Blind Channel Estimate Via Per Survivor Processing

In many applications signal acquisition and data decoding must be carried out in the presence of unknown, possibly time-varying channel parameters. When the Signal-to-Noise Ration (SNR) is medium or rapidly varying, ad hoc, methods for acquisition and data decoding do not perform well and there is a need for more sophisticated solutions. PSP allows unaided estimation of the channel where the channel is unknown due to geometry, doppler, interference/jamming or noise. In essence a multitude of estimates are tracked via a Viterbi trellis type algorithm to determine the optimum channel estimate in conjunction with the optimum communications decision. This, in turn, can provide substantial gains over ad hoc methods in the performance of communication systems. We propose to: 1) look in greater depth at PSP application to real world blind channel estimation problems, 2) size implementation of PSP and assess it's tractability today, and 3) extend the PSP theory in new areas as required. Areas to be investigated include co-channel interference mitigation, antijam enhancements, improved acquisition and tracking, and increased data rates in time varying bandlimited channels. ViaSat will apply its extensive experience in real world communications systems to evaluate the implementations necessary to reduce the PSP theory to practice.