Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
381377 | Engineering Applications of Artificial Intelligence | 2008 | 9 Pages |
Digital signal-type identification is an important issue in communication intelligence system. Most of the proposed identifiers (techniques) can only identify a few kinds of digital signal and/or low order of digital signals. They usually require high levels of signal-to-noise ratio (SNR). This paper presents an intelligent technique that includes a variety of digital signals. In this technique, a combination of the higher-order moments and the higher-order cumulants up to eighth are proposed as the effective features for representation of the considered digital signals. A multilayer perceptron neural network with resilient back propagation learning algorithm is utilized to determine class of the received signal. The numbers of nodes in the hidden layer, along with the selection of input features are optimized using a genetic algorithm. Simulation results show that the proposed technique has high performance for identification the different digital signal types even at very low SNRs. This high performance is achieved with only seven selected features and the least possible number of nodes in the hidden layer, which have been optimized using the genetic algorithm.