Intelligent reflector-assisted large-scale connection activation detection and channel estimation method
By using non-orthogonal pilots for activation detection and orthogonal pilots for channel estimation in mMTC scenarios, the difficulty of RIS activation detection and channel estimation in large-scale MIMO systems is solved, reducing the computational complexity and error, and improving the System performance and efficiency.
Patent Information
- Application Number
- CN202411164182.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-23
- Publication Date
- 2024-11-29
AI Technical Summary
In large-scale machine communication (mMTC) scenarios, intelligent reflective surface (RIS) increases the difficulty of activation detection and channel estimation of IoT devices due to its passive characteristics, especially in the scheduling-free access mode, resulting in channel estimation errors. and increase in computational complexity.
Non-orthogonal pilots are used as the preamble sequence for activation detection, and orthogonal pilots are used in the channel estimation stage. By dividing them into two stages: activation detection and channel estimation, they are processed separately to reduce the computational complexity and reduce channel estimation error.
It effectively reduces the computational complexity of activation detection, and improves system performance and efficiency by reducing channel estimation errors, and improves communication coverage and device battery life.
Smart Images

Figure CN119052031A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of wireless communication technology, and in particular to an activation detection and channel estimation method for large-scale connections assisted by an intelligent reflecting surface. Background Art
[0002] With the rapid development of the Internet of Things (IoT), massive machine-type communications (mMTC) has been regarded as one of the three core application scenarios of the fifth generation (5G) wireless networks. The main challenge facing mMTC is to support intermittent communication of short-duration, short data packets between the base station (BS) and a large number of IoT devices. In recent years, the unlicensed random access method has attracted much attention because it can significantly reduce signaling overhead and access latency. This method allows IoT devices to directly transmit pilot sequences and data when the device is activated without obtaining authorization from the base station. At the BS side, activation detection and channel estimation are performed through the pilot sequence, thereby achieving data decoding.
[0003] In order to further improve the propagation environment and enhance system performance to meet the needs of wireless communications, smart reflective surfaces (RIS) have received widespread attention in recent years. RIS consists of a large number of passive reflective units, which optimize the phase offset of each unit by utilizing instantaneous channel state information (CSI) to redirect the incident signal to enhance signal quality and improve system performance. However, since RIS is a passive device and cannot actively process information, it cannot independently utilize pilot estimation (CSI) and cannot transmit pilot signals. In addition, due to the large number of reflective units, the dimension of channel estimation is significantly increased. In order to fully realize the potential of RIS, the challenge of channel estimation must be addressed to ensure that the channel characteristics in the communication environment can be accurately grasped, thereby optimizing signal reflection and transmission and improving system performance and efficiency. Studies have shown that applying RIS to mMTC scenarios can significantly improve system capacity, expand communication coverage, and extend the battery life of IoT devices. However, the passive nature of RIS will increase the difficulty of activation detection and channel estimation of IoT devices.
[0004] The present invention provides an activation detection and channel estimation method suitable for RIS-assisted large-scale connections. In the mMTC scenario, due to the use of a scheduling-free access method, the BS side needs to identify activated users through pilots. However, due to the large number of devices, it is impossible to assign orthogonal pilots to each device, and the use of non-orthogonal pilots will inevitably lead to large channel estimation errors. In addition, the introduction of RIS further increases the dimension of the channel to be estimated. The present invention utilizes the structural characteristics of the pilot and proposes an effective activation detection and channel estimation method.
[0005] After searching, the application publication number CN115765900B, a method for uplink transmission of a large-scale MIMO system assisted by an intelligent reflector in a superimposed pilot mode, in order to solve the problem of mutual interference between pilot signals and data signals caused by the use of the superimposed pilot mode, the technical solution adopted by the present invention includes: establishing a model of a large-scale MIMO system assisted by an intelligent reflector in a superimposed pilot mode; based on the constructed model, performing uplink channel estimation and signal detection; analyzing the uplink achievable rate of the system; according to the uplink achievable rate, constructing an optimization problem formula for maximizing the system sum rate; according to the optimization problem formula, designing a strategy for jointly optimizing the power control coefficients of the user pilot and data signals and the phase of the IRS reflection unit. The present invention further improves the sum rate of the large-scale MIMO system assisted by the intelligent reflector in the superimposed pilot mode, on the premise of ensuring that the total transmission signal-to-noise ratio of the pilot and data signals of each user is not greater than the set maximum transmission signal-to-noise ratio and the phase of the IRS reflection unit meets the set range.
[0006] The above patent applies RIS technology to large-scale MIMO systems, in which users use authorized access and the pilot signal part of the superimposed pilot uses an orthogonal sequence. However, if this technology is applied to the mMTC scenario, the spectrum efficiency will be greatly reduced. In order to solve this problem, the present invention adopts a scheduling-free access method, uses non-orthogonal sequences as pilots, and divides activation detection and channel estimation into two stages. In the activation detection stage, RIS is set to the "off" state to greatly reduce the calculation complexity of the base station. At the same time, a non-orthogonal pilot composed of a linear combination of orthogonal pilots is used as the preamble sequence. The base station can determine the activated user by analyzing the covariance matrix elements of the received signal. In the channel estimation stage, since the number of activated users is much smaller than the total number of cell devices, orthogonal pilots can be used to reduce pilot pollution and reduce channel estimation errors. Summary of the invention
[0007] The present invention aims to solve the above problems of the prior art. A method for activation detection and channel estimation of large-scale connections assisted by an intelligent reflective surface is proposed. The technical solution of the present invention is as follows:
[0008] An intelligent reflective surface assisted activation detection and channel estimation method for large-scale connections, comprising the following steps:
[0009] Step 1: Build a system model;
[0010] Step 2: Generate an orthogonal pilot set and a non-orthogonal pilot set;
[0011] Step 3: Activation detection is performed through the construction characteristics of direct links and non-orthogonal pilots;
[0012] Step 4: Estimate the equivalent channel information of the direct link and the reflected link.
[0013] Furthermore, the step 1 of constructing the system model specifically includes:
[0014] A1. Assume that the number of base station antennas is M, K static IoT devices are sparsely activated, and κ represents the device set. The number of single RIS reflector elements is N; the activation state of the device is represented by α k It means that when α k =1, indicating that the device is activated, and the activation index set is used Represents; through prior information, it is assumed that the activation probability ε of the device is known at the base station;
[0015] A2. Establish channel model: Consider that the base station antenna and devices at different locations are sufficiently separated to generate spatially independent channels; represents the channel information from the kth device to the base station; where g k Indicates the corresponding large-scale fading and shadow fading; represents the Rayleigh fading part; represents the channel matrix from RIS to BS, represents the channel coefficient from the kth user to the RIS; all channels satisfy block fading, that is, within a coherent time block, the channel remains unchanged; define Θ = diag (θ) as the diagonal phase shift matrix of the RIS, where φ i is the phase of the i-th reflection unit, β i ∈[0,1] is the amplitude of the i-th reflection unit, 1≤i≤N; when β i = 0, indicating that the i-th reflector unit of RIS is not activated and is in the "off" state, β i =1 indicates that the i-th reflection unit of RIS is activated and is in the "on" state.
[0016] Furthermore, the step 2 of generating an orthogonal pilot set and a non-orthogonal pilot set specifically includes:
[0017] B1. Construct an L×L dimensional orthogonal pilot set Q, each orthogonal pilot uses q j Represents, where j∈[1,L]. Since the number of activated devices is much less than the number of cell devices, L can satisfy: L≥K×ε;
[0018] B2. Using q j A linear combination of to generate a non-orthogonal pilot sequence;
[0019] Assuming that the kth non-orthogonal pilot is composed of a linear combination of two orthogonal pilots, then u k It can be expressed as:
[0020]
[0021] |B k |<<L
[0022] Among them, B k is the orthogonal pilot set of the kth device, A j is the jth orthogonal pilot q j The device set, f(j,k) is a parameter to improve activation detection; the following matrix form can be further defined, where U represents the required non-orthogonal pilot matrix, and F represents the linear combination parameter matrix;
[0023]
[0024] Furthermore, the step three performs activation detection through the structural characteristics of the direct link and the non-orthogonal pilot, specifically including:
[0025] When C1 and RIS are in the OFF state, the non-orthogonal pilot signal transmitted by the direct link received by the BS can be expressed as:
[0026]
[0027] Where Γ=diag(γ), γ=(γ1,...,γ N ), γ n =α n g n , H d =[h d,1 ,...,h d,K ] T , N has a mean of 0 and a variance of σ 2 Gaussian white noise; H d represents the direct link matrix from K devices to the base station.
[0028] C2. Use a matrix consisting of an orthogonal sequence set for matched filtering:
[0029]
[0030] C3. Calculate the covariance matrix of S
[0031] C4. Determine the activation state of device k through the elements in the covariance matrix;
[0032] When M→∞, For γ n The consistent estimator of n When it is greater than or equal to the judgment threshold μ, it means that device k is activated, so the estimated set of activated users is obtained Where μ is the decision threshold, which can be obtained by making the false alarm and missed detection probabilities equal.
[0033] Furthermore, the step 4 estimates the equivalent channel information of the direct link and the reflected link, specifically including:
[0034] Assume that the set of activated users required in step 3 satisfies: The activated user randomly selects a pilot sequence from the orthogonal pilot resource pool Q for transmission. When RIS is in the ON state, the pilot reaches the BS end through the direct link and the reflected link, which can be expressed as:
[0035]
[0036] in, K a A direct link from user to BS, K a The channel information from user to RIS, F represents K a The pilot matrix of a user, let θ = [1,...,1] T , then the above formula can be expressed as:
[0037]
[0038] make represents the equivalent channel information. Due to the column orthogonality of F, the channel information H eq It can be expressed as:
[0039]
[0040] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the activation detection and channel estimation method for large-scale connection assisted by an intelligent reflecting surface as described in any one of the items is implemented.
[0041] A non-transitory computer-readable storage medium stores a computer program thereon, wherein when the computer program is executed by a processor, the activation detection and channel estimation method for large-scale connection assisted by an intelligent reflecting surface as described in any one of the items is implemented.
[0042] A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the activation detection and channel estimation method for large-scale connection assisted by an intelligent reflective surface as described in any one of the items is implemented. The advantages and beneficial effects of the present invention are as follows:
[0043] The present invention utilizes the characteristics of the total number of IoT devices and the number of activated devices, and divides activation detection and channel estimation into two stages, sending non-orthogonal pilots and orthogonal pilots respectively. With the structural characteristics of non-orthogonal pilots, the activated users can be effectively estimated through the covariance matrix of the received signal, thereby greatly reducing the computational complexity of activation detection. In the channel estimation stage, due to the use of orthogonal pilots, the channel estimation error caused by non-orthogonal pilots can be effectively reduced.
[0044] The present invention divides activation detection and channel estimation into two stages, and considers the difference in the total number of devices and the number of activations, and respectively adopts non-orthogonal and orthogonal pilots. In the activation detection stage, the RIS is set to the "off" state, and a non-orthogonal pilot composed of a linear combination of orthogonal pilots is used as a leading sequence. The BS determines whether the user is activated by the elements of the covariance matrix of the received signal. This method can greatly reduce the computational complexity of the BS end. In the channel estimation stage, since the number of activated users is much smaller than the total number of cell devices, an orthogonal pilot is used, thereby effectively reducing pilot pollution and reducing channel estimation errors. BRIEF DESCRIPTION OF THE DRAWINGS
[0045] Figure 1 The present invention provides a preferred embodiment of the intelligent reflective surface assisted large-scale connection activation detection and channel estimation flow chart. DETAILED DESCRIPTION
[0046] The following will describe the technical solutions in the embodiments of the present invention in detail in conjunction with the accompanying drawings in the embodiments of the present invention. The described embodiments are only part of the embodiments of the present invention.
[0047] The technical solution of the present invention to solve the above technical problems is:
[0048] The present invention combines the characteristics of the total number of IoT devices and the number of activations thereof, divides activation detection and channel estimation into two stages, and uses non-orthogonal and orthogonal pilot signals for processing respectively. In the activation detection stage, RIS is set to the off state. At this time, the BS receives the non-orthogonal pilot signal transmitted through the direct link and calculates the covariance matrix of the received signal. Utilizing the structural characteristics of the non-orthogonal pilot, the activated user can be effectively detected. In the channel estimation stage, RIS is set to the on state. Although the introduction of RIS increases the dimension of the channel to be estimated, due to the orthogonality of the pilot, the influence of pilot pollution can be effectively reduced, thereby accurately estimating the equivalent channels of the direct link and the reflection link. The specific scheme is as follows.
[0049] Step 1: Build a system model
[0050] 1. Assume that the number of base station antennas is M, and K static IoT devices are sparsely activated. Indicates a device set, with The number of single RIS reflector elements is N. The activation state of the device is represented by α k It means that when α k =1, indicating that the device is activated, and the activation index set is used Represents. Through prior information, it is assumed that the base station knows the activation probability ε of the device.
[0051] 2. Establish a channel model: Consider that the base station antenna and devices at different locations are sufficiently separated to generate spatially independent channels. represents the channel information from the kth device to the base station. k Indicates the corresponding large-scale fading and shadow fading. represents the Rayleigh fading part. represents the channel matrix from RIS to BS, represents the channel coefficient from the kth user to the RIS. All channels satisfy block fading, that is, within a coherent time block, the channel remains unchanged. Define Θ as the phase shift matrix of RIS, Θ = diag(θ),
[0052] Step 2: Generate orthogonal pilot sets and non-orthogonal pilot sets
[0053] 1. Construct an L×L dimensional orthogonal pilot set Q, each orthogonal pilot uses q j Represents, where j∈[1,L]. Since the number of activated devices is much less than the number of cell devices, L can satisfy: L≥K×ε.
[0054] 2. Use q j A linear combination of to generate a non-orthogonal pilot sequence;
[0055] Assuming that the kth non-orthogonal pilot is composed of a linear combination of two orthogonal pilots, then u k It can be expressed as:
[0056]
[0057] |B k |<<L
[0058] Among them, B k is the orthogonal pilot set of the kth device, A j is the jth orthogonal pilot q j The device set of , f(k,j) is the parameter to improve activation detection.
[0059]
[0060] Step 3: Activation detection through direct link and non-orthogonal pilot construction characteristics
[0061] 1. When RIS is in the OFF state, the non-orthogonal pilot signal transmitted by the BS receiving the direct link can be expressed as:
[0062]
[0063] Where Γ=diag(γ), γ=(γ1,...,γ N ), γ n =α n g n , H d =[h d,1 ,...,h d,K ] T , N has a mean of 0 and a variance of σ 2 Gaussian white noise.
[0064] 2. Use a matrix consisting of an orthogonal sequence set for matched filtering:
[0065]
[0066] 3. Calculate the covariance matrix of S
[0067] 6. Determine the activation status of device k by elements in the covariance matrix
[0068] When M→∞, For γ n The consistent estimator of When , device k is activated. Where μ is the decision threshold, which can be obtained by making the false alarm and missed detection probabilities equal. The final activated user set is
[0069] Step 4: Estimate the equivalent channel information of the direct link and the reflected link
[0070] Assume that the set of activated users required in step 3 satisfies: The activated user randomly selects a pilot sequence from the orthogonal pilot resource pool Q for transmission. When RIS is in the ON state, the pilot reaches the BS end through the direct link and the reflected link, which can be expressed as:
[0071]
[0072] in, K a A direct link from user to BS, K a The channel information from user to RIS, F represents K a The pilot matrix of a user, let θ = [1,...,1] T , then the above formula can be expressed as:
[0073]
[0074] make represents the equivalent channel information. Due to the column orthogonality of F, the channel information H eq It can be expressed as:
[0075]
[0076] The systems, devices, modules or units described in the above embodiments may be implemented by computer chips or entities, or by products with certain functions.
[0077] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.
[0078] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.
[0079] The above embodiments should be understood to be only used to illustrate the present invention and not to limit the protection scope of the present invention. After reading the contents of the present invention, technicians can make various changes or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.
Claims
1. A method for activation detection and channel estimation of large-scale connections assisted by intelligent reflective surfaces, characterized in that: The following steps are involved: Step 1: Build a system model; Step 2: Generate an orthogonal pilot set and a non-orthogonal pilot set; Step 3: Activation detection is performed through the construction characteristics of direct links and non-orthogonal pilots; Step 4: Estimate the equivalent channel information of the direct link and the reflected link.
2. The method for activation detection and channel estimation of large-scale connections assisted by intelligent reflective surfaces according to claim 1, characterized in that: The step 1 of building a system model specifically includes: A1. Assume that the number of base station antennas is M, and K static IoT devices are sparsely activated. Indicates a device set, with The number of single RIS reflector elements is N; the activation state of the device is represented by α k It means that when α k =1, indicating that the device is activated, and the activation index set is used Represents; using prior information, the activation probability ε of the known device at the base station; A2. Establish channel model: Consider that the base station antenna and devices at different locations are sufficiently separated to generate spatially independent channels; represents the channel information from the kth device to the base station; where g k Indicates the corresponding large-scale fading and shadow fading; represents the Rayleigh fading part; represents the channel matrix from RIS to BS, represents the channel coefficient from the kth user to the RIS; all channels satisfy block fading, that is, within a coherent time block, the channel remains unchanged; define Θ = diag (θ) as the diagonal phase shift matrix of the RIS, where φ i is the phase of the i-th reflection unit, β i ∈[0,1] is the amplitude of the i-th reflection unit, 1≤i≤N; when β i = 0, indicating that the i-th reflector unit of RIS is not activated and is in the "off" state, β i =1 indicates that the i-th reflection unit of RIS is activated and is in the "on" state.
3. The method for activation detection and channel estimation of large-scale connections assisted by intelligent reflective surfaces according to claim 2, characterized in that: The step 2 of generating an orthogonal pilot set and a non-orthogonal pilot set specifically includes: B1. Construct an L×L dimensional orthogonal pilot set Q, each orthogonal pilot uses q j Represents, where j∈[1,L]. Since the number of activated devices is much less than the number of cell devices, L can satisfy: L≥K×ε; B2. Using q j A linear combination of to generate a non-orthogonal pilot sequence; Assuming that the kth non-orthogonal pilot is composed of a linear combination of two orthogonal pilots, then u k It can be expressed as: |B k |<<L Among them, B k is the orthogonal pilot set of the kth device, A j is the jth orthogonal pilot q j The device set, f(j,k) is a parameter to improve activation detection; the following matrix form can be further defined, where U represents the required non-orthogonal pilot matrix, and F represents the linear combination parameter matrix; 4. The method for activation detection and channel estimation of large-scale connections assisted by intelligent reflective surfaces according to claim 3, characterized in that: The step three performs activation detection through the construction characteristics of the direct link and the non-orthogonal pilot, specifically including: When both C1 and RIS elements are in the "off" state, the non-orthogonal pilot signal transmitted by the direct link received by the BS can be expressed as: Where Γ=diag(γ), γ=(γ1,...,γ N ), γ n =α n g n , H d =[h d,1 ,...,h d,K ] T , N has a mean of 0 and a variance of σ 2 Gaussian white noise; H d represents the direct link matrix from K devices to the base station. C2. Use a matrix consisting of an orthogonal sequence set for matched filtering: C3. Calculate the covariance matrix of S C4. Determine the activation state of device k through the elements in the covariance matrix; When M→∞, For γ n The consistent estimator of n When it is greater than or equal to the judgment threshold μ, it means that device k is activated, so the estimated set of activated users is obtained Where μ is the decision threshold, which can be obtained by making the false alarm and missed detection probabilities equal.
5. The method for activation detection and channel estimation of large-scale connections assisted by intelligent reflective surfaces according to claim 4, characterized in that: The step 4 estimates the equivalent channel information of the direct link and the reflected link, specifically including: Assume that the set of activated users required in step 3 satisfies: The activated user randomly selects a pilot sequence from the orthogonal pilot resource pool Q for transmission. When RIS is in the ON state, the pilot reaches the BS end through the direct link and the reflected link, which can be expressed as: in, K a A direct link from user to BS, K a The channel information from user to RIS, F represents K a The pilot matrix of a user, let θ = [1,...,1] T , then the above formula can be expressed as: make represents the equivalent channel information. Due to the column orthogonality of F, the channel information H eq It can be expressed as:
6. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the activation detection and channel estimation method for large-scale connection assisted by an intelligent reflecting surface is implemented as described in any one of claims 1 to 5.
7. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the activation detection and channel estimation method for large-scale connection assisted by an intelligent reflecting surface is implemented as claimed in any one of claims 1 to 5.
8. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the activation detection and channel estimation method for large-scale connection assisted by an intelligent reflecting surface is implemented as claimed in any one of claims 1 to 5.
