User association and mode selection joint optimization method for network-assisted free duplex cellular-free wireless access network

Abstract

The invention discloses a user association and mode selection joint optimization method for a network-assisted free duplex cellular-free wireless access network. The method comprises the following steps: establishing a network-assisted free duplex architecture system; establishing user QoS constraint, user antenna mode selection and EDU-UE association problems by taking maximization of the total spectrum efficiency of the system as a target; determining an association relationship between the user and the EDU through a matching game algorithm; dynamically scheduling uplink and downlink working modes of a UE antenna by adopting a quantum genetic algorithm; the method has good expandability.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless communication transmission, and in particular to a method for joint optimization of user association and mode selection in a network-assisted free duplex cellular-free wireless access network. Background Art

[0002] Time division duplex (TDD) and frequency division duplex (FDD) are commonly used fixed duplex modes. To increase system resource utilization, in-band full-duplex (IBFD) and same-frequency full-duplex (CCFD) technologies enable base stations to simultaneously transmit and receive data within the same frequency band, overcoming the limitations of traditional half-duplex and doubling spectrum efficiency. However, in practice, the base station's uplink transmit antenna can cause significant interference to the downlink receive antenna, resulting in performance loss. The in-band wireless full-duplex coordinated multipoint (CoMPflex) system uses two spatially separated and coordinated half-duplex (HD) base stations to simulate a full-duplex (FD) base station, reducing the impact of self-interference by separating the uplink and downlink antennas. However, in large-scale communication scenarios, where base stations are densely deployed, these duplex technologies still face the challenge of inter-antenna interference.

[0003] In the Network-Assisted Full-Duplex (NAFD) architecture, APs can flexibly operate in full-duplex, CCFD, and mixed-duplex modes. Within the same time and frequency domains, different APs can flexibly select uplink and downlink operating modes, enabling simultaneous signal transmission and reception and mitigating self-interference in CCFD. However, NAFD primarily considers duplex mode scheduling on the AP side, ignoring duplex mode selection on the user side. Furthermore, NAFD systems lack scalability. Summary of the Invention

[0004] Purpose of the invention: The purpose of the present invention is to provide a joint optimization method for user association and mode selection in a network-assisted free duplex cellular-free wireless access network to solve the problem of lack of scalability in the prior art.

[0005] Technical solution: The network-assisted free duplex cellular-free wireless access network user association and mode selection joint optimization method of the present invention includes the following steps: (1) Establish a network-assisted free duplex architecture system, including a central processing unit (CPU), multiple edge distributed units (EDUs), access points (APs) with fixed uplink and downlink working states, and multi-antenna user equipment (UEs); (2) To maximize the total system spectrum efficiency, establish user QoS constraints, user antenna mode selection and EDU-UE association issues; (3) Determine the association between users and EDUs through matching game algorithms; (4) Use quantum genetic algorithm to dynamically schedule the uplink and downlink working modes of UE antennas.

[0006] Furthermore, in step (1), each EDU is equipped with several uplink APs and downlink APs, and the antenna working mode of the UE can be dynamically scheduled.

[0007] Furthermore, in step (2), the total spectrum efficiency is composed of the superposition of the downlink total spectrum efficiency and the uplink total spectrum efficiency; wherein the downlink total spectrum efficiency includes the useful signal of the user, the interference signals of other downlink users, the interference signal of the uplink antenna to the downlink antenna, and the channel noise; the uplink total spectrum efficiency includes the useful signal of the target user, the interference signals of other uplink users, the interference signal of the downlink antenna to the uplink antenna, and the channel noise.

[0008] Furthermore, in step (2), the problem is modeled as: ; in, Downlink total spectrum efficiency; is the total uplink spectrum efficiency; and is the uplink and downlink working vector of user k; Indicates the association between user k and EDUx; EDUx represents the association coefficient between user and EDU, where 1 indicates association and 0 indicates no association.

[0009] Furthermore, the constraints include: ; ; ; in, is the QoS constraint for downlink user k, is the QoS constraint for uplink user k; ; They represent user k, The nth antenna works in uplink and downlink; D,k represents the downlink spectrum efficiency of user k; r U,k represents the uplink spectrum efficiency of user k.

[0010] Furthermore, in step (3), the matching game algorithm includes the following steps: (31) The user and EDU generate a two-way preference list based on the utility function, where the utility function is the total uplink and downlink spectrum efficiency of the system; (32) The free UE initiates a matching request to the EDU at the top of the preference list, and the EDU accepts or rejects the request based on the associated threshold; (33) The surplus UE initiates a secondary matching request, and the EDU determines whether to associate based on the spectrum efficiency improvement; (34) Perform exchange matching and exchange blocking to associate users; determine whether to exchange based on the utility function, which is the system uplink and downlink speed.

[0011] Furthermore, in step (4), the quantum genetic algorithm includes the following steps: (41) Initialize the quantum population: All amplitude pairs of each individual Initialized as: ; in, Represent the probability of each quantum bit taking the ground state and excited state respectively, and satisfy ; (42) Measuring the population state: Measure each individual in the population once to obtain the individual state; (43) Calculate fitness: Calculate the fitness of each individual, using the system and rate as the fitness function; (44) Quantum revolving gate evolution: generating new populations based on quantum revolving gates; (45) Quantum NOT gate mutation: Introducing quantum NOT gate to perform quantum mutation operation on the population; (46) Repeat steps (42) to (45) until the algorithm converges or the maximum number of iterations is reached.

[0012] The network-assisted free duplex cellular-free wireless access network user association and mode selection joint optimization system of the present invention comprises: Duplex architecture module: used to establish a network-assisted free duplex architecture system, including a central processing unit (CPU), multiple edge distributed units (EDUs), access points (APs) with fixed uplink and downlink working states, and multi-antenna user equipment (UEs); Matching game algorithm module: used to determine the association between users and EDUs through the matching game algorithm to meet backhaul constraints and user QoS requirements; Quantum genetic algorithm module: used to dynamically schedule the uplink and downlink working modes of UE antennas using quantum genetic algorithm; Joint optimization module: used to jointly optimize EDU-UE association and UE antenna mode selection with the goal of maximizing the system's total spectrum efficiency.

[0013] An electronic device according to the present invention includes a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any one of the methods when executing the program.

[0014] The computer-readable storage medium of the present invention stores a computer program, which implements the steps of any one of the methods when executed by a processor.

[0015] Beneficial effects: Compared with the existing technology, the present invention has the following significant advantages: The present invention studies a network-assisted free duplex architecture based on a non-cellular wireless access network. In the uplink, the AP receives the transmission signal from the UE uplink antenna, and then sends it to the edge distributed unit (EDU). The EDU is responsible for signal demodulation and sends the processed data to the central processing unit (CPU) for merging. At the same time, the CPU sends the downlink data to the EDU for precoding processing, and then sends it to the downlink UE through the associated AP. The UE receives the signal using the downlink antenna. The user can flexibly adjust the uplink and downlink working modes of the antenna, make full use of system resources, and improve the spectrum utilization efficiency of the system. At the same time, considering the user's service quality and the system's backhaul constraints, flexible matching and association of EDU and UE are performed. Therefore, this architecture can achieve a compromise between system performance and computational complexity and has good scalability. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a schematic diagram of a communication system model of the present invention; Figure 2 is a performance gain diagram of the EDU-UE association method of the present invention; Figure 3 It is a performance gain diagram of the user antenna mode selection method of the present invention. DETAILED DESCRIPTION

[0017] The technical solution of the present invention will be further described below with reference to the accompanying drawings.

[0018] like Figure 1 As shown, an embodiment of the present invention provides a network-assisted free duplex cellular-free wireless access network user association and mode selection joint optimization method, including the following steps: S1 establishes a network-assisted free duplex architecture system model: let the number of EDUs in the system be X, and the number of APs mounted on each EDU be The APs in the system have fixed uplink and downlink. Uplink AP, Downlink APs, where the number of AP antennas is M. For a single , The number of uplink APs mounted on it, The number of downlink APs mounted on it.

[0019] Consider the case of multiple user antennas. Assume that there are K users in total, the number of user antennas is 2, and all users are in full-duplex state, that is, one antenna of each user works in the uplink and the other antenna works in the downlink. , Represents a user The nth antenna works in the uplink (downlink) and must satisfy , and is the uplink and downlink working vector of user k.

[0020] The specific process of downlink transmission is as follows: According to the local channel state information, the precoding vector is ,in for Mounted The precoding vector between the downlink AP and user k. for The downlink channel between user k, where for Mounted A downlink channel between AP and user k. The signal received by user k is: ; in, ;in, ; for The signal sent to downlink user k is for The transmission power for user k. For users The channel between user k. ,in, , For users The transmission power. is additive Gaussian white noise; is the interference of the uplink antenna of the full-duplex user to the downlink antenna, and the residual error after the interference is eliminated. In the final equation, the first term on the right side of the equal sign is the useful signal of user k, the second term is the downlink interference signal of other users, the third term is the uplink interference signal of other users, the fourth term is the channel noise, and the fifth term is the residual interference error from the uplink antenna to the downlink antenna after interference cancellation for full-duplex user k.

[0021] Therefore, the downlink spectrum efficiency calculated at user k is: ; ; Downlink total spectrum efficiency for: .

[0022] The specific process of uplink transmission is as follows: The uplink signal received at is: ;in, for The uplink channel between user k, for Mounted The uplink channel between AP and user k. express The downlink signal sent by the mounted downlink AP is The interference generated by the uplink AP, matrix express Mounted downlink AP and Channel between mounted uplink APs. is additive white Gaussian noise.

[0023] Considering that some EDUs can share data, define To be able to The shared EDU set and its complement For not being able to EDU collection to be shared. Can be used with For data sharing, Know the signal , so IAI interference cancellation can be performed. However, considering the channel estimation error, the interference cannot be completely eliminated. is the channel estimation error of IAI, and follows Therefore, the received uplink signal can be rewritten as: ; exist Design receiver , the demodulated user The uplink signal sent is: , so the spectrum efficiency of uplink user k integrated at the CPU is: ;

[0024] ; Uplink total spectrum efficiency for: .

[0025] To maximize the total system spectrum efficiency, the EDU-UE association and UE antenna mode selection are jointly optimized as follows: The optimization problem is modeled as: ; in, Downlink total spectrum efficiency; is the total uplink spectrum efficiency; and is the uplink and downlink working vector of user k; Indicates the association between user k and EDUx; EDUx represents the association coefficient between user and EDU, where 1 indicates association and 0 indicates no association.

[0026] Constraints include: ; ; ; in, is the QoS constraint for downlink user k, is the QoS constraint for uplink user k; ; They represent user k, The nth antenna works in uplink and downlink; D,k represents the downlink spectrum efficiency of user k; r U,k represents the uplink spectrum efficiency of user k.

[0027] S2 determines the association between users and EDUs through a matching game algorithm. The specific algorithm steps are as follows: (21) Each user will form a preference list based on the utility function value of each EDU in descending order, where the utility function is the total spectrum efficiency of uplink and downlink. Similarly, each EDU will form a preference list based on the utility function value of each user in descending order.

[0028] (22) Define users who are not associated with any EDU as free UEs. All free UEs send matching requests to the first EDU in their respective preference lists, form an application list for each EDU, and delete these EDUs from the corresponding UE's preference list.

[0029] (23) Traverse all EDUs. If the total number of UEs associated with the current EDU plus the number of UEs applying for matching does not exceed the association threshold, the EDU will accept all matching applications and update the matching list. Otherwise, the EDU will select (threshold number - associated number) UEs that are ranked before the EDU's preference list from the UEs applying for matching and associate them, and update the matching list. If the number of UEs associated with the EDU has reached the threshold, all applications will be rejected.

[0030] (24) A UE associated with an EDU is deleted from the free UE list.

[0031] (25) Repeat (22)-(24) until the free UE list is cleared.

[0032] (26) At this point, all UEs have been associated with an EDU. UEs whose preference lists have not been cleared are defined as redundant UEs.

[0033] (27) Similarly, all the surplus UEs send matching requests to the first EDU in their respective preference lists, forming a request list for each EDU and deleting it from the corresponding UE's preference list.

[0034] (28) Traverse all EDUs that have received applications. If the total number of UEs associated with the current EDU plus the number of UEs applying for matching does not exceed the association threshold, the EDU will make a judgment in sequence: if association can improve spectrum efficiency, then choose to associate; otherwise, reject and update the matching list. Otherwise, the EDU will select (threshold number - number of associated UEs) UEs that are ranked before the EDU's preference list from the UEs applying for matching and make judgments in sequence, and update the matching list. If the number of EDU associations has reached the threshold, all applications will be rejected.

[0035] (29) The UEs whose preference lists have been cleared are deleted from the surplus UE list. Repeat steps (27)-(29) until the surplus UE list is cleared. Perform exchange matching and exchange blocking to associate users. Define the utility function as the system uplink and downlink rate. If the utility is improved, the exchange operation is allowed to be performed, otherwise it is not performed.

[0036] like Figure 2 As shown in the figure, the performance gain graph of the matching theory algorithm for EDU-UE association in the present invention shows that the matching theory algorithm proposed in the present invention shows significant advantages in the EDU-UE association problem. When there are 7 users, the spectrum efficiency achieved by the matching theory algorithm is improved by 7.4%, 9.7%, and 62% compared with DCC, K-means, and random association, respectively. Its performance advantage mainly comes from the dynamic and stable bilateral matching mechanism. This mechanism optimizes the association relationship between users and EDUs in real time, effectively reducing inter-user interference while ensuring service quality. The matching theory algorithm achieves a good balance between computational complexity and performance, and is particularly suitable for the real-time resource scheduling requirements in large-scale user scenarios, providing an effective solution for the scalability of non-cellular MIMO systems.

[0037] S3 uses a quantum genetic algorithm to dynamically schedule the uplink and downlink operating modes of UE antennas. The specific algorithm flow is as follows: (31) Initialize the quantum population: There are a total of quantum individuals, since each user has two antennas, each antenna only works in uplink or downlink and the working modes of the two antennas are complementary, that is, Therefore, when one antenna mode is determined, the other antenna mode is also determined. Consider the first antenna working mode of each user as an individual. Therefore, each individual is composed of the amplitude pair with the dimension of the total number of users K. Initialize all amplitude pairs of each individual to ,Right now ; in, Represent the probability of each quantum bit taking the ground state and excited state respectively, and satisfy ; (32) Measure a population once: measure each individual in the population once and obtain the individual's state ;state The measurement satisfies the rule: ,in, is a random variable uniformly distributed between 0 and 1.

[0038] (33) Fitness calculation: According to Calculate the fitness of each individual and take the system and rate as the fitness function, that is, ,in, Therefore, the state of the individual with the highest fitness value can be identified as the target optimal value for the evolution of the population individuals. .

[0039] (34) Quantum revolving gate evolution: A new population is generated based on the quantum revolving gate. The quantum revolving gate function is defined as , the updated population is ,in, ; is the quantum rotation angle, The specific quantum rotation strategy is shown in Table 1.

[0040] Table 1 Quantum rotation strategy ; Among them, in Table 1, Represents the antenna state of each individual k-th user in the population; The best state of the current k-th user; Expressed as the adaptation function of the current state; Indicates the current overall state of the individual; Expressed as the adaptation function of the current best state; The current best performance state; rotation quantum angle; 、 are the probabilities of the current individual taking the ground state and excited state respectively.

[0041] (35) Quantum mutation: In order to prevent the population from falling into the local optimal solution, the quantum NOT gate is introduced to perform quantum mutation operation on the population. The quantum NOT gate is defined as , the mutation rule is: if , Represents the probability of mutation; then a pair of amplitude pairs of individuals are randomly selected And implement the mutation operation, after the mutation, the amplitude pair of the individual is .

[0042] (36) Repeat steps (32) to (35) until the algorithm converges or reaches the maximum number of iterations. .

[0043] like Figure 3 Figure 2 shows the performance gain of using a greedy genetic algorithm to select user antenna operating modes, demonstrating the algorithm's superior performance in UE mode selection. The QGA algorithm consistently outperforms the random assignment method in terms of system performance metrics and is able to approach the performance of the computationally demanding exhaustive method. Specifically, in typical scenarios, the QGA algorithm achieves approximately 95% of the performance of the exhaustive method while reducing the computational complexity from the exponential level of the exhaustive method to the polynomial level, making it suitable for large-scale communication scenarios with a large number of users and antennas.

Claims

1. A joint optimization method for user association and mode selection in a network-assisted free duplex cellular-free wireless access network, characterized in that: The following steps are involved: (1) Establish a network-assisted free duplex architecture system, including a central processing unit (CPU), multiple edge distributed units (EDUs), access points (APs) with fixed uplink and downlink working states, and multi-antenna user equipment (UEs); (2) To maximize the total system spectrum efficiency, establish user QoS constraints, user antenna mode selection and EDU-UE association issues; (3) Determine the association between users and EDUs through matching game algorithms; (4) Use quantum genetic algorithm to dynamically schedule the uplink and downlink working modes of UE antennas.

2. The method for joint optimization of user association and mode selection in a network-assisted free duplex cellular-free wireless access network according to claim 1, characterized in that: In step (1), each EDU is equipped with several uplink APs and downlink APs, and the antenna working mode of the UE can be dynamically scheduled.

3. The method for joint optimization of user association and mode selection in a network-assisted free duplex cellular-free wireless access network according to claim 1, characterized in that: In step (2), the total spectrum efficiency is composed of the superposition of the downlink total spectrum efficiency and the uplink total spectrum efficiency; wherein the downlink total spectrum efficiency includes the useful signal of the user, the interference signals of other downlink users, the interference signal of the uplink antenna to the downlink antenna, and the channel noise; the uplink total spectrum efficiency includes the useful signal of the target user, the interference signals of other uplink users, the interference signal of the downlink antenna to the uplink antenna, and the channel noise.

4. The method for joint optimization of user association and mode selection in a network-assisted free duplex cellular-free wireless access network according to claim 3, characterized in that: In step (2), the problem is modeled as: ； in, Downlink total spectrum efficiency; is the total uplink spectrum efficiency; and is the uplink and downlink working vector of user k; Indicates the association between user k and EDUx; EDUx represents the association coefficient between user and EDU, where 1 indicates association and 0 indicates no association.

5. The method for joint optimization of user association and mode selection in a network-assisted free duplex cellular-free wireless access network according to claim 4, characterized in that: Constraints include: ； ； ； in, is the QoS constraint for downlink user k, is the QoS constraint for uplink user k; ; They represent user k, The nth antenna works in uplink and downlink; D,k represents the downlink spectrum efficiency of user k; r U,k represents the uplink spectrum efficiency of user k.

6. The method for joint optimization of user association and mode selection in a network-assisted free duplex cellular-free wireless access network according to claim 1, characterized in that: In step (3), the matching game algorithm includes the following steps: (31) The user and EDU generate a two-way preference list based on the utility function, where the utility function is the total uplink and downlink spectrum efficiency of the system; (32) The free UE initiates a matching request to the EDU at the top of the preference list, and the EDU accepts or rejects the request based on the associated threshold; (33) The surplus UE initiates a secondary matching request, and the EDU determines whether to associate based on the spectrum efficiency improvement; (34) Perform exchange matching and exchange blocking to associate users; determine whether to exchange based on the utility function, which is the system uplink and downlink speed.

7. The method for joint optimization of user association and mode selection in a network-assisted free duplex cellular-free wireless access network according to claim 1, characterized in that: In step (4), the quantum genetic algorithm includes the following steps: (41) Initialize the quantum population: All amplitude pairs of each individual Initialized as: ； in, Represent the probability of each quantum bit taking the ground state and excited state respectively, and satisfy ; (42) Measuring the population state: Measure each individual in the population once to obtain the individual state; (43) Calculate fitness: Calculate the fitness of each individual, using the system and rate as the fitness function; (44) Quantum revolving gate evolution: generating new populations based on quantum revolving gates; (45) Quantum NOT gate mutation: Introducing quantum NOT gate to perform quantum mutation operation on the population; (46) Repeat steps (42) to (45) until the algorithm converges or the maximum number of iterations is reached.

8. A network-assisted free duplex cellular-free wireless access network user association and mode selection joint optimization system, characterized in that: include: Duplex architecture module: used to establish a network-assisted free duplex architecture system, including a central processing unit (CPU), multiple edge distributed units (EDUs), access points (APs) with fixed uplink and downlink working states, and multi-antenna user equipment (UEs); Matching game algorithm module: used to determine the association between users and EDUs through the matching game algorithm to meet backhaul constraints and user QoS requirements; Quantum genetic algorithm module: used to dynamically schedule the uplink and downlink working modes of UE antennas using quantum genetic algorithm; Joint optimization module: used to jointly optimize EDU-UE association and UE antenna mode selection with the goal of maximizing the system's total spectrum efficiency.

9. An electronic device, characterized in that: The method comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that A computer program is stored, and when the program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images