Research On Vertical Handoff Algorithm For Heterogeneous Wireless Networks

The development of the wireless communication technology has dramatically promoted the emergence of new radio access technologies.Different wireless networks have different advantages on providing services to users,and they could not replace each other.It is hard to satisfy the increasingly diversified requirements of users,depending on a certain wireless network.The future wireless networks would evolve towards the heterogeneous wireless networks.As the main technology of mobile management in heterogeneous wireless networks,vertical handoff is the process,whereby users could switch between different wireless networks to satisfy their requirements in heterogeneous wireless networks.And many scholars have paid attention to and researched the vertical handoff.So far,many achievements about the vertical handoff have been made.However,there are some problems needing further study,mainly including the following two points.Firstly,prior studies have failed to consider the benefit of the system in the long term,when making handoff decisions.Secondly,the existing algorithms could not be applied to the scene where the users move fast due to their high complexity or less flexibility.In order to solve the problem of vertical handoff in heterogeneous wireless networks that the existing algorithms do not consider the benefit of the system in the long term,the multi-armed bandit model(MAB)based vertical handoff algorithm is proposed in this thesis.Firstly,the problem of vertical handoff is formulated as the multi-armed bandit problem.Each network is one arm of the MAB.At each decision epoch,a user only chooses one network to access.The selected network would provide reward,the network does not be selected would not provide reward,and the state of the selected network would evolve according to the homogeneous markov chain.Then,considering the diversity of users' services,the users' services are divided into real-time services and non-real-time services.As different types of services have different requirements about the network parameters,combined weights of network parameters are assigned by the analytic hierarchy process(AHP)and the entropy method.The reward functions,which are on the basis of network parameters and the combined weights,are constructed for the real-time services and non-real-time services respectively.Finally,computing the Gittins indices of the networks in different states.Based on the Gittins theorem,the handoff strategy for maximizing the benefit of the system is obtained.The simulation results indicate that the proposed algorithm could select the optimal networks for users with different types of services,reduce blocking probability of the system,thereby improving performance of the system.In order to solve the problem that the existing vertical handoff algorithms could not be applied to the scene where the users move fast,the bayesian decision and the multi-criteria decision making(MCDM)method based vertical handoff algorithm is proposed in this thesis.Firstly,based on the users' movement tendency and bayesian decision,the handoff target network set would be constructed and refreshed.Secondly,the weights of parameters,influencing the handoff decision,are assigned by the fuzzy analytic hierarchy process(FAHP).Then,considering the diversity of terminal services,the reward functions are constructed for different services respectively.Finally,based on the maximum benefit rule,the optimal network for a user to access is the network with the highest benefit.The simulation results show that the proposed algorithm could satisfy the requirements of the high-speed users about the delay,and decrease the number of handovers.