Spectrum Access Strategy And Performance Analysis Of Cognitive Radio Networks

Through the method that the secondary user utilizes the idle licensed spectrum by chance in temporal and spatial dimensions without interference to primary user, cognitive radio can obviously improve the spec-trum efficiency and further to remit the shortage problem of wireless spectrum resource. Dynamic spectrum access, which is an effective method to improve the performance of secondary users and guarantee the QoS of primary users, is one of the key technologies to realize cognitive radio. In real network scenarios for ap-plication of cognitive radio, with the affects of multiple cross-layer factors, user mobility and diversity and so on, the modeling and analysis for cognitive radio network are very difficult. Searching efficient and rea-sonable spectrum access strategies in real scenarios has been important issue to further research cognitive radio networks. Based on the above issues, this dissertation researches several application scenarios of cog-nitive radio networks. Each scenario is modeled and analyzed firstly and then the spectrum access strategy is proposed. The working mainly includes:analysis of spectrum access strategy with multiple cross-layer considerations in multi-channel cognitive radio networks, spatial-temporal spectrum access opportunity and strategy for mobile cognitive radio networks, density-adaptive spectrum access for cognitive radio sensor networks, handoff delay analysis of reactive spectrum handoff for classified secondary users, and method on target channel sequence selection based on proactive spectrum handoff for classified secondary users.First, a novel spectrum access strategy minimizing data packet delay of secondary users is proposed for the multi-channel cognitive radio network with multiple cross-layer considerations. Under interweave spectrum access scheme, the impacts of multiple factors crossing physical layer, data link layer and MAC layer on transmit delay and successful probability of data packet are synthetically considered. And these cross-layer considerations are integrated into M/M/1/K queuing model to deduce the average packet delay and loss rate of secondary users. Then a nonlinear optimal problem with the object of minimizing secondary users’packet delay is proposed. By solving this optimal problem using a modified genetic algorithm, the optimal spectrum access vector is obtained. Simulation results verify the proposed spectrum access strategy outperforms equiprobability, inverse proportion and random access strategies. The numerical results further indicate that the limited tolerable retransmission attempt for ARQ and buffer capacity can both bring opposite impacts on data packet delay or on packet loss rate, respectively. It is necessary to make a tradeoff to design the two factors.Second, a multi-channel spectrum access strategy maximizing the spectrum access opportunity of sec-ondary users is proposed for mobile cognitive radio network. As the mobility of users, the spectrum hole may happen not only in temporal dimension but also in spatial dimension. Possion point process is used to model for two spectrum utilizations separately produced when the secondary user locates in and outside the primary protect region. The analytical expression of spatial-temporal spectrum access opportunity is obtained in a certain area region. Then the optimal guard distance of primary user is found through maximizing the spatial-temporal spectrum access opportunity. Upon the guard distance, the throughput of secondary user is derived on Nakagami fading channel. The optimal spectrum access strategy is proposed by maximizing the throughput. Simulation results indicate the throughput can be improved using the proposed strategy when the density of secondary users is low, while the improvement is not obvious when the density of secondary users is higher than the amount of channels.Third, an adaptive spectrum access strategy based on secondary node density is proposed in cognitive radio sensor networks. Considering the facts that the energy of secondary node is limited and the random sleeping mechanism can be adopted in wireless sensor networks, and the real-time transmission requirement of burst data traffic, the transmitting and receiving process of secondary node is modeled. For different node densities, the energy efficiency and the transmitting failure probability of data packets are analyzed under interweave and underlay spectrum access schemes during a channel switch state, respectively. A log function is used to design a comprehensive utility index which integrates energy efficiency and transmitting failure probability of data packets. Through maximizing the utility index, a node-density adaptive spectrum access strategy is proposed. According to the strategy, the interweave spectrum access scheme should be used when the node density is low, otherwise, the underlay scheme should be used. Simulation results verify the adaptive node density threshold exists and the proposed strategy can improve the network performance obviously.Fourth, based on reactive spectrum handoff mode, a multiple handoff process model is established for classified secondary users and the cumulative multiple handoff delay is analyzed. According to different delay requirements of various communication services in cognitive radio networks, secondary users are divided into two classes with different priorities of accessing to spectrum. Markov process is introduced to model the multiple handoff process of classified secondary users. In order to facilitate the analysis of cumulative handoff delay, a two channels network is used as an example to match a preemptive resume priority M/M/2/ queue. Numerical results provide the admission access region for the classified secondary users.At last, based on proactive spectrum handoff mode, an adaptive selection strategy of target channel sequence for classified secondary users is proposed. Considering the interweave spectrum access scheme and proactive spectrum handoff mode, the multiple handoff process of classified secondary users is modeled as a preemptive resume priority M/G/1 queue. Two types of cumulative handoff delay are derived for two different types of channel selection strategy, "always stay" and "always change". After minimizing the cumulative delay, the adaptive channel selection strategy is found for classified secondary users. Numerical results indicate the cumulative handoff delay can be reduced evidently if "always stay" is adopted when the channel probability occupied by primary user is low, and in turn "always change" is used.