基于平均互信息量的物理层抽象算法

摘要

针对物理层抽象技术缺乏理论模型以及等效指数信噪比映射(EESM)物理层抽象算法依赖调整参数和通用性较差的缺点,依据信息论、信号检测和概率理论,提出了物理层抽象的概率模型,并据此推导得出基于平均互信息量的物理层抽象算法--块平均接收信息率(RBIR)算法.基于采用MIMO-OFDM技术和最小均方误差(MMSE)检测算法的WiMax Ⅱ系统的仿真结果表明,对于ITU PedB 3kmph和ITU VA 30kmph信道模型、多种调制编码方式,该算法都能够获得与EESM算法相当的性能,并且不需要相关的调整参数,从而使得该算法更具一般性,能够较容易地扩展到不同的无线通信系统中,实现物理层抽象.该算法的有效性进一步验证了本文提出的物理层抽象概率模型.%In consideration of the problems that the physical layer abstraction technique lacks theoretical model support and the exponential effective SNR mapping (EESM) algorithm has a limited universality and is dependent on tuning parameters, the paper proposes a theoretical model to describe the physical layer abstraction. Then, the received block information rate (RBIR) algorithm, a mean mutual information based algorithm for the physical layer abstraction, is deduced with this proposed model. The results of the simulation from the WiMax Ⅱ down link system, with MIMO-OFDM and MMSE detections, for ITU PedB 3kmph and ITU VA 30kmph channel models and various modulation and coding schemes, show that the proposed algorithm can obtain the same accuracy as the EESM algorithm, without any tuning parameters. Thus the proposed algorithm is more general and easily extended to various wireless systems to implement the physical layer abstraction. The validation of the proposed algorithm also proves that the proposed model is accurate.