Research On Key Technique Of Resource Management In Wireless Ad Hoc Network

With the rapid development of networking and wireless communication technologies,wireless ad hoc network has been a hot research topic. Nowadays, wireless ad hoc net-work has been utilized not only in military but also the commercial area. The dimensionsand the application ranges are rapidly extended. Hence, how to manage the resource rea-sonably, improve the network performance as well as service quality, and ensure the secureand reliable information transmissions has been an important task of the wireless ad hocnetwork. In order to realize the efficient resource management, this dissertation focuses onthe research of heterogeneous backbone architecture, joint topology control and routing, andthe multi-rate aware transmission power control problem. Firstly, a scalable multi-channelmulti-radio wireless backbone architecture is proposed in this dissertation. Based on that,a ?exible joint channel assignment, transmission power control and routing adjustment pro-tocol is designed to further improve the network throughput. Finally, the multi-rate awareoptimal transmission power control problem is analyzed theoretically. Major contributionsand the originalities of this dissertation are as follows:(1) A scalable multi-radio multi-channel heterogeneous wireless ad hoc network archi-tecture is proposed in this dissertation. This architecture can be utilized as the backbone ofwireless ad hoc network and aims to improve the network performance and scalability.Due to the in?uence of contention and interference on the wireless broadcast channel,the network throughput is far from satisfactory. Early simulation experience also suggeststhat the throughput decreases sharply with the increase of hops, therefore the network is notscalable with network size. The proposed scalable multi-radio multi-channel heterogeneousarchitecture offers different transmission capabilities to super nodes and router nodes. By theaid of the long-range transmission of super nodes, the hop count is reduced and consequentlythe scalability is enhanced. Meanwhile, super nodes and router nodes have different num-ber of radios and channel assignment. The utilization of multi-channel multi-radio greatly reduces the interference of super nodes to other nodes and alleviates the in?uence of hiddenterminals caused by asymmetric links.(2) A novel joint topology control and routing protocol is proposed in this dissertation.It resides between MAC and network layer and aims to reduce the interference and conse-quently improve the network throughput by coordinating the transmission power, channelassignment and route selection among multiple nodes in a distributed way.The number of available channels in IEEE 802.11 is limited, and the radios equippedon each node are generally less than channels. Besides, due to the channel dependencyand connectivity requirement among nodes, the co-channel interference can only be reducedto a certain degree. This dissertation focuses on considering both channel diversity andspatial reusability to reduce co-channel interference, and presents a joint topology controland routing protocol. Based on the traffic information measured and exchanged among two-hop neighbor nodes, this protocol quantifies the difference of various adjustment candidatesby the aid of ECATM, then selects feasible adjustment candidate with the smallest ECATMvalue, and finally coordinates affected nodes to adjust the transmission power at each node,the channel selection on each wireless interface, and route selection between nodes.(3) The multi-rate aware optimal transmission power control problem is theoreticallymodeled and analyzed in this dissertation. This model illustrates the tradeoff between spatialreuse and physical layer data rate. The optimal transmission power can be derived based onthis model which maximizes the network throughput.Previous works mainly focus on the tradeoff between transmission power and hop counton path while neglecting the impact of multiple rates, or they are based on the assumption oftime division MAC so that their results can hardly be applied to wireless ad hoc network. Inthis dissertation, the behavior of basic IEEE 802.11 DCF is approximated by the p-persistentCSMA through a Markov chain model. Various factors, such as hidden terminals, multi-hopdata ?ow and concurrent interference are incorporated to shed light on the rate-aware optimaltransmission power choice problem. From these models, the average per-node throughput,the optimal transmission power and expected transmission delay for given scenario can bederived. Finally the correctness of this model is analyzed by numerical results.