基于免疫算法的球形译码算法研究

在多输入多输出(MIMO)系统的信号检测算法中,球形译码算法的性能最接近最大似然检测算法,是目前应用最为广泛的一种信号检测方法,但传统的球形译码算法在低信噪比时计算复杂度较高,在高信噪比时误码率性能不够理想.该文提出了一种改进的球形译码方法,利用免疫算法的寻优能力来获得最佳的初始搜索半径,从而达到在保证译码性能的同时有效地降低计算复杂度的目的.实验仿真结果表明,与传统的球形译码算法相比,基于免疫算法的球形译码方法在高信噪比下可以达到最大似然译码的性能,同时能有效降低低信噪比时的运算复杂度,从而更符合现代无线通信实时性的要求.

Sphere decoding algorithm based on an immune algorithm

Among all multiple-input multiple-output (MIMO) signal detection algorithms, the performance of the sphere decoding (SD) algorithm is closest to the performance of the maximum likelihood (ML) decoding algorithm. At present, the SD algorithm is the most widely used as a signal detection method. Unfortunately the calculation complexity of the conventional SD algorithm is still high at low signal-to-noise ratios (SNRs) and its performance is not ideal at high SNRs. This paper presents a novel SD method where the initial search radius can be obtained using an immune algorithm. Based on results of experiments and simulations, it was concluded that compared with the conventional SD algorithm, the new sphere decoding method not only has the same bit error rate performance as the ML decoding algorithm at high SNRs, but also effectively reduces the calculation complexity at low SNRs. It should satisfy real-time requirement in wireless communication.