LBO Home IndoChina | About Us | To Advertise | Contact Us rss LBO Mobil rss rss rss rss rss
Tue, 28 April 2015 15:48:43
Sri Lanka Dialog broadband speeds rated best
05 Nov, 2010 18:08:35
Nov 05, 2010 (LBO) - Broadband download speeds delivered by Dialog Axiata, Sri Lanka's dominant mobile operator, exceed expectations although its main rival Sri Lanka Telecom (SLT) does better on other measurements, a report said.
Dialog's 1Mbps (megabits per second) post-paid broadband package delivers more than promised in download speeds, recent Quality of Service Experience (QoSE) tests by regional think tank LIRNEasia said.

It also offers better value for money (kbps per US dollar) when compared to similar offerings by SLT, the dominant fixed line operator, and its Mobitel mobile phone subsidiary.

In its latest QoSE report, LIRNEasia compared download speeds of three similar broadband packages in Sri Lanka.

These were Dialog's post-paid broadband (1Mbps), Mobitel’s Zoom890 (1Mbps) and SLT’s Office (2 Mbps).

The underlying technologies of the different packages would most certainly impact their performance, said Shazna Zuhyle, research manager at LIRNEasia.

But the focus of LIRNEasia’s broadband QoSE benchmarking was on the comparison between what is promised or advertised versus the actual speeds delivered by the operators and how they measure up on other measurements.

These include round trip time (RTT), or the time delays in data transfer, as well as jitter, the variation in time between the arrival of data packets, and packet loss thresholds.

"In this light, the Dialog broadband package performed above expectations, while both SLT and Mobitel, below the advertised speeds," the LIRNEasia statement said.

"However, on other measures SLT performed considerably better than the Mobitel and Dialog packages on Round Trip Time as well as jitter."

The QoSE methodology compared upload and download speeds of data transfer to different server locations, within the Internet Service Provider, within Sri Lanka and internationally, as well as other important indicators of quality.

LIRNEasia said the tests were done at six time intervals on two weekdays in eleven locations across seven countries in South and Southeast Asia.

Rather than comparing download speeds between packages in different locations, LIRNEasia analyzes package speed compared to what was promised by the operator, as well as the "value for money" or Kbps per US dollar.

"LIRNEasia’s QoSE methodology offers much needed independent information for consumers to make informed decisions on broadband packages," said Zuhyle.

"The (telecom regulator) should promote such transparency in order to boost competition in the broadband market and improve service quality."

Corrected - Megabits para 2

Bookmark and Share
Your Comment
Your Name/Handle
Your Email (Your email will not be displayed)
Your Email
Receivers Email
Your Comment
17. Mahisha Nov 09
I will probably read the report over the weekend, time permitting. Sir (Rohan Samarajiva), I don't see why you need to "defend" yourself when it is obvious that people have clearly not read your report. If they have read your report, and if they challenge it, then it is a different story.

Tharindu Maligaspe: may I humbly request that you refer to Rohan Samarajiva simply as "Rohan Samarajiva" as he seems to call him self (without placing Dr in front) and not "Mr" Rohan Samarajiva as I believe he has read for a PhD? He probably doesn't give a rat's *** with the prefix Mr, but I feel it is due - in a public forum like this. A few people are trying to do something, lets encourage them with constructive criticism - and not for the mere sake of doing so. Shall we?

16. Rohan Samarajiva Nov 09
Page 4 of the report has the list of cities (Column Name: City). We also have the list of test cities documented in our methodology document which is on the Benchmarks page ( under 'Fixed BB QoSE'.

The test sites are:

- Colombo, Sri Lanka
- Chennai, India
- Bengaluru, India
- Mumbai, India
- New Delhi, India
- Dhaka, Bangladesh
- Thimphu, Bhutan
- Bangkok, Thailand
- Chiang Mai, Thailand
- Jakarta, Indonesia
- Manila, Philippines

Please note that our primary objective is to push the regulatory agencies (who have the resources and the duty to regulate QoSE) to act. We do not have the resources to test in multiple sites in Sri Lanka, which is one of the smaller countries we work in.

For those who want to test their home locations, the software is freely downloadable at, though there is a need to register (information on the webpage). They can then publicize the results on any of these threads (LBO, if permitted, or LIRNEasia, for sure) or start their own blog.

Broadband performance varies according to location, whether provided on wireless or wireline. Therefore, decentralized solutions are ideal. Random sampling is a solution, but it's too expensive for us. We have explored crowdsourcing as a solution, but could not take it to fruition.

I am curious why people do not approach the Special Committee on Broadband ( which has access to the billions of rupees collected by the TRC, and can afford even to do random sampling (though we think that may be overkill).

15. RS Nov 09
To all the commenters going postal on this report, have you actually read it?

The report doesn't say Dialog is merely states that Dialog performs better on delivery against promise i.e. they deliver close to what they promise. It doesn't mean Dialog is faster.

If you're all going to spam this forum with your comments, please at least have the decency to read the report first!

Also, try knocking the argument and not the author...

Man the comments on LBO are going the way of Xxxxx Mxxxxx.

14. Tharindu Maligaspe Nov 09
Mr. Samarajiva,
I will be grateful if you answer this simple question. (which I could not find in the link you gave)

How many locations you have tested before decision Dialog is superior in performance? Just give me the number. Thanks.

13. Suraweera Nov 09
What about performance of Airtel?
12. J Nov 08
I guess one needs to take the study in it's context and not as a rule. A line on a three wheeler today said "One sun shines on all". However as one may die of high temperatures caused by the Sun, it does not have a similar effect on someone else at another location.

Similarly this report seems to be from a few locations on 2 weekdays, which may be from one or two locations as the report says eleven locations across seven countries . I guess it would be good for LIRNEasia to publish the locations to compliment the study and not make it a general rule that Dialog is great from all locations in Sri Lanka.

We all have our individual opinions on Dialog. I am thankful to LBO for publishing the names of the service providers, as this may now call for other studies by others, whereby “we” the users may benefit with better quality.

11. Rajive de Silva Nov 08
To Rohan Samarajiva,

Whatever you say, the broadband quality in Sri Lanka is absoultely dismal. Local providers have to go a long way before they can come up to acceptable standards.

10. Serala De Alwis Nov 08
This is like Megastar. :-)
I have a Mobitel data line and I am happy about it, while some of my friends frequently complain about their Dialog lines. I do not know what exceeding customer satisfaction means, to me it looks marketing mumbo jumbo. I am not saying Mobitel exceeds my satisfaction but it is good enough for me.

I had a problem with my line only once, but when I complained they took prompt action. I had a call on the same day evening and they corrected the problem immediately. So I think I have to write this as an appreciation of Mobitel’s good service.

Is it too much to ask LBO how much it received from each company as advertisements for the last three months?

9. Rohan Samarajiva Nov 08
Mr Maligaspe,
We do many things other than broadband QoSE work and we do not monitor LBO comments on a minute-by-minute basis. So it was not avoidance, it was other work.

If you took the trouble to check out the URL I had given in my previous comment, you would have seen that we have given all the methodological details (actually, you can find this by using Google too).

The method was developed with IIT Madras and presented in beta at multiple meetings at IESL before we started releasing reports to the media in 2009. The operators participated in these consultations and agreed that the method, while not perfect (nothing is) was acceptable.

We stopped benchmarking mobile voice prices when they converged on low numbers and the ITU started reporting basket prices.

In the same way, we'd like to stop benchmarking broadband QoSE when quality improves to acceptable levels and the TRC takes on the task of exerting pressure on operators. Or someone. I understand that someone has posted a comment on our site saying he has developed a new method for testing quality. Check it out too.

8. Tharindu Maligaspe Nov 08
Can LIRNEasia just tell me how many locations you have tested?

Why you avoid this question?

7. aj Nov 08
Check this report. Shows providers have come a long way in a short time after LirneAsia started studying the issue.

Sri Lankan broadband users getting less: study

6. D Nov 08
Yes...Well said "Opinion" and "Raj"...
5. Opinion Nov 08
Well if anything I think what should be said is lirneasia is "doing" a marketing campaign for Dialog...
4. XFighter Nov 08
Dialog speeds are all a gimmick- this test should have been after the so called not so 'fair user policy' kicks in. SLT in particular should be appreciated for not imposing such ridiculous constrains on their broad band connections. This article also missing the fact that mobitel has far better broad band coverage than dialog
3. Rohan Samarajiva Nov 07
I kindly invite the reader who thinks this is advertising for Dialog to visit where the details of the study may be found. The study is much broader in scope than is suggested by the news report, covering multiple countries and products on a range of quality dimensions.

As an independent research organization, we do not promote one company or another. We did debate the pros and cons of identifying the operators, but it seems that association of performance with brand names is what is useful for consumers, not Company A does this and Company B does that.

We ran comparative performance ads in newspapers last November and received no complaints from operators or from consumers. We were gratified that those ads awakened the Telecom Regulatory Commission from its slumber on the subject of broadband quality.

Also please note that our report (and also what the reporter picked up) shows that SLT is superior to Dialog in certain dimensions of performance.

As the research manager says in the last para, our objective is that of getting all the companies to improve their quality of service. If readers have suggestions on how this can be done without mentioning company names, we welcome suggestions.

2. Tharindu Maligaspe Nov 07
Dialog exceeds user expectations!

Before I conclude this is a thinly disguised advertisement, please tell how many locations you have tested before drawing this conclusion. 100? 200? 500?

Have you tested anywhere near Homagama where I live?

1. Raj Nov 06
How ridiculous! LBO is doing marketing campaign for Dialog