Direct Positioning Method Combining Adaptive LASSO and Block Sparse Bayesian
In passive localization,the advantage of direct localization method is that it is applicable to low signal-to-noise ratio and independent parameters.However,when the radiation source is far from the passive detection system,due to the influence of low signal-to-noise ratio,some unknown parameters in the received signal model significantly reduce the algorithm's positioning performance for the radiation source.To effectively solve this problem,a radiation source direct positioning method combining adaptive LASSO prior and block sparse Bayesian is proposed.A hierarchi-cal sparse model is constructed through Bayesian theory,combining different prior distributions to give elements in the signal independent adaptive LASSO.At the same time,the block structure and intra-block correlation of the signal are explored.The dictionary of different base stations with shared sparsity is jointly reconstructed to complete the dictionary and achieve long-distance radiation source positioning.The simulation results show that at long distances,when the number of snapshots is small and the signal-to-noise ratio is low,the proposed algorithm is obviously superior to the tra-ditional direct positioning algorithms such as MUSIC,Laplace prior methods,and block sparse Bayesian methods in terms of radiation source positioning performance.
direct positioningadaptive LASSO priorblock sparse Bayesian Learning(BSBL)overcomplete dictionary
国家科技期刊平台
NSTL
万方数据
