Modeling, Analysis And Application Of The Class I Sleep Mode In IEEE 802.16e

With the rapid developments and amalgamation of computer technologies and commu-nication technologies, the development trends of computer communication are broadband mobilized and mobile combining broadband implementation. IEEE 802.16e is the latest standard for mobile wireless broadband access systems of next generation networks. In mo-bile networks, power saving is one of the most significant issues for the battery-powered mobile stations to extend their lifetime before recharging. IEEE 802.16e introduces sleep mode to save energy of mobile stations. According to the operation mechanism of type I sleep mode in IEEE 802.16e, a set of systemic vacation queue models with close-down time and multiple inhomogeneous vacations are built for different kinds of network traffics. Sys-temic research on the type I sleep mode in IEEE 802.16e is performed by theory analysis, numerical examples and simulation experiments.Firstly, for memoryless property of data frame arrivals initiated by users, a Geom/G/1 queue model with close-down time and multiple inhomogeneous vacations is built. Under steady state conditions, energy saving ratio, system handover ratio for a mobile station sup-porting sleep mode and average delay time of data frame are derived using the method of embedded Markov chain. The dependency relationships between system performance mea-sures and system configuration parameters are addressed by numerical examples. Theory analysis results is validated by simulation experiments and the examples of applying theory analysis results to configure system parameters are displayed.Secondly, considering the bursty property of data frame arrivals, batch arrival mech-anism is introduced and batch size X is supposed to be Pareto(c,α) distribution, a GeomX/G/1 queue model with close-down time and multiple inhomogeneous vacations is built. System performance measures are derived by embedded Markov chain method. The dependency relationships of energy saving ratio and system handover ratio on system con-figuration parameters are discussed by numerical examples. For the variance of bath size is infinite, we show the dependency relationships of data frame average delay on system con-figuration parameters through simulation experiments, moreover, the examples of applying theory analysis results and simulation results are given. Thirdly, considering correlation and bursty property of network traffic, data frame arrival process is extended to D-BMAP (Discrete-Time Batch Markovian Process), a D-BMAP/G/1 queue model with close-down time and multiple inhomogeneous vacations is built. To avoid all the matrix-analytic procedures, a discrete time factorization principles with general vacations is proposed and proved. Applying the factorization principles, we de-rive the system performance measures. Taking a special case of D-BMAP into account, the dependency relationships of system performance measures on system configuration parame-ters are discussed by numerical examples, and some examples of applying theory analysis results to configure system parameters are displayed.Finally, considering both downlink traffics and uplink traffics, a queue model of two servers sharing vacations with close-down time and multiple inhomogeneous vacations is built. System performance measures are derived by two-dimension embedded Markov chain method. The dependency relationships of system performance measures on system config-uration parameters are discussed by numerical examples and simulation experiments, and some examples of applying theory analysis results to configure system parameters are dis-played.