Architectures for Real-Time Sequential Detection
Date: June 09 - June 19, 1986
We present the sequential detection for diffusion type signals in both the fixed probability of error and the Bayesian formulation. The optimal strategy in both cases is a threshold policy with explicitły computable thresholds. We provide numerical schemes for approximating the relevant likelihood ratio and provide an architecture for real-time signal processing.