Bio-Inspired Integrated Chips for Telecommunications S/W Defect-Tracking

Abstract

Defect tracking is important in evaluating the reliability of the software used in telecommunication networks. Bio-inspired integrated approaches and embedded chips have been developed and implemented to track improvements in the software reliability. In this paper, the integrated model for the failure discovery during testing is combined with bio-inspired approaches using the recurrent dynamic neural network (RDNN) with parametric adjustments and wavelets as basis; and the adaptive parameters RDNN (ARDNN) where the criterion is to minimize the error in failure intensity estimation, subject to the model constraints. Simulation results favor our adaptive recurrent dynamic neural network, with reduced error from 88% to 1.25 -to- 8% based on the number of iterations in the training phase.. The ARDNN approach provides optimum solution to the dynamic problem at hand since it iterates on the shape of the wavelet basis and provide adequate recovery of the data in the form of piecewise linear differential.