Narrow-band signal detection using neural network pre-processor

The neural-network-based- ALE system has the capability of extracting periodic components automatically from signals consisting of mixed relatively weak periodic and relatively strong broad-band components. The authors use this ALE to configurate a neural-network-based adaptive narrow-band signal detection system, and present the simulation results performed in comparison with the detection performance of a conventional detection system and an Adaline-based detection system with two processing mechanisms of the neural-network-based detection system. This comparison lends support to the neural-network-based ALE as a candidate component in an adaptive detection system.