SNIR estimation and statistics for DS-SS CDMA communications

Estimation of signal quality is critical in high-capacity spread spectrum systems for several reasons, e.g., power control and multiuser detection. In some techniques, signal amplitude is used as the measure of signal quality, but this is not as good a measure as the signal-to-noise-plus-interference ratio (SNIR). In this work we investigate several SNIR estimators based upon correlator outputs, and show by simulation that they are both accurate in terms of estimating bit error probability, and rapid in that only several tens of bits are required. We also derive the SNIR probability density function for the equal-received-energy case and validate it by computer simulations.