On the estimation of the loop topology with SELT

Telecom operators base their DSL-tariffs mainly on the offered bit rate. The maximal achievable bit rate is different for each customer, because it depends on two parameters. On the one hand, the transfer function of the ‘subscriber loop’, which depends on the network topology (the cabling between the central office and the subscriber) and on the other hand the Power Spectral Density of the noise. Equipment able to measure these two parameters already exists. However, the presence of a technician at the customer’s home is still needed. This is the main disadvantage of the so-called “Dual-Ended Line Testing”. In contrast, Single-Ended Line Testing (SELT) aims at identifying the subscriber loop by measuring the S11 parameter at the central office and using a noise model. This eliminates the dispatching of a technician, but is of course much more complex and intrinsically less accurate than with DELT. A useful tool in this is Time Domain Reflectometry [1].