Known-Interference Cancellation Over Time-Varying Rayleigh-Fading Channels

Abstract

Security and covertness are both essential properties of wireless tactical communications that can be introduced at the physical layer of these systems through cooperative jamming. The concept relies on sharing in advance the jamming waveform or means to generate it with cooperative receivers, and using that knowledge to cancel out the jammer’s impact through known-interference cancellation (KIC) while non-cooperative receivers cannot do the same. A promising KIC method is the frequency offsets-compensated least mean squares (FO-LMS) algorithm that iteratively updates its estimates of the wireless channel as well as carrier and sampling frequency offsets proportionally to user-defined step sizes. The iterative updates inevitably cause misadjustments, the extent of which the step sizes affect. In time-varying conditions, selecting the step sizes becomes a trade-off between the algorithm’s tracking performance and the extent of misadjustments. In this work, we characterize that trade-off and the optimal step size over Rayleigh-fading channels with Jakes’ Doppler spectrum through theoretical expressions. Supporting simulation and measurement results are provided, which demonstrate the accuracy and advantage of the optimal step size expression compared to exhaustive search methods.