Abstract-Electrical cables of all types are subject to aggressive environments that can create defects or accelerate aging. Many application domains require diagnosis methods and tools. Among many methods, reflectometry has proven to be the best candidate and can be easily applied to the detection and localization of hard defects, while only requiring one access point to the wire. But soft defects are more difficult to track and require new powerful methods. This paper presents a review of the recent state of the art in the field of wired network diagnosis and shows the evolution of future activities in this domain. It provides new perspectives and new research domains are proposed.

International audienceA new technique is proposed to reconstruct faulty wiring networks from the time-domain reflectometry (TDR) response. The developed method is also for characterization of defects in the branches of the network. The direct problem (propagation along the cables) is modeled by RLCG circuit parameters computed by finite element method (FEM) and the finite-difference time domain (FDTD) method. Genetic algorithms (GAs) are used to solve the inverse problem. The proposed method allows to accurately locate wire faults. Some examples are presented to validate and illustrate the ability of this reconstruction method

The design of dielectric materials with artificial molecules formed by electrically small dipole antennas loaded with passive electrical circuit elements is considered. Variations in the antenna loads lead to known and generalizations of known dielectric material models. These artificial dielectrics are characterized in terms of their equivalent susceptibilities and polarization vectors both in the frequency and time domains. With suitable choices in the antenna loads one can design the loss and dispersion properties of the resultant materials.