Training-Assisted Channel Estimation for Low-Complexity Squared-Envelope Receivers

Squared-envelope receivers, also known as energy detectors, are, due to their simplified circuitry, low-cost and low-complexity receivers. Hence they are attractive implementation structures for future Internet-of-Things (IoT) applications. Even though there is considerable work on the wider research area of squared-envelope receivers, a comprehensive comparison and statistical characterization of training-assisted channel estimators for squared-envelope receivers appear to be absent from the literature. A detailed description of practical channel estimation schemes is necessary for the optimal training design of latency-constrained IoT applications. In this paper, various channel estimators are derived, their bias and variance are studied, and their performance is numerically compared against the Cramer-Rao lower bound.