a terminal provided in an embodiment of the present application is shown in FIG. Figure 3 As shown, the system comprises a processor, an input device, an output device and a memory, which are connected to each other, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions, and the program includes instructions for executing the following steps;

[0188] Based on the massive multiple-input multiple-output (mMIMO) model, a satellite communication channel model for the uplink low-Earth orbit (LEO) satellite IoT system is established.

[0189] Determine a second satellite communication pilot signal received by a satellite side based on a satellite communication channel model of the uplink LEO satellite IoT system and a first satellite communication pilot signal sent by a user side;

[0190] performing sparse feature mining processing on the second satellite communication pilot signal to obtain a sparse feature and a quantized received signal;

[0191] Performing sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix;

[0192] Perform multi-user detection based on channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data;

[0193] In a case where the speed reliability relationship data satisfies a first preset condition, the reference channel estimation matrix is determined to be a target channel estimation matrix.

[0194] In this example, a satellite communication channel model of an uplink low-Earth orbit (LEO) satellite IoT system is established based on a massive multiple-input multiple-output (mMIMO) model. A second satellite communication pilot signal received by the satellite side is determined based on the satellite communication channel model of the uplink LEO satellite IoT system and a first satellite communication pilot signal sent by the user side. Sparse feature mining is performed on the second satellite communication pilot signal to obtain sparse features and a quantized received signal. Sparse signal recovery is performed based on the sparse features and the quantized received signal to obtain a reference channel estimation matrix. Multi-user detection is performed based on the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data. When the speed reliability relationship data meets a first preset condition, the reference channel estimation matrix is determined to be the target channel estimation matrix. This is beneficial for simplifying the complexity of calculations in satellite communication scenarios with limited computing resources and obtaining a more accurate target channel estimation matrix.

[0195] The above mainly introduces the scheme of the embodiment of the present application from the perspective of the execution process on the method side. It is understandable that, in order to implement the above functions, the terminal includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the units and algorithm steps of the various examples described in the embodiments provided herein, the present application can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in a hardware or computer software driven hardware manner depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0196] The embodiment of the present application can divide the terminal into functional units according to the above method example. For example, each functional unit can be divided according to each function, or two or more functions can be integrated into one processing unit. The above integrated unit can be implemented in the form of hardware or in the form of software functional units. It should be noted that the division of units in the embodiment of the present application is schematic and is only a logical functional division. In actual implementation, there may be other division methods.

[0197] In line with the above, please see Figure 4 , Figure 4 The present invention provides a schematic diagram of the structure of a low-complexity multi-user detection device in a satellite Internet of Things scenario. Figure 4 As shown, the device includes:

[0198] The first processing unit 101 is configured to establish a satellite communication channel model for an uplink low earth orbit (LEO) satellite IoT system based on a massive multiple-input multiple-output (mMIMO) model.

[0199] The first determining unit 102 is configured to determine a second satellite communication pilot signal received by the satellite side based on a satellite communication channel model of the uplink LEO satellite IoT system and a first satellite communication pilot signal sent by the user side;

[0200] The second processing unit 103 is configured to perform sparse feature mining processing on the second satellite communication pilot signal to obtain a sparse feature and a quantized received signal;

[0201] A third processing unit 104 is configured to perform sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix;

[0202] A fourth processing unit 105 is configured to perform multi-user detection based on the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data;

[0203] The second determining unit is configured to determine, when the speed reliability relationship data satisfies a first preset condition, that the reference channel estimation matrix is a target channel estimation matrix.

[0204] In one possible implementation, the first determining unit 102 is configured to determine the second satellite communication pilot signal received by the satellite side based on the satellite communication channel model of the uplink LEO satellite IoT system and the first satellite communication pilot signal sent by the user side, specifically:

[0205] Determine the satellite uplink transmission spatial domain channel impulse response (CIR) based on the satellite communication channel model of the uplink LEO satellite IoT system;

[0206] Determine a time domain representation of the first satellite communication pilot signal according to the frequency domain representation of the first satellite communication pilot signal sent by the user side;

[0207] performing compensation processing on the time domain representation of the first satellite communication pilot signal to obtain a first compensated satellite communication pilot signal;

[0208] A second satellite communication pilot signal received by the satellite side is determined according to the first compensated satellite communication pilot signal and the satellite uplink transmission spatial domain CIR.

[0209] In one possible implementation, the second processing unit 103 is configured to perform sparse feature mining on the second satellite communication pilot signal to obtain a sparse feature and a quantized received signal, specifically:

[0210] Determining satellite-side received signals of K users based on the second satellite communication pilot signal;

[0211] Performing angle domain conversion processing on the satellite side received signal to obtain an angle domain received signal;

[0212] determining an angle domain channel matrix according to the angle domain received signal;

[0213] Determining, according to the angle-domain channel matrix, a sparse characteristic of the K users and the channel matrix in the angle domain;

[0214] A quantized received signal is determined according to the angle domain channel matrix.

[0215] In one possible implementation, the third processing unit 104 is configured to perform sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix, specifically for:

[0216] Initializing the quantized received signal as an initial residual signal;

[0217] Selecting a current active user using a maximum correlation criterion based on the initial residual signal and the measurement matrix;

[0218] The current active users and historical active users are combined into an active user set;

[0219] Based on the active user set, selecting corresponding column vectors from the measurement matrix to obtain a first matrix;

[0220] Determine, by using a least squares method, a first channel estimation matrix corresponding to the active user set based on the first matrix;

[0221] performing updating processing on the initial residual signal according to the first channel estimation matrix to obtain a first residual signal;

[0222] When the first residual signal satisfies a second preset condition, the first channel estimation matrix is determined to be a reference channel estimation matrix.

[0223] In one possible implementation, the fourth processing unit 105 is configured to perform multi-user detection based on the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data, specifically to:

[0224] Determining initial gradient data according to the reference channel estimation matrix, channel uncertainty and the satellite-side received signal;

[0225] Determining initial search direction data according to the initial gradient data;

[0226] Determining a minimum mean square error filter matrix based on noise power, channel hardening characteristics, the reference channel estimation matrix, and the channel uncertainty;

[0227] updating the iteration step size according to the minimum mean square error filter matrix, the initial search direction data, and the initial gradient data to obtain a first iteration step size;

[0228] Update the iteration point according to the first iteration step size, the initial search direction data and the initial iteration point to obtain a first iteration point;

[0229] If a third preset condition is met, determining the first iteration point to be a transmitted signal after multi-user detection;

[0230] Speed reliability relationship data is determined based on the transmitted signal after multi-user detection, the minimum mean square error filter matrix and the transmission power of the user data signal.

[0231] An embodiment of the present application also provides a computer storage medium, wherein the computer storage medium stores a computer program for electronic data exchange, and the computer program enables a computer to execute some or all steps of the low-complexity multi-user detection method in any satellite Internet of Things scenario as described in the above method embodiments.

[0232] An embodiment of the present application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program, and the computer program enables a computer to execute some or all steps of the low-complexity multi-user detection method in any satellite Internet of Things scenario as described in any of the above method embodiments.

[0233] It should be noted that for the aforementioned method embodiments, for the sake of simplicity, they are all expressed as a series of action combinations, but those skilled in the art should be aware that this application is not limited by the order of the actions described, because according to this application, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by this application.

[0234] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0235] In the several embodiments provided in this application, it should be understood that the disclosed devices can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is merely a logical function division. In actual implementation, there may be other division methods, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, and the indirect coupling or communication connection of devices or units can be electrical or other forms.

[0236] The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected to achieve the purpose of this embodiment according to actual needs.

[0237] In addition, the functional units in the various embodiments of the application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software program modules.

[0238] If the integrated unit is implemented in the form of a software program module and sold or used as an independent product, it can be stored in a computer-readable memory. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a memory, including a number of instructions for enabling a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned memory includes: various media that can store program codes, such as a U disk, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk or an optical disk.

[0239] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable memory, which may include: a flash drive, a read-only memory, a random access memory, a magnetic disk or an optical disk, etc.

[0240] The above is a detailed introduction to the embodiments of the present application. Specific examples are used herein to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method and core idea of the present application. At the same time, for those skilled in the art, according to the idea of the present application, there may be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A low-complexity multi-user detection method in a satellite Internet of Things scenario, characterized in that: The method comprises: Based on the massive multiple-input multiple-output (mMIMO) model, a satellite communication channel model for the uplink low-Earth orbit (LEO) satellite IoT system is established. Determining a second satellite communication pilot signal received by a satellite side based on the satellite communication channel model and the first satellite communication pilot signal sent by the user side; performing sparse feature mining processing on the second satellite communication pilot signal to obtain a sparse feature and a quantized received signal; Performing sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix; Perform multi-user detection based on channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data; In a case where the speed reliability relationship data satisfies a first preset condition, the reference channel estimation matrix is determined to be a target channel estimation matrix.

2. The low-complexity multi-user detection method in the satellite Internet of Things scenario according to claim 1 is characterized in that: The determining, based on the satellite communication channel model of the uplink LEO satellite IoT system and the first satellite communication pilot signal sent by the user side, the second satellite communication pilot signal received by the satellite side includes: Determine the satellite uplink transmission spatial domain channel impulse response (CIR) based on the satellite communication channel model of the uplink LEO satellite IoT system; Determine a time domain representation of the first satellite communication pilot signal according to the frequency domain representation of the first satellite communication pilot signal sent by the user side; performing compensation processing on the time domain representation of the first satellite communication pilot signal to obtain a first compensated satellite communication pilot signal; A second satellite communication pilot signal received by the satellite side is determined according to the first compensated satellite communication pilot signal and the satellite uplink transmission spatial domain CIR.

3. The low-complexity multi-user detection method in the satellite Internet of Things scenario according to claim 2, characterized in that: The performing sparse characteristic mining processing on the second satellite communication pilot signal to obtain a sparse characteristic and a quantized received signal includes: Determining satellite-side received signals of K users based on the second satellite communication pilot signal; Performing angle domain conversion processing on the satellite side received signal to obtain an angle domain received signal; determining an angle domain channel matrix according to the angle domain received signal; Determining, according to the angle-domain channel matrix, a sparse characteristic of the K users and the channel matrix in the angle domain; A quantized received signal is determined according to the angle domain channel matrix.

4. The low-complexity multi-user detection method in the satellite Internet of Things scenario according to claim 3 is characterized in that: The performing sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix includes: Initializing the quantized received signal as an initial residual signal; Selecting a current active user using a maximum correlation criterion based on the initial residual signal and the measurement matrix; The current active users and historical active users are combined into an active user set; Based on the active user set, selecting corresponding column vectors from the measurement matrix to obtain a first matrix; Determine, by using a least squares method, a first channel estimation matrix corresponding to the active user set based on the first matrix; performing updating processing on the initial residual signal according to the first channel estimation matrix to obtain a first residual signal; When the first residual signal satisfies a second preset condition, the first channel estimation matrix is determined to be a reference channel estimation matrix.

5. The low-complexity multi-user detection method in the satellite Internet of Things scenario according to claim 4 is characterized in that: The performing multi-user detection according to the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data includes: Determining initial gradient data according to the reference channel estimation matrix, channel uncertainty and the satellite-side received signal; Determining initial search direction data according to the initial gradient data; Determining a minimum mean square error filter matrix based on noise power, channel hardening characteristics, the reference channel estimation matrix, and the channel uncertainty; updating the iteration step size according to the minimum mean square error filter matrix, the initial search direction data, and the initial gradient data to obtain a first iteration step size; Update the iteration point according to the first iteration step size, the initial search direction data and the initial iteration point to obtain a first iteration point; If a third preset condition is met, determining the first iteration point to be a transmitted signal after multi-user detection; Speed reliability relationship data is determined based on the transmitted signal after multi-user detection, the minimum mean square error filter matrix and the transmission power of the user data signal.

6. A low-complexity multi-user detection device in a satellite Internet of Things scenario, characterized in that: The device comprises: The first processing unit is configured to establish a satellite communication channel model for an uplink low earth orbit (LEO) satellite IoT system based on a massive multiple-input multiple-output (mMIMO) model. A first determining unit is configured to determine a second satellite communication pilot signal received by a satellite side based on a satellite communication channel model of the uplink LEO satellite Internet of Things system and a first satellite communication pilot signal sent by a user side; A second processing unit is configured to perform sparse feature mining processing on the second satellite communication pilot signal to obtain a sparse feature and a quantized received signal; a third processing unit, configured to perform sparse signal recovery processing based on the sparse characteristic and the quantized received signal to obtain a reference channel estimation matrix; a fourth processing unit, configured to perform multi-user detection based on the channel uncertainty and the reference channel estimation matrix to obtain speed reliability relationship data; The second determining unit is configured to determine, when the speed reliability relationship data satisfies a first preset condition, that the reference channel estimation matrix is a target channel estimation matrix.

7. The low-complexity multi-user detection device in the satellite Internet of Things scenario according to claim 6, characterized in that: The first determining unit is configured to determine the second satellite communication pilot signal received by the satellite side based on the satellite communication channel model of the uplink LEO satellite Internet of Things system and the first satellite communication pilot signal sent by the user side, specifically for: Determine the satellite uplink transmission spatial domain channel impulse response (CIR) based on the satellite communication channel model of the uplink LEO satellite IoT system; Determine a time domain representation of the first satellite communication pilot signal according to the frequency domain representation of the first satellite communication pilot signal sent by the user side; performing compensation processing on the time domain representation of the first satellite communication pilot signal to obtain a first compensated satellite communication pilot signal; A second satellite communication pilot signal received by the satellite side is determined according to the first compensated satellite communication pilot signal and the satellite uplink transmission spatial domain CIR.

8. The low-complexity multi-user detection device in the satellite Internet of Things scenario according to claim 7, characterized in that: The second processing unit is configured to perform sparse feature mining on the second satellite communication pilot signal to obtain a sparse feature and a quantized received signal, specifically for: Determining satellite-side received signals of K users based on the second satellite communication pilot signal; Performing angle domain conversion processing on the satellite side received signal to obtain an angle domain received signal; determining an angle domain channel matrix according to the angle domain received signal; Determining, according to the angle-domain channel matrix, a sparse characteristic of the K users and the channel matrix in the angle domain; A quantized received signal is determined according to the angle domain channel matrix.

9. A terminal, characterized in that: The method comprises a processor, an input device, an output device and a memory, wherein the processor, the input device, the output device and the memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the method according to any one of claims 1 to 5.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor is caused to perform the method according to any one of claims 1 to 5.

Images