Research On Channel Assignment And Routing In Multi-radio Wireless Networks

With the rapid development of integrated circuits and communication technologies, there are several new research trends arising in the area of the wireless communication network. Firstly, in the network architecture, addressing on the limitations in convergence and deployment of the traditional single-hop network, multi-hop relay based network architecture is widely applied in the practical network construction; secondly, in the network resource management, due to the scarcity of the wireless spectrum and the same time with popular trend of network deployment, spectrum management technologies represented by multi-channel, cognitive radio are becoming a hot topic; finally, in the character of communication nodes, with the wide emergence of small-scale, low-energy-consume and low-price wireless transceivers, it is now popular to deploy several transceivers on one node. Therefore, multi-radio multi-channel multi-hop wireless network is becoming one of the key research directions in wireless communication area.The resource allocation of multi-radio multi-channel multi-hop wireless network is more complex than the one in traditional single-hop wireless network. On one hand, the general resource many related to many aspects like time domain (link scheduling), space domain (routing) and spectrum domain (channel assignment); on the other hand, with the development of pervasive computing, nodes which may belonged to other networks might access to the current network temporarily, the resulting sudden flows may let the burst character become more significant and the relation between nodes are more complex. Under this situation, the network resource allocation problem and the corresponding research purpose are diverse in different scenarios. In this paper, the author considers the allocation of resources in the multi- radio multi-channel multi-hop wireless network under three different scenarios and studies each of them respectively.The first scenario is multi-hop wireless access network, in which the resource between nodes can be allocated by the Network Manager, and the goal of resource allocation which includes channel and routing is to maximize the network throughput. Addressing on the flow burst, we propose a joint resource allocation optimization model under the constraints of transmission flows, channels resource and interferences, then study the routing, channels allocation and link schedule strategy and propose a link schedule approach based on the optimal solution of resource allocation with uncertain input traffic.The second scenario is multi-radio Ad Hoc Networks. There is no manager and scheduler in this scenario, where the traffic transmission is done by the cooperation between nodes. And the resource allocation goal is to quickly allocate the transmission link for the coming flow. In this paper, we apply discrete rule and propose a distributed link schedule mechanism and a routing protocol separately. More specifically, we propose a channel switch based interface management approach addressing on nodes interface queue resources, an on-demand routing framework to minimize the end to end delay. The routing metric of this protocol is designed based on the consideration of queuing delay, channel switch cost, retransmission delay and etc.The third scenario is the mutual area cooperative network. In such scenario, nodes in the same area will compete for the spectrum resource of multi-interface multi-channel non-cooperatively. The study purpose of resource allocation is to verify the steady-state performance of the network when nodes are selfish. In this paper, we model the multi-interface channel allocation problem under heterogeneous channels with non-cooperative game theory; we then discuss the existence condition of Nash equilibrium and propose a distributed algorithm to realize Nash equilibrium. Since the solution of Nash equilibrium is not unique, we propose a refined Nash equilibrium model to maximize the network utility, and the same time design a particle swarm optimization (PSO) based algorithm to find the optimal solution.