An Intelligent Call Admission Control For Heterogeneous Mobile Networks

In recent years, technologies of communication have been developed rapidly. New wireless communication system emerges one after another, and all of these systems comprise the heterogeneous networks. It is well accepted by both the industrials and the academics that the next generation network (NGN) should be an open heterogeneous wireless network (HWN) environment, which can intelligently integrate the existing and the future radio access systems. In this dissertation, under the environment of HWN, the problems of radio resource management are intensively studied. The contents of this work are listed as follows:CAC algorithm in the HWN is studied in this work. A BP_CAC algorithm is proposed for an integrated cellular/WLAN network, in which the neural network with back propagation (BP) algorithm is used to adjust the amount of due resource. The function of BP algorithm is to intelligently predict the due resource in the next instant, and to make decision whether to receive the call or to reject the call according to the output. On the basis of BP_CAC algorithm, the problem of bandwidth degradation is studied next. Be combined with BP_CAC algorithm, a BPBDA_CAC algorithm is composed. The function of BDA is to decrease the allocated bandwidth of on-going calls for new coming calls when there is no enough bandwidth to accept the new coming calls. This policy can improve the serviceability of system. And this is the second key point that we studied. The function of BUA is to improve the quality of service (QoS) when there is enough resource available. On the basis of BPBDA_CAC algorithm, a BPBM_CAC algorithm is composed in consideration of BUA. The function of BUA is to execute bandwidth upgrade when the loading is alleviated, in order to provide better QoS.In simulation with MATLAB, the conclusion is worked out. First, the BP_CAC algorithm makes better performance compared with Fixed_CAC and Dynamic_CAC. It decreases NCBP to a great extent, at the same time it maintains HCDP in an acceptable level. It also increase resource utilization rate and improve the grade of system (GoS). Second, the BPBDA_CAC algorithm outperforms in every aspect than the BP_CAC algorithm, such as in NCBP, HCDP, GoS. Last, the BPBM_CAC algorithm outperforms than the BPBDA_CAC algorithm in improving GoS. For this policy executing bandwidth upgrade for on-going calls, and improving the calls' QoS level, the BPBM_CAC algorithm improves the satisfaction of users.