Research On Accrss Selection And Device Management In Cognitive Wireless Networks

In the scenario of heterogeneous networks, different wireless access technologies provide different performance indicators such as transmission range, transmission rate, and QoS. Terminals in the networks can use cognitive technology to select and access the best network, to achieve a high utilization of the network resources. Cognitive technology is one of hot spots in the research fields of "Green Communications.Referring to the principles and technologies of Artificial Neutral Networks (ANN), this thesis examines the application of neutral networks in cognitive wireless networks. Firstly the basic principles, the current statues of research, and key technologies of the cognitive wireless networks and ANN were studied in this thesis; secondly, the conventional RSS handover decision algorithm is analyzed and a mobile terminal moving speed-based wireless access network selection mechanism is proposed; finally, this thesis uses the probabilistic neutral network for mathematical modeling, and presents the method of sorting the target access networks of the terminal based on its speed of mobility. In this way, the terminals of the same class could always select one kind of access network, so that the degradation of QoS caused by the frequent handover could be avoided.Besides, the application of cognitive technology in the area of "Green Communications" is also examined, and a network flow prediction-based dynamic device energy consumption management technology was built. Based on the analysis of network flow modeling, this thesis puts forward a dynamic management method for the device energy consumption by using the BP neutral network to model and simulate the network flow to achieve the goal of "Green Communications".