The Research Of Using LP Algorithm With Cross-Spectrum Matrix To Beam Forming

LP algorithm is a high resolution algorithm which is advanced long before. But its using field in sonar is limited because LP algorithm does not use the statistic specialities of noise. This article advanced two methods named cross-spectrum field linearity predict (CSAR) and sub-band decompose methods of AR (SBDAR) to improve this shortcoming.The article firstly reviews the principle of the conventional beam-forming , inverse beam-forming , AR algorithms ; secondly, we advanced the CSAR algorithm . The theory analysis and simulation indicate that :the performance of CSAR algorithm is better than that of CBF, IBF and AR algorithms when the SNR is high enough. When the SNR is little, the performance of CSAR is no better than the purely AR algorithms .In order to improve the performance of spacial spectrum estimating algorithm at low level of SNR, bring out the sub-band decompose methods of AR (SBDAR) algorithms. This method first shift the phase of signals to zeros, and then make the new serial lowpass the filter. By the operatioin of lowpass,wo preserve the information of signals and restrain the part of noise . The theory analysis and simulation indicate SBAR algorithm improve the performance of spectrum estimation.