Summary

  • TRAI's 26 October release reports mobile-network measurements made around Vijayanagara district on 11–14 August, including a 74-kilometre stretch of NH-50 and 13 hotspots.
  • On that highway route, its annexure records 481 no-service samples out of 4,759 for BSNL. The release's district poor-signal count for the same operator differs by 100 samples between the prose and the annexure.

A signal bar is a poor witness. TRAI brought a ruler instead: drive a car, place calls, send data and count what actually happens. Its 26 October original report covers 439.8 kilometres of district roads, 74 kilometres from Hosapete towards Banavikallu on NH-50, 13 hotspots and 7.6 kilometres walked. The itinerary runs from Hampi monuments and Hosapete station to a court, health centre, bus stand and dam entrance. This is a test of places where people need a connection, not a carrier's coloured coverage map.

The highway table makes one hole unusually visible. TRAI's Annexure 2 records 481 BSNL observations classified as limited service or no coverage among 4,759 samples. That is 10.11% of this operator's samples on this route. Its poor-signal total is a different measure, 1,512 out of 4,759. The same table records no no-service samples for Airtel or RJIL, and eight among 5,303 for VIL. These are observations on a defined road at a defined time, with separate denominators for each operator. A percentage without its route and denominator would be a much worse map.

Calls tell another, smaller story. Three of 23 successfully established BSNL calls on the highway dropped; VIL recorded three of 28, while Airtel and RJIL recorded none among 27 each. Twenty-three calls are a thin reed for a district-wide slogan. The report's broader district drive records 1,938 BSNL no-service samples out of 48,469, or 4.00%. The fact that the road and district numbers differ is useful: a network can look respectable across a mixed itinerary while a particular road remains awkward.

Then the ruler needs checking. The report's main paragraph says BSNL had 18,287 district samples below the technology-specific signal threshold out of 48,469; Annexure 1 says 18,387. Both appear in TRAI's own PDF. The 100-sample difference changes the calculated share by about 0.21 percentage points. It does not erase the highway gap. It does mean the public record should say which count is right, especially when the point of publishing is to turn network claims into checkable numbers.

TRAI says it has shared findings with the providers for action. The publication contains no completed repair result. The next useful document would pair repeat measurements at the same weak spots with a reconciled table. A regulator that asks operators to answer for a signal must make its own arithmetic answerable too.

Original source