Reflectometry, a wire diagnosis technique, allows to detect and localize electrical defects in networks efficiently. In order to achieve a very precise online diagnosis without altering the functioning of a network, reflectometry methods should use specific wideband Multi- Carrier signals, whose generation requires dedicated numerical tools. The underlying architecture of any reflectometry system for the injection of the signal requires appropriate Digital-to-Analog Converters. Yet, measuring the reflected signal should be performed using Analog-to-Digital Converters capable of reaching high sampling frequencies together with sufficient resolution. Such converter are either extremely expensive or beyond nowadays state of the art. In that respect, this study addresses a new architectural approach for designing such reflectometry systems based on Compressive Sampling method bypassing the Nyquist rate. Thus, thanks to the compressed acquisition and its analog encoder, we succeed in reconstructing the reflectogram with a lower sampling frequency and also in identifying nearby defects.