On the Performance of Equalizers in an Aeronautical Telemetry Environment

Abstract

This paper examines the bit error rate (BER) performance of four equalization techniques with Forney and matched filter observation models for BPSK operating over multipath channels measured at Edwards AFB. The maximum likelihood sequence estimator is able to eliminate all the ISI induced by the multipath channel. For the filter-based equalizers (zero-forcing, minimum mean-squared error, and decision-feedback), the depth of the nulls defines the performance of the filter-based equalizers. Generally speaking, the matched filter observation model produces a channel with deeper nulls than the Forney observation model. But this advantage is achieved at the cost of a noise whitening filter. For flight-line channels the Forney observation model is the best choice for any filter-based equalizer. For low-elevation angle up-and-away channels the decision-feedback equalizer with either observation model yields excellent performance.