Range limited adaptive pulse compression via linear Bayes estimation

The original adaptive pulse compression (APC) algorithm and its modified version, the Gain-Constraint APC (GCAPC), have the processing window decreasing effect, i.e. the length of the estimating output will decrease by a constant on each iteration stage. As a result, the iteration number is limited by the available input length and the extra computational load is inevitable for large initial processing window. Our analysis indicates that the estimating methods in APC and the GCAPC algorithms are special cases of the linear Bayes estimation. We derive the edge extended APC (EAPC) and GCAPC (EGCAPC) and an iterative linear Bayesian (ILB) algorithms that free from the processing window decreasing effect. We also proposed a modified iterative linear Bayesian (MILB) algorithm based on the Bayes theory and the original iterative algorithm that further improves the estimating performance. The MILB is compared with the EAPC, EGCAPC and ILB via simulations and is shown to be superior.