A Low-Cost Radar Processing—The Pulse Pair Compression Technique

The goal of a Radar or Sonar system is to detect a delayed and frequency-shifted signal in noise, and to estimate these time and frequency parameters. Under the popular Gaussian additive noise hypothesis, this estimation and detection scheme results in a generalized cross-correlation computation, usually conducted using Fast Fourier Transforms. This so-called Range-Doppler mapping is a computationally intensive task that may prevent low-cost designs or high-rate systems. In this paper, we present a simple and low-cost signal processing solution to both detect a possible target and estimate its range. Moreover, for the common case of Linear Frequency Modulated pulses, this processing is simplified even further. Interestingly enough, this simplified radar architecture does not exhibit any performance loss. This range-Doppler radar processing could be particularly interesting for applications where hardware constraints are strong or when one wants to increase the processing rate.