Entropy Measures Based Complexity Analysis of Impulse Noise Sources for xDSL/PLC Systems

Abstract

The main objective of this paper is to do entropy measures based complexity analysis of impulse noises encountered in wireline communication systems. We have demonstrated in this paper that sample entropy, which is an information theory based statistical measure, is able to capture the time-domain complexity (in terms of variability and regularity) of impulse noises arising from most commonly used domestic electronic appliances. Results show that sample entropy features based analysis of impulse noises may also aid in identifying the primary disturber (active noise source) and thus can be helpful in troubleshooting the performance issues in wireline communication systems.