Indoor heterogeneous network resource allocation method based on improved cuckoo algorithm

Abstract

The application discloses an indoor heterogeneous network resource allocation method based on an improved cuckoo algorithm, and is implemented according to the following steps: based on user and user network access points, an indoor VLC heterogeneous network system model is constructed; in the indoor VLC heterogeneous network system model, an improved cuckoo algorithm is used to perform layer-by-layer allocation on VLC resource blocks; the application combines the advantages of two communication technologies of VLC and WiFi, constructs an indoor VLC heterogeneous network, and can provide complete indoor wireless network coverage and high system throughput; aiming at the problems existing in the existing cuckoo algorithm, the improved cuckoo algorithm is used, so that the convergence speed of the algorithm is effectively improved, and the problems of the existing scheme, such as limited use scene, high algorithm complexity and easy falling into a local optimal solution, are solved.

Description

Technical Field

[0001] This invention relates to the field of communication network technology, specifically to an indoor heterogeneous network resource allocation method based on an improved cuckoo algorithm. Background Technology

[0002] In recent years, with the rapid development of internet technology and the widespread application of various mobile terminals, people's demand for high-speed services has been increasing. Traditional radio frequency communication, due to the scarcity of spectrum resources, is unable to meet this growing demand. Visible light communication (VLC) has become a high-speed wireless access solution with great development potential due to its advantages such as high transmission rate, resistance to electromagnetic interference, low cost, high security, abundant spectrum, and no need for authentication. VLC, based on existing indoor white LED lighting, can simultaneously provide lighting and communication, reducing costs while offering great convenience. However, in practical scenarios, a single wireless communication system composed solely of VLC suffers from problems such as small coverage area, susceptibility to optical link obstruction, and difficulties in uplink transmission. Therefore, VLC networks can be combined with wide-coverage WiFi networks to form an indoor VLC heterogeneous network.

[0003] Resource allocation in indoor VLC heterogeneous networks is a discrete optimization problem. Existing solutions mainly include traditional resource allocation algorithms, graph coloring-based resource allocation algorithms, deep learning-based resource allocation algorithms, and swarm intelligence-based resource allocation algorithms. Traditional resource allocation algorithms are generally implemented based on greedy algorithms and linear programming, which are easy to implement and maintain, but lack flexibility and robustness. When faced with limited resources and a large number of users, the resource allocation problem becomes non-convex and highly complex, making it difficult for traditional algorithms to handle. Graph coloring-based resource allocation algorithms are relatively mature. They represent the resource allocation problem as a perturbation graph and allocate users according to different rules. The algorithm is simple to implement and can handle large-scale resource allocation problems, but because it is an NP-complete problem, it may consume a lot of time and computational resources when dealing with large-scale and complex problems. In recent years, with the development of deep learning models, deep learning-based resource allocation technology has attracted widespread attention and research. This method can learn and predict network resource needs through a large amount of training data to optimize allocation decisions. However, its disadvantage is that it requires a large amount of sample data, and its accuracy is highly dependent on the quality and quantity of the data used. Swarm optimization algorithms are commonly used to simulate physical phenomena or biological evolution processes in nature, and are frequently used to solve complex engineering optimization problems, such as genetic algorithms, particle swarm optimization, and firefly algorithms. Resource allocation algorithms based on swarm optimization algorithms can perform fast global searches, have good adaptability and the ability to solve complex multi-constraint problems, and can effectively solve highly complex NP-complete problems. However, due to the randomness and global search characteristics, they inevitably suffer from slow convergence speed, low solution accuracy, and a tendency to get trapped in local optima. The Cuckoo Algorithm is an emerging swarm optimization algorithm. In the Cuckoo Algorithm, each bird egg corresponds to a nest location, representing a candidate solution to the optimization problem. This algorithm has fewer parameters, is easy to adjust and implement, and has advantages such as strong global search capability and fast convergence speed. In recent years, it has been applied to resource allocation problems. However, the existing Cuckoo Algorithm also suffers from the problem of getting trapped in local optima, making it difficult to apply well to solving resource allocation problems in complex indoor VLC heterogeneous networks. Summary of the Invention

[0004] The purpose of this invention is to provide an indoor heterogeneous network resource allocation method based on an improved cuckoo algorithm. This method combines the advantages of both VLC and WiFi communication technologies to construct an indoor VLC heterogeneous network, which can provide complete indoor wireless network coverage and high system throughput. To address the problems of existing cuckoo algorithms, an improved cuckoo algorithm is adopted, which effectively improves the convergence speed of the algorithm and solves the problems of limited application scenarios, high algorithm complexity, and easy getting trapped in local optima in existing solutions.

[0005] The technical solution adopted in this invention is an indoor heterogeneous network resource allocation method based on an improved cuckoo algorithm, which is implemented according to the following steps: Step 1: Construct an indoor VLC heterogeneous network system model based on the network access points between users; Step 2: Within the indoor VLC heterogeneous network system model, the improved Cuckoo algorithm is used to allocate VLC resource blocks layer by layer; The VLC heterogeneous network system includes both VLC and WiFi communication networks; Step 2 is as follows: Step 2.1: Number all resource blocks at all access points uniformly, treat the resource block allocation scheme of each layer of users as a bird's nest, set the fitness function to a function related to the system throughput of the VLC network, and initialize the population; Step 2.2: Update the nest location via Levi's flight; Step 2.3: Generate random numbers Compare it with the improved adaptive discovery probability If a comparison is made, If the nest position is positive, the nest position is updated; otherwise, it remains unchanged. After the position update is completed, the fitness value of each nest is calculated and compared with the previous fitness value. The better nest position is retained. Step 2.4: Determine if the maximum number of iterations has been reached. If the optimal solution has been achieved, output the global optimal solution, i.e., the optimal resource block allocation scheme; otherwise, use the location of the bird's nest in this generation as the initial bird's nest for the next generation, and return to step 2.2 for the next iteration. Step 2.5: Calculate the rate allocated to users based on the allocated resource blocks, and determine whether the rate requirements of all users have been met. If they have been met, output the resource allocation scheme in Step 2.4; if they have not been met, determine whether there are still VLC resource blocks that can be allocated to users whose requirements have not been met. If so, continue with the next resource allocation according to Steps 2.1 to 2.5. If there are no allocated VLC resource blocks, allocate WiFi resource blocks to users. The population initialization described in step 2.1 specifically involves: Set population size Search space range Maximum number of iterations The optimal fitness value is The low-difference Halton sequence from the quasi-Monte Carlo method is introduced to initialize the location of the bird's nest, generating... The system identifies the locations of each nest and calculates its corresponding fitness value. Nests are then sorted by fitness value, and the nests with the best fitness values ​​are selected. One bird nest was used as the initial population; Step 2.2 specifically involves: employing an improved adaptive step size factor during Levi's flight. The nest position is updated according to the update formula, and its corresponding fitness value is calculated and compared with the fitness value corresponding to the nest position of the previous generation. Nest positions with better fitness are retained to realize nest position update. The location of the bird's nest mentioned in step 2.3 has been updated to:

[0006] in, The updated location of the Bird's Nest. Location of the abandoned bird's nest and Improved adaptive discovery probability for two randomly selected bird nest locations , , , For the first t The fitness value of the best individual in the population. For the first t The first generation of the population i The fitness value of each individual.

[0007] The invention is further characterized by: Step 1 is as follows: The user set is User access point set Among them, VLC access point set WiFi access point set .

[0008] The fitness function is calculated using the following formula: The formula for calculating the throughput of a VLC network system is as follows:

[0009] Location of the i-th bird's nest The fitness function is expressed as:

[0010] In the formula, Indicates user and VLC Connected, VLC arrive The signal-to-interference-plus-noise ratio (SINR) is given by e = 2.718, where e is the Euler number.

[0011] The formula for initializing the bird's nest location using the low-difference Halton sequence introduced in the quasi-Monte Carlo method is as follows:

[0012]

[0013] in, This represents the location of the i-th bird's nest in generation t. As the lower bound of the search space, For the middle value of the search space, This is the upper limit of the search space. It is a Halton sequence.

[0014] The formula for updating the Bird's Nest location is as follows:

[0015] in, This refers to the location of the next generation of the Bird's Nest stadium, where α is the step size adjustment factor. This represents point-to-point product. It is a random search path for Levi's flight. It is an improved adaptive step size factor. , .

[0016] The beneficial effects of this invention are: (1) This invention applies the improved cuckoo algorithm to resource allocation in indoor VLC heterogeneous networks, which improves the performance of the algorithm. The parameter settings are simple, the convergence speed is fast, it is not easy to get trapped in local optima, and it is easy to implement. It can find the optimal resource allocation scheme during the iteration process. (2) The present invention has strong adaptability to the scene. It can provide a good allocation scheme for different lamp layouts, sizes and access capabilities, realize resource allocation in complex indoor scenes, and has more practical application significance. (3) This invention aims to maximize system throughput, considers global optimal allocation, prioritizes the allocation of VLC resources, and then uses WiFi as a supplement to allocate resources to users whose needs are not met, thus obtaining the optimal allocation scheme. This can better maximize the benefits of indoor VLC heterogeneous networks, and significantly improve system throughput, user satisfaction and fairness. Attached Figure Description

[0017] Figure 1 This is a flowchart of the indoor heterogeneous network resource allocation method based on the improved cuckoo algorithm in this invention; Figure 2 This is a schematic diagram of the indoor VLC heterogeneous network system model structure in this invention; Figure 3 This is a flowchart of the improved cuckoo algorithm in this invention.

[0018] Figure 4This is a schematic diagram of the lamp layout in Embodiment 2 of the present invention; Figure 5 for Figure 4 Schematic diagram of the lighting performance of the three layouts; Figure 6 This is a schematic diagram comparing the convergence of the ICS algorithm, CS algorithm, and PSO algorithm in an embodiment of the present invention. Figure 7(a) shows the simulation results of the system throughput under the three lamp layouts in Example 3; Figure 7(b) shows the simulation results of user satisfaction under the three lamp layouts in Example 3; Figure 7(c) shows the service fairness simulation results under the three lamp layouts in Example 3. Detailed Implementation

[0019] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0020] Example 1 This invention combines a VLC network with a wide-coverage WiFi network to form an indoor VLC heterogeneous network. This heterogeneous network offers higher communication speeds and larger coverage areas, providing flexible solutions for different types of data services. Furthermore, since the frequency bands of VLC and WiFi do not overlap, there is no co-channel interference, significantly improving the system throughput of this heterogeneous network. Simultaneously, by rationally utilizing resource allocation methods to plan user connection schemes, system throughput can be greatly enhanced.

[0021] This invention proposes an indoor heterogeneous network resource allocation method based on an improved cuckoo algorithm. Aiming to maximize system throughput, this method provides an optimal resource allocation scheme for indoor VLC heterogeneous networks, even with varying indoor lighting layouts, AP (access point) capabilities, and user needs. This improves the performance and reliability of the indoor VLC heterogeneous network. Figure 1 As shown, please follow these steps: Step 1: Based on the network access points between users, construct an indoor VLC heterogeneous network system model, such as... Figure 2 As shown; the specific process is as follows: The user set is User access point set Among them, VLC access point set WiFi access point set .

[0022] Step 2: Within the indoor VLC heterogeneous network system model, an improved cuckoo algorithm is used to allocate VLC resource blocks layer by layer. The improved cuckoo algorithm offers high performance in exploring the solution space, can flexibly escape local optima, and has a simple structure with fewer parameters, making it easy to implement. The specific process is as follows: Step 2.1: Number all resource blocks at all access points uniformly, treat the resource block allocation scheme of each layer of users as a bird's nest, set the fitness function to a function related to the system throughput of the VLC network, and initialize the population; The formula for calculating the throughput of a VLC network system is as follows:

[0023] Location of the i-th bird's nest The fitness function is expressed as:

[0024] In the formula, Indicates user and VLC Connected, VLC arrive The signal-to-interference-plus-noise ratio (SINR) is given by e = 2.718, where e is the Euler number.

[0025] The specific process of population initialization is as follows: Set population size Search space range Maximum number of iterations The optimal fitness value is The low-difference Halton sequence from the quasi-Monte Carlo method is introduced to initialize the location of the bird's nest, generating... The system identifies the locations of each nest and calculates its corresponding fitness value. Nests are then sorted by fitness value, and the nests with the best fitness values ​​are selected. One bird nest was used as the initial population.

[0026] The formula for initializing the bird's nest location using the low-difference Halton sequence from the quasi-Monte Carlo method is as follows:

[0027]

[0028] in, This represents the location of the i-th bird's nest in generation t. As the lower bound of the search space, For the middle value of the search space, This is the upper limit of the search space. It is a Halton sequence.

[0029] This invention uses Halton sequences to improve the generation of the initial population, employs an adaptive step size factor in the Levy flight, and uses an adaptive discovery probability in the biased random walk phase, thereby improving the convergence speed of the algorithm.

[0030] Step 2.2: Update the nest position via Levy flight; the specific process is as follows: During Levy flight, an improved adaptive step size factor is used. The nest position is updated according to the update formula, and its corresponding fitness value is calculated and compared with the fitness value corresponding to the nest position of the previous generation. Nest positions with better fitness are retained, thus realizing the nest position update.

[0031] The formula for updating the location of the Bird's Nest is as follows:

[0032] in, This is the location of the next generation Bird's Nest. It is the step size adjustment factor. This represents point-to-point product. It is a random search path for Levi's flight. It is an improved adaptive step size factor. , .

[0033] Step 2.3: Generate random numbers Compare it with the improved adaptive discovery probability If a comparison is made, If the nest position is positive, the nest position is updated; otherwise, it remains unchanged. After the position update is completed, the fitness value of each nest is calculated and compared with the previous fitness value. The better nest position is retained. The process of updating the bird's nest location using random walk is as follows:

[0034] in, The updated location of the Bird's Nest. Location of the abandoned bird's nest and Improved adaptive discovery probability for two randomly selected bird nest locations , , , For the first t The fitness value of the best individual in the population. For the first t The first generation of the population i The fitness value of each individual.

[0035] Step 2.4: Determine if the maximum number of iterations has been reached. If the optimal solution has been achieved, output the global optimal solution, i.e., the optimal resource block allocation scheme; otherwise, use the location of the bird's nest in this generation as the initial bird's nest for the next generation, and return to step 2.2 for the next iteration. Step 2.5: Calculate the rate allocated to users based on the allocated resource blocks, and determine whether the rate requirements of all users have been met. If they have been met, output the resource allocation scheme in Step 2.4; if not, determine whether there are still VLC resource blocks available for allocation to users whose requirements have not been met. If so, continue with the next resource allocation according to Steps 2.1 to 2.5. If there are no allocated VLC resource blocks, allocate WiFi resource blocks to users.

[0036] In this invention, VLC resources are allocated first when allocating resources. If the user's needs are still not met after all available VLC resources have been allocated, WiFi resources are then allocated as a supplement.

[0037] This invention is applicable to resource allocation in complex indoor scenarios, such as different lamp layouts, lamp sizes, lamp access capabilities, and user needs, and can provide the best allocation scheme in different scenarios.

[0038] Example 2 Simulation scene settings: The layout of the lights is as follows Figure 4 The diagrams show three layouts: honeycomb, square, and uniform. Users are randomly distributed within a room measuring 8m × 8m × 3m.

[0039] Figure 5 These are schematic diagrams illustrating the lighting performance of three different layouts:

[0040] Different lamp layouts can affect VLC performance. Figure 5 The table above shows the illuminance distribution and specific illuminance values ​​under the three layout settings, and the indoor lighting needs can be met under all three conditions.

[0041] Communication performance: System throughput: Figure 7(a) shows the relationship between system throughput and the number of UEs under three different lamp layouts. It can be seen from the figure that system throughput increases with the number of UEs. This is because as the number of UEs increases, the data transmission demand in the network also increases, and each AP is allocated more resource blocks. Compared with the greedy algorithm and the PSO algorithm, the ICS algorithm exhibits better performance under different lamp layouts.

[0042] User satisfaction: Figure 7(b) shows the relationship between user satisfaction and the number of UEs under three different lamp layouts. The figure reveals that UE satisfaction decreases as the number of UEs increases. This is because an increase in the number of UEs leads to an increase in overall data rate demand, while the access capacity of each AP is limited. Compared to the greedy algorithm and the PSO algorithm, the ICS algorithm exhibits better performance under different lamp layouts.

[0043] Fairness of service: As shown in Figure 7(c), this figure illustrates the relationship between service fairness and the number of UEs under three different light layouts. As the number of UEs increases, the demand for resources also increases, leading to more intense competition. The figure shows that SFI decreases as the number of UEs increases. Compared to other algorithms, the ICS algorithm exhibits better performance under different light layouts.

[0044] Verification of the algorithm's convergence speed: The following is a verification performed using a cellular layout with 15 users.

[0045] Compare the convergence of the ICS, CS, and PSO algorithms, such as... Figure 6 As shown, from Figure 5 As can be seen, the convergence speed and accuracy of the ICS and CS algorithms are significantly higher than those of the PSO algorithm. Furthermore, the ICS algorithm converges faster than the CS algorithm. In addition, the initial solution generated by the ICS algorithm is of high quality, thus demonstrating that the optimization method used in this invention has a significantly faster convergence speed.

[0046] Example 3 Comparison between the traditional cuckoo algorithm and the improved cuckoo algorithm in this invention:

[0047] Compared to the traditional cuckoo algorithm, the improved cuckoo algorithm shows little change in time complexity. Although the time complexity of the population initialization operation increases after introducing the Halton sequence... However, the time complexity of position updates during the iteration process is the same for both algorithms. Therefore, the time complexity of the entire algorithm is basically the same, while the improved cuckoo algorithm in this invention can produce more stable results.

[0048] Figures 7(a), 7(b), and 7(c) show the simulation results of throughput, user satisfaction, and service fairness under three different lamp layouts. It can be seen that the improved cuckoo algorithm can achieve the best communication performance under different lamp layouts for resource allocation.

[0049] Through the above methods, this invention provides an indoor heterogeneous network resource allocation method based on an improved cuckoo algorithm. This method combines the advantages of both VLC and WiFi communication technologies to construct an indoor VLC heterogeneous network, providing complete indoor wireless network coverage and high system throughput. Addressing the problems of the existing cuckoo algorithm, an improved version is adopted, refining the initial population settings, Lévy flight, and biased random walks. This effectively increases the algorithm's convergence speed and solves problems such as limited applicability, high algorithm complexity, and susceptibility to local optima in existing solutions. Furthermore, to address the difficulties of the aforementioned resource allocation algorithms in providing good allocation schemes and high complexity under complex scenarios, high user density, and high data rate requirements, an improved cuckoo algorithm with fewer parameters and faster convergence speed is used. This algorithm is well-suited for scenarios with different light layouts, access capabilities, and user needs, providing the optimal allocation scheme for each scenario, maximizing system throughput, and thus improving the performance and reliability of the indoor VLC heterogeneous network.

Claims

1. A method for resource allocation in indoor heterogeneous networks based on an improved cuckoo algorithm, characterized in that, The specific steps are as follows: Step 1: Construct an indoor VLC heterogeneous network system model based on the network access points between users; Step 2: Within the indoor VLC heterogeneous network system model, the improved Cuckoo algorithm is used to allocate VLC resource blocks layer by layer; The VLC heterogeneous network system includes both VLC and WiFi communication networks; Step 2 is as follows: Step 2.1: Number all resource blocks at all access points uniformly, treat the resource block allocation scheme of each layer of users as a bird's nest, set the fitness function to a function related to the system throughput of the VLC network, and initialize the population; Step 2.2: Update the nest location via Levi's flight; Step 2.3: Generate random numbers Compare it with the improved adaptive discovery probability If a comparison is made, If the location of the Bird's Nest is updated, then the location will be changed; otherwise, it will remain unchanged. After completing the location update, calculate the fitness value of each nest, compare it with the previous fitness value, and retain the better nest location; Step 2.4: Determine if the maximum number of iterations has been reached. If the optimal solution has been achieved, output the global optimal solution, i.e., the optimal resource block allocation scheme; otherwise, use the location of the bird's nest in this generation as the initial bird's nest for the next generation, and return to step 2.2 for the next iteration. Step 2.5: Calculate the rate allocated to users based on the allocated resource blocks, and determine whether the rate requirements of all users have been met. If they have been met, output the resource allocation scheme in Step 2.

4. If not, determine whether there are still VLC resource blocks available for allocation to users whose needs are not met. If so, continue with the next resource allocation according to steps 2.1 to 2.

5. If there are no VLC resource blocks available for allocation, allocate WiFi resource blocks to the users. The population initialization described in step 2.1 specifically involves: Set population size Search space range Maximum number of iterations The optimal fitness value is The low-difference Halton sequence from the quasi-Monte Carlo method is introduced to initialize the location of the bird's nest, generating... The system identifies the locations of each nest and calculates its corresponding fitness value. Nests are then sorted by fitness value, and the nests with the best fitness values ​​are selected. One bird nest was used as the initial population; Step 2.2 specifically involves: employing an improved adaptive step size factor during Levi's flight. The nest position is updated according to the update formula, and its corresponding fitness value is calculated and compared with the fitness value corresponding to the nest position of the previous generation. Nest positions with better fitness are retained to realize nest position update. The location of the bird's nest mentioned in step 2.3 has been updated to: in, The updated location of the Bird's Nest. Location of the abandoned bird's nest and Improved adaptive discovery probability for two randomly selected bird nest locations , , , For the first t The fitness value of the best individual in the population. For the first t The first generation of the population i The fitness value of each individual.

2. The indoor heterogeneous network resource allocation method based on the improved cuckoo algorithm according to claim 1, characterized in that, Step 1 is as follows: The user set is User access point set Among them, VLC access point set WiFi access point set .

3. The indoor heterogeneous network resource allocation method based on the improved cuckoo algorithm according to claim 1, characterized in that, The fitness function is calculated using the following formula: The formula for calculating the throughput of a VLC network system is as follows: Location of the i-th bird's nest The fitness function is expressed as: In the formula, Indicates user and VLC Connected, VLC arrive The signal-to-interference-plus-noise ratio (SINR) is given by e = 2.718, where e is the Euler number.

4. The indoor heterogeneous network resource allocation method based on the improved cuckoo algorithm according to claim 1, characterized in that, The formula for initializing the bird's nest location using the low-difference Halton sequence introduced in the quasi-Monte Carlo method is as follows: in, This represents the location of the i-th bird's nest in generation t. As the lower bound of the search space, For the middle value of the search space, This is the upper limit of the search space. It is a Halton sequence.

5. The indoor heterogeneous network resource allocation method based on the improved cuckoo algorithm according to claim 1, characterized in that, The formula for updating the Bird's Nest location is as follows: in, This refers to the location of the next generation of the Bird's Nest stadium, where α is the step size adjustment factor. This represents point-to-point product. It is a random search path for Levi's flight. It is an improved adaptive step size factor. , .

Images