论文部分内容阅读
Turbo码在信道编码中通过迭代译码的方式可以较好的逼近香农限.本文以其译码算法中的迭代次数作为时间轴,译码输出作为状态变量,信噪比SNR及信息比特数N作为系统参数建立动力学模型,研究Turbo译码输出与迭代次数之间的关系.通过大量计算机仿真和理论分析发现随着信噪比SNR由小到大,译码算法先后经历了不确定不动点、奇异区和清晰不动点三个阶段,其中在由不确定不动点过渡到奇异区时发生了分岔现象.通过改变信息比特数N的方法得到了离散时间动力学中的切分岔、倍周期分岔和Neimark-Sacker分岔.在奇异区内观察到倍周期、准周期、周期三、混沌等不同的相空间轨迹.奇异区的出现给Turbo码在低信噪比下的应用带来了一定困难,本文通过延迟反馈控制的方法将相空间轨道稳定到不动点上,仿真结果表明,本算法可以使Turbo码在低信噪比奇异区内获得0·1—0·3dB的增益.
Turbo codes can approximate the Shannon limit by iterative decoding in channel coding.In this paper, the number of iterations in the decoding algorithm is taken as the time axis, the output of decoding is used as the state variable, SNR and the number of information bits N As a system parameter to establish a dynamic model to study the relationship between the output of Turbo decoding and the number of iterations.Through a large number of computer simulation and theoretical analysis found that with the signal to noise ratio SNR from small to large decoding algorithm has gone through an uncertain Point, singularity and clear fixed point, in which the bifurcation phenomenon occurs when the transition from the uncertain fixed point to the singularity zone is achieved. The segmentation in discrete time dynamics is obtained by changing the number of information bits N Bifurcation, bifurcation bifurcation and Neimark-Sacker bifurcation.Under the singularity zone, different phase space trajectories such as doubling period, quasiperiodicity, periodicity three, and chaos are observed.The appearance of singularity gives turbo codes with low signal-to-noise ratio The application brings about some difficulties. In this paper, the phase space orbit is stabilized to the fixed point by the method of delayed feedback control. The simulation results show that this algorithm can make the Turbo code obtain 0 · 1-0 3dB gain.