An Adaptive Load Balancing Algorithm Based On Predicting The Future Load In Heterogeneous Wireless Networks

With the rapid development of wireless communication technology, the integration of heterogeneous networks has become the trend of the development of future wireless network. Different wireless access technologies gives users a variety of service experience and faster data transfer speeds. Since different access networks have different characteristics, such as the characteristics of some networks’ covering range are wide, and some networks are real-time system, and some networks’ data transfer rate is high, different characteristics of these networks have good complementary. However, because of the diversification of wireless access networks and they are not compatible with each other,which makes the mobile users and operators facing many problems.The Integration of heterogeneous wireless network system is very complex, and its core is the joint radio resource management of heterogeneous networks, and to deal with the joint management of radio resources, many key technologies must be addressed, such as spectrum sharing, joint admission control, network selection and load balancing, and many other key technologies. Load balancing technology can able to solving the uneven distribution of heterogeneous wireless network traffic, it can improving system capacity, balancing the traffic load, and effectively improve the radio resource utilization of heterogeneous networks.Currently, there are a lot of research on load balancing algorithms in the heterogeneous networks. however, most of these algorithms take advantage of the current network load value as network selection criteria, these algorithms can not well balanced the networks’ load to reduce the frequent handover times, also these algorithms can not reduce the times of trigger load balancing algorithm to reduce the waste of system wireless resources.In response to these problems, in order to avoid the frequent network selection caused by the load fluctuation of networks, an adaptive load balancing algorithm with the future load predicting is proposed for wireless heterogeneous networks. The probabilities of networks’ future load states are predicted by the Markov-Chain, then the predicted probabilities are mapped into load tread values by the load trend function. The trend values are used for network selection and also used for the adaptive adjustment of the trigger threshold for the network selection. The simulation results show that the proposed algorithm can reduce the access blocking probability and the times of vertical handover.