Quasi-optimum detection results using a neural network

We study some particularities for the application of a neural network to binary detection. Using a modeled input (the classical J. Marcum model for pulsed radar detection), we optimize the design of the network, evaluating its performance by Monte Carlo trials. After comparing the detection curves with the theoretical optimum ones, it is found that the number of pulses integrated for each detection is critical for a quasi-optimum performance of the neural network.