eprintid: 715
rev_number: 8
eprint_status: archive
userid: 17
dir: disk0/00/00/07/15
datestamp: 2018-05-27 16:44:12
lastmod: 2018-05-27 16:44:12
status_changed: 2018-05-27 16:44:12
type: report
metadata_visibility: show
creators_name: Leese, R.
creators_name: Mondal, Anirban
creators_name: Smith, Warren
title: Dynamic Line Management
ispublished: pub
subjects: telecom
studygroups: esgi73
companyname: Talk Talk Technology
full_text_status: public
abstract: Talk Talk Technology has an ongoing programme of Dynamic Line Management (DLM), aimed at improving the performance of ADSL broadband connections. Choosing the most appropriate configuration, or "profile", for an individual line allows a balance to be reached between achieving a high line speed and a stable customer experience. Among the objectives of DLM are to determine the likely line speed of new connections and to identify lines that might benefit from an updated profile.

Addressing the questions raised by TalkTalk regarding DLM requires three complementary types of modelling: physical modelling of the electrical characteristics of ADSL lines (these are existing copper telephone lines); information modelling of the associated communication channel, taking into account the levels of noise and the coding of ADSL signals; and finally statistical modelling, which enables data gathered from the exchange to be used to construct relationships between key variables and to identify lines that might be faulty or on poor profiles. The work of the Study Group 2010 contributed mainly to developing the statistical modelling. The report summarises the progress that was made and the opportunities for developing these techniques further.
problem_statement: Prior to taking a new customer, Talk Talk Technology provides an estimate on the likely line speed. This is an important part of the sale and the estimate needs to be as accurate as possible. Broadband speeds have been the subject of much consumer and Ofcom interest. One of the challenges with providing accurate speed estimates is that at the pre-sales stage only the line length, resistance and capacitance are made available. Currently the line length is used to determine the likely speed, but it is known that there are quality issues with the BT line length data.
(1.1.2) Talk Talk Technology posed three questions to the study group.
• Can a better estimate of speed against line length, resistance and capacitance be made?
• Is it possible to identify systematically the errors in supplied param- eters (e.g. line length)?
• Can we distinguish between lines that are configured incorrectly (on the wrong profile) and faulty lines?
(1.1.3) The Study Group concentrated on two aspects of these problems. The first issue was to determine an improved estimate of line length at the pre-sales stage. The relationship between capacitance (from the wire to earth) and the line length was the focus of this study. The second issue was to investigate better estimates of the speed.
(1.1.4) The group’s work involved three related strands: physical modelling, in- formation modelling and the statistical modelling of the dataset provided by Talk Talk Technology. Section 2 below deals with physical and infor- mation modelling, Section 3 looks at relevant statistical modelling tech- niques, and Section 4 provides an initial application of these techniques to the Talk Talk dataset. Finally, Section 5 draws some conclusions and makes recommendations for further work.
date: 2010
citation:   Leese, R. and Mondal, Anirban and Smith, Warren  (2010) Dynamic Line Management.  [Study Group Report]     
document_url: http://miis.maths.ox.ac.uk/miis/715/1/ESGI73-TalkTalk_CaseStudy.pdf