Skip to Main Content
This paper presents an automated means to predicatively detect service degradation signatures or symptoms of VDSL (Very High Bit-rate Digital Line Subscriber) based on performance monitoring measurements. It describes a management structure and methodology for trouble sectionalization and isolation to verify and confirm problem locations. This paper also provides new computing techniques to determine loop and tap lengths based on loop attenuation at different tones. These computing techniques effectively lower the computing time so that the computing program can be deployed to support real time applications and process large numbers of lines in backend during automated processes. To improve the fault detection, we also include the methodology of SELT (Single Ended Loop Testing) as part of the performance monitoring process. With our methodology and management structure, the detection rate is 95% and false positive is ~0.01%. As a result, the implementation of automated performance monitoring capabilities to support Service Assurance (SA) operations has been generally satisfactory.