功率受限和自动重传请求条件下系统吞吐量最大化的研究  

On Maximization under the Constraint of Power with ARQ Considering the Fading Channels and Buffer

作　　者：彭烈新[1] 朱光喜[1] 卢小峰[1] 宁国勤[1] 

机构地区：[1]华中科技大学电信系宽带无线通信与多媒体技术研究中心武汉光电国家实验室,武汉430074

出　　处：《计算机科学》2006年第8期46-49,共4页Computer Science

基　　金：国家自然科学基金重大项目"未来移动通信系统基础理论与技术研究"(No.60496315);国家高技术研究发展计划(863计划)(No.2003AA12331005)

摘　　要：本文研究了具有ARQ功能的基于衰落信道和数据链路层缓冲区队列状态的资源最优分配问题,目标是通过自适应调整功率分配和调制方式,在系统平均功率的限制下,使系统的吞吐量达到最大。在这个系统中并不限制ARQ的重发次数,所以最大化系统的吞吐量等效于使链路层的缓冲区溢出的数据包最小。本文把这样一个优化问题构造为马尔可夫决策过程,并提出了用动态规划解决该问题的方法。出于实用性的考虑,本文还提出了一种简单的次优资源分配方法,仿真结果显示这种方法与最优的调度方法性能非常接近。The problem of adaptive power allocation and modulation mode selection with ARQ over fading channels and finite buffer size in the link layer is considered. In the model, the paper first assumes that knowledge of the buffer occupancy and the fading state is available at the transmitter. The objective is to vary the transmission power and modulation mode according to the buffer and channel conditions so that the long-term system throughput is maximized under some average and peak transmission power constraint. Here, the retransmission times of ARQ is not constrained, so maximization of system throughput is equivalent to minimizing packet loss due to buffer overflow regardless of trans- mission error. The paper formulates this optimization problem as a Markov decision process（MDP） and use dynamic programming techniques to obtain the solution. A simplified scheduling scheme is also presented for package transmission and the results show that the minimum of the package loss is very close to that achieved by the optimal policy.

关 键 词：MDP 动态规划 吞吐量 ARQ POMDP 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术]