Performance analysis of a frame detection algorithm for burst mode communications

In burst mode communication, the beginning of a burst is often estimated by comparing the power in a sliding window to a threshold. The simplicity of the algorithm makes it very attractive. The probability density function of the detection variable for this simple algorithm is the subject of (H. H. Nguyen et al., April 2003) where it is shown to be approximately Gamma distributed. This paper provides further analysis of this low complexity edge/burst detection DSP algorithm. The analysis yields two equations that relate "the probability of false detection" and "the probability of missing a detection" to window size and the decision threshold.