In cognitive radio networks, the spectrum utilization can be improved by cognitive users opportunistically using the idle channels licensed to the primary users. However, the new arrived cognitive users may not be abl...In cognitive radio networks, the spectrum utilization can be improved by cognitive users opportunistically using the idle channels licensed to the primary users. However, the new arrived cognitive users may not be able to use the channel immediately since the channel usage state is random. This will impose additional time delay for the cognitive users. Excessive waiting delay can make cognitive users miss the spectrum access chances. In this paper, a discrete-time Markov queuing model from a macro point of view is provided. Through the matrix-geometric solution theory, the average sojourn time for cognitive users in the steady state before accessing the spectrum is obtained. Given the tolerant delay of cognitive users, the macro-based throughput is derived and an access control mechanism is proposed. The numerical results show the effects of service completion probability on average sojourn time and throughput. It is confirmed that the throughput can be obviously improved by using the proposed access control mechanism. Finally, the performance evaluations based on users are compared to that based on data packets.展开更多
为了验证当前网络节点失效而产生的流量衰减问题,提出一种基于离散时间的新型网络实际流量预测算法(Prediction algorithm based on Discrete-Time for FARIMA model,PDF).该算法首先利用离散时间理论推导节点数据包的排队情况,计算出...为了验证当前网络节点失效而产生的流量衰减问题,提出一种基于离散时间的新型网络实际流量预测算法(Prediction algorithm based on Discrete-Time for FARIMA model,PDF).该算法首先利用离散时间理论推导节点数据包的排队情况,计算出存在失效节点时流量平均对长的数学公式,并结合FARIMA模型建立预测方法.最后,通过数学仿真深入分析平均对长与服务率等影响因素之间的关系.仿真结果表明,PDF算法具有较好的适应性,与原始流量标准差为10.23.展开更多
文摘In cognitive radio networks, the spectrum utilization can be improved by cognitive users opportunistically using the idle channels licensed to the primary users. However, the new arrived cognitive users may not be able to use the channel immediately since the channel usage state is random. This will impose additional time delay for the cognitive users. Excessive waiting delay can make cognitive users miss the spectrum access chances. In this paper, a discrete-time Markov queuing model from a macro point of view is provided. Through the matrix-geometric solution theory, the average sojourn time for cognitive users in the steady state before accessing the spectrum is obtained. Given the tolerant delay of cognitive users, the macro-based throughput is derived and an access control mechanism is proposed. The numerical results show the effects of service completion probability on average sojourn time and throughput. It is confirmed that the throughput can be obviously improved by using the proposed access control mechanism. Finally, the performance evaluations based on users are compared to that based on data packets.
文摘为了验证当前网络节点失效而产生的流量衰减问题,提出一种基于离散时间的新型网络实际流量预测算法(Prediction algorithm based on Discrete-Time for FARIMA model,PDF).该算法首先利用离散时间理论推导节点数据包的排队情况,计算出存在失效节点时流量平均对长的数学公式,并结合FARIMA模型建立预测方法.最后,通过数学仿真深入分析平均对长与服务率等影响因素之间的关系.仿真结果表明,PDF算法具有较好的适应性,与原始流量标准差为10.23.
