Heterogeneous network multi-objective optimization interference method, apparatus and terminal device

Abstract

The application is suitable for the field of information communication technology, and provides a heterogeneous network multi-target optimization interference method, device and terminal equipment. The method comprises the following steps: constructing an ideal signal state information model; constructing a non-ideal channel state information model based on the ideal signal state information model and a preset error, wherein the preset error comprises a delay error, an estimation error and a fading error; constructing a first mathematical model with system throughput and system energy efficiency as objective functions based on the non-ideal signal state information model; converting the first mathematical model into a non-cooperative game model of power allocation; and solving the non-cooperative game model of power allocation by using a recurrent neural network algorithm to obtain the optimal game result of system throughput and system energy efficiency when the signal-to-interference-and-noise ratio is maximum. The application can ensure the communication quality while improving the system throughput and system energy efficiency of the communication system under the condition of fast movement of the mobile communication equipment.

Description

Technical Field

[0001] This application belongs to the field of information and communication technology, and in particular relates to a method, apparatus and terminal equipment for multi-target optimized interference in heterogeneous networks. Background Technology

[0002] In recent years, people have increasingly higher demands for mobile data services in terms of communication speed and stability. At the same time, traditional communication networks are struggling to provide fast and stable mobile communication services for mobile terminals in fast-moving environments. With the deployment of 5G communication systems, the wireless communication network architecture will gradually shift from traditional cellular networks to heterogeneous cellular networks, greatly improving network capacity and transmission speed.

[0003] However, with the increase in the density of base stations and communication links, as well as the increasing complexity of network layers, dynamic interference occurs between many different types of base stations, which seriously affects the throughput and energy efficiency of the network system.

[0004] Therefore, how to achieve capacity gain and reduce system power consumption of communication network while ensuring wireless communication quality has become a key problem that mobile terminals inevitably face in communication during rapid movement. Summary of the Invention

[0005] To overcome the problems existing in related technologies, embodiments of this application provide a method, apparatus and terminal equipment for multi-target optimized interference in heterogeneous networks, which can achieve capacity gain and improve system energy efficiency while ensuring wireless communication quality.

[0006] This application is achieved through the following technical solution:

[0007] In a first aspect, embodiments of this application provide a multi-objective optimization interference method for heterogeneous networks, including:

[0008] An ideal signal state information model is constructed. Based on the ideal signal state information model and preset errors, a non-ideal channel state information model is constructed, whereby the preset errors include delay error, estimation error, and fading error inherent in the ideal signal state information. Based on the non-ideal signal state information model, a first mathematical model is constructed with the objective function of optimizing system throughput and system energy efficiency. The first mathematical model is transformed into a non-cooperative game model for power allocation. A recursive neural network algorithm is used to solve the non-cooperative game model for power allocation, obtaining the optimal game result for system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized.

[0009] In one possible implementation of the first aspect, constructing the ideal signal state information model includes: setting the environment of the ideal signal state information model, setting each transmitter in the MIMO finite feedback channel of M target users to correspond to one receiver, and ensuring that the transmitter and receiver frequencies do not interfere with each other, and the number of transmitter antennas is N. t There are N receivers with N antennas. r One, base station transmission power is P trans .

[0010] Calculate the received signal y at the j-th receiver. j The expression is:

[0011]

[0012] In the formula, P j η represents the signal power received by the j-th receiver from the i-th transmitter. ij p represents the path loss of the j-th receiver receiving the signal from the i-th transmitter. i H represents the transmit power allocated by the base station to the i-th transmitter. ji H represents the channel matrix from the i-th transmitter to the j-th receiver. ji The dimension is N r ×N t H ji The elements of H follow a complex Gaussian distribution with mean 0 and variance 1. jj H represents the channel matrix from the j-th transmitter to the j-th receiver. jj The dimension is N r ×N t H jj The elements of V follow a complex Gaussian distribution with mean 0 and variance 1. i V represents the precoding matrix of the i-th transmitter. i The vector dimension is N r ×d i V j V represents the precoding matrix of the j-th transmitter. j The vector dimension is N r ×d i x i Let x represent the transmitted signal of the i-th transmitter. i The vector dimension is d i ×1, data stream is d i x j Let x represent the transmitted signal of the j-th transmitter. j The vector dimension is d j ×1, data stream is d j n jThis represents additive complex white Gaussian noise in the channel, with an average value of 0.

[0013] Calculate the received signal based on interference suppression matrix processing The expression is:

[0014]

[0015] In the formula, V represents the suppression matrix V that interferes with the j-th receiver. j The precoding matrix represents the ideal channel. Let V represent the interference suppression matrix of an ideal channel, where V is the precoding matrix of the ideal channel. j and ideal channel

[0016] Interference suppression matrix The qualified expression is:

[0017]

[0018] The MAXSINR IA algorithm is used to process the received signal based on the interference suppression matrix. The signal-to-interference-plus-noise ratio (SIR) received by the j-th receiver is obtained by the following expression:

[0019]

[0020] In the formula, B j This represents the matrix containing interference and noise.

[0021] The system throughput is obtained based on the signal-to-interference-plus-noise ratio (SIR) received by the j-th receiver, expressed as:

[0022]

[0023] In one possible implementation of the first aspect, a non-ideal signal state information model is constructed, including: constructing a channel matrix based on delay error, estimation error, and fading error, expressed as:

[0024]

[0025] In the formula, E ji E represents the estimation error matrix between the i-th transmitter and the j-th receiver. ji It follows a complex Gaussian distribution with mean 0 and variance 1, where ε represents the error factor, ε∈[0,1], ε=0 indicates no fading error, ε=1 indicates the transmitter did not obtain channel information, H ω Let ρ represent the normalized white Gaussian noise matrix, δ represent the channel matrix correlation coefficient, and ρ = J0(2πf dτ) denotes the channel matrix correlation function, where J0 denotes the zeroth-order Bessel function of the first kind, f d τ represents the maximum Doppler offset, and τ represents the channel matrix delay.

[0026] Based on the channel matrix considering delay error, estimation error, and fading error, the signal-to-interference-plus-noise ratio (SIR) received by the j-th receiver is obtained, expressed as:

[0027]

[0028] In the formula, This represents the interference plus noise matrix of a non-ideal channel.

[0029] The system throughput is obtained based on the signal-to-interference-plus-noise ratio (SIR) received by the j-th receiver, expressed as:

[0030]

[0031] In one possible implementation of the first aspect, the expression for the first mathematical model is:

[0032]

[0033] In the formula, P represents the total power of the system. total P represents the total energy consumption of the system. trans γ represents the total power transmitted by the BS. min p represents the minimum signal-to-interference-plus-noise ratio (SIR) required for normal communication by the target user. j p represents the power allocated to target user j. c This indicates the fixed losses of the receiver's components.

[0034] In one possible implementation of the first aspect, the basic elements of the non-cooperative game model for power allocation include players, a policy space, and a utility function. Players are M target users within a pre-defined time period of the system's data stream, assuming d... j Let D be the matrix of k independent data streams sent by the j-th target user within a preset time period. The policy space represents the available power allocation schemes for the M target users in a non-cooperative game model of power allocation, assuming p... j Let be the policy vector for the j-th target user, then Let Π be the power of the k-th independent data stream sent to the j-th target user, where the system's decision matrix is ​​represented by Π. The utility function, taking system throughput as the payoff and system energy consumption as the cost, is the utility function of the j-th target user, which is the payoff gained by target user j to improve system performance minus the system cost. Its expression is:

[0035]

[0036] In the formula, f j (·) = 1 indicates that the j-th target user receives the service from the base station, f j (·) = 0 indicates that the j-th target user does not receive the base station's service, α is the signal-to-interference-plus-noise ratio weight factor, α > 0, β is the system energy consumption weight factor, β > 0, λ is the weight factor of the target user j's input function, λ > 0, where the j-th target user's policy vector p j It is a negative term.

[0037] In one possible implementation of the first aspect, after transforming the first mathematical model into a non-cooperative game model of power allocation, the expression for the non-cooperative game model of power allocation is:

[0038]

[0039] In one possible implementation of the first aspect, a recurrent neural network algorithm is used to solve the non-cooperative game model of power allocation to obtain the optimal game result of system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized, including:

[0040] Step 1: Set up the initial environment for the recurrent neural network algorithm:

[0041] Let i be the i-th receiver of the target user, M be the total number of receivers for all target users, and N be the maximum number of iterations.

[0042] Initialize the precoding matrix as follows The initial interference suppression matrix is ​​as follows: The initial interference noise matrix is ​​as follows Initialize the signal power matrix received by the i-th receiver as follows: Each receiver receiving the initial power from the base station is used as an initial recurrent neuron;

[0043] Define system utility function F (0) =0, calculate each target user u i Effect function;

[0044] Step 2: Starting from the nth iteration, the signal power matrix for the current iteration is: Will be Substituting into the signal-to-interference-plus-noise ratio (SINR) calculation formula, the interference-noise matrix is ​​obtained as follows: The formula for calculating the signal-to-interference-plus-noise ratio is:

[0045]

[0046] Step 3: The interference noise matrix obtained in Step 2 is as follows: Calculate the interference suppression matrix

[0047] Step 4: In the reciprocity channel, calculate the corresponding matrix interference noise matrix based on the signal-to-interference-plus-noise ratio (SINR) calculation formula.

[0048] Step 5: Employ the maximum signal-to-interference-plus-noise ratio (SINR) interference alignment algorithm, combined with the matrix interference noise matrix. The corresponding interference suppression matrix is ​​calculated as follows:

[0049] Step 6, Order

[0050] Step 7: Calculate the signal power matrix for the next iteration using a power allocation algorithm.

[0051] Step 8: End the nth iteration, and repeat steps 2 to 7 until the maximum number of iterations N or the power signal matrix P is reached. i convergence.

[0052] Secondly, embodiments of this application provide a heterogeneous network multi-target optimized interference device, comprising:

[0053] The channel model construction module is used to construct an ideal signal state information model, and also to construct a non-ideal channel state information model based on the ideal signal state information model and preset errors, wherein the preset errors include delay error, estimation error, and fading error. The mathematical model construction module is used to construct a first mathematical model based on the non-ideal signal state information model, with the objective function of optimizing system throughput and system energy efficiency, and also to transform the first mathematical model into a non-cooperative game model for power allocation. The model solving module is used to solve the non-cooperative game model for power allocation using a recurrent neural network algorithm to obtain the optimal game result for system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized.

[0054] Thirdly, embodiments of this application provide a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the heterogeneous network multi-target optimization interference method as described in any of the first aspects.

[0055] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the heterogeneous network multi-target optimization interference method as described in any of the first aspects.

[0056] Fifthly, embodiments of this application provide a computer program product that, when run on a terminal device, causes the terminal device to execute the heterogeneous network multi-target optimization interference method described in any of the first aspects above.

[0057] It is understood that the beneficial effects of the second to fifth aspects mentioned above can be found in the relevant descriptions in the first aspect mentioned above, and will not be repeated here.

[0058] The beneficial effects of the embodiments in this application compared with the prior art are:

[0059] In this embodiment, a non-ideal signal state information model is constructed based on an ideal signal state information model and a preset error. A first mathematical model is established with system throughput and system energy efficiency as objective functions. The first mathematical model is then transformed into a non-cooperative game model. Finally, by solving the non-cooperative game model, the optimal game result for system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized is obtained. This ensures the communication quality of mobile communication devices in a fast-moving state, achieves a gain in communication system throughput, and improves system energy efficiency.

[0060] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this specification. Attached Figure Description

[0061] To more clearly illustrate the technical solutions in the embodiments of this application, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0062] Figure 1 This is a flowchart illustrating a multi-objective optimization interference method for heterogeneous networks provided in an embodiment of this application;

[0063] Figure 2 This is a schematic diagram illustrating an application scenario of the multi-objective optimization interference method for heterogeneous networks provided in an embodiment of this application;

[0064] Figure 3 This is a schematic diagram of an application scenario of a high-speed rail mobile communication network architecture based on a MATLAB simulation platform, provided in one embodiment of this application.

[0065] Figure 4(a) is a comparison of the system throughput of different algorithms at different train speeds provided in an embodiment of this application;

[0066] Figure 4(b) is a comparison of the system energy efficiency of different algorithms at different train speeds provided in an embodiment of this application;

[0067] Figure 4(c) is a comparison of the system bit error rates of different algorithms at different train speeds provided in an embodiment of this application;

[0068] Figure 5(a) is a comparison of the system throughput of different algorithms under different numbers of trackside devices provided in an embodiment of this application;

[0069] Figure 5(b) is a comparison of the system energy efficiency of different algorithms under different numbers of trackside devices provided in an embodiment of this application;

[0070] Figure 5(c) is a comparison of the system bit error rates of different algorithms under different numbers of trackside devices provided in an embodiment of this application;

[0071] Figure 6(a) is a comparison of the system throughput of different algorithms under different numbers of train users according to an embodiment of this application;

[0072] Figure 6(b) is a comparison of the system energy efficiency of different algorithms under different numbers of train users according to an embodiment of this application;

[0073] Figure 6(c) is a comparison of the system bit error rates of different algorithms under different numbers of train users provided in an embodiment of this application;

[0074] Figure 7(a) shows the convergence of the utility function provided in an embodiment of this application at different speeds;

[0075] Figure 7(b) shows the convergence of the utility function provided in an embodiment of this application under different numbers of orbital square devices;

[0076] Figure 7(c) shows the convergence of the utility function provided in an embodiment of this application under different numbers of train users;

[0077] Figure 8 This is a comparison of the different algorithm complexities provided in the embodiments of this application;

[0078] Figure 9 This is a schematic diagram of the structure of the heterogeneous network multi-target optimization interference device provided in the embodiments of this application;

[0079] Figure 10 This is a schematic diagram of the structure of the terminal device provided in the embodiments of this application. Detailed Implementation

[0080] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0081] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0082] It should also be understood that the term “and / or” as used in this application specification and the appended claims means any combination of one or more of the associated listed items and all possible combinations, and includes such combinations.

[0083] As used in this application specification and the appended claims, the term "if" may be interpreted, depending on the context, as "when," "once," "in response to determination," or "in response to detection." Similarly, the phrase "if determined" or "if detected [the described condition or event]" may be interpreted, depending on the context, as meaning "once determined," "in response to determination," "once detected [the described condition or event]," or "in response to detection [the described condition or event]."

[0084] Furthermore, in the description of this application and the appended claims, the terms "first," "second," "third," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0085] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0086] With the deployment of fifth-generation (5G) communication systems, the wireless communication network architecture will gradually shift from traditional cellular networks to heterogeneous cellular networks, greatly improving network capacity and transmission speed. However, with the increase in base station and communication link density and the increasing complexity of network layers, dynamic interference occurs between many different types of base stations, severely affecting the system's network throughput.

[0087] Interference Alignment (IA) technology amplifies the spatial size of the desired signal by compressing the size of the interference space. It can quickly separate the interference signal from the desired signal at the receiver, achieve effective interference management, improve the channel capacity of the wireless network and the reliability of the communication system, and thus meet the quality of service requirements of wireless mobile users.

[0088] With the rapid development of my country's transportation network and the increasing speed of transportation, people's daily travel has become more convenient and frequent, and the number of terminal devices used to receive communication has also increased. As a result, traditional IA technology is easily affected by network topology and channel status information. Changes in key factors such as user location and number of terminals significantly reduce the effectiveness of traditional IA solutions, making it impossible for mobile communication devices to guarantee stable communication quality in fast-moving environments.

[0089] Based on the above problems, and taking into full account system throughput, user communication transmission reliability and base station energy efficiency, this application provides a multi-target optimized interference method for heterogeneous networks.

[0090] Figure 1 This is a schematic flowchart of a heterogeneous network multi-objective optimization interference method provided in an embodiment of this application, with reference to... Figure 1 The method is described in detail below:

[0091] In step 101, an ideal signal state information model is constructed.

[0092] For ease of explanation, this application uses only the 5G Ultra-Dense Network (UDN) providing communication for high-speed rail as an example model, and a mobile phone receiving 5G communication signals as an example terminal device, and abbreviates Channel State Information as CSI.

[0093] Step A1: Set up the ideal CSI model environment

[0094] For example, embodiments of this application can be applied to, for example, Figure 2 In the exemplary scenario shown, the 5G communication system model is exemplarily divided into three layers: the control panel layer, the physical layer, and the user panel layer. The control panel layer can be an LTE-R base station, the physical layer can include communication equipment, trackside equipment, and the railway environment, etc., and the user panel layer can include WLAN, LTE base stations, 5G networks, etc.

[0095] For example, consider M target users' mobile phones using MIMO finite feedback channels, with each transmitter corresponding to one receiver, and neglecting adjacent frequency interference. Also, assume the number of transmitter antennas is N. tThere are N receivers with N antennas. r One, base station transmission power is P trans .

[0096] Step A2: Calculate the received signal y of the j-th receiver. j

[0097] For example, y j The expression is:

[0098]

[0099] In the formula, P j η represents the signal power received by the j-th receiver from the i-th transmitter. ij H represents the path loss of the j-th receiver receiving the signal from the i-th transmitter, and pi represents the transmit power allocated by the base station to the i-th transmitter. ji H represents the channel matrix from the i-th transmitter to the j-th receiver. ji The dimension is N r ×N t H ji The elements of H follow a complex Gaussian distribution with mean 0 and variance 1; jj H represents the channel matrix from the j-th transmitter to the j-th receiver. jj The dimension is N r ×N t H jj The elements of V follow a complex Gaussian distribution with mean 0 and variance 1; i V represents the precoding matrix of the i-th transmitter. i The vector dimension is N r ×d i V j V represents the precoding matrix of the j-th transmitter. j The vector dimension is N r ×d i ;x i Let x represent the transmitted signal of the i-th transmitter. i The vector dimension is d i ×1, data stream is d i x j Let x represent the transmitted signal of the j-th transmitter. j The vector dimension is d j ×1, data stream is d j n j This represents additive complex white Gaussian noise in the channel, with an average value of 0.

[0100] Step A3: Calculate the received signal based on interference suppression matrix processing.

[0101] For example, The expression is:

[0102]

[0103] In the formula, V represents the interference suppression matrix of the j-th receiver in an ideal channel. j Let represent the precoding matrix of the j-th receiver in an ideal channel.

[0104] For example, to achieve spatial alignment of interference signals, the precoding matrix V in an ideal channel needs to be adjusted. j and the interference suppression matrix in an ideal channel To impose a constraint, the constraint expression is:

[0105]

[0106] Step A4: Obtain the signal-to-interference-plus-noise ratio (SINR) of the signal received by the j-th receiver.

[0107] For example, the MAXSINR IA algorithm is used to process the received signal based on the interference suppression matrix processing. The SINR of the received signal from the j-th receiver is obtained. j The expression is:

[0108]

[0109] In the formula, B j This represents the matrix containing interference and noise.

[0110] Step A5: SINR obtained in step A4 j An example of obtaining system throughput is expressed as follows:

[0111]

[0112] In step 102, a non-ideal channel state information model is constructed based on the ideal signal state information model and the preset error.

[0113] In communication systems requiring channel state information feedback, due to limitations in system hardware performance and the propagation time of CSI feedback, the transmitter's CSI typically has a certain delay and estimation error. Simultaneously, signal fading occurs during transmission. Therefore, for example, this application sets the preset error to include delay error, estimation error, and fading error.

[0114] Step B1: Construct a channel matrix based on delay error, estimation error, and fading error.

[0115] For example, channel matrix The expression is:

[0116]

[0117] In the formula, E ji E represents the estimation error matrix between the i-th transmitter and the j-th receiver. ji It follows a complex Gaussian distribution with mean 0 and variance 1; ε represents the error factor, ε∈[0,1], where ε=0 indicates no fading error, and ε=1 indicates that the transmitter has not obtained channel information; H ω Let ρ represent the normalized white Gaussian noise matrix, δ represent the channel matrix correlation coefficient, and ρ = J0(2πf d τ) denotes the channel matrix correlation function, where J0 denotes the zeroth-order Bessel function of the first kind, f d H represents the maximum Doppler offset, τ represents the channel matrix delay; ji H represents the channel matrix from the i-th transmitter to the j-th receiver. ji The dimension is N r ×N t H ji The elements follow a complex Gaussian distribution with a mean of 0 and a variance of 1.

[0118] Step B2: Obtain the signal SINR′ received by the j-th receiver. j

[0119] For example, SINR′ j The expression is:

[0120]

[0121] In the formula, This represents the interference plus noise matrix of a non-ideal channel. This represents the suppression matrix for non-ideal channel interference at the j-th receiver. This represents the precoding matrix for a non-ideal channel.

[0122] Step B3: Based on the signal SINR′ received by the j-th receiver in step B2. j Calculation system throughput

[0123] For example, The expression is:

[0124]

[0125] In step 103, based on the non-ideal signal state information model, a first mathematical model is constructed with the objective function of optimizing system throughput and system energy efficiency.

[0126] In some embodiments, the total system power and the received SINR are used as constraints, and the system throughput and system energy consumption are used as objective functions to construct a first mathematical model. The expression is:

[0127]

[0128] In the formula, P represents the total power of the system. total P represents the total energy consumption of the system. trans γ represents the total power transmitted by the BS. min p represents the minimum signal-to-interference-plus-noise ratio (SIR) required for normal communication by the target user. j pc represents the power allocated to target user j, and pc represents the fixed losses of the receiver components.

[0129] In step 104, the first mathematical model is transformed into a non-cooperative game model for power allocation.

[0130] In some embodiments, the first mathematical model can be transformed into a non-cooperative game model for power allocation.

[0131] For example, a non-cooperative game model has three basic elements: players, policy space, and utility function.

[0132] In the non-cooperative game model of power allocation, the players are M mobile phones and the data streams within a preset time period of the system. Assume d... j Let D be the matrix of k independent data streams sent by the j-th mobile phone within a preset time period.

[0133] In the non-cooperative game model of power allocation, the strategy space consists of M mobile phones with selectable power allocation schemes. Assume p... j Let be the policy vector for the j-th target user, then The power of the k-th independent data stream sent to the j-th target user is represented by the system's decision matrix Π.

[0134] In the non-cooperative game model of power allocation, the utility function takes system throughput as the payoff of the game process and system energy consumption as the cost. Therefore, the utility function of the j-th phone is the payoff gained by the j-th phone to improve system performance minus the system cost, expressed as:

[0135]

[0136] In the formula, f j(·) = 1 indicates that the j-th target user receives the service from the base station, f j (·)=0 indicates that the j-th target user does not receive the base station's service, α is the weight factor of the signal-to-interference-plus-noise ratio, α>0, β is the weight factor of the system energy consumption, β>0, and λ is the weight factor of the target user j's input function, λ>0.

[0137] It should be noted that p j It is added as a negative term in the utility function to encourage users to reduce transmission power to obtain greater utility, thereby increasing throughput, reducing energy consumption, and improving energy efficiency.

[0138] Therefore, the expression for the non-cooperative game model of power allocation based on the transformation of the first mathematical model can be:

[0139]

[0140] In step 105, a recurrent neural network algorithm is used to solve the non-cooperative game model of power allocation to obtain the optimal game result of system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized.

[0141] For example, this application uses a recurrent neural network (RNN) algorithm to solve the non-cooperative game model of power allocation. The specific steps are as follows:

[0142] Step 1: Set up the initial environment for the RNN, including:

[0143] Let i be the i-th mobile phone of the target user, M be the total number of mobile phones of all target users, and N be the maximum number of iterations;

[0144] Initialize the precoding matrix as follows The initial interference suppression matrix is ​​as follows: The initial interference noise matrix is ​​as follows Initialize the signal power matrix received by the i-th mobile phone as follows: Each receiver receiving the initial power from the base station is used as an initial recurrent neuron;

[0145] Define system utility function F (0) =0, calculate the effect function u for each target user's mobile phone. i .

[0146] Step 2: Starting from the nth iteration, the signal power matrix for the current iteration is: Will be Substituting into the SINR calculation formula, the interference noise matrix is ​​obtained as follows: The formula for calculating SINR is:

[0147]

[0148] Step 3: The interference noise matrix obtained in Step 2 is as follows: Calculate the interference suppression matrix

[0149] Step 4: In the reciprocity channel, calculate the corresponding interference noise matrix based on the SINR calculation formula.

[0150] Step 5: Employ the maximum SINR interference alignment algorithm, combined with the interference noise matrix as follows: The corresponding interference suppression matrix is ​​calculated as follows:

[0151] Step 6, Order

[0152] Step 7: Calculate the signal power matrix for the next iteration using a power allocation algorithm.

[0153] Step 8: End the nth iteration, and repeat steps 2 to 7 until the maximum number of iterations N is reached or the power signal matrix P is reached. i convergence.

[0154] The calculated result is the optimal trade-off between network system throughput and system energy efficiency when SINR is maximized.

[0155] To improve the energy efficiency of the network system, this step adopts an interference alignment algorithm based on the maximum signal-to-interference-plus-noise ratio (SINR). By iteratively selecting a precoding matrix and a mobile phone suppression matrix that satisfy the constraints, the SINR of the mobile phone is maximized, thereby optimizing the maximum SINR of the mobile phone and ensuring the communication quality of the mobile phone.

[0156] This application constructs a non-ideal CSI model based on an ideal CSI model and a preset error. It establishes a first mathematical model with system throughput and system energy efficiency as objective functions. The first mathematical model is then transformed into a non-cooperative game model. Finally, by solving the non-cooperative game model, the optimal game result of system throughput and system energy efficiency when SINR is maximized is obtained. This ensures the communication quality of mobile communication devices in a fast-moving state, achieves the gain of communication system throughput, and improves system energy efficiency.

[0157] To verify the feasibility and beneficial effects of the proposed solution, a comparative analysis was conducted using the MATLAB simulation platform and traditional interference optimization algorithms. The specific analysis process is as follows:

[0158] Simulation platform parameter settings

[0159] A high-speed rail mobile communication network architecture based on the MATLAB simulation platform was constructed, such as... Figure 3 As shown. In the high-speed rail environment, single-hop communication between train users and base stations is considered, and the train will quickly pass through the heterogeneous network covered by these base stations.

[0160] In this architecture, a 500-meter-long high-speed rail line is deployed, with seven base stations, including 5G and LTE-R base stations, located on both sides of the track. The train is 25 meters long, and 20 users are randomly distributed in each cell. Assuming the communication services of the train users follow a Poisson distribution, and referring to the standard services defined by 3GPP on the LTE network, the service type is randomly selected. The train travels along the track at a constant speed. Furthermore, a certain number of trackside devices are evenly distributed on both sides of the track, generating communication services as the train passes through, with the service type being the same as that of ordinary users. In the proposed simulation scenario, random users are served by the nearest base station, and train users are also served by the nearest base station; however, as the train moves, the serving base station for the train users changes accordingly.

[0161] The experiment comprehensively considered the communication performance of three types of users—randomly distributed users, trackside devices, and train users—to reflect the impact of the proposed interference alignment algorithm on the overall network characteristics of the system. System throughput and energy efficiency were used to represent network effectiveness, and bit error rate was used to represent network reliability. The experiment was conducted in three scenarios. In the first scenario, the number of trackside devices and train users was set to fixed values, and the impact of train speed on the performance of the two algorithms was compared. In the second scenario, train speed and the number of train users were set to fixed values, and the impact of the number of trackside devices on the performance of the two algorithms was analyzed. In the third scenario, train speed and the number of trackside devices were set to fixed values, and the impact of the number of train users on the performance of the two algorithms was analyzed. Furthermore, the convergence speed of the algorithm was tested in the three scenarios, thus proving the existence of Nash equilibrium points in the game model and the effectiveness of the algorithm.

[0162] In the experiment, the system channel parameter values ​​were represented by the average values ​​obtained from 60 runs of the system executing each experimental algorithm. The channel parameters were set as follows: ρ = 0.9966, ε = 0.4, β = 1, λ = 0.1, p c =1%P t Other parameters involved in the simulation are shown in Table 1.

[0163] Table 1 Simulation Parameters

[0164] parameter value Carrier frequency / GHz 5.25 Bands / MHz 20 Number of antennas 8 Number of receiving antennas 4 Channel Model WINNER II eNb power / dBm 30 Path loss Friis propagation loss model recession model Friis Spectral Propagation Loss Model Antenna height of ENb / m 16 Antenna gain of ENb / dBi 12 UE antenna height / m 1.5 UE antenna gain / dBi 0 Simulation cycle / s 2

[0165] Analysis of the impact of train speed on algorithm performance

[0166] Two trackside devices were set up on the network platform to study the impact on system performance at four different operating speeds: 50 km / h, 100 km / h, 200 km / h, and 300 km / h. The results are shown in Figure 4.

[0167] As shown in Figure 4, the system throughput shows a decreasing trend with increasing train speed. According to Figures 4(a) and 4(b), the system throughput and energy efficiency increase with the increase of downlink SINR, and tend to stabilize when SINR ≥ 15dB. However, with the increase of train speed, the stability and imperfections of the channel will have a significant impact on the reception quality, and the number of cell users will surge in a short period of time, reducing the overall system throughput and energy efficiency. As shown in Figure 4(c), the system bit error rate gradually increases with the increase of speed. The bit error rate is relatively high when the signal-to-noise ratio is below 5dB, and the bit error rate decreases rapidly below 5dB. Under high-speed movement, due to the large radial velocity between the base station and the user, there is a large Doppler frequency shift in the user's received signal, which increases the bit error rate. In addition, when the channel transmission quality is poor, i.e., SNR ≤ 5dB, the bit error rate performance will deteriorate significantly with the increase of train speed, and tend to stabilize when SNR ≥ 15dB.

[0168] The following describes the effect of the algorithm proposed in this application based on the algorithms in references [1] and [2]. For reference [1], please refer to: XU Z, HOFER M, ZEMEN TA time-variant channel prediction and feedback framework for interference alignment[J]. IEEE Transactions on Vehicular Technology, 2020, 66(07): 5961–5973. For reference [2], please refer to: ZHAO F, WANG W, CHEN HB, et al. Interference alignment and game-theoretic power allocation in MIMO heterogeneous sensor networks communications[J]. Signal Processing, 2019, 126(09): 1375-1383.

[0169] Compared with the algorithms in references [1] and [2], the algorithm proposed in this application significantly improves the system's throughput, energy efficiency, and bit error rate. Figure 4(a) and 4(b)As can be seen, with the increase of SINR, the proposed algorithm has a significant advantage in improving throughput and energy efficiency. This is mainly because the proposed algorithm designs a bi-objective game model for power allocation, which can ensure the maximum system throughput and adopt a compromise method to reduce system energy consumption. In addition, under the above-mentioned non-ideal channel state information conditions, the proposed algorithm is significantly better than the algorithms in references [1] and [2] in terms of throughput and energy efficiency in low SINR environment. As shown in Figure 4(c), the system bit error rate performance is shown. It can be seen from the figure that regardless of the speed, the proposed algorithm obtains a lower system bit error rate than the algorithms in references [1] and [2]. This is mainly because the proposed algorithm reasonably allocates the transmission power of the data stream, thereby greatly improving the communication reliability of the system, and this is more significant when SINR ≥ 10dB. When SINR is 30dB and the train speed is 300km / h, the throughput increases by 48.95%, the energy efficiency increases by 53.02%, and the bit error rate decreases by 14.37%. Therefore, it can be seen that using a bi-objective optimization model to improve the power allocation scheme can effectively and reasonably solve the interference management problem of high-speed rail mobile communication systems.

[0170] Impact of the number of trackside equipment

[0171] In practical high-speed rail communication network scenarios, train operation control systems achieve train information exchange and moving block functions through vehicle-to-ground wireless communication. This vehicle-to-ground communication requires collaborative communication between onboard communication equipment and trackside equipment. Therefore, the deployment of trackside equipment will have a certain impact on the performance of the railway communication network. To further verify the adaptability of the proposed algorithm in the railway environment, a train speed of 200 km / h and trackside equipment (N) were tested. TD The scenarios with quantities of 5, 7, 9, 11, 13, and 15 were evaluated.

[0172] Experiments were conducted to evaluate the effectiveness and reliability of the proposed algorithm and the algorithms in references [1] and [2]. The results are shown in Figure 5. Figure 5 shows that the number of different trackside devices has a significant impact on system performance. With N... TD With the increase of N, the system throughput and energy efficiency increase, but the average bit error rate also increases. This is mainly because during the experimental time, N... TD The increase in spectrum will lead to an increase in the number of communication services in the system. With sufficient spectrum resources, this improves resource utilization, thereby changing the system's throughput and energy efficiency. However, the number of interference signals received by users during signal transmission also increases, affecting the SINR of each received signal and thus reducing the receiver's bit error rate performance.

[0173] By comparing the performance curves of the three algorithms, it can be seen that under four different N... TDCompared with the algorithms in references [1] and [2], the proposed algorithm significantly improves system performance. This is because the proposed algorithm proposes a dual-objective optimization model for power allocation throughput and energy efficiency, and uses an RNN network to solve the model. This algorithm provides a more reasonable power allocation scheme for communication terminals, including trackside equipment, to ensure that the maximum SINR interference alignment process can be matched, thereby obtaining a more optimized precoding matrix and interference reception suppression matrix. When SINR is 30dB, N TD When the value is 5, the proposed algorithm achieves a throughput of 88.03%, an energy efficiency of 71.58%, and a bit error rate of 30.84%. Experiments show that, compared with the algorithms in references [1] and [2], the proposed algorithm has outstanding anti-interference performance and robust performance in the high-speed rail mobile communication network environment.

[0174] Impact of the number of train users

[0175] For different numbers of train users (N) TU Experimental analysis was conducted. When the number of trackside devices was 2, the train speed was 200 km / h, and N... TU The values ​​are 5, 10, 15, 20, 25, and 30, respectively. Figure 6 shows the impact of the two algorithms on system performance. As can be seen from the figure, the increase in interference signal sources will affect the received SINR of each channel signal, thereby reducing the system's bit error rate performance. When N... TU When the frequency is greater than 15, the available spectrum resources of the system are limited, and the increase in the number of users will gradually reduce the performance of the system. Therefore, the selection and scheduling of users are very important. In addition, the performance comparison of the two algorithms shows that the proposed algorithm has a significant performance advantage over the algorithms in references [1] and [2].

[0176] Algorithm convergence analysis

[0177] To verify the optimization efficiency of the game theory model, the convergence of the proposed algorithm was evaluated using a utility function, and the results are shown in Figure 7. Figure 7(a) shows the convergence curves of the utility function when the number of trackside devices is a fixed value of 2, and the train speeds are 50 km / h, 100 km / h, 150 km / h, and 200 km / h. As can be seen from the figure, the train speed has a relatively small impact on the convergence of the algorithm, and the utility function tends to converge after 30 iterations. Figure 7(b) shows the convergence curves of the utility function when the train speed is 200 km / h, and the number of trackside devices is 5, 10, 15, and 20. According to Figure 7(b), the algorithm converges quickly, and the utility function value tends to stabilize after 10 iterations. Figure 7(c) shows the convergence curves of the utility function when the train speed is 200 km / h, and the number of train users is 10, 15, 20, and 25. As shown in the figure, the utility function value is basically stable after 35 iterations.

[0178] Therefore, the convergence of the utility function indicates the existence of a Nash equilibrium in the proposed game model, and the optimization process consumes relatively little time. Furthermore, the convergence speed of the algorithm is minimally affected by train speed, the number of communication devices in the environment, and the number of concurrent services. This demonstrates that the proposed algorithm possesses certain adaptability and feasibility in high-speed rail wireless communication networks.

[0179] Algorithm Complexity Analysis

[0180] Algorithm complexity is another key metric for evaluating algorithm performance. In terms of complexity, this paper employs the penalty function method for optimization.

[20] Because of V i It is N r ×d i A 3D matrix, where each column represents a precoding vector, and the number of data streams transmitted is d1 = d2 = ... d... i =d, therefore the transmitter needs to optimize Md precoding vectors, while the receiver needs to calculate M interference suppression matrices using the zero-forcing algorithm. Therefore, the system complexity F is as follows:

[0181] F = O(Mtr 3.5 )+O(rt 3 )

[0182] like Figure 8 The figure shows the relationship between the number of floating-point operands and the number of iterations for the algorithms in reference [1], reference [2], and the algorithm proposed in this paper. As can be seen from the figure, the algorithm in reference [1] has the highest complexity, while the algorithm proposed in this paper has the lowest complexity, which further indicates that the algorithm proposed in this paper is faster in terms of running speed.

[0183] in conclusion

[0184] To address the communication interference problem in high-speed rail wireless networks, a multi-objective optimization interference algorithm based on game theory is proposed, and the following conclusions are drawn through simulation experiments:

[0185] (1) When the signal-to-interference-plus-noise ratio (SINR) is 30dB and the train speed is 300km / h, the proposed algorithm increases the throughput by 48.95%, increases the energy efficiency by 53.02%, and reduces the bit error rate by 14.37%, which can effectively solve the interference management problem of the communication system.

[0186] (2) When the signal-to-interference-plus-noise ratio (SINR) is 30 dB and the trackside equipment N TD When the value is 5, the proposed algorithm achieves a throughput of 88.03% and an energy efficiency of 71.58%, demonstrating good anti-interference and robustness.

[0187] (3) Compared with the algorithms in references [1] and [2], the proposed algorithm has significant advantages in terms of system throughput, energy efficiency and bit error rate. Moreover, the convergence speed is less affected by train speed, communication equipment and number of concurrent services, and has good environmental adaptability.

[0188] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0189] Corresponding to the heterogeneous network multi-objective optimization interference method described in the above embodiments, Figure 9 The diagram shows a structural block diagram of a heterogeneous network multi-target optimized interference device provided in an embodiment of this application. For ease of explanation, only the parts related to the embodiments of this application are shown.

[0190] See Figure 9 The heterogeneous network multi-target optimization interference device in this application embodiment may include: a channel model construction module 201, a mathematical model construction module 202, and a model solving module 203.

[0191] Channel model construction module 201 is used to construct an ideal signal state information model.

[0192] Optionally, constructing an ideal signal state information model includes: setting the environment of the ideal signal state information model, setting each transmitter in the MIMO finite feedback channel of M target users to correspond to one receiver, and ensuring that the transmitter and receiver frequencies do not interfere with each other, and the number of transmitter antennas is N. t There are N receivers with N antennas. r One, base station transmission power is P trans .

[0193] Calculate the received signal y at the j-th receiver. j The expression is:

[0194]

[0195] In the formula, P j η represents the signal power received by the j-th receiver from the i-th transmitter. ij p represents the path loss of the j-th receiver receiving the signal from the i-th transmitter. i H represents the transmit power allocated by the base station to the i-th transmitter. ji H represents the channel matrix from the i-th transmitter to the j-th receiver. ji The dimension is N r ×N t H jiThe elements of H follow a complex Gaussian distribution with mean 0 and variance 1. jj H represents the channel matrix from the j-th transmitter to the j-th receiver. jj The dimension is N r ×N t H jj The elements of V follow a complex Gaussian distribution with mean 0 and variance 1. i V represents the precoding matrix of the i-th transmitter. i The vector dimension is N r ×d i V j V represents the precoding matrix of the j-th transmitter. j The vector dimension is N r ×d i x i Let x represent the transmitted signal of the i-th transmitter. i The vector dimension is d i ×1, data stream is d i x j Let x represent the transmitted signal of the j-th transmitter. j The vector dimension is d j ×1, data stream is d j n j This represents additive complex white Gaussian noise in the channel, with an average value of 0.

[0196] Calculate the received signal based on interference suppression matrix processing The expression is:

[0197]

[0198] In the formula, V represents the suppression matrix V that interferes with the j-th receiver. j The precoding matrix represents the ideal channel. Let V represent the interference suppression matrix of an ideal channel, where V is the precoding matrix of the ideal channel. j Interference suppression matrix of ideal channel The qualified expression is:

[0199]

[0200] The MAXSINR IA algorithm is used to process the received signal based on the interference suppression matrix. The signal-to-interference-plus-noise ratio (SIR) received by the j-th receiver is obtained by the following expression:

[0201]

[0202] In the formula, B j This represents the matrix containing interference and noise.

[0203] The system throughput is obtained based on the signal-to-interference-plus-noise ratio (SIR) received by the j-th receiver, expressed as:

[0204]

[0205] The channel model construction module 201 is also used to construct a non-ideal channel state information model based on the ideal signal state information model and the preset error, wherein the preset error includes delay error, estimation error and fading error.

[0206] Optionally, constructing a non-ideal signal state information model includes: constructing a channel matrix based on delay error, estimation error, and fading error, with the expression:

[0207]

[0208] In the formula, E ji E represents the estimation error matrix between the i-th transmitter and the j-th receiver. ji It follows a complex Gaussian distribution with mean 0 and variance 1, where ε represents the error factor, ε∈[0,1], ε=0 indicates no fading error, ε=1 indicates the transmitter did not obtain channel information, H ω Let ρ represent the normalized white Gaussian noise matrix, δ represent the channel matrix correlation coefficient, and ρ = J0(2πf d τ) denotes the channel matrix correlation function, where J0 denotes the zeroth-order Bessel function of the first kind, f d τ represents the maximum Doppler offset, and τ represents the channel matrix delay.

[0209] Based on the channel matrix considering delay error, estimation error, and fading error, the signal-to-interference-plus-noise ratio (SIR) received by the j-th receiver is obtained, expressed as:

[0210]

[0211] In the formula, This represents the interference plus noise matrix of a non-ideal channel.

[0212] The system throughput is obtained based on the signal-to-interference-plus-noise ratio (SIR) received by the j-th receiver, expressed as:

[0213]

[0214] The mathematical model construction module 202 is used to construct a first mathematical model based on the non-ideal signal state information model, with the objective function being to optimize system throughput and system energy efficiency.

[0215] Optionally, the expression for the first mathematical model is:

[0216]

[0217] In the formula, P represents the total power of the system. total P represents the total energy consumption of the system. trans γ represents the total power transmitted by the BS. min p represents the minimum signal-to-interference-plus-noise ratio (SIR) required for normal communication by the target user. j pc represents the power allocated to target user j, and pc represents the fixed losses of the receiver components.

[0218] The mathematical model building module 202 is also used to transform the first mathematical model into a non-cooperative game model for power allocation.

[0219] Optionally, the basic elements of a non-cooperative game model for power allocation include players, policy space, and utility function. Players are M target users within a preset time period of the system's data stream, assuming d... j Let D be the matrix of k independent data streams sent by the j-th target user within a preset time period. The policy space represents the available power allocation schemes for the M target users in a non-cooperative game model of power allocation, assuming p... j Let be the policy vector for the j-th target user, then Let Π be the power of the k-th independent data stream sent to the j-th target user, where the system's decision matrix is ​​represented by Π. The utility function, taking system throughput as the payoff and system energy consumption as the cost, is the utility function of the j-th target user, which is the payoff gained by target user j to improve system performance minus the system cost. Its expression is:

[0220]

[0221] In the formula, f j (·) = 1 indicates that the j-th target user receives the service from the base station, f j (·) = 0 indicates that the j-th target user does not receive the base station's service, α is the signal-to-interference-plus-noise ratio weight factor, α > 0, β is the system energy consumption weight factor, β > 0, λ is the weight factor of the target user j's input function, λ > 0, where the j-th target user's policy vector p j It is a negative term.

[0222] Furthermore, after transforming the first mathematical model into a non-cooperative game model of power allocation, the expression for the non-cooperative game model of power allocation is:

[0223]

[0224] The model solving module 203 is used to solve the non-cooperative game model of the power allocation using a recurrent neural network algorithm to obtain the optimal game result of system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized.

[0225] Optionally, the steps for solving the non-cooperative game model of power allocation using a recurrent neural network algorithm include: Step 1, setting up the initial environment for the recurrent neural network algorithm:

[0226] Let i be the i-th receiver of the target user, M be the total number of receivers for all target users, and N be the maximum number of iterations.

[0227] Initialize the precoding matrix as follows The initial interference suppression matrix is ​​as follows: The initial interference noise matrix is ​​as follows Initialize the signal power matrix received by the i-th receiver as follows: Each receiver receiving the initial power from the base station is used as an initial recurrent neuron;

[0228] Define system utility function F (0) =0, calculate each target user u i Effect function;

[0229] Step 2: Starting from the nth iteration, the signal power matrix for the current iteration is: Will be Substituting into the signal-to-interference-plus-noise ratio (SINR) calculation formula, the interference-noise matrix is ​​obtained as follows: The formula for calculating the signal-to-interference-plus-noise ratio is:

[0230]

[0231] Step 3: The interference noise matrix obtained in Step 2 is as follows: Calculate the interference suppression matrix

[0232] Step 4: In the reciprocity channel, calculate the corresponding matrix interference noise matrix based on the signal-to-interference-plus-noise ratio (SINR) calculation formula.

[0233] Step 5: Employ the maximum signal-to-interference-plus-noise ratio (SINR) interference alignment algorithm, combined with the matrix interference noise matrix. The corresponding interference suppression matrix is ​​calculated as follows:

[0234] Step 6, Order

[0235] Step 7: Calculate the signal power matrix for the next iteration using a power allocation algorithm.

[0236] Step 8: End the nth iteration, and repeat steps 2 to 7 until the maximum number of iterations N or the power signal matrix P is reached. i convergence.

[0237] It should be noted that the information interaction and execution process between the above-mentioned devices / units are based on the same concept as the method embodiments of this application. For details on their specific functions and technical effects, please refer to the method embodiments section, and they will not be repeated here.

[0238] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is merely an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiments can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit. Furthermore, the specific names of the functional units and modules are only for easy differentiation and are not intended to limit the scope of protection of this application. The specific working process of the units and modules in the above system can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0239] This application also provides a terminal device, see [link to relevant documentation] Figure 10 The terminal 300 may include: at least one processor 310, a memory 320, and a computer program stored in the memory 320 and executable on the at least one processor 310. When the processor 310 executes the computer program, it implements the steps in any of the above method embodiments, for example... Figure 2 Steps 101 to 105 in the illustrated embodiment. Alternatively, when the processor 310 executes the computer program, it implements the functions of each module / unit in the above-described device embodiments, for example... Figure 9 The functions of modules 201 to 203 are shown.

[0240] For example, a computer program may be divided into one or more modules / units, one or more of which are stored in memory 320 and executed by processor 310 to complete this application. The one or more modules / units may be a series of computer program segments capable of performing specific functions, which describe the execution process of the computer program in terminal device 300.

[0241] Those skilled in the art will understand that Figure 10This is merely an example of a terminal device and does not constitute a limitation on the terminal device. It may include more or fewer components than shown, or combine certain components, or different components, such as input / output devices, network access devices, buses, etc.

[0242] The processor 310 can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor.

[0243] The memory 320 can be an internal storage unit of the terminal device or an external storage device, such as a plug-in hard drive, a smart media card (SMC), a secure digital card (SD), or a flash card. The memory 320 is used to store the computer program 321 and other programs and data required by the terminal device. The memory 320 can also be used to temporarily store data that has been output or will be output.

[0244] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0245] The heterogeneous network multi-target optimization interference method provided in this application embodiment can be applied to terminal devices such as computers, wearable devices, vehicle devices, tablet computers, laptops, netbooks, personal digital assistants (PDAs), augmented reality (AR) / virtual reality (VR) devices, and mobile phones. This application embodiment does not impose any restrictions on the specific type of terminal device.

[0246] This application also provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the various embodiments of the heterogeneous network multi-objective optimization interference method described above.

[0247] This application provides a computer program product that, when run on a mobile terminal, enables the mobile terminal to implement the steps in the various embodiments of the heterogeneous network multi-objective optimization interference method described above.

[0248] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the methods of the above embodiments of this application can be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include at least: any entity or device capable of carrying computer program code to a photographing device / terminal device, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. Examples include USB flash drives, portable hard drives, magnetic disks, or optical disks. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electrical carrier signals or telecommunication signals.

[0249] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0250] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0251] In the embodiments provided in this application, it should be understood that the disclosed apparatus / network devices and methods can be implemented in other ways. For example, the apparatus / network device embodiments described above are merely illustrative. For instance, the division of modules or units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0252] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0253] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application, and should all be included within the protection scope of this application.

Claims

1. A multi-objective optimization interference method for heterogeneous networks, characterized in that, include: Construct an ideal signal state information model; Based on the ideal signal state information model and the preset error, a non-ideal channel state information model is constructed. The preset error includes delay error, estimation error and fading error. The construction of the non-ideal channel state information model includes: Construct a channel matrix based on the delay error, the estimation error, and the fading error, expressed as: In the formula, Indicates the first i The transmitter and the first j The estimation error matrix between the receivers It follows a complex Gaussian distribution with mean 0 and variance 1. Indicates the error factor. ,in, This indicates that there is no fading error. This indicates that the transmitter did not receive channel information. This represents the normalized Gaussian white noise matrix. Represents the correlation coefficient of the channel matrix. ρ = J 0(2 πf d τ ) represents the channel matrix correlation function, where, J 0 denotes the zeroth-order Bessel function of the first kind. f d Indicates the maximum Doppler shift. τ Indicates the channel matrix delay. Indicates from the first i The transmitter to the first j Channel matrix for each receiver, The dimension is N r × N t , The elements of follow a complex Gaussian distribution with mean 0 and variance 1; Based on the channel matrix derived from the delay error, the estimation error, and the fading error, the first... j The signal-to-interference-plus-noise ratio (SIR) of the signal received by each receiver is expressed as: In the formula, This represents the interference plus noise matrix of a non-ideal channel. The suppression matrix represents the interference suppression matrix for non-ideal channels. The precoding matrix represents a non-ideal channel; Based on the first j The system throughput is obtained by calculating the signal-to-interference-plus-noise ratio (SIR) of the signals received by each receiver, expressed as: Based on the aforementioned non-ideal channel state information model, a first mathematical model is constructed with the objective functions of optimizing system throughput and system energy efficiency; the expression of the first mathematical model is: In the formula, Indicates the total power of the system. This represents the total energy consumption of the system. This indicates the total power transmitted by the BS. This represents the minimum signal-to-interference-plus-noise ratio (SIR) required for normal communication by the target user. Indicates target user j The allocated power This indicates the fixed losses of the receiver's components; The basic elements of the non-cooperative game model of power allocation include players, strategy space, and utility function; The players are data streams from M target users within a preset system time period, assuming... d j For the first j Send by the target user within the preset time period k Each independent data stream, then D A data stream matrix sent by M target users within the preset time period; The strategy space represents the power allocation schemes available to M target users in the non-cooperative game model of power allocation, assuming... For the first j The strategy vector for each target user, then For the first j The first target user sent the k The power of each independent data stream is represented by the system's decision matrix as Π; The utility function is defined as taking system throughput as the payoff of the game process and system energy consumption as the game cost. Then, the... j The utility function for a target user is: j The expression for the benefit gained from improved system performance minus the system cost is: In the formula, Indicates the first j Each target user receives services from the base station. Indicates the first j The target user did not receive the base station's service. The weight factor for signal-to-interference-plus-noise ratio (SINR). , The weight factor of system energy consumption. , For target users j The weighting factor of the input function , among which, the j Strategy vector for each target user It is a negative term; After transforming the first mathematical model into a non-cooperative game model for power allocation, the expression of the non-cooperative game model for power allocation is: The first mathematical model is transformed into a non-cooperative game model for power allocation; A recurrent neural network algorithm is used to solve the non-cooperative game model of the power allocation, and the optimal game result of system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized is obtained.

2. The method as described in claim 1, characterized in that, The construction of the ideal signal state information model includes: In the ideal channel state information model environment, each transmitter in the MIMO finite feedback channel for M target users corresponds to one receiver, and the transmitter and receiver frequencies do not interfere with each other. The number of transmitter antennas is... N t The number of antennas for the receiver is [number missing]. N r There are [number] base stations with a transmission power of [power]. P trans ; Calculate the first j Received signal of each receiver y j The expression is: In the formula, Indicates the first j The receiver receives the first i The signal power of each transmitter Indicates the first j The receiver receives the first i The transmission path loss of a transmitter signal. Indicates that the base station is assigned to the first i The transmission power of each transmitter, Indicates from the first i The transmitter to the first j Channel matrix for each receiver, The dimension is N r × N t , The elements follow a complex Gaussian distribution with mean 0 and variance 1. Indicates from the first j The transmitter to the first j Channel matrix for each receiver, The dimension is N r × N t , The elements follow a complex Gaussian distribution with mean 0 and variance 1. Indicates the first i The precoding matrix of each transmitter, V i The dimension of the vector is N r × d i , Indicates the first j The precoding matrix of each transmitter, The dimension of the vector is N r × d i , Indicates the first i The signal transmitted by each transmitter The dimension of the vector is d i ×1, data stream is d i , Indicates the first j The signal transmitted by each transmitter The dimension of the vector is d j ×1, data stream is d j , This represents additive complex white Gaussian noise in the channel, with an average value of 0. Calculate the received signal based on interference suppression matrix processing The expression is: In the formula, Indicates the first j The interference suppression matrix of a receiver in an ideal channel. Indicates the first j The precoding matrices of a receiver in an ideal channel, wherein the precoding matrices in the ideal channel and the interference suppression matrix in the ideal channel The qualified expression is: The received signal based on the interference suppression matrix processing is processed using the MAXSINR IA algorithm. , obtained the j The signal-to-interference-plus-noise ratio (SIR) of the signal received by each receiver is expressed as: In the formula, A matrix representing interference plus noise; Based on the first j The system throughput is obtained by calculating the signal-to-interference-plus-noise ratio (SIR) of the signals received by each receiver, expressed as: 。 3. The method as described in claim 1, characterized in that, The method of using a recurrent neural network algorithm to solve the non-cooperative game model of power allocation to obtain the optimal game result of system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized includes: Step 1: Set up the initial environment for the recurrent neural network algorithm: set up i For the target user's first i There are N receivers, where M is the total number of receivers for all target users, and the maximum number of iterations is N. Initialize the precoding matrix as follows The initial interference suppression matrix is: The initial interference noise matrix is ​​as follows: Initialize the first i The signal power matrix received by each receiver is as follows: Each receiver receiving the initial power of the base station is used as an initial recurrent neuron; Define system utility functions F (0) =0, calculate the effect function for each target user. u i ; Step 2, from the... n If the iteration begins at the next iteration, then the signal power matrix for the current iteration number is: , will be Substituting into the signal-to-interference-plus-noise ratio (SINR) calculation formula, the interference-noise matrix is ​​obtained as follows: The formula for calculating the signal-to-interference-plus-noise ratio is: Step 3: Based on the interference noise matrix obtained in Step 2, Calculate the interference suppression matrix Step 4: In the reciprocal channel, calculate the corresponding interference noise matrix based on the aforementioned signal-to-interference-plus-noise ratio (SINR) calculation formula. ; Step 5: Employ the maximum signal-to-interference-plus-noise ratio (SINR) interference alignment algorithm, combined with the aforementioned matrix interference noise matrix. The corresponding interference suppression matrix is ​​calculated as follows: ; Step 6, Order ; Step 7: Calculate the signal power matrix for the next iteration using a power allocation algorithm. ; Step 8, End of Section n For each iteration, repeat steps 2 through 7 until the maximum number of iterations N is reached or the power signal matrix is ​​reached. convergence.

4. A multi-target optimized interference device for heterogeneous networks, characterized in that, include: The channel model construction module is used to construct an ideal signal state information model and to construct a non-ideal channel state information model based on the ideal signal state information model and preset errors, wherein the preset errors include delay error, estimation error and fading error. The construction of the non-ideal channel state information model includes: Construct a channel matrix based on the delay error, the estimation error, and the fading error, expressed as: In the formula, Indicates the first i The transmitter and the first j The estimation error matrix between the receivers It follows a complex Gaussian distribution with mean 0 and variance 1. Indicates the error factor. ,in, This indicates that there is no fading error. This indicates that the transmitter did not receive channel information. This represents the normalized Gaussian white noise matrix. Represents the correlation coefficient of the channel matrix. ρ = J 0(2 πf d τ ) represents the channel matrix correlation function, where, J 0 denotes the zeroth-order Bessel function of the first kind. f d Indicates the maximum Doppler shift. τ Indicates the channel matrix delay. Indicates from the first i The transmitter to the first j Channel matrix for each receiver, The dimension is N r × N t , The elements of follow a complex Gaussian distribution with mean 0 and variance 1; Based on the channel matrix derived from the delay error, the estimation error, and the fading error, the first... j The signal-to-interference-plus-noise ratio (SIR) of the signal received by each receiver is expressed as: In the formula, This represents the interference plus noise matrix of a non-ideal channel. The suppression matrix represents the interference suppression matrix for non-ideal channels. The precoding matrix represents a non-ideal channel; Based on the first j The system throughput is obtained by calculating the signal-to-interference-plus-noise ratio (SIR) of the signals received by each receiver, expressed as: The mathematical model construction module is used to construct a first mathematical model based on the non-ideal channel state information model, with the objective functions of optimizing system throughput and system energy efficiency. It is also used to transform the first mathematical model into a non-cooperative game model for power allocation. The expression of the first mathematical model is: In the formula, Indicates the total power of the system. This represents the total energy consumption of the system. This indicates the total power transmitted by the BS. This represents the minimum signal-to-interference-plus-noise ratio (SIR) required for normal communication by the target user. Indicates target user j The allocated power This indicates the fixed losses of the receiver's components; The basic elements of the non-cooperative game model of power allocation include players, strategy space, and utility function; The players are data streams from M target users within a preset system time period, assuming... d j For the first j Send by the target user within the preset time period k Each independent data stream, then D A data stream matrix sent by M target users within the preset time period; The strategy space represents the power allocation schemes available to M target users in the non-cooperative game model of power allocation, assuming... For the first j The strategy vector for each target user, then For the first j The first target user sent the k The power of each independent data stream is represented by the system's decision matrix as Π; The utility function is defined as taking system throughput as the payoff of the game process and system energy consumption as the game cost. Then, the... j The utility function for a target user is: j The expression for the benefit gained from improved system performance minus the system cost is: In the formula, Indicates the first j Each target user receives services from the base station. Indicates the first j The target user did not receive the base station's service. The weight factor for signal-to-interference-plus-noise ratio (SINR). , The weight factor of system energy consumption. , For target users j The weighting factor of the input function , among which, the j Strategy vector for each target user It is a negative term; After transforming the first mathematical model into a non-cooperative game model for power allocation, the expression of the non-cooperative game model for power allocation is: The model solving module is used to solve the non-cooperative game model of the power allocation using a recurrent neural network algorithm, and obtain the optimal game result of system throughput and system energy efficiency when the signal-to-interference-plus-noise ratio is maximized.

5. A terminal device, comprising a memory and a processor, wherein the memory stores a computer program executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1 to 3.

6. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1 to 3.

Images