New Delhi: The Cellular Operators Association of India (COAI) has written to the Telecom Regulatory Authority of India (Trai), requesting it to rectify what it calls discrepancies in an official app measuring wireless data speeds.
In a letter addressed to Trai secretary S.K. Gupta, the telecom industry lobby urged the regulator to also meet operators to address concerns with respect to the MySpeed app. Mint has seen a copy of the letter.
Trai had released a white paper on the measurement of wireless data speeds on 5 February, detailing the working of the MySpeed app.
“Careful analysis of the data presented in the white paper reveals certain discrepancies within the results of the MySpeed app, which primarily question the level of accuracy of this data and may also defeat the purpose of creating such an app," the letter said.
“We are getting multiple results in multiple speed testing apps for operators. What is causing these anomalies? We have to streamline this issue," COAI director general Rajan Mathews said.
Apart from MySpeed, companies such as Ookla and OpenSignal too measure data speeds.
The letter follows Trai publishing data on MySpeed portal wherein Reliance Jio topped the chart of fastest 4G telecom operators with an average peak download speed of 21.3mbps on its network in January, over two times rival Airtel’s average peak speed of 8.8mbps in the reported month.
Vodafone and Idea Cellular registered download speeds of 7.2mbps and 6.8mbps, respectively.
Meanwhile, Airtel on 23 March rolled out a new ad campaign Sab Kuch Try Karo, Fir Sahi Chuno, urging users to decide which network is the best by trying all options wherein it also said that Ookla has rated Airtel as India’s fastest network.
The COAI letter also said the app collects both manual and automatic samples for testing speeds and for one operator, the ratio of manual and automatic samples was starkly different from that of the other three operators.
The lobby group has urged the regulator to re-examine the current methodology and process of test logs being used in the app, apart from considering a larger sample size and an appropriate filter mechanism to discard spurious samples from the data.