Unlicensed random access method for joint pilot frequency data transmission

Abstract

The invention relates to the technical field of wireless communication, in particular to a joint pilot frequency data transmission authorization-free random access method, which comprises the following steps of: constructing a joint pilot frequency data transmission system model in cellular-free large-scale MIMO (Multiple Input Multiple Output), deducing a user spectrum efficiency closed expression, and establishing a maximum and minimum user power allocation optimization problem; reconstructing an objective function and constraint conditions in the optimization problem, constructing an inter-AP power distribution optimization problem model, converting the inter-AP power distribution optimization problem model into an equivalent convex feasibility problem by adopting a bisection method, and performing optimal power distribution; and constructing a constraint condition by using the optimally allocated spectrum efficiency, and solving by using a successive approximation method and an iterative algorithm model by using a maximum non-conflict user number as a target function to obtain a configuration parameter of maximum user number access of the joint pilot frequency data transmission system model. According to the method, multiple scenes of simultaneous access of large-scale users are met, the signal processing overhead is low, the time delay is low, the signal processing complexity is low, the optimal configuration scheme can be optimized, and the feasibility is good.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and more particularly to a method for random access to joint pilot data transmission without authorization. Background Art

[0002] At present, due to the advantages of non-cellular massive MIMO, such as being able to communicate with the central processor through the backhaul network, improving space utilization, and increasing spectrum utilization, it is considered to be a key candidate technology in 6G and future wireless communications. Patents related to non-cellular massive MIMO are also emerging in an endless stream. However, most of the related patents consider an idealized environment. In future communication systems, the number of communication users per square kilometer may be as high as 10^7, and large-scale access issues will become increasingly important. Traditional random access protocols are mainly divided into two categories: authorized access protocols and unauthorized access protocols. Authorized random access relies on a handshake mechanism to allocate a dedicated channel for each accessed user device and perform authentication. It is limited by limited channel resources and is unable to cope with large-scale random access scenarios. At the same time, it brings an extremely high signal processing burden and is difficult to meet the needs of future communication systems. In view of this, unauthorized random access has received widespread attention based on its advantages such as low signal processing overhead and low latency.

[0003] However, due to the large scale of users, the sequences are no longer orthogonal, which introduces additional multiple access interference, and further increases the processing complexity of the receiver. Usually, the receiver processing process often uses compressed sensing technology to recover sparse signals. However, although compressed sensing performs well in signal recovery, the processing complexity and resource overhead it brings cannot be ignored.

[0004] Therefore, how to provide a new data transmission solution to solve the shortcomings of the traditional unlicensed random access method in non-cellular large-scale MIMO, solve the problem that some unlicensed random access methods using orthogonal joint pilot data transmission are difficult to implement, and remove the problem of increased receiver signal processing complexity caused by compressed sensing and other methods are issues that urgently need to be solved by people in this field. Summary of the invention

[0005] In view of this, the present invention provides a method for random access without authorization for joint pilot data transmission to solve the technical problems in the above technical background. This method meets the needs of large-scale random access in various application scenarios in the era of the Internet of Things. The technical solution has the characteristics of easy implementation, good versatility, low signal processing overhead, low latency, and low complexity.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] S1. Based on the non-cellular massive MIMO system, a joint pilot data transmission system model is constructed by combining the backoff mechanism and frequency division multiplexing;

[0008] S2. Based on the constructed joint pilot data transmission system model, a closed-form expression of user spectrum efficiency is derived and a maximum-minimization user power allocation optimization problem is established;

[0009] S3. Reconstruct the objective function and constraints in the user power allocation optimization problem, build the power allocation optimization problem model between APs, and use the dichotomy method to transform the power allocation optimization problem model into multiple equivalent convex feasibility problems to perform optimal power allocation;

[0010] S4. Constraints are established based on the user spectrum efficiency after optimal power allocation. The maximization of the number of non-conflicting users is used as the objective function to obtain the optimization problem of the joint pilot data transmission system model when it is optimally configured. The optimization problem at the time of optimal configuration is converted into multiple geometric programming problems using a successive approximation algorithm and solved using an iterative algorithm to obtain the parameter configuration of the joint pilot data transmission system model with the largest number of users accessed.

[0011] Furthermore, the joint pilot data transmission system model constructed by S1 includes: a CPU, L access points AP and N single-antenna users, and two-way communication between the three; wherein each access point AP contains M antennas, M*L>>N, and the M antennas are uniformly linear arrays, distributed in geographical locations, and are used to collect signals from users, and send user data to the CPU through the backhaul network, and the CPU completes data processing; all users complete one pilot and Q data packets uplink transmission through the access point within the coherence time T, the pilot length in the pilot pool is τ, and there are τ groups of orthogonal pilots; within the pilot duration of length τ, the user attempts to activate according to the backoff probability broadcast by the access point during random access, and the activated user randomly selects a pilot, and users who select the same pilot sequence are defined as conflicting users, and users who select different pilot sequences from other users are defined as non-conflicting users;

[0012] The joint pilot data transmission system model divides the time T into W sub-time slots according to the pilot length, where S user sets will use their corresponding S sub-time slots for pilot transmission, S = WQ; in the first sub-time slot, the AP broadcasts a random access backoff probability, all users try to select the pilot and send the pilot signal to the AP, after receiving the pilot signal, the AP feeds back information to notify silent users and conflicting users to wait for the second sub-time slot to try the pilot; after the pilot transmission in the first sub-time slot, the AP broadcasts a new random access backoff probability, and the user entering the second sub-time slot repeats the same operation as in the first sub-time slot, and all users are divided into different user sets according to the backoff probability issued every τ sub-time slot length, and users in different user sets select different sub-time slots for pilot transmission; if the CPU detects the user and the AP estimates the user's channel, the user continuously and successfully transmits Q data packets in subsequent sub-time slots, and the data packet length is τ;

[0013] During the pilot duration of length τ, the received pilot matrix expression at the l end of the access point AP is:

[0014]

[0015] Among them, ρ p and ρ u is the normalized pilot power and normalized data power, g lk is the channel vector between user k and access point AP l, ν k and μ k are the pilot power control coefficient and data power control coefficient of user k, υ t is the orthogonal pilot sequence sent by user t, x k is the orthogonal data sequence sent by user k; N l,s is the receive noise matrix, is the set of non-conflicting users in sub-slot s, represents the integration of the set, where q s Represents the number of data packets transmitted in s sub-time slots.

[0016] Furthermore, the joint pilot data transmission system model includes a Rice channel model, which uses uplink orthogonal pilots to estimate the channel and derive the minimum mean square error estimation channel. The specific process is:

[0017] The channel vector g from user n to access point l ln It is expressed as:

[0018]

[0019] Among them, h ln is the signal component from the access point AP l to user n, β lnis the large-scale fading coefficient of the channel from access point AP l to user n, K ln represents the Rice K factor, which measures the power ratio between line-of-sight and non-line-of-sight signals; g ln,LOS is the line-of-sight channel vector, g ln,NLOS is the non-line-of-sight channel vector;

[0020]

[0021] h ln,NLOS represents the non-line-of-sight signal component, whose elements are independent and identically distributed Gaussian random variables with a mean of 0 and a variance of 1; h ln,LOS Represents the line-of-sight signal component;

[0022] Assume that all users in the joint pilot data transmission system model are randomly distributed in space and have different arrival angles, h ln,LOS The mth element of is represented as:

[0023]

[0024] Where d is the antenna spacing, λ is the wavelength, and θ ln is the arrival angle from user n to access point AP1; h ln,LOS , changes slowly in time and remains constant in a sufficient number of receiving stages; the Rice K factor and arrival angle in the LOS component are effectively estimated through channel feedback and sample mean estimation methods, and other parameters are known and accurate in the system, h ln,LOS is the known part of the system model;

[0025] According to the received pilot matrix at the l end of the access point AP, we can obtain the least squares estimated channel as:

[0026]

[0027] in, is the least squares estimated channel error vector, which is expressed as follows:

[0028]

[0029] and,

[0030] The channel estimation error introduced in the past sub-time slot will affect the channel estimation stage of the current sub-time slot. Assume that in the user set of the previous sub-time slot, the channel estimation error of user k satisfies:

[0031]

[0032] in, represents the variance of background noise and interference;

[0033] Then the least squares estimated channel error expression of the non-line-of-sight portion between user n and access point AP l in this sub-time slot is Satisfies the following distribution;

[0034]

[0035] Among them, s The distribution of users is consistent;

[0036] According to the approximate time series recurrence relationship:

[0037]

[0038] exist Assuming that t is a stationary process, σ 2 (t)→σ 2 The sufficient condition is:

[0039]

[0040] And, it is satisfied with

[0041]

[0042] in is the expectation operation, represents the average value of all possible sub-time slots s, and the formula is:

[0043]

[0044] Based on the above least squares estimation channel, the minimum mean square error estimation MMSE is used to obtain the minimum mean square estimation channel vector of the non-line-of-sight part between user n and the l end of the access point AP, and its expression is:

[0045]

[0046] in,

[0047] β ln is the large-scale fading coefficient of the channel from access point AP l to user n;

[0048] The integrated expression of the least squares estimation channel and the minimum mean square error estimation channel is:

[0049]

[0050] Where n is the serial number of the service user, l is the port number of the access point AP, K represents the estimation of the channel vector between user n and access point AP l, lnrepresents the Rice K factor, which measures the power ratio between line-of-sight and non-line-of-sight signals; g ln,LOS is the line-of-sight channel vector, is the estimate of the non-line-of-sight channel vector.

[0051] Furthermore, in the channel estimation phase of the joint pilot data transmission system model, in the first sub-time slot, only the pilot vector is used for channel estimation to successfully decode the data packet in the second sub-time slot; after decoding, the data packet is removed and the remaining signal is used to estimate the channel vector of the user sending the pilot in the second sub-time slot;

[0052] The first S sub-slots are all performed by repeating the operations in the first sub-slot to gradually decode the data packets of all users;

[0053] Each user has a pilot signal and Q data packets, and it takes S+Q sub-time slots for all users to be detected and complete the uplink transmission of data packets.

[0054] Furthermore, in S2, based on the joint pilot data transmission system model, the process of deriving the closed-form expression of the user spectrum efficiency is:

[0055] All users send data to the AP immediately after the pilot signal is sent. Set the uplink data symbol sent by the nth user to x. k1 , satisfying |x k1 | 2 =1, the first symbol duration of the system in the s+εth sub-time slot, the quantized received data vector at the l-th end of the access point AP is expressed as:

[0056]

[0057] in, It is the set of pilot presence decision survivors defined in the s+ε sub-time slot, specifically defined as:

[0058]

[0059] Among them, F v,2 ={S+1,...,S+Q} represents the sub-slots where the pilot is lost;

[0060] Based on the maximum ratio joint receiver, the access point AP uses the estimated channel to construct a decoding matrix.

[0061] Will With r l,s+ε After the inner product is performed, it is sent to the CPU via the fronthaul link. Then, for the first symbol duration of the s+εth sub-time slot, the total demodulated signal received by the CPU for user n is:

[0062]

[0063] Using the worst incoherent noise theory method, for the first symbol duration of the s+εth sub-time slot, the signal-to-noise ratio expression of user n is:

[0064]

[0065] Based on the characteristics of massive MIMO, its closed-form expression is:

[0066]

[0067] in,

[0068]

[0069] η ln =1 indicates LS situation, otherwise η ln =ξ ln Indicates MMSE situation;

[0070] The system throughput is affected by the signal-to-noise ratio. For the first symbol duration of the s+εth sub-time slot, the closed-form expression of the spectrum efficiency of the nth user is expressed as

[0071] R n,s+ε =log2(1+SINR n,s+ε ).

[0072] Furthermore, in S2, based on the closed-form expression of total spectrum efficiency, the optimization problem of user power allocation is established as:

[0073]

[0074]

[0075]

[0076] s=1,...,S,ε=1,...,Q

[0077] The constraints are conditions that the power control factors should satisfy.

[0078] Furthermore, S3 includes:

[0079] S31. Introducing slack variables The objective function and constraints in the user power allocation optimization problem in S2 are reconstructed as follows:

[0080]

[0081] When p ν nWhen fixed, the objective function in S31 is quasi-convex and is further optimized using the bisection method, specifically:

[0082] S321. Setting the objective function The upper and lower bounds of and set up Solve the feasibility problem in the above geometric programming problem, specifically

[0083] S322. If the problem is feasible, then improve the lower bound, that is,

[0084] S323. If the problem is infeasible, reduce the upper bound, that is, Return to S33 and re-order Determine new feasibility issues again until End the iteration, where ∈ represents the maximum error;

[0085] S324, the central processor obtains the optimal parameter solution for power allocation between APs, and outputs {μ n}and

[0086] Furthermore, the specific steps of the maximum and minimum AP power control algorithm include:

[0087] L1, input τ, N, T, Q, p a ;

[0088] L2, Initialization ∈ and ∈>0;

[0089] L3. Settings Solving feasibility issues and

[0090] L4, using binary update

[0091] L5. If the problem is feasible, set

[0092] L6. If the problem is not feasible, set

[0093] L7, Output {μ n},

[0094] Furthermore, in S4, the step of obtaining the parameter configuration when the user access of the joint pilot data transmission system model is maximum is specifically:

[0095] S41, taking the number of non-conflicting users accessible to the system as the objective function, converting the user spectrum efficiency after power compensation into a constraint condition, and obtaining the optimization problem when the system access is maximum, which is expressed as:

[0096]

[0097] S42. Considering the stability of the system, the number of non-conflicting users in each sub-time slot is set to be equal, and the set system model tends to be steady-state. The number of non-conflicting users in each set is controllable through the backoff probability, which satisfies the function:

[0098]

[0099] Due to N s will decrease as s increases. Consider N s If the backoff probability is too low, it will fail to obtain the specified number of non-conflicting users. The minimum number of non-conflicting users in the last user set is 0, and its value is continuous. The system only considers the maximum value.

[0100] Without considering When the constraints are met, the optimization problem of the optimal solution of system configuration is:

[0101]

[0102] Among them, for The optimization problem is

[0103]

[0104] subject to 0≤p a ≤1

[0105] The problem is solved using the idea of ​​successive upper bound minimization;

[0106] After the above sub-problems are solved, the remaining system parameters τ and Q are integers with upper bounds and are obtained through traversal.

[0107] It can be seen from the above technical solutions that, compared with the prior art, the present invention has the following beneficial effects:

[0108] (1) The present invention utilizes an access point to broadcast a backoff probability in different sub-time slots to implement a pilot for frequency division multiplexing of all users. The data uplink transmission method does not require a communication handshake and is a method of random control of user access without authorization, which meets various scenarios of large-scale random access of users. Compared with the authorized random access method, it has the characteristics of low signal processing overhead and low latency. Compared with the one-to-one correspondence between pilots and users in the traditional solution, the present invention can use a small number of pilots to meet the scenarios of a large number of users, and can obtain feasible transmission capacity after optimization.

[0109] (2) Traditional unlicensed random access methods often use methods such as compressed sensing, which have complex overhead caused by signal processing. The method for uplink transmission of a large amount of user data provided by the present invention eliminates the complexity brought to the receiver by methods such as compressed sensing, benefits from the characteristics brought by non-cellular large-scale MIMO, and uses a local orthogonal method for signal processing to reduce complexity, which is obviously lower in complexity than methods such as compressed sensing. In addition, compared with traditional systems that do not use joint pilot data transmission, the present invention uses an orthogonal method for signal processing, which can achieve higher spectrum efficiency.

[0110] (3) Compared with other joint pilot data transmission authorization-free random access methods that only provide method descriptions but fail to provide effective system optimal configuration solutions, the technical solution of the present invention can implement the system's optimal configuration solution, can adapt to different system usage environments, and can obtain a complete system configuration solution through a specific optimization processing method, and has good implementability and versatility.

[0111] (4) The present invention proposes a new transmission model and allocation algorithm to address the shortcomings of existing unlicensed random access cellular-free large-scale MIMO scenarios, such as difficulty in resolving user conflicts and difficulty in evaluating spectrum efficiency. The main parameter in the algorithm of the present invention is the large-scale fading coefficient, which remains unchanged in many adjacent coherent time intervals. Therefore, the optimization problem of the present invention is simplified and has good operability. BRIEF DESCRIPTION OF THE DRAWINGS

[0112] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0113] Figure 1 It is an overall method flow chart of the method for random access of joint pilot data transmission without authorization according to an embodiment of the present invention;

[0114] Figure 2 is a transmission flow chart of pilot data in a method for random access without authorization for joint pilot data transmission according to an embodiment of the present invention;

[0115] Figure 3 It is an algorithm flow chart of the random access method for joint pilot data transmission without authorization according to an embodiment of the present invention;

[0116] Figure 4 It is a formula derivation and verification of the method for random access of joint pilot data transmission without authorization of the present invention and a comparison diagram with the traditional method;

[0117] Figure 5 A comparison table of the number of non-conflicting users and the number of required pilots in the algorithm simulation optimization of the embodiment of the present invention and the traditional method;

[0118] Figure 6 This is a comparison table of the minimum SE before system optimization and the minimum SE after optimization in the algorithm simulation optimization of an embodiment of the present invention and a traditional method. DETAILED DESCRIPTION

[0119] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0120] like Figure 1 As shown, an embodiment of the present invention discloses a method for random access of joint pilot data transmission without authorization, the method comprising the following steps:

[0121] Step 1: Based on the non-cellular massive MIMO system, a joint pilot data transmission system model in non-cellular massive MIMO is constructed in combination with backoff mechanism and frequency division multiplexing; the uplink orthogonal pilot is used to estimate the channel error and derive the minimum mean square error estimation channel;

[0122] Step 2: Calculate the expression structure of user spectrum efficiency through the joint pilot data transmission system model in non-cellular massive MIMO, obtain the closed-form expression of user spectrum efficiency under random access according to the characteristics of non-cellular massive MIMO, and establish the maximum and minimum user power allocation optimization problem;

[0123] Step 3: Reconstruct the objective function and constraints in the user power allocation optimization problem, build a power allocation optimization problem model between APs, and use the dichotomy method to transform the power allocation optimization problem model into multiple equivalent convex feasibility problems to perform optimal power allocation.

[0124] Step 4: Establish constraints based on the user spectrum efficiency after optimal power allocation, take maximizing the number of non-conflicting users as the objective function, and obtain the optimization problem when the joint pilot data transmission system model is optimally configured. Use the successive approximation algorithm to convert it into multiple geometric programming problems, and use the iterative algorithm to obtain the parameter configuration in the joint pilot data transmission system model when the number of users accessed is the maximum.

[0125] In step 1 of this embodiment, the constructed joint pilot data transmission system model in non-cellular massive MIMO includes: a CPU, L access points AP and N single-antenna users, and two-way communication between the three; each access point AP includes M antennas, M*L>>N, and the M antennas are uniformly linear arrays, distributed in geographical locations, and are used to collect signals from users, and send user data to the CPU through a backhaul network, and the CPU completes data processing; all users complete one pilot and Q data packets uplink transmission through the access point within a coherent time T, the pilot length in the pilot pool is τ, and there are τ groups of orthogonal pilots; within the pilot duration of length τ, the user attempts to activate according to the backoff probability broadcast by the access point during random access, and the activated user randomly selects a pilot, and users who select the same pilot sequence are defined as conflicting users, and users who select different pilot sequences from other users are defined as non-conflicting users;

[0126] like Figure 2 As shown, in the joint pilot data transmission method, the joint pilot data transmission system model divides the coherence time T into W sub-time slots of length τ according to the pilot length, wherein S user sets will use their corresponding S sub-time slots for pilot transmission, S=WQ; in the first sub-time slot, the AP broadcasts a random access backoff probability, all users try to select a pilot and send a pilot signal to the AP, and the AP feeds back information after receiving the pilot signal so that silent users and conflicting users can try the pilot while waiting for the second sub-time slot; after the pilot transmission in the first sub-time slot, the AP broadcasts a new random access backoff probability, and the user entering the second sub-time slot repeats the same operation as in the first sub-time slot, and finally, all users are divided into different user sets according to the backoff probability issued every τ sub-time slot length; users in different user sets transmit pilots according to the sub-time slots during the pilot transmission, and if the CPU can detect the user and the AP can estimate the user's channel, these users can successfully transmit data packets in subsequent sub-time slots, assuming that each user transmits Q data packets of length τ.

[0127] During the pilot duration of length τ, the received pilot matrix expression at the l end of the access point AP is:

[0128]

[0129] Among them, ρ p and ρu is the normalized pilot power and normalized data power, g lk is the channel vector between user k and access point AP l, ν k and μ k are the pilot power control coefficient and data power control coefficient of user k, υ t is the orthogonal pilot sequence sent by user t, x k is the orthogonal data sequence sent by user k; in addition, N l,s is the receive noise matrix, is the set of non-conflicting users in sub-slot s, represents the integration of the set, where q s Represents the number of data packets transmitted in s sub-time slots;

[0130] In order to fit the actual application, the Rice channel model is considered in channel processing. Since the line-of-sight part is known in large-scale MIMO, the known part is removed and the least squares method is used. It is assumed that the channel estimation error of user k in the previous user set satisfies At this time, the least squares estimated channel expression of the non-line-of-sight portion between user n and the l-end of the access point AP is:

[0131]

[0132] in, is the least squares estimated channel error vector, which is expressed as follows:

[0133]

[0134] and,

[0135] for In order to prevent the estimation error variance from increasing arbitrarily over time, which would lead to data transmission failure, we need to effectively control the joint pilot data transmission system model. The specific control constraints are as follows:

[0136]

[0137] Based on the least squares estimation channel expression of the non-line-of-sight part, the least mean square estimation channel expression of the non-line-of-sight part between user n and the l end of the access point AP is derived using the minimum mean square error criterion:

[0138]

[0139] in β ln is the large-scale fading coefficient of the channel from access point AP l to user n.

[0140] Finally, the least squares and least mean squares estimated channel expressions between the integrated user n and the access point AP l are:

[0141]

[0142] Furthermore, in step 2, the specific process of deriving the closed-form expression of user spectrum efficiency is as follows:

[0143] All users send data to the AP immediately after the pilot signal is sent, regardless of whether a conflict occurs. Set the uplink data symbol sent by the nth user to x k1 , satisfying |x k1 | 2 =1, considering the generality, taking the first symbol duration in the s+εth sub-time slot, the quantized received data vector at the l-th end of the access point AP is expressed as:

[0144]

[0145] in, It is the set of pilot presence decision survivors defined in the s+ε sub-time slot, specifically defined as:

[0146]

[0147] F v,2 ={S+1,...,S+Q} represents the sub-slots where the pilot is lost;

[0148] Based on the maximum ratio joint receiver, the access point AP uses the estimated channel to construct a decoding matrix. With r l,s+ε After the inner product is performed, it is sent to the CPU via the fronthaul link. Then, for the first symbol duration of the s+εth sub-time slot, the total demodulated signal received by the CPU for user n is:

[0149]

[0150] Using the worst incoherent noise theory method, for the first symbol duration of the s+εth sub-time slot, the signal-to-noise ratio expression of user n is:

[0151]

[0152] Based on the characteristics of massive MIMO, its closed-form expression is:

[0153]

[0154] The system throughput is affected by the signal-to-noise ratio. For the first symbol duration of the s+εth sub-time slot, the closed-form expression of the spectrum efficiency of the nth user is expressed as

[0155] R n,s+ε =log2(1+SINR n,s+ε )

[0156] Furthermore, in step 2 of the present invention, based on the closed-form expression of total spectrum efficiency, the optimization problem of user power allocation is established as follows:

[0157]

[0158]

[0159] Among them, the constraints are the conditions that the power control factors should satisfy;

[0160] Obviously, the optimization problem of user power allocation is not a standard optimization problem, and it is difficult to judge its convexity, so it is difficult to find its optimal solution. In order to effectively solve the optimization problem of user power allocation, the slack variable is introduced. The objective function and constraints in the optimization problem of user power allocation are reconstructed, and the reconstructed optimization problem is:

[0161]

[0162] When p ν n When fixed, the objective function of the reconstructed optimization problem model is quasi-convex, so the bisection method is used to solve the model optimization problem.

[0163] Set the problem objective function The upper and lower bounds of and And settings Solve the feasibility issues mentioned above If the problem is feasible, then improve the lower bound If the problem is infeasible, then reduce the upper bound to Re-order Determine feasibility; until End the iteration, where ∈ represents the maximum error.

[0164] The specific steps of the maximum and minimum AP power control algorithm are shown in sub-algorithm 1:

[0165]

[0166]

[0167] Furthermore, in step 4 of the present invention, obtaining the configuration parameters of the model when the user access is maximum in the joint pilot data transmission system model specifically includes:

[0168] The number of users that can be connected to the joint pilot data transmission system model is taken as the objective function, and the optimal solution for the configuration of the joint pilot data transmission system model is obtained. The optimization problem is expressed as follows:

[0169]

[0170] Considering the stability of the joint pilot data transmission system model, the number of non-conflicting users in each sub-time slot should be roughly equal. Since the set joint pilot data transmission system model should be more stable, we can assume that the number of non-conflicting users in each set should be roughly equal. In addition, since the number of non-conflicting users in each set can be controlled by the backoff probability, it satisfies the function:

[0171]

[0172] Due to N s will decrease as s increases. Consider N s Too low will result in the backoff probability failing to obtain the specified number of non-conflicting users. Therefore, the range of the number of non-conflicting users in the last user set is the range of the number of non-conflicting users in each set of the joint pilot data transmission system model. Considering that the minimum value of this range is 0 and continuous, only the maximum value needs to be considered. When the constraints are met, the optimization problem of the optimal solution for the joint pilot data transmission system model configuration can be optimized as follows:

[0173]

[0174] Among them for The optimization problem is

[0175]

[0176] The problem can be solved using the idea of ​​successive upper bound minimization.

[0177] The specific steps are shown in Sub-Algorithm 2:

[0178]

[0179] like Figure 3 As shown, τ and Q are both integers with upper bounds. To this end, the required original data is obtained by traversing the feasible configuration of the joint pilot data transmission system model. Specifically, traversing τ and Q to obtain 's previous term, and traverse Get the number of users in the last sub-time slot Input it into sub-algorithm 2 and continuously obtain Use with Storage. If Save the corresponding parameter information of the joint pilot data transmission system model; otherwise discard it. Sort the stored information from large to small, and input the largest data into sub-algorithm 1 to obtain {μ n}and Determine the feasibility of the joint pilot data transmission system model and If it is true, the optimal parameters τ, Q, {μ n},{p a,s}; Otherwise, transmit the next data.

[0180] The above iteration is expressed as comprehensive algorithm 1:

[0181]

[0182] Figure 5 and Figure 6 The two tables in show the optimal number of non-collision users and their corresponding pilot requirements for the new scheme and the traditional scheme in the LS and MMSE cases, respectively.

[0183] In the joint pilot data transmission system model, we assume that the minimum rate R min =5bit / s / Hz, the number of antennas M=100, and considering five cases, namely N=10, 20, 30, 50, 80, the maximum number of pilots allowed by the joint pilot data transmission system model is set to 30.

[0184] Figure 5 The table in compares the number of non-conflicting users and their corresponding pilot requirements for the traditional method and the PDSDI method in the LS and MMSE cases. The results show that the PDSDI method outperforms the traditional method in terms of increasing the number of non-conflicting users and reducing the pilot requirements, especially in user-dense scenarios. Whether in the LS or MMSE case, the PDSDI method outperforms the traditional method in various configurations, showing higher efficiency and robustness.

[0185] Figure 6 The table in compares the optimized (Opt.) and unoptimized (Unopt.) results of the minimum SE values ​​obtained by the user under the traditional method and the PDSDI method, and analyzes the LS and MMSE cases respectively. Figure 6The table in shows that the optimized PDSDI results are significantly better than the unoptimized results in both LS and MMSE cases, providing higher SE values. This shows that after optimization, the PDSDI method improves the user SE value under LS and MMSE configurations, demonstrating that its optimization performance is no less than that of the traditional method.

[0186] In this embodiment, during the estimation phase of the joint pilot data transmission channel, in the first sub-time slot, only the pilot vector is used for channel estimation, so that the data packet in the second sub-time slot can be successfully decoded; after decoding, the data packet will be removed, and the remaining signal can be used to estimate the channel vector of the user sending the pilot in the second sub-time slot. In this way, all data packets can be decoded step by step.

[0187] This embodiment is a transmission model suitable for unlicensed random access to non-cellular large-scale MIMO scenarios, and a system parameter optimization strategy is designed for this model to meet different application scenarios. The main body includes: transmitting pilot data in the same frame and automatically dividing users so as to obtain pilots from the same pilot pool multiple times without conflict, greatly reducing the system conflict rate; since the overall spectrum efficiency of the system is difficult to optimize, considering the threshold of the spectrum efficiency of a single user, compensation is made through power optimization, and the spectrum efficiency of all users meets the minimum requirements. At this time, the optimization target is converted to the number of non-conflicting users in the system, thereby indirectly obtaining a relatively good system spectrum efficiency.

[0188] This embodiment proposes a new transmission model and allocation algorithm to address the shortcomings of the existing unlicensed random access non-cellular large-scale MIMO scenario, which is difficult to resolve user conflicts and difficult to evaluate spectrum efficiency. The main parameter in the algorithm involved in the present invention is the large-scale fading coefficient, which remains unchanged in many adjacent coherent time intervals. Therefore, the optimization problem involved in this embodiment is simplified and has good operability.

[0189] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0190] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for random access without authorization for joint pilot data transmission, characterized in that: The steps include: S1. Based on the non-cellular massive MIMO system, a joint pilot data transmission system model is constructed in combination with backoff mechanism and frequency division multiplexing; S2. Derivation of a closed-form expression for user spectrum efficiency through a joint pilot data transmission system model, and establishment of a maximum-minimum user power allocation optimization problem; S3. Reconstruct the objective function and constraints in the user power allocation optimization problem, build a power allocation optimization problem model between APs, and use the dichotomy method to transform the power allocation optimization problem model into multiple equivalent convex feasibility problems to perform optimal power allocation. S4. Constraints are established based on the user spectrum efficiency after optimal power allocation. The maximization of the number of non-conflicting users is used as the objective function to obtain the optimization problem of the joint pilot data transmission system model when it is optimally configured. The successive approximation algorithm is used to transform the optimization problem at the optimal configuration into multiple geometric programming problems. The iterative algorithm is used to obtain the parameter configuration of the joint pilot data transmission system model when the number of users accessed is the maximum.

2. The method for grant-free random access of joint pilot data transmission according to claim 1, characterized in that: The joint pilot data transmission system model constructed by S1 includes: a CPU, L access points AP and N single-antenna users, and two-way communication between the three; each access point AP contains M antennas, M*L>>N, and the M antennas are uniformly linear arrays, distributed in geographical locations, used to collect signals from users, send user data to the CPU through the backhaul network, and the CPU completes data processing; all users complete one pilot and Q data packets uplink transmission through the access point within the coherence time T, the pilot length in the pilot pool is τ, and there are τ groups of orthogonal pilots; within the pilot duration of length τ, the user attempts to activate according to the backoff probability broadcast by the access point during random access, and the activated user randomly selects the pilot, and the user who selects the same pilot sequence is defined as a conflicting user, and the user who selects a different pilot sequence from other users is defined as a non-conflicting user; The joint pilot data transmission system model divides the coherent time T into W sub-timeslots according to the pilot length, where S user sets will use their corresponding S sub-timeslots for pilot transmission, S = WQ; in the first sub-timeslot, the AP broadcasts a random access backoff probability, all users try to select the pilot and send the pilot signal to the AP, after receiving the pilot signal, the AP feeds back information to notify silent users and conflicting users to wait for the second sub-timeslot to try the pilot; after the pilot transmission in the first sub-timeslot, the AP broadcasts a new random access backoff probability, and the users entering the second sub-timeslot repeat the same operation as in the first sub-timeslot, and all users are divided into different user sets according to the backoff probability issued every τ sub-timeslot length, and users in different user sets select different sub-timeslots for pilot transmission; if the CPU detects the user and the AP estimates the user's channel, the user continuously and successfully transmits Q data packets in subsequent sub-timeslots, and the data packet length is τ; During the pilot duration of length τ, the received pilot matrix expression at the l end of the access point AP is: Among them, ρ p and ρ u is the normalized pilot power and normalized data power, g lk is the channel vector between user k and access point AP l, ν k and μ k are the pilot power control coefficient and data power control coefficient of user k, υ t is the orthogonal pilot sequence sent by user t, x k is the orthogonal data sequence sent by user k; N l,s is the receive noise matrix, is the set of non-conflicting users in sub-slot s, represents the integration of the set, where q s Represents the number of data packets transmitted in s sub-time slots.

3. The method for grant-free random access of joint pilot data transmission according to claim 2, characterized in that: The joint pilot data transmission system model includes the Rice channel model, which uses the uplink orthogonal pilot to estimate the channel and derive the minimum mean square error estimation channel. The specific process is: The channel vector g from user n to access point l ln It is expressed as: Among them, h ln is the signal component from the access point AP l to user n, β ln is the large-scale fading coefficient of the channel from access point AP l to user n, K ln represents the Rice K factor, which measures the power ratio between line-of-sight and non-line-of-sight signals; g ln,LOS is the line-of-sight channel vector, g ln,NLOS is the non-line-of-sight channel vector; h ln,NLOS represents the non-line-of-sight signal component, whose elements are independent and identically distributed Gaussian random variables with a mean of 0 and a variance of 1; h ln,LOS Represents the line-of-sight signal component; Assume that all users in the joint pilot data transmission system model are randomly distributed in space and have different arrival angles, h ln,LOS The mth element of is represented as: Where d is the antenna spacing, λ is the wavelength, and θ ln is the arrival angle of user n to access point l; h ln,LOS , changes slowly in time and remains constant in a sufficient number of receiving stages; the Rice K factor and arrival angle in the LOS component are effectively estimated through channel feedback and sample mean estimation methods, and other parameters are known and accurate in the system, h ln,LOS is the known part of the system model; According to the received pilot matrix expression at the l end of the access point AP, the least squares estimation channel expression of the non-line-of-sight part is derived as follows: in, is the least squares estimated channel error vector, which is expressed as follows: and, The channel estimation error introduced in the past sub-time slot will affect the channel estimation stage of the current sub-time slot. Assume that in the user set of the previous sub-time slot, the channel estimation error of user k satisfies: in, represents the variance of background noise and interference; Then the least squares estimated channel error expression of the non-line-of-sight portion between user n and access point AP l in this sub-time slot is Satisfies the following distribution; Among them, s The distribution of users is consistent; According to the approximate time series recurrence relationship: exist is a stationary process of t, σ 2 (t)→σ 2 The sufficient condition is: And, it is satisfied with in is the expectation operation, represents the average value of all possible sub-time slots s, and the formula is: Based on the above least squares estimation channel, the minimum mean square error estimation MMSE is used to obtain the minimum mean square estimation channel vector of the non-line-of-sight part between user n and the l end of the access point AP, and its expression is: in, β ln is the large-scale fading coefficient of the channel from access point AP l to user n; The integrated expression of the least squares estimation channel and the minimum mean square error estimation channel is: Where n is the serial number of the service user, l is the port number of the access point AP, K represents the estimation of the channel vector between user n and access point AP l, ln represents the Rice K factor, which measures the power ratio between line-of-sight and non-line-of-sight signals; g ln,LOS is the line-of-sight channel vector, is the estimate of the non-line-of-sight channel vector.

4. The method for grant-free random access of joint pilot data transmission according to claim 1, characterized in that: In the channel estimation phase of the joint pilot data transmission system model, in the first sub-time slot, only the pilot vector is used for channel estimation to successfully decode the data packet in the second sub-time slot; After decoding, the data packet is removed and the remaining signal in the second sub-slot is used to estimate the channel vector of the user that sent the pilot in the second sub-slot; The first S sub-slots are all performed by repeating the operations in the first sub-slot to gradually decode the data packets of all users; Each user has a pilot signal and Q data packets, and it takes S+Q sub-time slots for all users to be detected and complete the uplink transmission of data packets.

5. The method for grant-free random access of joint pilot data transmission according to claim 3, characterized in that: In S2, based on the joint pilot data transmission system model, the process of deriving the closed-form expression of user spectrum efficiency is: All users send data to the AP immediately after the pilot signal is sent. Set the uplink data symbol sent by the nth user to x. k1 , satisfying |x k1 | 2 =1, the first symbol duration of the system in the s+εth sub-time slot, the quantized received data vector at the l-th end of the access point AP is expressed as: in, It is the set of pilot presence decision survivors defined in the s+ε sub-time slot, specifically defined as: F v,2 ={S+1,...,S+Q} represents the sub-slots where the pilot is lost; Based on the maximum ratio joint receiver, the access point AP uses the estimated channel to construct a decoding matrix. With r l,s+ε After the inner product is performed, it is sent to the CPU via the fronthaul link. Then, for the first symbol duration of the s+εth sub-time slot, the total demodulated signal received by the CPU for user n is: Using the worst incoherent noise theory method, for the first symbol duration of the s+εth sub-time slot, the signal-to-noise ratio expression of user n is: Based on the characteristics of massive MIMO, its closed-form expression is: in, η ln =1 indicates LS situation, otherwise η ln =ξ ln Indicates MMSE situation; The system throughput is affected by the signal-to-noise ratio. For the first symbol duration of the s+εth sub-time slot, the closed-form expression of the spectrum efficiency of the nth user is expressed as R n,s+ε =log2(1+SINR n,s+ε )。 6. The method for grant-free random access of joint pilot data transmission according to claim 5, characterized in that: In S2, based on the closed-form expression of total spectrum efficiency, the optimization problem of user power allocation is established as: s=1,...,S,ε=1,...,Q The constraints are conditions that the power control factors should satisfy.

7. The method for grant-free random access of joint pilot data transmission according to claim 6, characterized in that: S3 includes: S31. Introducing slack variables The objective function and constraints in the user power allocation optimization problem in S2 are reconstructed as follows:

8. The method for grant-free random access of joint pilot data transmission according to claim 7, characterized in that: When p ν n When fixed, the objective function in S31 is quasi-convex and is further optimized using the bisection method, specifically: S321. Setting the objective function The upper and lower bounds of and set up Solve the feasibility problem in the above geometric programming problem, specifically S322. If the problem is feasible, then improve the lower bound, that is, S323. If the problem is infeasible, reduce the upper bound, that is, Return to S33 and re-order Determine new feasibility issues again until End the iteration, where ∈ represents the maximum error; S324, the central processor obtains the optimal parameter solution for power allocation between APs, and outputs {μ n }and 9. The method for grant-free random access of joint pilot data transmission according to claim 1, characterized in that: The power allocation optimization problem model between APs includes a maximum and minimum AP power control algorithm, and the specific steps of the maximum and minimum AP power control algorithm include: L1, input τ, N, T, Q, p a ; L2, Initialization ∈ and ∈>0; L3. Settings Solving feasibility issues and L4, using binary update L5. If the problem is feasible, set L6. If the problem is not feasible, set L7, while Output {μ n }, 10. The method for grant-free random access of joint pilot data transmission according to claim 8 or 9, characterized in that: In S4, the steps of obtaining the maximum system configuration for system access are as follows: S41, taking the number of non-conflicting users accessible to the system as the objective function, converting the user spectrum efficiency after power compensation into a constraint condition, and obtaining the optimization problem when the system access is maximum, which is expressed as: S42. Considering the stability of the system, the number of non-conflicting users in each sub-time slot is set to be equal, and the set system model tends to be steady-state. The number of non-conflicting users in each set is controllable through the backoff probability, which satisfies the function: Due to N s will decrease as s increases. Consider N s If the backoff probability is too low, it will fail to obtain the specified number of non-conflicting users. The minimum number of non-conflicting users in the last user set is 0, and its value is continuous. The system only considers the maximum value. Without considering When the constraints are met, the optimization problem of the optimal solution of system configuration is: Among them, for The optimization problem is subject to 0≤p a ≤1 The problem is solved using the idea of ​​successive upper bound minimization; After the above sub-problems are solved, the remaining system parameters τ and Q are integers with upper bounds and are obtained through traversal.

Images