【作者】 张燕；

【导师】 金顺福；

【作者基本信息】 燕山大学 ， 计算机软件与理论， 2013， 硕士

【摘要】 随着通信产业的快速发展，如何实现可持续发展的绿色通信已成为当今社会一个亟待解决的问题。针对长期演进项目LTE(Long Term Evolution)，提出一种带有休眠延迟机制的改进的非连续接收DRX(Discontinuous Reception)节能策略。建立系统的数学模型，并结合数值实验对改进的DRX节能策略进行性能评估和系统优化。首先，为了在减小能量消耗的同时提高系统的响应性能，对传统的DRX节能策略进行改进，使系统在从工作状态进入休眠状态之前经历一个休眠延迟期。为了综合评价改进的DRX节能策略，构建系统性能指标体系。考虑系统在工作状态和休眠状态之间的切换过程，建立了一个带有休假延迟期和启动期的两阶段多重休假排队模型。其次，针对远程通信、简单文件传输等无后效数据业务，设数据帧的到达过程遵循伯努利分布，数据帧的传输时间服从一般分布。基于无限容量，采用嵌入马尔科夫链方法，对排队模型进行稳态分析。基于有限容量，采用补充变量方法，给出排队模型的稳态解。导出系统切换率、能量节省率、数据帧平均延迟和系统阻塞率等性能指标的表达式。再次，针对多媒体应用等相依性数据业务，将系统输入设为离散时间成批马尔科夫链到达过程D-BMAP(Discrete Time-Batch Markovian Arrival Process)。基于无限容量，利用一般休假的分解定理，给出稳态下系统任意时刻的数据帧数及数据帧平均等待时间的分布。导出相依业务下系统切换率、能量节省率及有效利用率等性能指标的表达式。最后，通过数值实验与系统仿真定量刻画短休眠窗口数量阈值，系统缓存容量和休眠延迟定时器长度等系统参数对DRX节能策略性能的影响，并验证节能策略改进的有效性。考虑不同性能指标之间的折衷关系，构建成本函数，对改进的DRX节能策略进行优化设计。更多还原

【Abstract】 With the rapid development of the communications industry, how to build a greencommunication with sustainable development has become one of the significant issues tobe addressed urgently in today’s society. We propose an enhanced energy saving strategywith a sleep-delay period for Discontinuous Reception (DRX) in Long Term Evolution(LTE) systems. Based on the enhanced energy saving strategy for DRX, a systemmathematical model is developed. With numerical results, the system performance isinvestigated and the system parameter is optimized.Firstly, in order to improve the response performance of system while reducing theenergy consumption, by introducing a sleep-delay preiod before the system enters into asleep state from a working state, we proposed an enhanced energy saving strategy forDRX. We contract a performance index system for the comprehensive evalution of theenhanced energy saving strategy. By considering the procedure for the system switching toa sleep state from a working state, a two-stage multiple vacations queueing model withsleep-delay and set-up is built.Secondly, considering the memoryless traffic of data frames, such astelecommunications network and file transfer protocol, we assume the arrival process ofdata frames to follow a Bernoulli distribution and the transmission time of data frames tofollow a general distribution. Based on the infinite buffer, by using an embedded Markovchain, we present an exact analysis to numerically evaluate the performance of the systemmodel. Based on the finite buffer, we analyzed the system model in the steady state byusing the supplemental variable method. Then the formulas for the performance measures,in terms of the handover ratio, energy consumption, average latency of data frames andsystem blocking ratio, are calculated.Thirdly, considering the correlation traffic of data frames, such as multimediaapplications, we suppose the system input to be a Discrete-Time Batch Markov ArrivalProcess (D-BMAP). Based on the infinite buffer, the number of data frames at any instantand the average waiting time are derived by using the factorization principles forgeneralized vacations. Then the system performance measures, such as the system handover ratio, the energy saving ratio and the effective utilization rate, are given.At last, we provide numerical experiments with analysis and simulation toquantitatively describe the influence of the system parameters, such as the threshold of theshort DRX stages, the system capacity and the time length of the sleep-delay timer, on thesystem performance. The effictiveness of the enhanced energy saving mechanism isvalidated. Considering the trade-off between different performance measures, we builtcost functions and optimize the system parameters for the enhanced energy saving strategyfor DRX.更多还原