Network Selection In Cognitive-Heterogeneous Network

Heterogeneous network is one of the characteristics of next-generation mobile communication systems. The future mobile communication networks will include a variety of wireless access technologies, and which will offer different coverage, throughput and QoS and so on. The complement of these technologies in performance makes heterogeneous network convergence an inevitable trend. In the heterogeneous network environment, the network access selection is to be an important research topic.The cognitive network is self-awareness, adaptive, self-configuring, self-awareness and self-learning, so each communication node in the network is no longer a totally controlled node, but with varying degrees initiative. So the function of traditional network will be optimized.Furthermore, the cognitive network is self-awareness, adaptive, self-configuring, self-awareness, self-learning and other intelligent features, so each communication node in the network is no longer a totally controlled node, but with varying degrees initiative,. Each node can take the initiative to determine communication behaviors independently according to the surrounding environment and network status, which makes the behavioral model of the cognitive network will largely different from the traditional networks. So the function of traditional network will be optimized.So we first introduce the present development in the key technologies in cognitive network, heterogeneous network and cognitive-heterogeneous network. Then investigated several network selection algorithms,such as AHP, TOPSIS and fuzzy gray correlation method, in which the application of fuzzy logic is to determine some measure of uncertainty and non-uniform parameter values; the application of AHP is to determine the weight of each attribute based on different service request; for multi-attribute decision-making algorithm that often appears sorted disorder issue, the improved TOPSIS is proposed; also a fuzzy gray correlation method is proposed for different occasions and service requestsIn addition, analyzed the handover algorithm and improved the algorithms from different parts. The SP algorithm based on Kalman filtering is to decrease the handover delay and the VA algorithm is to improve the performance.The two algorithms were verified using MATLAB. However, the network environment is changing always and the best set of parameters will change, so the final paper studied on the relationship between the handover failure ratio, ping-pong handover ratio and the handover parameters. The optimization model and optimization algorithm was proposed and designed based on the results, which help to reduce the complexity of human modification of handover parameters, making the networks more intelligent.