An Improved Symbol Detection Algorithm Based on Expectation-Maximum

In this paper, the problem of symbol detection of linearly modulated signals in the presence of unknown carrier phase is studied. We proposed a signal detection algorithm based on EM, which is different from conventional solution that not need estimation carrier phase, the symbol information can be obtained directly, thus reducing the performance loss caused by hierarchical processing of the traditional receiver. And resolved the problems of SER degradation that caused by phase ambiguity and the integral interval. The deterioration of the symbol detection performance was improved by sub-regional integration and adjusts the integration interval. Simulation results show that the symbol detection performance improved significantly.