Blind diagnosis of a black-boxed fully-loaded wiring network for configuration structuring and fault monitoring

摘要

Although enormous efforts have been exerted to limit the extensive reckoning on wired networks amid nowadays wirefree revolution, dispensing cables is still out of reach. Accordingly, guaranteeing a trustworthy and reliable usage of cables is necessary. Despite the fact that many techniques have been deployed for this sake, several restrictions hindered their proper application. In fact, a prior knowledge of the network's topology, disconnecting loads, and controlled measurement setups are among many other conditions that can greatly affect the performance of monitoring the cable health. On the other hand, the promising non-destructive testing abilities of reflectometry methods emphasized exploring helpful means to bypass the afore-mentioned restrictions. In this paper, we will propose an approach based on the standard reflectometry technique jointly integrated with the graph theory and the particle swarm optimization tool in order locate defects in a fully-loaded unknown-topology transmission line network. Practical experimental results are accomplished to validate the suggested method.