Research On Underwater Acoustic Channel Identification

With the development of signal processing technology, research on application of new arisen channel identification and equalization technology provides significant path for high quality information transfer underwater. This paper carries out concrete study on Underwater Acoustic channel identification technology.This paper established Underwater Acoustic channel models in both shallow sea and deep sea, on the basis of which a great deal of simulation research and actual sea experiment data verification of several identification algorithms are carried out based on Single-input Single-output (SISO) and Multi-input Multi-output (MIMO) Underwater Acoustic communication system. Study under the condition of SISO system indicates that the second-order blind identification algorithm put up distinct advantages on arithmetic speed, convergence rate and sensitivity of estimating channel order, but not as good as high-order statistical algorithm for anti-noise nature. Research on MIMO system demonstrates that outer-product decomposition algorithm (OPDA) have good anti-noise nature, small sensitivity of estimating channel order and need fewer observation data, but it can't identify channels exactly. With bad anti-noise nature and high sensitivity of estimating channel order, subspace algorithm can identify channels almost completely if SNR high enough and order estimation precise, so that source signal can be recovered through simple equalization algorithm.