Joint PSK Data Detection and Channel Estimation Under Frequency Selective Sparse Multipath Channels

Bursty data links can benefit directly from the removal of pilot symbol transmission for channel estimation by improving the spectral efficiency. For such networking scenarios including data or paging signals, blind equalization for joint data detection and channel estimation with few or no pilot can improve spectrum efficiency. Though some existing works typically have attempted to take advantage of the sparsity of multipath channels, substantial performance improvement remains elusive. In this work, we develop an iterative Markov chain Monte Carlo algorithm based on Gibbs sampling designed for sparse channels. We incorporate the channel sparsity in the form of an $l_{1}$ type prior probability distribution, and derive the posterior channel distribution via stochastic sampling. Furthermore, we propose transmitter and receiver structures that could resolve unknown phase ambiguity in frequency-selective channels. This algorithm is also generalizable to non-sparse channels.