Error rates for hard-limited, spread-spectrum correlators

The authors present a methodology for permitting comparisons between linear and hard-limited receiver systems used in the noncoherent detection of orthogonal signals. A simple analytical procedure is developed for the error-rate evaluations and shown to be useful at the low-to-moderate input signal-to-interference ratios (SIRs) of most concern in spread-spectrum applications. Results for two important interference types-Gaussian white noise and random phase CW (continuous wave)-are developed. Monte-Carlo simulations indicate that the theoretical results obtained are accurate to within 0.3 dB for a 32-ary system in the presence of random-phases CW interference. Additional performance curves for M=4 to M=1024, together with loss factor curves, allow simple comparisons to be made between linear and hard-limited system designs