A novel Bayesian detection approach for ack/nack signalling in third generation high speed packet data systems

In this paper we develop a novel framework for Bayesian detection that is generic in its formulation, and is referred to as Bayesian constrained false alarm detection. It is shown that this formulation serves as a bridge between classical and Bayesian detection approaches, and can be thought of as a generalization of both the classical Neyman Pearson detection framework as well as of the minimum risk Bayesian detection strategy. This problem formulation is motivated by an application of signaling (ack/nack) detection in third generation wireless packet data systems, such as HSDPA. In these systems, signaling information in the form of ack/nack are critical to be detected with high fidelity in order to ensure that the performance of a packet data system is not affected adversely. We also propose a provably convergent adaptive algorithm to estimate the a priori probabilities. These concepts and algorithms have widespread utility in any application of statistical signal processing. Simulation results are presented for their application to ack/nack detection in a realistic UMTS simulator.