A synaesthesia time resource allocation method and system for spectrum map construction

Abstract

The present application discloses a synaesthesia time resource allocation method for spectrum map construction, which belongs to the field of electromagnetic spectrum, including: establishing a time resource allocation model for sensors; setting information age based on timeliness indicators, establishing the relationship between information age and service rate, data arrival rate and packet loss rate in the data transmission stage as constraints of the time resource allocation model; transforming the resource optimization problem in the time resource allocation model after setting the information age constraint into a convex optimization problem; solving the convex optimization problem using the multi-strategy improved dung beetle optimization algorithm MSDBO to obtain the data acquisition T c , channel detection T s and data transmission time T t The optimal time resource allocation scheme for three time periods; MSDBO improves the dung beetle optimization algorithm DBO by setting chaotic pseudo-random, tangent flight search, and vertical and horizontal cross strategies. Addressing the incoherence between communication and perception in existing technologies, this application achieves efficient allocation of synaesthesia time resources during spectrum map construction.

Description

Technical Field

[0001] The present application relates to the field of electromagnetic spectrum, and more specifically, to a method and system for allocating synaesthesia time resources for spectrum map construction. Background Art

[0002] With the continuous advancement of wireless communication technology, electromagnetic spectrum maps have been widely used in spectrum resource management, electromagnetic pollution monitoring, and wireless communication network regulation. Electromagnetic spectrum maps, also known as radio environment maps, electromagnetic environment maps, and radio frequency radio maps, provide visual information on the utilization of different frequency bands, signal strength, and spectrum vacancy areas, helping users better understand and manage spectrum resources. The construction of spectrum maps relies on convenient, accurate, and rapid acquisition of electromagnetic spectrum data. Conventional approaches use one or more spectrum monitoring sensors to capture and analyze signals in the radio spectrum. These sensors include both dedicated spectrum monitoring sensors and non-dedicated sensors based on communication equipment. Dedicated sensors may excel in spectrum monitoring, but they are typically costly and have limited use cases. In contrast, non-dedicated sensors, such as common communication devices (e.g., smartphones, computers, battlefield wireless communication radios), are widely used in practice. Furthermore, these devices often integrate certain sensing capabilities alongside communication functions, providing electromagnetic environment awareness information for spectrum map construction. However, coordination between communication and sensing needs to be addressed.

[0003] Currently, when optimizing spectrum map sensing units, a common approach is to use intelligent optimization algorithms to optimize sensor placement and data sampling. Sensors can be categorized as dedicated and non-dedicated. Dedicated sensors are designed specifically for a specific task, while non-dedicated sensors offer greater flexibility. In the context of integrated communication and perception, non-dedicated sensors are particularly important because they combine communication and perception functions, enabling more efficient utilization of hardware and spectrum resources. However, coordination between communication and perception must be addressed.

[0004] Currently, the main approaches to integrating perception and communication include unified waveform design, power allocation, space division, and frequency division. However, these technologies face their own challenges in practical applications: unified waveform design may lead to compatibility issues with existing communication systems; power allocation methods may be inconvenient in interference control; space division technology requires high beam orthogonality; and frequency division technology may result in a high peak-to-average power ratio (PAPR). Therefore, to more comprehensively optimize sensors, save costs, and increase flexibility, a sensor communication perception resource allocation method for spectrum map construction is urgently needed. Summary of the Invention

[0005] 1. Technical problems to be solved

[0006] In response to the communication and perception inconsistency problem existing in the existing technology, this application provides a synaesthesia time resource allocation method and system for spectrum map construction. By establishing a time resource allocation model for sensors, introducing the information age indicator as a constraint condition, converting the resource optimization problem into a convex optimization problem, and using the multi-strategy improved dung beetle optimization algorithm for solution, the efficient allocation of synaesthesia time resources in the spectrum map construction process can be achieved.

[0007] 2. Technical solution

[0008] The purpose of this application is achieved through the following technical solutions.

[0009] One aspect of the present application provides a synaesthesia time resource allocation method for spectrum map construction, comprising: S1, establishing a time resource allocation model of a sensor for allocating data acquisition T c , channel detection T s and data transmission time T t ; S2, in the time resource allocation model, based on the timeliness index, set the information age, establish the relationship between the information age and the service rate, data arrival rate and packet loss rate in the data transmission stage, and use the established relationship as the constraint condition of the time resource allocation model; S3, transform the resource optimization problem in the time resource allocation model after setting the information age constraint into a convex optimization problem, and obtain the convex optimization model of synaesthesia time resource allocation constructed by spectrum map; S4, use the multi-strategy improved dung beetle optimization algorithm MSDBO to solve the convex optimization problem of the convex optimization model, and obtain the data acquisition T c , channel detection T s and data transmission time T t The optimal time resource allocation scheme for three time periods; among them, MSDBO improves the dung beetle optimization algorithm DBO by setting chaotic pseudo-random, tangent flight search and vertical and horizontal cross strategies to avoid the algorithm falling into local optimality.

[0010] Among them, the spectrum map: A spectrum map represents the spatial distribution of spectrum resource usage within a specific area. It collects and analyzes spectrum data within the area, obtaining information such as the occupancy status, signal strength, and modulation method of each frequency band, and presents it in a visual format. The spectrum map can help identify idle and congested areas of spectrum resources, optimize spectrum allocation and sharing schemes, and improve spectrum utilization efficiency. It has extensive applications in fields such as wireless communications and electronic warfare. Perceptual time resources: Perceptual time resources refer to the time resources used for tasks such as data collection, channel detection, and data transmission in wireless sensor networks for sensing purposes. Unlike traditional communication networks, wireless sensor network nodes typically have limited energy and computing power. Therefore, it is necessary to rationally allocate limited time resources to ensure the quality of sensing tasks while minimizing energy consumption and latency. Perceptual time resource allocation is a key issue in wireless sensor networks, involving multiple aspects such as perception, communication, and computing. Timeliness metrics: Timeliness metrics measure the real-time and effectiveness of data transmission in wireless sensor networks. Because sensor data is time-sensitive, outdated data may have lost its application value. Therefore, data must be transmitted from sensor nodes to sink nodes or control centers as quickly as possible. Common timeliness metrics include latency, deadline miss rate, and data throughput. In synaesthesia temporal resource allocation, timeliness metrics are often modeled as constraints or objective functions in the temporal resource allocation model to ensure real-time transmission of sensor data. Age of Information (AoI): Age of Information is a novel timeliness metric defined as the random delay from the generation of information at the source node to its successful reception at the destination node. Unlike traditional latency, AoI depends not only on the data's transmission time but also on the data's generation time and update frequency. AoI characterizes the freshness of information from its generation to its use and is of great significance in real-time applications such as condition monitoring, process control, and autonomous driving. By minimizing the average or peak AoI, the system's timeliness performance can be optimized. In a given synaesthesia temporal resource allocation scheme, information age is used to model data timeliness and introduced as a constraint in the temporal resource allocation model to ensure real-time spectrum map construction.

[0011] Furthermore, S1, a time resource allocation model of sensors is established, including: setting the sensors to be divided into data collection T c , channel detection T s and data transmission time T t Three non-overlapping time periods; data collection T c Used to obtain raw data, channel detection T sUsed to judge the channel status, data transmission time T t Used to transmit the collected data to the receiving end; establish position m i Sensors at data acquisition T c The relationship between the total amount of data and the false alarm probability and detection probability is calculated by taking the amount of collected original data as input and calculating the corresponding false alarm probability and detection probability; establishing the position m i The sensor at the data transmission time T t The relationship between the amount of data and the channel detection result is s The channel status is judged as the basis to determine the data transmission time T t The amount of data transmitted; according to data collection T c The relationship between the total amount of data and the false alarm probability and detection probability, as well as the data transmission time T t The amount of data and channel detection T s The relationship between them is used to establish a time resource allocation model for sensors.

[0012] False Alarm Probability: In the field of signal detection, a false alarm refers to the misinterpretation of noise as a target signal. The false alarm probability quantifies the probability of this occurring and is defined as the probability of determining that a target signal (H1) is present when no target signal is present (H0 is true). False alarm probability, typically denoted by Pf, is a key performance metric for signal detection systems, reflecting the frequency of false alarms generated by the system. Detection Probability quantifies the probability of correctly determining that a target signal (H1) is present when the target signal is indeed present (H1 is true). Detection Probability, typically denoted by Pd, is another key performance metric for signal detection systems, reflecting the system's ability to correctly detect target signals.

[0013] Further, the corresponding false alarm probability and detection probability are calculated, including: calculating the position m by the following formula i Sensors at data acquisition T c The total amount of data R c : Among them, T c,i Indicates position m i The data collection time period of the sensor; Indicates position m i The channel capacity of the sensor at position m is calculated by the following formula i The false alarm probability of the sensor at: Among them, P f,i (T s,i ) indicates position T s,i The false alarm probability of the sensor at Ts,i Indicates position m i The channel detection time of the sensor at ; Q is the complementary distribution function of the standard Gaussian function; γ i Indicates position m i The signal-to-noise ratio of the received signal of the sensor at Q -1 represents the inverse function of Q; represents the preset target detection probability; f s Represents the sampling frequency; the position m is calculated by the following formula i The detection probability of the sensor at: Among them, P d,i Indicates position m i The detection probability of the sensor at represents the target false alarm probability; τ represents the detection threshold; the amount of data transmitted in the data transmission time period is determined by the following formula: R t,i =R i (T t,i ,T s )×T t,i , where R t,i Indicates position m i The amount of data from the sensor at the data transmission stage; R i Indicates position m i The transmission rate of the sensor at t,i Indicates position m i The data transmission time of the sensor at

[0014] Among them, the complementary cumulative distribution function of the standard Gaussian function (Complementary Cumulative Distribution Function of Standard Gaussian): The standard Gaussian function (also known as the standard normal distribution) is a Gaussian function with mean 0 and variance 1. Its probability density function is: f(x) = 1 / sqrt(2π)*exp(-x^2 / 2). The complementary distribution function is defined as the probability that a random variable is greater than a certain value. For a standard Gaussian random variable Z, its complementary distribution function is: Q(x) = P(Z>x) = 1 / sqrt(2π)∫x∞exp(-t^2 / 2)dt. The Q function is widely used in the fields of communications and signal detection, and is commonly used to calculate performance indicators such as symbol error rate and bit error rate. In the given method, the Q function is used to calculate the false alarm probability. The Q-1 function is the inverse function of the Q function, that is, if Q(x) = y, then x = Q-1(y).

[0015] Furthermore, a time resource allocation model for sensors is established, including: c The total data volume, false alarm probability, detection probability, and data transmission time T tThe amount of transmitted data is used as a constraint, the time resource allocation of the three time periods is used as the optimization variable, and the objective function is to minimize the total time of data acquisition, channel detection and data transmission. The time resource allocation model of the sensor is constructed by the following formula:

[0016] max R t,i

[0017] st

[0018] T c +T s +T t,i ≤T

[0019]

[0020] R c ≤R t

[0021] in, Indicates the minimum sensor acquisition time; T indicates the total time; P d represents the target detection probability of the sensor; represents the set target detection probability; P(H0) represents the probability of channel idleness; 1-P f (T s ) represents the probability that the channel is idle and the sensor successfully detects that the channel is idle; R i (T t,i ) indicates position m i The transmission rate of the sensor; Indicates position m i The minimum transmission rate of the sensor at R c Indicates the amount of data collected by the sensor; R t Indicates the amount of data transmitted by the sensor;

[0022] Furthermore, in the time resource allocation model, S2 sets the information age based on the timeliness index, establishes the relationship between the information age and the service rate, data arrival rate and packet loss rate in the data transmission phase, and uses the established relationship as the constraint condition of the time resource allocation model, including: for each sensor node i, for each sensor node i, set t1, t2, ..., t n As the time when the data packet is successfully uploaded, we can define Δ(t) = tu(t), where u(t) is the timestamp of the last data packet upload and t is the current time. Δ(t) represents the information age at the current time, that is, the time elapsed since the last data packet upload. Define the time interval for status updates as X, and the service time of status updates in the system as T. Calculate the average information age Δ within the observation interval. τ : Where τ represents the length of the observation interval; according to the average information age Δ τ Calculate Δ, where Δ represents the information age related to the data arrival rate and time interval; Where Λ is the data arrival rate of the system, E[X 2 ] represents the expected value of the square of the packet arrival interval; E[XT] represents the expected value of the product of the interval X and the service time T; according to the service time and service rate μ of each packet, the sensor upload model is modeled as an M / M / 1 queue model through exponential distribution, and the average information age ΔM / M / 1 of an M / M / 1 sequence in the queue model is calculated: Where ρ represents the load factor, Using the packet loss rate ρ' to correct the average information age ΔM / M / 1 of the M / M / 1 sequence, we can get the information age Δ'M / M / 1 of the sensor: Based on the information age Δ'M / M / 1 of the sensor, the information age constraint is set as the constraint of the time resource allocation model: Among them, Δ'M / M / 1 represents the information age of the sensor network; Indicates the maximum allowed value of the information age.

[0023] Status Update Interval: During data transmission, sensor nodes periodically send their latest collected data to the receiver. The time interval between two consecutive data updates is called the Status Update Interval, denoted by Y. The Status Update Interval reflects the frequency of data updates. The smaller the interval, the more frequent the data updates and the higher the timeliness of the data. Service Time of Status Update in the System: When a sensor node sends the latest data to the receiver, the data experiences a certain delay in the transmission channel. This delay is called the Service Time of Status Update in the system, denoted by T. The service time depends on the size of the data packet and the transmission link rate, reflecting the delay incurred during data transmission. Queuing Model: The queuing model describes the process of a customer arriving to receive service, waiting for service if service is not immediately available, and then leaving the system after receiving service. In the field of data transmission, data packets can be considered customers and the transmission channel can be considered servers. The process of data packets waiting for transmission at the sender and waiting for processing at the receiver can be characterized by the queuing model. Common queue models include M / M / 1, M / G / 1, and G / M / 1, where M represents Poisson arrival (memoryless), G represents general distribution (general), and 1 represents a single server. In the given method, the sensor upload model is modeled as an M / M / 1 queue, meaning that packet arrival follows a Poisson distribution, service time follows an exponential distribution, and there is only one transmission channel.

[0024] Furthermore, S3 converts the resource optimization problem in the time resource allocation model after setting the information age constraint into a convex optimization problem, thereby obtaining a convex optimization model for synaesthesia time resource allocation constructed by the spectrum map, including: using the established time resource allocation model as the objective function and the information age constraint as the constraint condition to construct the resource optimization problem for time resource allocation; performing a convexity analysis on the resource optimization problem; according to the result of the convexity analysis, if the objective function is a convex function and the constraint condition is a convex set, converting the resource optimization problem into a convex optimization problem; and establishing a convex optimization model for synaesthesia time resource allocation constructed by the spectrum map based on the converted convex optimization problem;

[0025] Convexity analysis is a method used to determine whether an optimization problem is convex. In the field of optimization, convex optimization problems are a class of problems with well-defined properties, possessing globally optimal solutions and solvable using algorithms with polynomial time complexity. Therefore, converting a general optimization problem into a convex one can greatly simplify its solution. An optimization problem is called a convex optimization problem if the objective function is a convex function. A function f defined on a convex set is called a convex function if, for any two points x1, x2, and 0≤θ≤1, f(θx1+(1-θ)x2)≤θf(x1)+(1-θ)f(x2). Intuitively, a convex function lies above the line connecting the two points. A constraint set is called a convex set. A set S is called a convex set if, for any two points x1, x2, and 0≤θ≤1, the point θx1+(1-θ)x2 is also in S. Intuitively speaking, the line connecting any two points inside a convex set is still inside the set. The process of convexity analysis is to test whether the objective function and constraints of the optimization problem satisfy the properties of convex functions and convex sets. Common methods include: Direct method: Based on the definition of convex functions and convex sets, directly verify whether the objective function and constraint set satisfy the conditions. First-order condition: For a differentiable function, if its gradient satisfies Then the function is a convex function. Second-order condition: For a quadratic differentiable function, if its Hessian matrix is ​​semi-positive definite, then the function is a convex function. Convexity-preserving operations: Common functions such as polynomial functions and exponential functions are still convex functions after linear operations, maximum operations, summation operations, etc. In this application, convexity analysis is used to determine whether the established resource optimization model can be converted into a convex optimization model. If the objective function (time resource allocation model) is a convex function with respect to the decision variables, and the constraints (information age constraints) constitute a convex set, the original problem can be equivalently converted into a convex optimization problem for solution. There are many ready-made solution algorithms for convex optimization problems, such as the interior point method, subgradient method, ellipsoid method, etc., which will greatly reduce the difficulty of solving the time resource allocation problem and improve the efficiency of spectrum map construction.

[0026] Furthermore, the expression of the convex optimization model is as follows:

[0027]

[0028] st

[0029] T c +T s +T t,i ≤T

[0030]

[0031] R c ≤R t

[0032] in, α i represents the coefficient based on the detection probability; γ i represents the signal-to-noise ratio (SNR); β represents the time parameter T s The coefficient, Q -1 Represents the inverse cumulative distribution function of the standard normal distribution.

[0033] Furthermore, in S4, the multi-strategy improved dung beetle optimization algorithm MSDBO is used to solve the convex optimization problem of the convex optimization model and obtain the data collection T c , channel detection T s and data transmission time T t The optimal time resource allocation scheme for three time periods includes: using a chaotic pseudo-random strategy to perform chaotic pseudo-random updates on the positions of individuals in the MSDBO; using a tangent flight search strategy to update the positions of individuals in the algorithm processed by the chaotic pseudo-random strategy; wherein the update includes concentrated search and exploratory search; using a vertical and horizontal cross strategy to optimize the positions of individuals in the optimization algorithm processed by the tangent flight search strategy; using the convex optimization model of synaesthesia time resource allocation constructed by the spectrum map as the optimization target of the multi-strategy improved dung beetle optimization algorithm MSDBO, obtaining the optimal solution of the convex optimization model through multiple iterations, and obtaining the optimal time resource allocation scheme t for the three time periods of data acquisition, channel detection and data transmission. i,c , t i,s , t i,t , as the final solution for allocating synaesthesia time resources for spectrum map construction. A chaotic pseudo-random strategy is used to update the position of individuals. The formula is as follows:

[0034]

[0035] Among them, X i,d (t+1) represents the position of the individual at time t+1; X i,d (t) represents the position of individual i at time t; η and μ represent the chaotic coefficients; mod represents the modulo division (remainder) operation; when η and μ are between 0 and 1, the system is in a chaotic state, and r is a random number between 0 and 1;

[0036] For pre-selection, η=0.4 and μ=0.3.

[0037] Among them, the Chaotic Pseudo-random Strategy (CPSS): Chaos is a nonlinear dynamical phenomenon characterized by sensitivity to initial values, randomness, and ergodicity. This CPSS utilizes a chaotic system to generate pseudo-random number sequences for updating individual positions in optimization algorithms. Compared to traditional random number generation methods, CPSS exhibits improved randomness and ergodicity, enhancing the global search capabilities of optimization algorithms and preventing premature convergence. Commonly used chaotic systems include the Logistic map, the Tent map, and the Kent map.

[0038] Tangent Flight Search Strategy: Tangent flight search is a heuristic search strategy that mimics the search behavior of animals such as birds and insects. In a tangent flight search, an individual conducts a local search (focused search) along a tangent line in the neighborhood of the current optimal solution, while also making jumps to more distant areas (exploratory search) with a certain probability to search for possible better solutions. By balancing focused and exploratory search, the tangent flight search strategy can improve the convergence speed and global optimization capabilities of the optimization algorithm.

[0039] Vertical-Horizontal Crossover Strategy: Crossover is a common operation in evolutionary algorithms, generating new offspring individuals by exchanging some of the genes of two parent individuals. The vertical-horizontal crossover strategy is a special crossover method that first performs a vertical (dimensional) crossover on the two parent individuals, then performs a horizontal (within-dimensional) crossover on the resulting intermediate individuals to ultimately generate the offspring individuals. Compared to traditional single-point and multi-point crossover methods, vertical-horizontal crossover can more fully inherit the beneficial characteristics of the parent individuals while also introducing new genetic combinations, helping to maintain population diversity.

[0040] The Dung Beetle Optimization Algorithm (DBOA) is a novel swarm intelligence optimization algorithm that simulates the natural behavior of dung beetles pushing a ball to feed. In DBOA, each individual dung beetle represents a candidate solution. By simulating the beetle's rolling, attacking, and searching behaviors, the algorithm continuously updates its position and ultimately finds the global optimal solution. Compared to other swarm intelligence algorithms, DBOA has stronger development and local search capabilities and has been applied in various fields.

[0041] Improved Dung Beetle Optimization Algorithm (DBOA with Multiple Strategies, MSDBOA): Building on the standard DBOA, the algorithm further enhances performance, improving optimization efficiency and accuracy by introducing multiple strategies, including chaotic pseudo-randomization, tangent flight search, and crossover. The chaotic pseudo-randomization strategy is used to initialize the population and perturb individual positions; the tangent flight search strategy balances local exploitation and global exploration; and the crossover strategy generates new offspring individuals. By integrating multiple strategies, MSDBOA can more efficiently solve complex optimization problems.

[0042] Furthermore, the focused search is performed using the following formula: in represents the position of the individual at the next iteration, is the position of the individual at the current iteration, step is the step size, tan(θ) is the tangent function, and θ is a random angle, whose value is usually between 0 and between, is the location of the currently found optimal solution.

[0043] Furthermore, the exploration search is performed through the following formula: The randn() function generates a standard normal distribution random number. The vertical and horizontal cross strategy is used to optimize the position of the individual in the dung beetle optimization algorithm MSDBO after the tangent flight search strategy is processed, including: i and individual X j , the next generation individual position generated by horizontal crossover in the dth dimension is calculated by the following formula:

[0044]

[0045] in, and are the positions of the new offspring produced by the previous generation of individuals, and are the positions of the individuals of the previous generation in the nth dimension, r1 and r2 are random numbers in the interval [0, 1], c1 and c2 are random numbers in the interval [-1, 1]. The generated offspring are compared with the individuals of the previous generation, and the better individuals are retained;

[0046] For individual X i , k≠d is a different dimension, and the position of the next generation of individuals generated by vertical crossover is calculated by the following formula:

[0047]

[0048] in, Indicates the position of individuals in the next generation; The position of the current generation individual represented by dimension n1; Indicates the position of the current generation individual of dimension n2.

[0049] Another aspect of the present application further provides a synaesthesia time resource allocation system for spectrum map construction, which is used to execute a synaesthesia time resource allocation method for spectrum map construction of the present application.

[0050] 3. Beneficial effects

[0051] Compared with the existing technology, the advantages of this application are:

[0052] By establishing a time resource allocation model for sensors, treating data collection, channel detection, and data transmission as three non-overlapping time periods, and constructing the relationship between parameters such as data volume, false alarm probability, and detection probability in each time period, it is possible to achieve refined allocation of sensor resources in the time dimension, improve the utilization efficiency of time resources, and avoid the problem of incoordination between communication and perception.

[0053] Introducing the information age indicator as a constraint condition of the time resource allocation model and establishing the relationship between information age and the service rate, data arrival rate and packet loss rate in the data transmission stage can quantify the timeliness of the data, ensure the real-time and effectiveness of the data, and improve the accuracy and reliability of spectrum map construction.

[0054] The resource optimization problem in the time resource allocation model after introducing information age constraints is transformed into a convex optimization problem. The feasibility of the transformation is verified through convexity analysis. This can simplify the difficulty of solving the optimization problem, improve the computational efficiency of resource allocation, and achieve efficient allocation of synaesthesia time resources in the process of spectrum map construction.

[0055] A multi-strategy improved dung beetle optimization algorithm is designed. The randomness and diversity of the algorithm are enhanced through the chaotic pseudo-random strategy, the centralized search and exploratory search capabilities of the algorithm are balanced through the tangent flight search strategy, and the local search capability of the algorithm is enhanced through the vertical and horizontal cross strategy. This can improve the convergence speed and global optimization capability of the algorithm, avoid falling into local optimality, and obtain a better time resource allocation plan.

[0056] The constructed convex optimization model is solved using the multi-strategy improved dung beetle optimization algorithm. The optimal resource allocation scheme for the three time periods of data acquisition, channel detection and data transmission is obtained through iterative optimization. While ensuring the timeliness of data transmission, the utilization efficiency of spectrum resources is maximized, the performance requirements of communication and perception are balanced, and the optimal time resource allocation scheme is provided for the efficient construction of spectrum maps. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 This is a schematic diagram of the overall structure of this application;

[0058] Figure 2 is the example scene model;

[0059] Figure 3 Time resource allocation diagram for sensors;

[0060] Figure 4 The initial population distribution diagram is generated by chaotic pseudo-random;

[0061] Figure 5 It is the step length diagram of concentrated search and exploratory search obtained by simulating 500 iterations;

[0062] Figure 6 The influence of the time allocation ratio of data acquisition, channel detection and data transmission on the information age;

[0063] Figure 7 This paper conducts a comparative experiment to compare the performance of the multi-strategy improved dung beetle optimization algorithm (MSDBO) with the original dung beetle optimization algorithm (DBO), the improved dung beetle optimization algorithm (GODBO), the peacock optimization algorithm (POA) and the artificial rabbit algorithm (ARO).

[0064] Figure 8 Another comparative experiment is conducted to compare the performance of the multi-strategy improved dung beetle optimization algorithm (MSDBO) with the original dung beetle optimization algorithm (DBO), the improved dung beetle optimization algorithm (GODBO), the peacock optimization algorithm (POA) and the artificial rabbit algorithm (ARO);

[0065] Figure 9 This is a comparison chart of the time resource allocation algorithm based on MSDBO, common optimization algorithms, and random allocation methods. DETAILED DESCRIPTION

[0066] The present application is described in detail below with reference to the accompanying drawings and specific embodiments.

[0067] Figure 1 This is a schematic diagram of the overall structure of this application. In one embodiment, the problem scenario provided by this application is as follows:

[0068] In the area under consideration, multiple radiation sources with unknown positions and uncertain powers and a group of sensors are arranged. The main task of the sensors is to collect the signals emitted by the radiation sources and realize the effective return of data. Figure 2 The emission power of the radiation source can be obtained by the collection Indicates that the position of the radiation source can be represented by the set Represented by, where m is the number of radiation sources. The signal strength of the radiation source received by the sensors arranged in the area is represented by P(m i ) indicates that m i is the location of the sensor.

[0069] Position m i The received signal power of the sensor at is modeled as: Where K is the free space path loss factor, ε is the path loss exponent, and m j represents the location of the radiation source, is at position m i The shadow loss obeys the log-normal distribution with a mean of 0 and a standard deviation of σ.

[0070] The work of data acquisition sensors is clearly divided into three main stages: electromagnetic data acquisition, channel detection and data communication. Figure 3 As shown, time T is divided into three time periods: electromagnetic data acquisition, channel detection, and data transmission. These three phases do not overlap with each other. For a single sensor, it includes an electromagnetic data acquisition time T c , a channel detection time T s and a data communication time T t Based on this, in this embodiment, if Figure 3 As shown, the present application proposes a communication perception time resource allocation method for spectrum map construction, comprising the following steps: A communication perception time resource allocation method for spectrum map construction, comprising: establishing a time resource allocation model of a sensor for allocating time resources for data acquisition, channel detection, and data transmission, specifically, setting the electromagnetic data acquisition time T c , a channel detection time T s and data transmission time T t ; Set the sensor to be divided into three non-overlapping time periods: data acquisition, channel detection, and data transmission; among them, the data acquisition time period is used to obtain the original data, the channel detection time period is used to judge the channel status, and the data transmission time period is used to transmit the collected data to the receiving end.

[0071] Establish position m i The relationship between the total amount of data collected by the sensor at the location m in the data collection period and the false alarm probability and detection probability is calculated by taking the collected original data as input and calculating the corresponding false alarm probability and detection probability. i The relationship between the sensor's data volume during the data transmission period and the channel detection results is determined, and the channel state determined during the channel detection period is used as a basis to determine the data volume transmitted during the data transmission period. Based on the relationship between the total data volume during the data collection period and the false alarm probability and detection probability, as well as the relationship between the data volume during the data transmission period and the channel detection results, a time resource allocation model for the sensor is established. The corresponding false alarm probability and detection probability are calculated, including the following:

[0072] The position m is calculated by the following formula iThe total amount of data R of the sensor in the data collection period c : Among them, T c,i Indicates the data collection period; Indicates position m i The channel capacity of the sensor at position m is calculated by the following formula i The false alarm probability of the sensor at: Among them, P f,i (T s,i ) indicates position T s,i The false alarm probability of the sensor at position m i The channel detection time of the sensor at ; Q is the complementary distribution function of the standard Gaussian function; γ i Indicates position m i The signal-to-noise ratio of the received signal of the sensor at Q -1 represents the inverse function of Q; Represents the preset target detection probability; represents the sampling frequency; the position m is calculated by the following formula i The detection probability of the sensor at: Among them, P d,i Indicates position m i The detection probability of the sensor at T s,i is the position m i The channel detection time of the sensor at ,γ i is the position m i The signal-to-noise ratio of the received signal of the sensor at f s is the sampling frequency, Q(*) is the complementary distribution function of the standard Gaussian function; represents the target false alarm probability; τ represents the detection threshold; the amount of data transmitted in the data transmission time period is determined by the following formula: R t,i =R i (T t,i ,T s )×T t,i , where R t,i Indicates position m i The amount of data from the sensor at the data transmission stage; R i Indicates position m i The transmission rate of the sensor at t,i Indicates position m i The data transmission time of the sensor at the location; T s Indicates the channel detection time.

[0073] A time resource allocation model for sensors is established, which includes taking the total data volume, false alarm probability, detection probability, and data volume of the data collection period as constraints, and the time resource allocation of the three time periods as optimization variables. The objective function is to minimize the total time for data collection, channel detection, and data transmission. The time resource allocation model for sensors is constructed using the following formula:

[0074] max R t,i

[0075] st

[0076] T c +T s +T t,i ≤T

[0077]

[0078] R c ≤R t

[0079] Among them, T c Indicates the sensor data collection time; Indicates the minimum sensor acquisition time; T indicates the total time; P d represents the target detection probability of the sensor; represents the set target detection probability; P(H0) represents the probability of channel idleness; 1-P f (T s ) represents the probability that the channel is idle and the sensor successfully detects that the channel is idle; R i (T t,i ) indicates position m i The transmission rate of the sensor; Indicates position m i The minimum transmission rate of the sensor at R c Indicates the amount of data collected by the sensor; Indicates the amount of data transmitted by the sensor.

[0080] In the time resource allocation model, based on the timeliness index, the information age is set, and the relationship between the information age and the service rate, data arrival rate and packet loss rate in the data transmission phase is established. The established relationship is used as the constraint condition of the time resource allocation model. Specifically, for each sensor node, for each sensor node, let t1, t2, ..., t n For the moment when a data packet is successfully uploaded, we can define Δ(t) = tu(t), where u(t) is the timestamp of the last data packet upload. Define the time interval between state updates as X, and the service time of state updates in the system as T. Then the average information age within the observation interval (0, τ) is expressed as: can be converted to: Where Λ is the data arrival rate of the system. The sensor upload model in this paper is modeled as a typical M / M / 1 queue model. It is assumed that the service time of each data packet is exponentially distributed with the service rate μ, and the load rate is Then the average information age of an M / M / 1 sequence can be expressed as: In actual sensor network operations, data packets may be lost due to various reasons (such as signal interference, poor channel quality, overload, etc.). The average information age ΔM / M / 1 of the M / M / 1 sequence is corrected using the packet loss rate ρ' to obtain the information age Δ'M / M / 1 of the sensor: Set the constraint formula: Among them, Δ'M / M / 1 represents the information age of the sensor network; Indicates the maximum allowed value of the information age.

[0081] The resource optimization problem in the time resource allocation model after setting the information age constraint is transformed into a convex optimization problem, resulting in a convex optimization model for synaesthesia time resource allocation constructed using a spectrum map. First, the established time resource allocation model is used as the objective function, and the information age constraint is used as a constraint to construct the resource optimization problem for time resource allocation. Specifically, based on the time resource allocation model, the information age constraint is introduced to construct the resource optimization problem for time resource allocation. Then, a convexity analysis is performed on the resource optimization problem. By analyzing the convexity of the objective function and the constraints, it is determined whether the resource optimization problem is convex. Specifically, the second-order derivative of the objective function is calculated. If its Hessian matrix is ​​semi-positive definite within the domain of definition, the objective function is convex. For the constraints, if they define a convex set, that is, if they satisfy the definition of a convex set, then the constraints are convex.

[0082] Based on the results of convexity analysis, if the objective function is convex and the constraints are convex sets, the resource optimization problem is converted into a convex optimization problem. Convex optimization problems have global optimal solutions and can be efficiently solved using existing convex optimization algorithms. Finally, based on the converted convex optimization problem, a convex optimization model for synaesthesia time resource allocation based on spectrum map construction is established; the expression of the convex optimization model is as follows:

[0083]

[0084] st

[0085] T c +T s +T t,i ≤T

[0086]

[0087] R c ≤R t

[0088] in:

[0089]

[0090] The convex optimization model of synaesthesia time resource allocation constructed by spectrum map is solved by using the multi-strategy improved dung beetle optimization algorithm (Multi-strategy Dung Beetle Optimization, MSDBO). A chaotic pseudo-random strategy is used to perform chaotic pseudo-random updates on the positions of individuals in MSDBO. Specifically, a chaotic map is introduced to generate pseudo-random numbers through a chaotic sequence to perturb the positions of individuals in the algorithm, thereby increasing the randomness and diversity of the search and preventing the algorithm from falling into local optimality. A tangent flight search strategy is used to update the positions of individuals in the algorithm after being processed by the chaotic pseudo-random strategy. The tangent flight search strategy includes two stages: concentrated search and exploratory search. By balancing local search and global search, the search efficiency and optimization ability of the algorithm are improved; among them, the update includes concentrated search and exploratory search; the concentrated search is achieved through the following formula: in represents the position of the individual at the next iteration, is the position of the individual at the current iteration, step is the step size, tan(θ) is the tangent function, and θ is a random angle, whose value is usually between 0 and between, is the location of the currently found optimal solution. The exploration search is performed using the following formula:

[0091] The random number generated by randn(*) is a standard normal distribution.

[0092] The vertical and horizontal cross strategy is used to optimize the position of individuals in the optimization algorithm after the tangent flight search strategy. The vertical and horizontal cross strategy enhances the local search ability of the algorithm and accelerates the convergence speed through information exchange between individuals. i and individual X j , the next generation individual position generated by horizontal crossover in the dth dimension is calculated by the following formula:

[0093]

[0094] in, and are the positions of the new offspring produced by the previous generation of individuals, and are the positions of the individuals of the previous generation in the nth dimension, r1 and r2 are random numbers in the interval [0, 1], c1 and c2 are random numbers in the interval [-1, 1], and the generated offspring are compared with the individuals of the previous generation, and the better individuals are retained; for individual X i , k≠d is a different dimension, and the position of the next generation of individuals generated by vertical crossover is calculated by the following formula: in, Indicates the position of individuals in the next generation; The position of the current generation individual represented by dimension n1; The position of the current generation individual represented as dimension n2.

[0095] A chaotic pseudo-random strategy is used to update the position of the individual. The formula is as follows:

[0096]

[0097] Among them, X i,d (t+1) represents the position of individual i at time t+1, X i,d (t) represents the position of individual i at time t; η and μ represent the chaotic coefficients; mod represents the modulo division (remainder) operation; when η and μ are between 0 and 1, the system is in a chaotic state, and r is a random number between 0 and 1; here we take η = 0.4, μ = 0.3.

[0098] The convex optimization model of synaesthesia time resource allocation constructed by the spectrum map is used as the optimization target of MSDBO, and the optimal solution of the convex optimization model is obtained through multiple iterations. Specifically, the objective function of the convex optimization model is used as the fitness function of MSDBO, and the individual positions are updated and optimized through the chaotic pseudo-random strategy, tangent flight search strategy and vertical and horizontal cross strategy. It is continuously iterated until the termination condition is met, and the optimal time resource allocation scheme is obtained. The optimal time resource allocation scheme t for the three time periods of data acquisition, channel detection and data transmission is obtained. i,c , t i,s , t i,t , serving as the final solution for synaesthesia time resource allocation for spectrum map construction. Through the above steps, the multi-strategy improved dung beetle optimization algorithm (MSDBO) was used to solve the convex optimization model for synaesthesia time resource allocation for spectrum map construction, resulting in the optimal time resource allocation scheme for the three time periods of data acquisition, channel detection, and data transmission. By introducing chaotic pseudo-randomness, tangent flight search, and vertical and horizontal crossover strategies, MSDBO improves the algorithm's search efficiency, optimization capability, and convergence speed, providing an optimized time resource allocation scheme for efficient spectrum map construction.

[0099] In this example, to verify the performance of the proposed time resource allocation algorithm, it is compared with other different algorithms. To make the experimental results fair and objective, the scale of all algorithms is set to 30, the maximum number of iterations is 200, and to eliminate the influence of randomness, all algorithms are independently run 50 times and the average is taken. The relevant simulation parameters are shown in the following table:

[0100]

[0101]

[0102] Figure 4 The initial population distribution diagram is generated by chaotic pseudo-random, where the horizontal axis represents the position of the individual and the vertical axis represents the fitness value of the individual. Figure 4 As can be seen, the initialization population generated by the chaotic pseudo-random strategy is more evenly distributed in the search space, with a higher proportion of high-quality solutions (individuals with higher fitness values). This distribution property helps the algorithm quickly locate potential high-quality areas in the early stages of optimization, accelerating convergence and improving optimization efficiency. Compared with random initialization, chaotic pseudo-random initialization provides a better starting point for the algorithm and increases the probability of finding the global optimal solution.

[0103] Figure 5 This is a graph showing the step size changes of the concentrated search and exploratory search obtained by simulating 500 iterations, where the horizontal axis represents the number of iterations and the vertical axis represents the size of the search step size. Figure 5 The following characteristics can be observed: In the early stages of the optimization (e.g., the first 100 iterations), the centralized search step size varies significantly, exhibiting significant fluctuations. This indicates that in the initial optimization phase, the algorithm actively explores the optimization space by varying the step size significantly, hoping to quickly find promising regions. This behavior helps the algorithm identify promising search directions early on, providing a good foundation for subsequent optimization. As the number of iterations increases, the changes in the centralized search step size gradually become more stable, and the fluctuations decrease. This indicates that as optimization progresses, the algorithm begins to focus on the currently found optimal region, performing a refined local search by reducing the step size. This strategy helps the algorithm further explore within promising regions in the hope of finding even better solutions. The exploratory search step size exhibits significant fluctuations throughout the iterations, with relatively strong variations. This reflects the algorithm's global search behavior, which involves exploring the entire search space through large step size variations to identify new possible regions. This global exploration helps the algorithm escape local optima and discover more optimal solutions.

[0104] The step sizes of the focused and exploratory searches alternate throughout the iterative process, reflecting the algorithm's characteristic of seeking a balance between exploration and exploitation. By properly adjusting the step sizes of the two search strategies, the algorithm is able to find a balance between expanding the search range and focusing on high-quality areas, improving optimization efficiency and solution quality. It can be seen that the initial population generated by the chaotic pseudo-random method is more evenly distributed and has a higher proportion of high-quality solutions, which helps the algorithm quickly find potential high-quality areas. The step size changes between the focused and exploratory searches reflect the algorithm's behavioral characteristics during the optimization process. By balancing exploration and exploitation, the algorithm's optimization ability and convergence speed are improved. These characteristics enable the MSDBO algorithm to efficiently solve the synaesthesia time resource allocation problem constructed by spectral maps and obtain an optimized time resource allocation solution.

[0105] Figure 6 The effect of the time allocation ratio of data acquisition, channel detection, and data transmission on the information age is shown. The horizontal and vertical axes represent the data acquisition time ratio and the data transmission time ratio, respectively. The colors in the figure represent the information age under the corresponding time allocation ratio. Figure 6 The following characteristics can be observed: Blue areas indicate lower information age, which typically occurs in regions with a small ratio of both data collection time and data transmission time. This indicates that when both data collection and transmission time are short, information timeliness is high, allowing the system to quickly acquire and update data, thereby maintaining a low information age and improving the system's real-time performance. Yellow areas indicate higher information age, which typically occur in regions with a large ratio of either data collection time or data transmission time. This indicates that when data collection or transmission time is long, information timeliness decreases, data updates slow down, and the information age significantly increases, reducing the system's real-time performance. Therefore, to ensure efficient system operation and data timeliness, it is necessary to rationally adjust data collection and transmission time to find the optimal time allocation ratio. By balancing the time overhead of data collection and transmission, we can ensure both the adequacy and accuracy of data collection and the timeliness and reliability of data transmission, thereby reducing information age and improving overall system performance.

[0106] Figure 7 and Figure 8 The performance comparison results of MSDBO algorithm and other optimization algorithms are shown, where the horizontal axis represents the number of iterations and the vertical axis represents the objective function value. Figure 7 and Figure 8It can be seen that the MSDBO algorithm converges best to the final objective function value, and the convergence curve shows an almost horizontal trend after 20 iterations. This demonstrates that the MSDBO algorithm can quickly find a near-optimal solution, possessing strong optimization capabilities and convergence speed. Compared to other algorithms, the MSDBO algorithm exhibits significant improvements in convergence speed. Compared to the fastest-converging ARO algorithm, MSDBO's convergence speed increased by approximately 33.33%; compared to the slowest-converging POA algorithm, MSDBO's convergence speed increased by approximately 73.68%. This demonstrates that the MSDBO algorithm, by introducing multiple optimization strategies, effectively accelerates convergence and shortens optimization time.

[0107] Figure 9 The performance comparison results of the MSDBO-based time resource allocation algorithm with other optimization algorithms and random allocation methods are shown. Figure 9 As can be seen, the MSDBO algorithm improves the optimization objective by approximately 12.82% compared to the random time allocation method. This demonstrates that the MSDBO algorithm can effectively optimize time resource allocation and achieve better system performance. It can be seen that the time allocation ratio for data acquisition, channel detection, and data transmission has a significant impact on information age, requiring reasonable adjustment of time allocation to reduce information age and improve system performance. The MSDBO algorithm demonstrates excellent optimization capabilities and convergence speed in optimizing time resource allocation, effectively optimizing time resource allocation and achieving better system performance, providing strong support for the efficient construction of spectrum maps.

[0108] This application proposes a sensor communication perception time resource allocation model for spectrum map construction, achieving a trade-off between node data acquisition, channel detection, and data transmission. Simulation results show that compared with other commonly used optimization algorithms, the proposed algorithm significantly accelerates convergence speed, achieves rapid spectrum map reconstruction, and improves spectrum map construction accuracy, while improving performance.

[0109] The above schematically describes the invention of the present application and its implementation methods. This description is not restrictive. Without departing from the spirit or basic features of the present application, the present application can be implemented in other specific forms. What is shown in the drawings is only one of the implementation methods of the invention of the present application. The actual structure is not limited to this. Any figure mark in the claims should not limit the claims involved. Therefore, if a person of ordinary skill in the art is inspired by it, without departing from the purpose of the present invention, a structural method and embodiment similar to the technical solution without creativity should fall within the scope of protection of this patent. In addition, the word "including" does not exclude other elements or steps, and the word "one" before an element does not exclude the inclusion of "multiple" elements. The multiple elements stated in the product claim can also be implemented by one element through software or hardware. Words such as first and second are used to indicate names and do not indicate any specific order.

Claims

1. A method for allocating synaesthesia time resources for spectrum map construction, comprising: S1, establish the sensor time resource allocation model to allocate data collection T c , channel detection T s and data transmission time T t ; S2, in the time resource allocation model, based on the timeliness index, sets the information age, establishes the relationship between the information age and the service rate, data arrival rate and packet loss rate in the data transmission phase, and uses the established relationship as the constraint condition of the time resource allocation model; S3, transforming the resource optimization problem in the time resource allocation model after setting the information age constraint into a convex optimization problem, and obtaining a convex optimization model for synaesthesia time resource allocation constructed by the spectrum map; S4, using the multi-strategy improved dung beetle optimization algorithm MSDBO, solve the convex optimization problem of the convex optimization model and obtain the data collection T c , channel detection T s and data transmission time T t The optimal time resource allocation scheme for three time periods; among them, MSDBO improves the dung beetle optimization algorithm DBO by setting chaotic pseudo-random, tangent flight search and vertical and horizontal cross strategies to avoid the algorithm falling into local optimality; S1, establishes a time resource allocation model for sensors, including: Set the sensor to collect data in time c , channel detection T s and data transmission time T t Three non-overlapping time periods; data collection T c Used to obtain raw data, channel detection T s Used to judge the channel status, data transmission time T t Used to transmit the collected data to the receiving end; Establish position m i Sensors at data acquisition T c The relationship between the total amount of data and the false alarm probability and detection probability is calculated by taking the amount of collected original data as input and calculating the corresponding false alarm probability and detection probability; Establish position m i The sensor at the data transmission time T t The relationship between the amount of data and the channel detection result is s The channel status is judged as the basis to determine the data transmission time T t The amount of data transmitted; According to data collection T c The relationship between the total amount of data and the false alarm probability and detection probability, as well as the data transmission time T t The amount of data and channel detection T s The relationship between them is used to establish a time resource allocation model for sensors; Calculate the corresponding false alarm probability and detection probability, including: The position m is calculated by the following formula i Sensors at data acquisition T c The total amount of data R c : Among them, T c,i Indicates position m i The data collection time period of the sensor; Indicates position m i The channel capacity of the sensor at The position m is calculated by the following formula i The false alarm probability of the sensor at: Among them, P f,i (T s,i ) indicates position m i The false alarm probability of the sensor at T s,i Indicates position m i The channel detection time of the sensor at ; Q is the complementary distribution function of the standard Gaussian function; γ i Indicates position m i The signal-to-noise ratio of the received signal of the sensor at Q -1 represents the inverse function of Q; represents the preset target detection probability; f s Indicates the sampling frequency; The position m is calculated by the following formula i The detection probability of the sensor at: Among them, P d,i Indicates position m i The detection probability of the sensor at represents the target false alarm probability; τ represents the detection threshold; The amount of data transmitted during a data transmission period is determined by the following formula: R t,i =R i (T t,i ,T s )×T t,i Among them, R t,i Indicates position m i The amount of data from the sensor at the data transmission stage; R i Indicates position m i The transmission rate of the sensor at t,i Indicates position m i Data transmission time of the sensor at Establish a time resource allocation model for sensors, including: Collect data T c The total data volume, false alarm probability, detection probability, and data transmission time T t The amount of transmitted data is used as a constraint, the time resource allocation of the three time periods is used as the optimization variable, and the objective function is to minimize the total time of data acquisition, channel detection and data transmission. The time resource allocation model of the sensor is constructed by the following formula: max R t,i T c +T s +T t,i ≤T R c ≤R t in, Indicates the minimum sensor acquisition time; T indicates the total time; P d represents the target detection probability of the sensor; represents the set target detection probability; P(H0) represents the probability of channel idleness; 1-P f (T s ) represents the probability that the channel is idle and the sensor successfully detects that the channel is idle; R i (T t,i ) indicates position m i The transmission rate of the sensor; Indicates position m i The minimum transmission rate of the sensor at R c Indicates the amount of data collected by the sensor; R t Indicates the data transmission volume of the sensor.

2. The method for allocating synaesthesia time resources for spectrum map construction according to claim 1, characterized in that: S2, establishes the constraints of the time resource allocation model, including: For each sensor node i, let t1, t2, ..., t n Define Δ(t) = tu(t) as the time when a data packet is successfully uploaded, where u(t) is the timestamp of the last data packet upload, t is the current time, and Δ(t) is the age of the information at the current time, that is, the time elapsed since the last data packet upload. Define the time interval for status updates as X, and the service time of status updates in the system as T. Calculate the average information age Δ within the observation interval (0,τ) τ : Where τ represents the length of the observation interval; According to the average information age Δ τ Calculate Δ, where Δ represents the information age related to the data arrival rate and time interval; Where Λ is the data arrival rate of the system, E[X 2 ] represents the expected value of the square of the packet arrival interval; E[XT] represents the expected value of the product of the interval X and the service time T; According to the service time and service rate μ of each data packet, the sensor upload model is modeled as an M / M / 1 queue model through exponential distribution, and the average information age ΔM / M / 1 of an M / M / 1 sequence in the queue model is calculated: Where ρ represents the load factor, Using the packet loss rate ρ' to correct the average information age ΔM / M / 1 of the M / M / 1 sequence, we can get the information age Δ'M / M / 1 of the sensor: Based on the information age Δ'M / M / 1 of the sensor, the information age constraint is set as the constraint of the time resource allocation model: Among them, Δ'M / M / 1 represents the information age of the sensor network; Indicates the maximum allowed value of the information age.

3. The method for allocating synaesthesia time resources for spectrum map construction according to claim 2, characterized in that: S3, transforming the resource optimization problem in the time resource allocation model after setting the information age constraint into a convex optimization problem, and obtaining a convex optimization model for synaesthesia time resource allocation constructed by the spectrum map, including: The established time resource allocation model is used as the objective function and the information age constraint is used as the constraint condition to construct the resource optimization problem of time resource allocation; Conduct convexity analysis on resource optimization problems; According to the results of convexity analysis, if the objective function is a convex function and the constraints are a convex set, the resource optimization problem is transformed into a convex optimization problem; According to the transformed convex optimization problem, a convex optimization model for synaesthesia time resource allocation constructed by spectrum map is established.

4. The method for allocating synaesthesia time resources for spectrum map construction according to claim 3, characterized in that: The expression of the convex optimization model is as follows: T c +T s +T t,i ≤T R c ≤R t in, α i represents the coefficient based on the detection probability; γ i represents the signal-to-noise ratio (SNR); β represents the time parameter T s The coefficient, Q -1 Represents the inverse cumulative distribution function of the standard normal distribution.

5. The method for allocating synaesthesia time resources for spectrum map construction according to claim 4, characterized in that: S4, using the multi-strategy improved dung beetle optimization algorithm MSDBO, solve the convex optimization problem of the convex optimization model and obtain the data collection T c , channel detection T s and data transmission time T t The optimal time resource allocation scheme for three time periods includes: Adopting chaotic pseudo-random strategy, the positions of individuals in MSDBO are updated in a chaotic pseudo-random manner; The tangent flight search strategy is used to update the position of individuals in the algorithm after being processed by the chaotic pseudo-random strategy. The update includes concentrated search and exploratory search. A vertical and horizontal cross strategy is used to optimize the position of individuals in the optimization algorithm after the tangent flight search strategy is processed; The convex optimization model of synaesthesia time resource allocation constructed by spectrum map is used as the optimization target of the multi-strategy improved dung beetle optimization algorithm MSDBO. The optimal solution of the convex optimization model is obtained through multiple iterations, and the optimal time resource allocation scheme t for the three time periods of data acquisition, channel detection and data transmission is obtained. i,c , t i,s , t i,t ,The final scheme of synaesthesia time resource allocation as spectrum map construction; A chaotic pseudo-random strategy is used to update the position of the individual. The formula is as follows: Among them, X i,d (t+1) represents the position of the individual at time t+1; X i,d (t) represents the position of individual i at time t; η and μ represent the chaotic coefficients; mod represents the modular division operation; when η and μ are between 0 and 1, the system is in a chaotic state, and r is a random number between 0 and 1.

6. The method for allocating synaesthesia time resources for spectrum map construction according to claim 5, characterized in that: The focused search is done by the following formula: in represents the position of the individual at the next iteration, is the position of the individual at the current iteration moment, step is the step size, tan(θ) is the tangent function, and θ is a random angle, which is usually between 0 and between, is the location of the optimal solution currently found; The exploration search is done through the following formula: The random number generated by randn() is a standard normal distribution. The vertical and horizontal cross strategy is used to optimize the position of individuals in the dung beetle optimization algorithm MSDBO after the tangent flight search strategy is processed, including: For individual X i and individual X j , the next generation individual position generated by horizontal crossover in the dth dimension is calculated by the following formula: in, and are the positions of the new offspring produced by the previous generation of individuals, and are the positions of the individuals of the previous generation in the nth dimension, r1 and r2 are random numbers in the interval [0, 1], c1 and c2 are random numbers in the interval [-1, 1]. The generated offspring are compared with the individuals of the previous generation, and the better individuals are retained; For individual X i , k≠d is a different dimension, and the position of the next generation of individuals generated by vertical crossover is calculated by the following formula: in, Indicates the position of individuals in the next generation; The position of the current generation individual represented by dimension n1; Indicates the position of the current generation individual of dimension n2.

7. A synaesthesia time resource allocation system for spectrum map construction, characterized in that: include: At least one processing unit PE; configured to execute instructions to implement the synaesthesia time resource allocation method for spectrum map construction according to any one of claims 1 to 6.

Images