Comparing Radar Receiver Pulse Deinterleaving Performance of Differing Window Functions for Bandpass FIR Filter Design

摘要

A radar receiver is designed to identify and process the desired echo from the radar transmission. Therefore, filter implementation is a key part of designing a receiver. This paper focuses on the performance of the Bartlett, Blackman-Harris, Chebyshev, Hamming, and Kaiser windows, which were analyzed by implementing each into a deinterleaving algorithm. Deinterleaving pulses are important because receivers often have signals interleaved with an abundance of noise. By using windows, it can reduce the abundance of noise and increase the performance of deinterleaving algorithms. The separation of the radar pulse trains from the pulse stream provides sufficient information to categorize and describe different signals with pulse descriptor words (PDW), such as Time Of Arrival (TOA), Time Of Departure (TOD), Pulse Width (PW), Pulse Repetition Interval (PRI), etc. This study presents the comparison of the five windowing techniques, the application of Hilbert transforms and envelope detection to analyze how each affects the deinterleaving algorithm. Each technique was analyzed by calculating their Signal-to-Noise Ratio (SNR), and the amount of information lost by measuring the error margin. From the data observed, the Blackman-Harris had the highest SNR, and the Kaiser window had the lowest percent error. The following sections will demonstrate the performance of each windowing technique.