Internet users in Sri Lanka can finally be quality complacent. This is great news for many, particularly the veterans of the dial-up days. (The first modem I used was a 14.4 kbps.)
The Telecom Regulatory Commission (TRC) has just released the fixed (ADSL/WiMax) and mobile broadband (3G/HSPA) performance ratings for January 2011. This is the first time the regulator released the mobile broadband performance results. They are far from conclusive but still reveal some important patterns.
TRC has done 3G performance tests within Colombo and that too only in four locations, namely Borella, Talahena (Battaramulla), Nawinna (Maharagama) and Polhengoda, on days TRC calls ‘random’. Statistically, the selection of neither the locations nor the dates can be called a random choice, as convenience seemed to have ruled. This is not a complaint; only an emphasis on the non-representative nature of the results to avoid a misreading.
Mobile broadband quality testing, unlike in the case of fixed broadband testing, needs a more comprehensive approach. The quality experienced by a consumer, depends on diverse factors. It varies according to the location (two points within even the same premises may give different readings depending upon the lines of sight), date and time of the day, weather conditions and even the number of concurrent users. This calls for repeated multiple location and multiple day testing. Results from a single day testing, as done by TRC, should be taken with a pinch of salt. It just can be a ‘bad day’. The steady-state, not the transitional, patterns matter.
Even with this snag, the following conclusions can be drawn:
a) The future of broadband in Sri Lanka will certainly be mobile. On average, the 3G packages have shown a far better promise than the fixed ones in value for money. The business sector may continue to use fixed broadband, but for residential use it looks 3G is the choice. This is not a surprise for a developing country without an advanced fiber network. The same trend can be seen elsewhere, most notably in Indonesia, a country that had faced serious difficulties in expanding its wired broadband network.
b) Roughly all three mobile broadband providers (Dialog, Mobitel and Airtel) have shown comparable performance. Minor inter-operator variations are seen, but the test results are not adequate to make hard conclusions. (Eg. One brand looks excellent in Borella but the quality drops at other locations; but more data is required before forming theories.)
c) The only disadvantage of using mobile broadband over fixed is the relatively long Return Trip Time (RTT). That has an impact only for interactive applications like online games, VoIP and chatting. For other business and residential uses it makes no difference. Same with jitter. RTT also makes no noticeable difference in audio and video applications.
This does not mean the end of challenges. The outstation performance may not be equally impressive. (My personal experience at a limited number of locations shows acceptable performance, but outside the Greater Colombo and few city areas mobile broadband is available only in pockets.
We still hear non-availability complaints from not-so-remote areas.) Heavy broadband users still face the so called ‘Fair Use Policy’ (FUP) rule by some operators, who drop the speed after a certain usage limit. Net neutrality, a condition most users in the west take for granted is only a dream in Sri Lanka. These are not serious issues. I have no doubts they will be addressed in due course, with the market expansion. Important is that we have achieved a goal unthinkable just few years back.
(Chanuka is an independent researcher in telecommunication. He has been actively engaged in testing broadband quality performance for the last three years.)