Underwater Acoustic DOA Estimation of Incoherent Signal Based on Improved GA-MUSIC

The problem of poor real-time capability, a large number of traversals and the long operation time of the MUSIC algorithm bring great difficulties to the application of DOA estimation. For the above problems, this paper fuses MUSIC algorithm and genetic algorithm and proposes an improved genetic MUSIC algorithm (GA-MUSIC). Also, improvement strategies are proposed for the three operators of selection, crossover, and variation of the genetic algorithm. The GA-MUSIC was simulated for DOA estimation to verify the effectiveness of this algorithm. The research shows that the GA-MUSIC proposed in this paper has the best operational performance at different precisions and SNRs. This algorithm can effectively reduce the iterative computation of the algorithm with better real-time performance, and the search success rate is above 95% while ensuring accuracy. Furthermore, this paper corroborates that the proposed improved genetic MUSIC algorithm can achieve better performance in single signals and incoherent multi-signal sources. It can be seen that the algorithm in this paper can effectively solve the problems of low success rates of DOA estimation searches and poor real-time performance and has a significant reference value for practical engineering applications.