Blind LTV shortening of doubly selective OFDM channels for UAS applications

This paper deals with the synthesis of a blind channel shortening algorithm for orthogonal frequency-division multiplexing (OFDM) systems operating over doubly selective wireless channels, a challenging scenario that is likely to happen in modern unmanned aircraft systems (UASs) data links. When the length of the OFDM cyclic prefix (CP) is smaller than the channel order, we propose to employ a blind linear time-varying (LTV) time-domain equalizer, which shortens the channel impulse response of the channel in the minimum mean-output energy (MMOE) sense, requiring only estimation of the second-order statistics of the received data. The equalizer design leverages on the complex-exponential (CE) basis expansion model (BEM) for the doubly selective channel, which naturally leads to a frequency-shift (FRESH) filter implementation. Monte Carlo computer simulations are carried out to assess the effectiveness of the proposed FRESH-MMOE channel shortener.