How to efficiently allocate spectrum resources is an important issue in cognitiveradio Ad Hoc networks. Because of Ad Hoc networks having distributed architecturesand dynamic topologies, however, available spectrum resources may change from timeto time and node to node. For this reason, a cross-layer algorithm of joint spectrumresources allocation and dynamic routing is demanded.In this dissertation, we study joint resource allocation and dynamic routing issuesfor multi-priority and delay-sensitive applications over cognitive radio Ad Hocnetworks. This can approximately be divided into the following three aspects:Firstly, we analyze the reasons of interference and construct a interference model.Based on the interference model, we use priority queuing theory to derive a solution ofaverage end-to-end delay. Unlike prior studies, the average end-to-end delay includespackets collision delay.Secondly, we study a dynamic routing algorithm in case of fixed transmissionspectrum and fixed transmission power. A load balancing routing algorithm, whichprobabilistically chooses relay nodes, is proposed. The simulation results demonstratethat this proposed routing algorithm can reduce average end-to-end delay, whileensuring high-priority applications’ quality of service.Thirdly, based on the proposed dynamic routing algorithm, we propose a jointresource allocation and dynamic routing algorithm in case of multiple transmissionspectrums and transmission powers. We theoretically demonstrate the efficiency andconvergence of this algorithm. |