Research On Network Access Decision Mechanism In Heterogeneous Wireless Networks

With rapid development of communication technologies, new wireless networks are continuously developed to provide better service quality. The next-generation wireless network will be the integration of heterogeneous wireless networks (HWNs).In this context, it is necessary to rationally use and manage the network resources. Network selection algorithm is one of the hot issues in radio resource management, and proper network decision algorithm will choose the best access network for users.The thesis firstly researches on single network selection algorithm with multiple attributes and proposes multi-attribute algorithms based on ANP (Analytical Networks Process) and based on AHP (Analytic Hierarchy Process) and synergetic theory respectively. Then we further research on network selection algorithms in multi-radio network scenario and propose network access selection algorithms based on multi-radio parallel transmission and based on joint utility optimization respectively. The thesis mainly researches on the following parts:(1) ANP method is applied to network selection scenario and the thesis proposes a new network selection algorithm based on ANP. The algorithm considers network attributes and target networks as elements of the net structure and classifies them. The judgment matrices within group and between groups are constructed according to the difference of elements. Then the unweighted matrix is composed by weight vectors which are calculated by judgment matrices within groups. In the same way, the weighted matrix is composed by weight vectors which are calculated by judgment matrices between groups. Then the weighted super-matrix is calculated and the limit matrix can be obtained. At last, the attribute weight and the network weight are calculated respectively and the target network is chosen according to the cost function value using the additive weighted method. The simulation results demonstrate that the proposed algorithm outperforms EW and TOPSIS algorithms in terms of packet delay, packet jitter, packet loss and the handoff number, so as to improve the service quality for users.(2) The thesis proposes a multiple attribute network selection algorithm based on AHP and synergetic theory. The algorithm applies synergetics to network selection, considering the candidate network as a compound system composed of multiple attribute subsystems, and combines the subsystem degree order with AHP weight to obtain entropy of the compound system, which means synergy degree between subsystems. The greater the synergy degree is, the better the network is. The algorithm takes not only the coordination of objective attributes but also different requirements into consideration, which ensures that users select the network with good performance. The simulation results reveal that the proposed algorithm can effectively reduce the handoff number and provide subscribers with satisfactory quality of service (QoS) according to different services.(3) A network access selection algorithm based on multi-radio parallel transmission is proposed in this thesis. The algorithm firstly determines all available wireless networks according to received signal power from base stations, and considers every subset of these networks as one network selection scheme. Then we calculate the aggregation attribute of every scheme to confirm the candidate network schemes which are limited by throughput and power consumption threshold. The matrix of multiple attribute decision making is built and the optimal scheme to access is determined based on cost function value of every network scheme by using TOPSIS algorithm with user preference. The distance between the optimal scheme and the ideal scheme is the nearest. Compared with single-access network algorithm, the simulation results show that the proposed multi-access network algorithm can make the best of idle network resources and obviously improve user quality of service, improve user throughput, reduce power consumption per bit and price cost per bit, and keep network load balanced simultaneously.(4) The thesis proposes a method of multi-access selection algorithm based on joint utility optimization in HWNs. The algorithm firstly adopts exponential smoothing method (ESM) to get measured values of received signal strength (RSS), utilizes the benefits and costs from user when accessing networks to establish network benefit and cost model, and then obtain network joint utility optimization model under the constraints of bandwidth and number of networks. Lagrange multiplier and dual optimization methods are used to solve the association matrix between user and network, namely the best scheme for users to access the networks. The simulation results show that the proposed algorithm can optimize joint utility as user access the networks, improve user throughput, effectively reduce vertical handoff number, and meet users’ service requirements.