基于不同解码类型的8PSK软解调算法研究

Research on 8PSK Soft Demodulation Algorithms Based on Different Types of Decoding

无线宽带系统常采用多进制调制来提高频谱利用率,同时根据应用的环境,选用卷积码或者Turbo码作为差错控制编码来对抗传输信道的噪声和衰落的影响。在信道解码的过程中,为了提高解码增益,译码时需要软解调信息,一般采用对数似然比LLR作为软解调的输出信息。LLR通常采用经典的欧式距离计算方法得到,但是这种方法运算量大、处理时间长、实现成本高。为了解决这个问题,设计了边界判决法和星座点合并法来计算LLR,最后通过仿真比较了简化前后算法的性能,并给出了实际应用的效果。

High-order modulation is often adopted in wireless broadband system to enhance spectrum utilization. In addition, error control coding, such as convolutional code or Turbo code is selected to reduce impacts of channel noise and fading. In the process of channel decoding, soft demodulation information is necessary to improve decoding gain. In general, LLR （Log Likelihood Ratio） is used as the output of soft demodulation. LLR is calculated according to classical Euclidean distance decision, but the algorithm suffers from large amounts of computation, long processing time and high implementation cost. In order to solve the problem, the combination of boundary decision with constellation point was designed to calculate LL1L Finally, the performances of the two algorithms were compared and the actual effect was presented.