Analysis of advantages of neuron-like systems in the procedure of signal comparison-measure calculation
Alexander Gennadievich Khobotov, Alexander Ivanovich Khil’ko, Vladimir G. Yakhno · Optical Memory and Neural Networks · 2008
We study the possibility of comparison of the results obtained by the conventional covariance method, the L 2 P method, and the proposed technique based on the use of neuron-like models for calculating comparison measures. It is shown that the proposed neuron-like technique has advantages over the covariance method for a decreased signal-to-noise ratio using one-dimensional chirp signals as an example.