Release of Speed Measuring Facility

Introductory Program held for Service Providers

An introductory program was held at TRCSL to demonstrate the functionalities of newly released broadband speed measuring tool to service providers.

The second version (Version 2.0) of speed measuring facility was unveiled by Mr. Anusha Palpita, Director General TRCSL on 16^th August 2012.

The TRCSL has introduced an on-line, installation-free speed measuring facility to Sri Lankan Internet community last year. The long standing requirement has been fulfilled by the regulator, after conducting an extensive level of research work in the field.

This is not just a tool but a facility built systematically with the adoption of unique methodology to meet exact requirements of broadband sector in the country.

To mitigate distinctive issues, a mechanism has been brought in and incorporated in the tool to measure end to end Internet download speeds accurately. This was the first known facility of such kind in the world which was to be made available on a website of a telecom regulator.

However with the launch of speedy services in the country with wider bandwidth exceeding 10Mbps in the downlink required the TRCSL to develop a facility which is capable of reading and measuring speeds precisely in high speed services. Version 2.0 was developed to meet such requirements.

This speed measuring facility of TRCSL works on two dedicated high-end servers which connect to the Internet at speeds above 1Gbps. This ensures that the server’s end is free of congestion. This is one of the prerequisites in measuring Internet speeds.

Version 2.0 incorporates number of features and options which would help the application to read high speed broadband connections correctly. In this context larger files have been included in the test, ranging between 35MB and 100MB.

The TRCSL has carefully chosen data centers in suitable geographic locations to meet specific requirements. Since majority of outbound data traffic accesses websites either in Europe or in USA, two data hubs in USA and in Europe which directly serves by major Internet exchanges has been selected for this purpose.

The German server is co-located in a data center in Düsseldorf directly peers with Equinix, one of the leading data centers and IX facility providers in the world. It connects to AMS-IX (Amsterdam Internet Exchange) and DE-CIX (Deutscher Commercial Internet Exchange) where the major backbone providers are present. The data center connects to the Internet through TeliaSonera, NTT, Level 3 and other Tier-1 backbone providers in Europe.

The Server in USA is located in New Jersey, in the East cost of USA which peers with Choopa Network, one of the leading network operators in USA. The data center links London Internet Exchange (LINX) directly through Trans Atlantic networks and nLayer Communications, Beyond Networks America, other Tier-1 operators will provide routing when accessing the data center.

By choosing these servers in locations where the presence of Tier-1 networks are available, TRCSL ensures that its speed measuring facility works on a system that provides best possible congestion free and reliable network path to service providers in Sri Lanka when accessing test servers.

Another unique feature has also been introduced in this latest release of speed measuring facility which incorporates a system which is capable of reading available bandwidth of the Internet connection correctly. This feature and the concept are novel and possibly the first regulator to implement such parameters of measurements.

The TRCSL has now developed a system based on download speeds, which is easier to explain when dealing with service providers since they advertise their bandwidth which has a direct relationship to data transfer speeds, thus providing the subscriber a clear picture about the quality of the Internet service. Hence, in domestic broadband services parameters such as RTT and packet loss has no significance.

It is a well- known fact that intended speeds could not be provided or maintained in a network where the value of RTT and packet loss is high. (Proper analysis must be done to determine the exact cause) But RTT values alone would be insufficient to identify network conditions to figure out speeds. For example RTT value of 2Mbps service and RTT value of 16Mbps are more or less equal. But speed variations could take place without reflecting the change on RTT value. Hence, an extensive level of analysis must be needed to identify complex situations like this. It is our view as a regulator that a subscriber’s right should be ensured by presenting clear, simple picture of the quality of service he receives instead of providing data leading complicated analysis. Therefore the level of analysis which must be performed by the subscriber in providing details such as RTT could not be expected from him or her.

A demonstration was carried out at the end of the session to show the functionality and the accuracy of speed testing facility developed by TRCSL (version 2.0) and to compare the same with other facilities available for subscriber’s access. In this regard SpeedTest.Net and AT TESTER have been used. To ensure that the test is conducting on a high speed platform, SLT ADSL16Mbps connection has been used. [Which can view HD streaming video with 1080p without the occurrence of data buffering] These testes were carried out on Win 7 Platform

Download Speed Results delivered by each utility/facility are as follows:

This program was conducted by Mr. Sanath Siriwardena, Consultant TRCSL. Bharti Airtel Lanka, Dialog Axiata, Sri Lanka Telecom, Etisalat Lanka, Mobitel and Lanka Bell (Day two) participated in this event.

This facility can be accessed through the following link: http://speedtest.trc.gov.lk/TRCSpeed/broadband.php

Broadband Speed Test Results - July 2012

Please note that download speed test results of HSPA+ services will not be published this time due to unavoidable circumstances

Mobile broadband download speed tests have been conducted in a specific area or in a township. Download speed results plotted in graphs shows the speeds available in the area or in the township where the test was conducted. It does not represent overall speeds of the operator, Island-wide. Therefore when reproducing or copying these speed result graphs or any of these statistics, other relevant information such as month of the test, location of the test, RSSI values must be published along with the test results, graphs for the knowledge of the public.