Investigation On Multi-Radio Access Selection In Heterogeneous Wireless Networks

In the near future, multitude of wireless communication networks based on a variety of radio access technologies (RATs) and standards will emerge and coexist. The availability of multiple access alternatives offers the capability of increasing the overall transmission capacity, providing better service quality and reducing the deployment costs for wireless access. These potential gains can be named multi-access gain. Multi-radio access selection (MRAS) that decide an appropriate radio access technology for a specific service is proved to be an efficient method to exploit the multi-radio access gain, which comes from two aspects: multi-radio access diversity (MRAD) and multi-radio access combining (MRAC). The thesis will investigate how to design optimum MRAS strategies to exploit the multi-radio access gain in heterogeneous wireless networks.Firstly, a description of the next generation of wireless communication network is given. As one of the trend of next generation wireless system, the key inter-working technologies of heterogeneous networks are analysed from the following point of view: core network, access network and terminal respectively. Then, the thesis emphasises on the inter-working strategy at the level of access network, and proposes two inter-working network architectures based on the concepts of generic link layer (GLL) and multi-radio resource management (MRRM).Secondly, the thesis divides the MRAS into two types: slow MRAS and fast MRAS, in terms of the time scale of implementation. On one hand, the access selection function will be triggered by some incidents including new service generation, inter-system handover or the degradation of QoS, where MRAS operates in the order of seconds and upwards, so it is named Slow MRAS. On the other hand, access selection can operate in the order of milliseconds to be able to react to rapid channel quality variations (fast fading), which is implemented by scheduling packets over multiple radio accesses and exploit more multi-radio access gain, so it is named fast MRAS. Based on the thorough analysis of implementation and the state of art of MRAS, the thesis presents the now-existed problems and challenges in the research on MRAS.For slow MRAS, the thesis first propose a theoretic analysis model for load balancing with multi-radio access selection by using multi-dimensions Markov chain. The analytical results reveal how the variation of main parameters affect the performance of load balancing, which will guide the designing of access selection algorithm. When considering the lack of MRAS algorithm for wireless packet network, the thesis gives a method to evaluate the required resource of service and network load based on the packet-level information, then propose a load balancing-based MRAS algorithm with consideration of various qualities of packet service. Moreover, MRAS actually depends on multiple deciding factors, which belong to different properties and are hard to be compared directly, including status of link, system capacity, network load, QoS providence, user preference, security and availability, etc. To settle this problem, the thesis proposes a multiple factors determined MRAS algorithm based on the cost function and analytic hierarchy process.For fast MRAS, both the selective and parallel multi-radio transmissions are investigated to exploit the gain of multi-radio transmission diversity and multi-radio transmission combining. Selective multi-radio transmission can be converted into the problem of packet scheduling in the condition of multi-user and multi-radio access, and can be implemented in two steps: multi-user scheduling and multi-radio accesses allocation. In the thesis, a novel selective multi-radio transmission algorithm is proposed, in which the two steps are jointly designed to achieve the optimum performance. Moreover, Parallel multi-radio transmission is used to increase the reliability of transmission in wireless environment, but at the cost of consuming more radio resources. In the thesis, the forward error correction (FEC) strategy and multi-radio transmission are combined to solve this problem. Two novel bit level FEC-based parallel multi-radio transmission algorithms: IPMRTD and CPMRTD are proposed. Furthermore, considering the complexity of implementation, the thesis also proposes a packet level FEC-based parallel multi-radio transmission algorithm, and discusses its applications in cooperative communications.Finally, the works in the thesis are summarized and future works are introduced.