Adaptive correlation techniques for spread spectrum communication systems

Spread spectrum communication systems can require significant computation to perform initial acquisition, searching across time and frequency to accurately detect and lock onto the desired signal. This initial acquisition problem is exacerbated in deeply spread non-binary signals lacking cyclostationary features or repetitive codes, leading to a desire for less computationally intensive approaches. This paper introduces adaptive correlation methods that employ iterative Bayesian estimation techniques to reduce the expected computational load of signal acquisition, yet still achieve desired detection probabilities. Improvements to the adaptive correlator technique are identified, permitting flexibility in the correlation search windows, making it configurable for different applications like GNSS, DSSS communications, and noise-like chaotic spread spectrum signals. This paper presents analytical predictions of adaptive correlator performance, validated by measured FPGA hardware results achieving savings of 80% over brute-force serial time-domain acquisition searches.