Performance Evaluation And Optimization Of The Sleep Mode In IEEE802.16m With Best Effort Traffic

As a next-generation mobile broadband wireless metropolitan area network, IEEE802.16m has been required to support more mobile transmission. With the development ofthe mobile Internet business, the energy consumption in mobile station becomes huge.How to reduce the energy consumption of the station battery and extend the life of themobile network has become the research focus in the related field. In this paper, aiming atthe Best Effort traffic, a stochastic queueing model is built. The system performance isevaluated and the system parameter is optimized for the sleep mode in IEEE802.16m.Firstly, based on the working principle of the sleep mode in IEEE802.16m with BestEffort traffic, according to the memoryless characteristic of Best Effort traffic, the packetsarrival process is supposed to follow a geometric distribution. Concerning the randomvariations of the first sleep length, the first vacation is supposed to be a stochastic variableand the subsequent vacations are assumed to increase binary-exponentially. Amultiple-vacation queueing model is built.Secondly, the transmission time of a data packet is assumed to follow a generaldistribution and the time length of the first vacation is obtained statistically withsimulation. The state transition probability matrix of the queueing system is given.Applying the method of a matrix geometric, the formulas for the energy saving ratio andthe average delay of packets are presented.Then, in order to characterize the time length of the first vacation, a PH distribution isintroduced. Moreover, the transmission time of a data packet is supposed to follow ageneral distribution and PH distribution, respectively. Multi-dimensional Markov processis constructed. By using the method of a matrix geometric, the formulas of relatedperformance measures of the sleep mode for IEEE802.16m are given.Finally, with numerical experiments and system simulation, the influence of the sleepcycle and the traffic load on the system performance of the sleep mode is investigatedquantitatively. By constructing cost functions, the trade-off between different performancemeasures is studied and the system parameter of the sleep cycle is optimized.