Research On Radio Allocation And Handover Mechanism In Wireless Relay Systems

The aspiration of communicating anywhere at any time boosts the evolution of wireless networks. The relay based wireless networks have solved many problems that exist in the traditional wireless networks, e.g. the lack of expansibility and robustness, thus the concept of relaying is a promising solution for the challenging throughput and high data rate coverage requirements of future wireless cellular networks. The dissertation focuses on wireless relay systems, taking the resource management as the key, and gives a thorough and comprehensive study on the resource allocation strategy and handover mechanism.Based on the relay link, the power allocation strategies for decode-and-forward (DF) and amplify-and-forward (AF) relaying method are respectively studied. The dissertation firstly deals with DF relaying scheme in multihop wireless networks. Through making full use of end-to-end system outage probability requirement in multihop route, an efficient optimal power allocation scheme that minimizing the aggregate power consumption is proposed. Theory analysis and numerical simulation indicate that compared with the traditional power allocation scheme which considering the each intermediate node's power constraint, the proposed scheme has superior performance in view of minimizing the power consumption on the condition of the same system outage guarantee. Based on the MIMO-OFDM AF relay system, from information theory point of view the channel capacity is analyzed and a new power allocation scheme is developed under a certain power constraint at the relay to achieve the maximum channel capacity. The proposed power allocation scheme can be computed with low computational complexity and achieve a higher system performance.In view of the frequency selective fading of wireless channel in broadband systems, Orthogonal Frequency Division Multiplexing (OFDM) will be widely adopted in the future communication systems. The problem of resource optimization becomes more complicated in OFDMA based relay systems. As dedicated relays are deployed in the traditional OFDMA system, two accompanying issues are envisioned, i.e., 1) the routing issue such that how to pair relay station (RS) with the mobile station (MS) and 2) the subcarrier allocation issue, such that how to assign subcarriers to these pairs. The dissertation addresses the joint routing and subcarrier allocation problem for two-hop OFDMA-based relay networks. An optimization problem that maximizes system capacity while guaranteeing minimum resource for each user is formulated, and two joint routing and subcarrier allocation strategies are proposed: the optimal solution and the suboptimal solution respectively.1) Optimal solution: The original optimization problem is a linear integer programming problem. A gereneral integer programming has been proved to be NP-complete. Fortunately, the original optimization problem can be tackled in virtue of gragh theory: we generate a new network gragh in a directed graph and then transform the original optimization problem into the equivalent "linear optimal distribution" problem in gragh theory. The original optimization problem can be tackled in virtue of the graph-theoretic approach. The proposed scheme gains the optimal solution with relatively low computational complexity. The process of searching circles in the generated network gragh includes the path selection and subcarrier allocation procedure, thus the joint routing and subcarrier allocation is realized.2) Suboptimal solution: From geometry point of view, the original optimization problem can be abstracted to a three-dimensional space (path, subcarrier and user). The resource allocation problem in a three-dimensional space is mapped into a two-dimensional plane (subcarrier and user). The mapping process deletes the excrescent elments, thus the proposed scheme can reduce the computational complexity significantly. The proposed scheme has some heuristic meaning and gains the suboptimal solution. In addition, the mapping process includes the path selection, so the joint routing and subcarrier allocation is realized.The dissertation addresses the problem of the packet scheduling design for OFDMA based relay networks with different traffic classes (delay-sensitive traffic and delay-insensitive traffic). In order to improve the system throughput and decrease the packet loss rate, three centralized packet scheduling schemes by adjusting transmission rate dynamically in response to the variation of channel quality and packet delay requirement are proposed. The scheduling strategies adopt the separated routing and resource allocation scheme and consider the constraints of user transmission, traffic quality of service (QoS) and user fairness. The proposed schemes exploit the independent subchannel fading of different users and acquire multiuser diversity gain, and satisfy the traffic QoS requirement. In addition, numerical results show that the OFDMA relay network outperforms the traditional (one hop) OFDMA network significantly in system performance.Handover is an indispensable problem in wireless resource management. In the traditional cellular networks, the BS is the unique handover target, while in relay networks, the handover target can be BS, RS in the same cell or RS in the neighbor cell and the handover node is MS or RS. So the handover process in single hop networks can't adapt to the relay networks. The dissertation proproses a whole set of handover process for the relay networks. In addition, a new solution is proposed in order to provide the continuity of data transmission during handover process. Our proposed handover mechanisms have applied patents and some content has been accepted and written in IEEE802.16j specification. At last, the diversity set update algorithm for macro diversity handover in Relay Networks is studied.