Approximation algorithm using: Neighbors, history and errors (NHE)

Abstract

In this work we design an approximation algorithm using neighbors, history and errors (NHE algorithm) to analyze and approximate the behavior of sensors readings after it fails. NHE algorithm computes and associates a specific equation to each sensor that has factors computed using neighbors, history data and the approximation error as a feedback factor. In some cases, it is difficult to reach each sensor to read its value either due to a connection problem or the sensor itself is down and not reachable. In those cases, NHE algorithm uses the evaluated equations to approximate this down sensor reading. The results of the experiments using this algorithm prove that it is powerful but with specific constraints.