Topology reconstruction of wiring networks using an iterative process based on Time-Domain Reflectometry and Forensic-Based Investigation algorithm

Time-Domain Reflectometry (TDR) generally consists of injecting a signal into the Network Under Test (NUT), and then collecting the multiple reflections that occur from each instance of junction or termination. However, since reflectometers combine primary reflections with multiple and intermediate reflections, pulses in the reflectometry response often overlap. Therefore, reconstructing wiring networks using only TDR responses is not feasible. In this paper, the Wiring Network Reconstruction (WNR) process is formulated as an optimisation problem. The optimisation algorithm used in this paper to solve the formulated problem is the Forensic-Based Investigation (FBI) algorithm. Solving the WNR problem consists of finding the topology and the length of branches of the NUT. In the objective function, the optimisation algorithm compares the TDR obtained from the NUT with the TDR of the predicted solutions, where all TDRs are generated analytically to save time. The obtained results using the FBI algorithm are tested against ten well-known optimisation algorithms over a set of seven experiments. In these experiments, five cases are simulation-based, while the last two are real-world cases. The results provided in the paper clearly show the effectiveness and resilience of the proposed approach for reconstructing wiring networks with different degrees of complexity.