Yesterday Offstumped had challenged Mr. Mukul Sinha and his Jan Sangharsh Manch to make public the phone records in their possesion and their analysis to determine if it will stand up to scientific scrutiny. There has been no response so far and it is not surprising. Today it appears from news reports was a crucial day at the Nanavati Commission hearings where the phone records issue came up. According this report Mr. Sinha claimed there was enough evidence to summon Mr. Modi. Mr. Sinha is also quoted as saying that if the Government is unable to provide its view on the phone records and his analysis, then the Commission must accept the JSM’s analysis and move forward.
Clearly they the JSM were worried on what surprises were in store for them at the hearing on the phone records, hence the interest in Offstumped’s piece from last month. When it became clear that the Government had not done its homework they probably dont see any immideate challenges to their version of the truth and hence, one must infer, this silence on Offstumped’s challenge.
While the Government of Gujarat and the JSM slug it out in the Nanavati Commission Offstumped has a few facts that should shed some scientific light on the usefulness of phone records and more specifically the Cell-ID in raw Call Detail Records, CDRs in pin pointing the location of a caller.
Fact#1 – The average cell site coverage radius globally is around 20km. (source: http://www.GSMWorld.com/universalaccessÂ Â The GSM Association’s Report on Universal Access)
Fact#2 – In a typical urban area coverage is not uniform and varies by location significantly. AlsoÂ the coverage data typicallyÂ available in the public domain is at a much higher level of granularity.Â As an example IDEA Cellular’s coverage maps for Ahmedabad and other urban areas of Gujarat from a few years back
Fact#3 – Using cell site identifier or Cell-ID is not particularly to establish the location is not particularly meaningful in scenarios of the kind under consideration as the cell site coverage may span an area of man square kilometres. For example in 2005 the TRAI conducted a quality of service analysis in the Lutyen’s Delhi area and the NDMC Area with the following findings. The report states that the area under consideration for Lutyens alone was 2800 hectares or 28 square kilometres, if you add the NDMC Area it will be upwards of that figure. For the Lutyen’s Area alone (28 square KM) Airtel had just 3 cell sites, Hutch had 4, Idea had 2, Reliance 3, Tata 1 and MTNL 5.
So in a city like Delhi as recently as 2005 on an average you had a cell site coverage area as small as 5 square Km (if you go with MTNL’s 5 sites) and as large as 28 square Km if you go by Tata. So consider areas of Gujarat which are not as densely populated as Delhi back in 2002 where the average cell site coverage area would have been many square kilometer.
So any kind of cell site based location identification of the caller could be off by as few as approx 2.5Kms if we go by a dense cell coverage or as large as 7 to 8 Km assuming the coverage area is a perfect square, which we know from the above coverage map is not the case.Â
So clearly Cell ID by itself is not conclusive. Some may argue that the signal strength could also be an indicator of distance from the cellular tower.
Fact#4 – In order to obtain a much more accurate picture of how signal strength varies by location requires a sophisticated calibration activity to be done based on integration of GPS hardware, GSM hardware and an appropriate software tool. This research paper by the Osmania University in Hyderabad in 2005 explains the complexity of such an activity as well as the issues with how coverage varies.
So unless Mr. Sinha and the JSM were able to magically go back in time to 2002 to conduct a calibration of this kind it is impossible for them to come up with any kind of model using signal strength as an indicator of the location of a caller from a cellular tower.
Meanwhile in other news Mr. Sinha in a free wheeling interview that appeared on Indian Muslims left no one in doubt about his political agenda which is reason enough to take his psuedo-science with a fistful of salt.
So Offstumped renews its challenge to the Jan Sangharsh Manch and Mr. Mukul Sinha to put their data and their analysis and let is stand up to scientific scrutiny.