Optimizing matched filters for GNSS receivers

Abstract

This paper presents methods for optimizing matched filter based GNSS acquisition units. Matched filters become computationally complex when the reference signal length is increased. They have a fixed length in time, which also limits the achievable integration time. In this paper, we show how to design an optimal adder for a large number of single-bit values, and how this can be utilized to build matched filters for multi-bit numbers and multiple inputs. By altering the incoming sample rates, it is possible to adapt the matched filter for changing signal frequency. Employing coherent and non-coherent integration after the matched filter allows enhancing the sensitivity of signal detection with modest implementation size increase.