Research On Resource Allocation Of Hybrid RF/FSO Mesh Networks

Wireless mesh network is one of the key technologies to construct the next generation broadband wireless access network.Due to its advantages of good scalability,wide coverage,and convenient deployment,it has been widely used in many fields.At present,wireless mesh networks are mainly based on radio frequency(RF),the capacity and performance of the RF-based wireless mesh network are limited,owing to the less available bandwidth and higher common channel interference of microwave and the limited transmission distance and coverage of millimeter-wave.Combined with the advantages of wireless mesh networks and FSO systems,hybrid RF/free space optical-communication(FSO)mesh networks have wider coverage,higher scalability,and larger network capacity while ensuring easy deployment and flexible networking.But with the increase of the node number,the resource of hybrid RF/FSO mesh network is rapidly consumed,and the network performance degrades.Resource allocation can improve the resource utilization rate,capacity,and performance of the network,which is one of the key contents to be studied urgently in hybrid RF/FSO mesh networks.Resource allocation solutions of RF-based wireless mesh networks are not completely suitable for solving the resource allocation problem of hybrid RF/FSO mesh networks.At present,there are few research works on resource allocation in hybrid RF/FSO mesh networks,and the following problems exist.(1)Without considering the joint optimization of multiple network resources,it is difficult to obtain the optimal solution of network resource allocation.(2)The effects of RF and FSO link attenuation on network resource allocation under various weather conditions are not fully considered,so the resource allocation solutions are not suitable for practical network deployment.(3)The effects of delay and network deployment cost on network resource allocation are not quantified,so it is difficult to provide reference for practical network resource allocation.In view of the shortcomings of the existing research work,this dissertation focuses on the resource allocation problem of hybrid RF/FSO mesh network,and it addresses the resource allocation problems of hybrid RF/FSO single-radio single-channel mesh networks,hybrid RF/FSO multi-radio multi-channel mesh network,and hybrid RF/FSO mesh fronthaul network.The main innovations are summarized as follows.(1)Considering RF and FSO link fading,a joint resource allocation scheme of the hybrid RF/FSO single-radio single-channel mesh network is proposed.FSO links are deployed in the backbone of infrastructure wireless mesh networks,and a hybrid RF/FSO single-radio single-channel mesh network is constructed.In consideration of RF and FSO link availability,with the goal of maximizing the throughput of the hybrid RF/FSO single-radio single-channel mesh network,the joint resource allocation problem in terms of power control,transmission slot assignment,topology control,and FSO link allocation is formulated as a mixed-integer linear programming(MILP).Gurobi and IBM CPLEX are used to solve the formulated MILP,and the optimal solution of the joint resource allocation problem is obtained.(2)Considering the effects of weather conditions,a joint resource allocation scheme of the hybrid RF/FSO multi-radio multi-channel mesh network is proposed.The bottleneck backbone links of the multi-radio multi-channel wireless mesh network are upgraded by FSO links,and a hybrid RF/FSO multi-radio multi-channel mesh network is constructed.Considering the weather effects on FSO link availability and the fading nature of RF links,the joint resource allocation problem of the hybrid RF/FSO multi-radio multi-channel mesh network is addressed.With the objective of maximizing the network throughput while guaranteeing fairness,and under the condition of satisfying the delay requirement,the joint resource allocation problem in terms of interface assignment,topology control,FSO link allocation,channel allocation,and routing is formulated as a two-stage optimization problem,and an improved iterated local search algorithm is proposed to solve the optimization problem.In the proposed algorithm,a heuristic method is adopted to rapidly obtain the initial solutions,an appropriate perturbation is adopted to improve the searching ability,and a kind of simulated annealing schedule is used as the acceptance criterion to filter the low-quality solutions.The improved iterated local search algorithm can obtain the near-optimal joint resource allocation solutions within polynomial time,and it can be applied in practice.(3)According to the ?-?-?-? and double generalized Gamma channel models,under sunny,rainy,and foggy weather conditions,the link availability of millimeter-wave and the link availability of FSO are deduced.On this basis,a joint resource allocation scheme of the hybrid RF/FSO mesh fronthaul network is proposed.Based on millimeter-wave and wireless mesh network technologies,a hybrid RF/FSO mesh fronthaul network is constructed.Aiming at the coexistence scenarios of massive machine type communications(m MTC),ultra-reliable and low latency communications(URLLC),and enhanced mobile broadband(e MBB),with the objective of maximizing the network throughput in a fair manner while reducing deployment cost,under the condition of satisfying the delay and reliability requirement,the resource allocation problem oriented to network slicing of hybrid RF/FSO mesh fronthaul network is formulated as an MILP.Based on Q-learning and krill herd algorithm,and using Q-learning to train the parameters of the krill herd algorithm adaptively,a resource allocation algorithm oriented to network slicing is proposed,which can respond to m MTC,URLLC,and e MBB slice demands intelligently and obtain the resource allocation solutions for hybrid RF/FSO mesh fronthaul networks.In this dissertation,we deeply research on resource allocation problems of hybrid RF/FSO mesh networks,and propose resource allocation schemes for hybrid RF/FSO single-radio single-channel mesh networks,hybrid RF/FSO multi-radio multi-channel mesh networks,and hybrid RF/FSO mesh fronthaul networks.These schemes conquer the limitations of current research,help improve the performance of hybrid RF/FSO mesh networks,and provide guidance for resource allocation of hybrid RF/FSO mesh networks.