Research On Vertical Handoff Algorithm For Multi-services In Heterogeneous Wireless Networks

With the rapid development of communication technology,many wireless access technologies with different transmission power,coverage and working mechanisms have emerged,such as cellular networks,Wireless Local Area Networks(WLAN),Worldwide Interoperability for Microwave Access(WiMAX)and Wireless Sensor Networks(WSN).With the drastic growth of smart devices and the emergence of new emerging multimedia applications worldwide,it is hard to meet the increasingly diverse needs of users,depending on a certain wireless network,heterogeneous wireless network that supports multi-network convergence and seamless roaming will become the inevitable trend of wireless network development in the future.As the key technology of mobility management in heterogeneous network,vertical handoff technology is used to ensure the continuity of the session when moving across heterogeneous networks,it has become one of the key technologies in heterogeneous wireless network and has important research significance.Recently,lots of effort has been made to vertical handoff algorithm in heterogeneous wireless network.However,with the continuous growth in user demand for various wireless services,traditional voice and SMS services have been unable to meet the increasingly diversified needs of users.Therefore,how to select the optimal networks for users with different types of services will be the research focus of vertical handoff technology for heterogeneous wireless networks in the future.However,there are relatively few researches on vertical handoff algorithm under multiple services circumstances currently,and traditional vertical handoff algorithms often rely on the current network status only and fail to consider the benefit of the system in the long term when making handoff decisions.In this regard,considering multi-service heterogeneous wireless network scenarios,two improved vertical handoff algorithms are proposed in this thesis to improve QoS of users.In order to solve the problem of vertical handoff in heterogeneous wireless networks that the existing algorithms do not consider the potential future network states and the benefit of the system in the long term,the Markov decision model based vertical handoff algorithm under multiple services circumstances is proposed in this thesis.Firstly,the problem of vertical handoff is formulated as the Markov decision process.Then,considering the diversity of users' services,wireless network servicesare divided into session services,interactive services,streaming services and background services.Considering the four QoS parameters of jitter,delay,packet loss rate and bandwidth,a weighted summation method is used to construct a QoS benefit function,while considering the impact of terminal movement speed on communication quality,a cost function is constructed according to user movement speed,and get the reward function of users with different types of services after executing the action according to the QoS benefit function and cost function.Finally,policy iteration algorithm is used to solve the Markov decision model,thereby obtaining a personalized vertical handover strategy that meets the needs of users with different types of services.The simulation results indicate that the proposed algorithm could make reasonable network selection for users with different types of services,and reduce the blocking rate and average number of handovers to a certain degree.However,in the actual vertical handover scenario,the state transition probability of the network is not easy to obtain,and the computational complexity of the handoff algorithm will increase as the number of candidate networks increases.Therefore,this thesis proposes a vertical handoff algorithm based on improved Q-learning.The algorithm divides the vertical handover process into two parts: handoff admission control and handoff decision.In the handoff admission control part,a handoff admission control algorithm based on Received Signal Strength(RSS)is proposed.The algorithm considers the RSS of the network,and sets a dynamic RSS threshold based on the current service's minimum requirements for communication quality and the user's movement speed.Before the handoff decision is made,networks that cannot meet the current service requirements are deleted from the candidate network set to reduce the computational complexity of the handoff decision algorithm.In the handoff decision part,a handoff decision algorithm based on improved Q-learning is proposed.The algorithm considers the diversity of users' services and multiple factors that affect network QoS.It improves the reward function in Q-learning using the Analytic Hierarchy Process(AHP)method,and based on the improved Q-learning algorithm to make handoff decisions,thus selecting the optimal network from the candidate network set.The simulation results indicate that the proposed algorithm could effectively reduce the blocking rate and avoid unnecessary handovers.