Detection and Characterization of Microsecond Intermittent Faults in Wired Networks

Intermittent faults (IFs), i.e., temporary malfunctions appearing at the random or periodic moments with normal behavior in between, are the measurable symptoms of the degradation of a system. Their frequency and severity increase over time, eventually leading to permanent faults. In order to optimally plan preventive maintenance, it is essential to catch IFs soon enough to repair them before they turn permanent. This paper introduces a new method, called continuous chaos time-domain reflectometry, which enables real-time monitoring of an electrical network and provides the detection and location information of microsecond IFs together with an estimation of their time of appearance and duration.