LIRNEasia releases ‘AshokaTissa’ methodology and preliminary results of broadband QoSE testing

Posted on March 19, 2008  /  28 Comments


At a well attended public seminar yesterday (March 18) at Institution of Engineers (Sri Lanka), LIRNEasia released its Broadband QoSE testing methodology (named ‘AshokaTissa’, after the greatest collaboration between India and Sri Lanka, the movement of Buddhism across the Palk Strait) and the preliminary test results of three of the most widely used broadband packages in Sri Lanka, SLT Office (2 Mbps / 512 kbps), SLT Home (512 kbps / 128 kbps) and Dialog (2 Mbps / 512 kbps) This was followed by the responses from SLT and Dialog Broadband. The event was jointly organised by LIRNEasia and Institution of Engineers. (Sri Lanka)

Speeches/Presentations available for downloading:

Comments from the Chair – Rohan Samarajiva

Introduction to broadband and Test Methodology – Timothy Gonsalves

Preliminary QoSE test results – Chanuka Wattegama


  1. Dear Dr Rohan Samarjiva

    Seminar-on QoS Benchmarking Project-The Quality of the Test Results

    For courtesies sake not to cause embarrassment I refrained from highlighting that the test you have conducted are not at all representative of the composition of the true traffic scenario and hence would be detrimental to the public interests . My implicit reference to the issue on traffic relations to draw attention to this issue was un-answered. The presentation though it is the accepted norm, failed to deal with the test limitations and its likely errors. It also failed to make any reference to the recognized international recommendations which are lingua franca of QoS management in IP networking. The presentation also did not refer to the well known sources conducting and publishing on the web, real-time QoS measurements which are highly reliable Furthermore any operator interested could permit access to the source to monitor their network performance real time and publish on the web. Alternately the operators may collectively provide access to a node like an IEX at which the group announces their presence. Hence to create the impression in the public that the results of your presentation are bona fide is unjust and is likely, as in the past provide collusive expertise opportunity to exploit the market for perpetuation of their lucrative job opportunities.

    The measures to boost the market competitiveness, the core aim of the white paper of 1985 on the sector reform that is enfeebled in particular after 2002, impairing price and quality of broadband services and thwarting its growth, was side tracked at the seminar. In this respect your response to my query on unbundling in India, you quite assertively in a rather condescending tone to convince the participants, conveyed the impression that unbundling of the local loop in India has been coffined , nailed buried and there is absolutely no possibility of resurrection and that Sri Lanka too should do the same Is it the truth, the whole truth and nothing but the truth ? In this regard I shall thank you to publish on your blog your assertions made on abandoning of local loop unbundling in Sri Lanka together with the attached report and your comments if any

    In the mean time kindly provide the contact particulars of the seminar panelist

    Kind regards

  2. Mr Gunawardene challenges the claim that I made in response to his claim that India had implemented local loop unbundling:

    Here is what the TRAI document says: “TRAI recommended unbundling of local loop of incumbent operators in its recommendations on “Accelerating Growth of Internet and Broadband Penetration” date 29th April 2004. However due to complexity in implementation, DoT did not accept.”

    TRAI’s Jan 2, 2008 Recommendations on the Growth of Broadband, p. 56. Report:;

    I agree that local loop unbundling is a theoretically optimal solution. But it is difficult to practically implement even in countries with good governance and strong regulatory agencies. It has been implemented in Japan, Korea, most of Europe, etc. But it has not been implemented in countries like Sri Lanka.

    Concepts that work in developed-market settings cannot be imported wholesale to countries that do not have the necessary conditions. Mr Gunawardene should know that these conditions do not exist in Sri Lanka, because he was the person who crafted the 1991 Sri Lanka Telecom Act, as the then Director of Telecommunications.

  3. Sanath Siriwardena

    Dear LirneAsia,

    As far as I know, the intention of the test is to get a snapshot about broadband speeds in Sri Lanka by taking SLT and Dialog. Good work indeed. What you have done, we must appreciate.
    The broacher provided at the presentation did not provide much about the exact testing methods. I will really appreciate if you could provide some of the important details about the testing methods for our reference.

    Little bit about ADLS speed test history

    In 2006, TRCSL conducted a test on ADSL platform, under the supervision of director legal ,Mrs Amarasiri. Mr. Priyanka Undugogage head of ADSL SLT(at that time) and Mr.Mahinda Herath, GM regulatory matters and MS Anusha from legal dep(all from SLT) were presented at the time of the test. (Test was conducted by the deputy director-technical, at TRCSL Mr. M.C.M Farook)
    The intention was to test the QoS in SLT ADSL from 0800-1700(in a weekday) using 512kbps and 2Mbps lines. For testing purposes, number of parameters were taken to maintain the accuracy of the test.
    We have used number of servers to measure the download speeds in both packages.
    If you need more details, on the same we can provide.
    Now I would like to share your methods of testing.
    If you could provide these details, I appreciate very much.
    1.Testing servers and their addresses
    2. Sample taking method (number of days?)
    3. Secondary parameters and details

    Thank you
    Sanath siriwardena

  4. The Broadband Quality of Service test plan and methodology is available at the following link.

  5. An interesting and relevant study of local loop unbundling experience worldwide is available at

  6. I thank Mr. K. K. Gunawardena for accepting my request to continue the discussion on-line. I am sure the points he make would immensely help us to further develop our methodology.

    Before responding, let me present the project background briefly.

    The absence of stringent broadband regulation (not limited to Sri Lanka) and the typical operator reluctance in responding to consumer issues, called for a simple method the consumers themselves can monitor performance of broadband and encounter the operators directly rather than depending on regulator action, which is an unusually long shot. Operators often take the ignorance of consumers’ and the lengthy complaint process to their advantage, and we wanted a more effective and straightforward answer. (We are still in phase I of developing the same)

    The idea was to develop a methodology to run from the consumer end, without operator/regulator knowledge. One key requirement was to take operators (and if possible the regulator) out of equation. The methodology should ensure (a) replication of tests by any consumer with basic PC literacy and (b) use of information available to any ordinary consumer. (For example, ‘contention ratio’, might be relevant but not used as it is not public knowledge)

    So what were our objectives? – The tests were not the end, but just means to the end. The raw data gathered were only to create the big picture. Once the picture is in public domain, in a competitive environment, operators have to take corrective action. (I think we have partially achieved this!). Then the raw data would be of no use.

    When Galileo dropped two iron balls from the tower of Pisa, he did not measure the time they took to hit ground to the micro second. He wanted to prove a point and he did. None of us know how long the balls took to hit earth, but we all know what he proved.

    We have no intention of telling the operators what they should do. Micromanaging their networks is not our job. Let them do what they can do best while we do what we are best at. So QoS Management was out.

    With this lengthy background (I hope I do not have to repeat) let me take the issues Mr. KKG raises one by one (about methodology. Rohan has already responded re LLU).

    Issue 1: (Mr. KKG says) Our tests are representative of the composition of the true traffic scenario.

    Yes, we agree, but that does not make significant difference to the outcome. To a layman, this is like me saying it takes one hour to reach Dehiwala by auto, because I know a car takes that long. Auto may travel slightly faster (because it can go through small spaces) but when creating the big picture we just ignore such minor details.

    I will be grateful if Mr. KKG educates us how an ordinary consumer finds this piece of information.

    Issue 2: The presentation failed to deal with the test limitations and its likely errors.

    Yes, we agree. Given the time constrains (20 mins for introduction and methodology and 20 mins for results) we explained what we did rather than the limits. We presumed audience/operators will bring the issues during Q&A.

    Neither operators nor audiences asked about limitations and it is not a surprise. Methodology is public and the tools are not new. Anyone with little knowledge in Electronics Engineering (and many even without that) can easily make out the limitations. (Anyway, I will make a seperate comment on this asap)

    Issue 3: The presentations(s) did not make any reference to the recognized international recommendations which are lingua franca of QoS management in IP networking.

    QoS management was out of our scope. Even if it is within scope, why repeat something ‘lingua franca’? :-)

    Issue 4: The presentation also did not refer to the well known sources conducting and publishing on the web, real-time QoS measurements which are highly reliable.

    We have briefly referred to a similar studies being conducted in Korea and UK. Think this is enough for our purpose.

    Further, the audience was there to hear something new. I am sure, in this Internet age, they would were not there to learn something on the web. So why bother?

    Finally, I am sorry that Mr. KKG could not learn some of the things he wanted by attending this event. I am further sorry that we do not have time to educate him on points out of the project scope. However, I am sure Mr. KKG could find the information either from web or IESL library.

    Any other suggestions by Mr. KGG to further improve our methodology are always welcome.

  7. For the benefit of Mr. KKG and anyone who is interested, let me state some limitations of our methodology and our explanations.

    1. It does not address the issues in the last mile for every consumer. (but from the ISP everybody goes in the same pipe) The last mile issues will be addressed in the next phases of our project when we test a large sample of links.

    2. The only way we could download a 5 MB file (we wanted this take care of fluctuations) was from a mail server. This could have probably created a packet loss between the Internet/mail servers. Still from our tests, we found it had not presented some packages to offer near 100% throughputs on selected times. So if one can, there is no reason why others cannot.

    3. The international server we have used was Of course, we could not test all servers and Yahoo is one of the most commonly used sites. Also out trace route records showed later there were no bottlenecks from the Yahoo end. Same issue with national/local servers.

    4. We did not check every possible application and also multiple downloads. However, we have to draw the line somewhere. We assume the impact is same on all packages, so taken care of.

    5. We used ping commands which depends only on ICMP packets. Other techniques such as HTTP ping that use IP packets involve processing at the server and thus would have higher RTT. Many routers have the capability to give lower priority to ICMP packets, but in practice this is not an issue except during denial of service attacks.

    6. We tested one package per day. There was a practical difficulty doing more than that because to take one set of readings we had to repeat each test three times and take average. On average, to test one reading it took 2-3 hours.

    We are aware of these limitations and will certainly take measures to address them when we proceed to the next phase.

  8. Seminar-on QoS Benchmarking Project & the Sidetracked Local Loop Unbundling (LLU) the vital prerequisite to Boost Market Competitiveness of Broadband Connectivity

    1.Reply of Dr Samrajiva and the Suppressed Truth

    Dr Samrajiva (Dr. S ) in his reply despite my request has not published the attachment to the email that shows the stance of the DOT as reported by the Times of India It is the just the reverse of what Dr S cited from the TRAI Report that DoT did not accept TRAI’s recommendation dated 29th April 2004.on the local loop of unbundling (LLU).

    The Times of India on 26th July 2006 however states “The DoT, with an eye on the consumer benefits, had formed a sub-group consisting of multiple stakeholders to explicitly push forward the LLU agenda… ” This incongruity arises from the fact Dr S though requested in my email; has not disclosed the whole truth as stated in the TRAI report in order to impress the seminar participants that India has abandoned LLU and Sri Lanka should do likewise.

    Prior to responding to comments of Dr.S I wish to therefore with due diligence apprise of the fact to the misinformed seminar participants and the public that the LLU in Sri Lanka was stalled not due to any constraints of the nature referred to by Dr. S. The LLU was stalled from he very incept of the start up of the ADLS service on the recommendation of the consultant who prepared around 2002 the External Gateway Operator (EGO) license for opening up of the international telephone service despite it was not in his terms of reference. An entity called the Public Interest Program Unit (PIPU) that engaged this consultant inveigled the then Govt to implement this consultant’s report that contained the above recommendation included rather inconspicuously. Thus LLU the most vital to speedily realize the goals of e Sri Lanka was never ever tried out in Sri Lanka

    Since then the ADSL service has continued to remain a monopoly on the strength of the aforementioned recommendation. Its outcome is the growing public discontent communicated by the public to the Regulator reporting the very high cost poor quality and very slow pace of growth of the broad band access service.

    1.1. Sri Lanka’ Telecom sector the first in the Region to Digitalize the Network and Liberalize its market -India Followed and Sought Guidance

    In this regard I wish to point that Sri Lanka was the first country to introduce digital switching technology in the Region It was also the first country in the Asia region to liberalize and privatize the sector It did not wait to follow India In fact India did consult Sri Lanka on these issues and even send out a team to seek opinion on digital switching So there is no rationale that in view of your purported claim Sri Lanka should follow India

    2. The Vast Untapped Latent Bandwidth of the Copper wire That has No Shelf Value

    This collective latent untapped latent bandwidth of the cooper wire is immense and the accessibility it offers nationwide is vast. This was also endorsed by the SLT speaker who made presentation at your seminar. Harnessing this capacity speedily and with very low marginal cost was made feasible with the advent of DSL technology over a decade ago Since then the capacities of these DSL devices have increased and the prices have diminished over twenty five fold

    There is much hue and cry to improve usage efficiency of Sri Lanka’s capital resources to cut down waste However up to now barley less than 1% of this available latent capacity has been harnessed. The fact that this idling capacity has no shelf value is ignored not withstanding the fact that activation of Collective latent Band width is of paramount to move to the next generation technology that offer low cost and feature abound services essential for the e Sri Lanka program. SLT is only now considering – nearly six years after introduction of ADSL – to deploy the capacity for IPTV But what would be the size of the IPTV market when majority cannot even afford the cheapest highly limited ADSL services offered by the SLT ?

    3.The comments of Dr.S unsubstantiated and not pertinent to the LLU

    The unsubstantiated comments Dr.S has made in the last two paragraphs of his response are devoid of rationale or not pertinent to LLU. For e g the comment that unbundling is only theoretical concept and not applicable to Sri Lanka is incorrect and besides it has not yet been tried out The unsubstantiated comments of the last paragraph blames the Telecom Act but Dr. S fail to cite necessary specific conditions and the provisions of Act which are stalling unbundling .he alos ignores the White Paper on sector Reform that preceded the Telecom Act

    4. Imperatives of the Cardinal Regulatory Functions

    These comments reminisce the vision of a Telecom sector reformist who said that the Telecom Acts have become the haven of employment – of lawyers and their like One of the key factors eroding market competitiveness is the failure of these expertise to recognize exhaustively the distinct roles of the cardinal regulatory functions. Interconnection is one of these cardinal functions another is unbundling a distinct sublet of Interconnection.

    Interconnection is not exclusive to the Telecom sector It is also featured prominently in several other the economic sectors But its application by these expertise fails to comply fully with the role of Interconnection as featured in these economic sectors including that of the monopoly era telecom sector This is the cause of the dismal failure for the interconnection as recommended by these experts

    5. ITU forewarns of the Imminent Failure of the EGO Licence & Offers Assistance Free Charge

    The interconnection conditions of the EGO license issued by the PIPU is a typical example that feature most notoriously these lapses mentioned above. The TRC, the sector regulator, apparently doubting its asymmetry harmful to new entrants referred it to the International Telecommunications Union (ITU) the UN’s specialized agency in telecommunications, for its opinion. The TRC was promptly forewarned that it would be an imminent failure, not to proceed with it and offered expertise free of charge to redo the EGO license. The ITU also offered assistance free of charge to address the more important issues such as price caps and unbundling The TRC responding on phone said that the PIPU has turned down the offer on the grounds it has awarded the consultancy on the approval of the PM It is gathered that the failure to accept this free offer made by the ITU has cost Government a consultancy fee of nearly Rs 26 million, excluding the monthly remuneration of US$ 25,000 paid to the head of the PIPU Subsequent to he dismal failure of the EGO licence TRC sent an emissary to the ITU The emissary was apprised of the solution and thereafter followed up with all the documentation of the solution But so far no action ahs been taken apparently due to the influence of vested interests of the cartel like market oligopoly and others supporting them The yet prevailing impact of this EGO Licencing unchecked by the regulator is evident from the problems encountered by Airtel recently licensed y the TRC

    6. LLU vital to boost broadband connectivity for realizing the goals of e Sri Lanka and Mahinda Chintana

    An excerpt of the Interconnection condition of the said EGO Licence is attached hereto I shall appreciate if Drs S the hailed telecom think tank of the region provides his invaluable observations in respect of these Interconnection conditions of the EGO Licence. Likewise Dr.S should enlighten the public on the necessity for the PIPU’s expert outside his terms of reference to recommend to block LLU notwithstanding its importance to boost competition in broadband connectivity which is of paramount importance for realizing the goals of e Sri Lanka and the Mahinda Chintana Benchmarking of broadband connectivity QoS with crude methods of testing without fully understanding the underlining services and its traffic relations will NOT help to boost affordable broad services even in the foreseeable future let alone in the long run. Its out come will be no different to that of the EGO license which barred entry of new operators to help the market oligopoly to sustain international calls charges much above the global market trend What is urgently needed is a boost in competitiveness of the market and not bench marking by amateurs

    Please publish my entire email on your web and also forward it to the panelist of the seminar

    Kind regards

    Copies to

    Director General TRC
    Nelum Jayaweera DirectorEconomics TRC

    The harmful impact of the EGO Licence unchecked by the TRC is yet prevailing. It is evident from the problems reported by AirTel the company recently licensed by the TRC and also the growing public discontent of the ADSL service. It is also affirmed by the response of Mr. Nelum Jayaweera, the Director Economics TRC when he was contacted on phone concerning an ADSL problem . He asked me “Why are you people always complaining about SLT tariffs and billing The tariffs are going down. Air Tel is offering calls at Rs 2 per minute but they are not getting interconnection to provide the service but you do not complain about interconnection ! ” This remark also affirms that TRC expects the general public to offer expert advice to solve the unresolved interconnection problems .

    Following the dismal failure of the EGO licence predicted by the ITU the TRC sent an emissary to the ITU to study the connected issues. The emissary was apprised of the solution and thereafter followed up with all the documentation on the solution But it appears from the above remark of Mr. Jayaweera that so far TRC has been unable to taken any correctives action to resolve the problem To avert another EGO like debacle TRC should warn the public of the pitfalls of the crude bench methodology presented at the QoS seminar

    President IESL, atten Eng R Gunawardana .

    Executive Secretary IESL Attn Eng Dhammika Welhenge

    President IEE AttnEng . M.Zubair

    Hony Secretary IET Sri Lanka Attn Eng Mr K B M I Perera

    Benchmarking of broadband connectivity QoS with crude methods of testing without any understanding of the underlining services and its traffic relations will NOT help to boost growth and affordability of broadband connectivity in the foreseeable future let alone in the long run. On the contrary such benchmarks could be readily exploited to serve the interests of the cartel like broadband service market and mislead the innocent public.

    The LLU has the potential to harness speedily, with minimal marginal cost, the nation wide spread, vast untapped bandwidth of the copper wire, that is most vital to boost competition in broadband connectivity which is of paramount impotence for realization of the goals of e Sri Lanka and the Mahinda Chintana

    It is best that IESL and IET jointly convenes a forum to identify the cause of and the speedy solution for the broadband access debacle to squash its exploitation of the end user and inflicting immense harm the national economy Jus to give and idea how much we are exploited the Current ADSL prices including free modem range in Japan from Y2,500( Rs 2600) to Y3,500 (Rs 3700) ($20-$28) for an 8 Megabits per second line, up to Y3,500 to Y4,000 for a 12 Mbps line, including the Internet access fee. We in Sri Lanka pay Rs 2225 that too without a free modem for a paltry speed of 0.5Mbps

  9. What’s all this fuss about a local loop when the bottleneck is in the International loop?

    I thought the outcome of the whole Lirneaisa study was that operators don’t buy enough International bandwidth?

    This is all very confusing.

  10. I am sorry to say this but some people are just like ‘tube lights’.

    Even after my presentation with empirical data and Mr. Madummage of SLT implicitly addressing this issue Mr. K. K. Gunawardana fails to understand where the bottleneck is.

    He keeps talking about Local Loop Unbundling. (LLU) What is LLU? Let us look at Ofcom definition.

    Local Loop Unbundling (LLU) is the process where the incumbent operator makes its local network (the copper cables that run from customers premises to the telephone exchange) available to other companies. Operators are then able to upgrade individual lines using DSL technology to offer services direct to the customer.

    According to Mr. Maddumage, the key reason why SL operators cannot offer high bandwidth (and/or lower prices) is INTERNATIONAL BANDWITH remaining a scarce resource. (There is a separate discussion on this at

    So creating local competition has little impact. International bandwidth prices are decided by big guys in USA, and NOT by local operators. To them SL market is nothing. Basically we are at the mercy of them.

    Of course, the situation in countries like Japan are entirely different, because most of the content users seek is local, either locally developed or international sites like Yahoo mirrored. So how a user in Japan sees international bandwidth is different the way I, a Sri Lankan user, see it.

    Again DSL is NOT the only way one can bridge the last mile. If someone is ready to bring additional international bandwidth there are hundred and one ways now to bridge the last mile. Cable, Wi-Max and HSPA are three already used options. So many other technologies are in pipeline.

  11. Mr K K Gunawardana

    Dear Mr Gunawardana

    This is only the second time since we started in 2004 that I am compelled to make a request from someone writing to our blog. The first time was when someone personally attacked late Professor V K Samaranayake. I requested he stop; and he did.

    My request to you this time is to please stick to professional matters on our blog as this is not a place for people to fabricate stories, just to sling mud. This is our blog and we can surely not publish your comments, but that is not how we work. We respect everyone’s right to disagree with us, but please do keep your arguments at a professional level.

    Thank you.

    Harsha de Silva, Ph.D
    Lead Economist, LIRNEasia

  12. In defense of MR KKG: the level of competition among operators has a bearing on quality of service.

    If there is only one supplier, there is little incentive to improve quality by buying adequate international bandwidth. Without competition, there is no alternative but to rely on regulatory pressure. But there is competitive pressure now, and it will increase in the coming months.

    Mr KKG believes that LLU will create competition. I agree, but argue that it’s not feasible at this time. LLU requires a very strong regulatory regime which we do not have. No developing country like Sri Lanka has succeeded in implementing LLU.

    He thinks that LLU was not implemented because of a consultant report in 2002. He is mistaken. That exercise was about ending the international gateway monopoly under a very tight timeline; LLU was not on the agenda. It would have been foolhardy to try to open up the local loop when we were dealing with a very large challenge.

    There is nothing preventing the TRC opening up the local loop if it so wishes. It is only Acts and court decisions that constrain regulatory action, not a single sentence in a consultant’s report.

    We have a second best solution in the form of WiMAX and HSPA beginning to provide competition to SLTL. I have long advocated removal of the obsolete restriction on other operators drawing copper. If that restriction is removed additional competition is likely to arise.

    With the existing and emerging competition and the publicity given to the large gap between delivery and promise with regard to international connectivity, we are likely to see improvements in quality. It will not happen overnight, but it will happen.

    I look for solutions that depend on competition and the reduction of information asymmetries; Mr KKG looks for command and control solutions.

    This is a honest difference of opinion and approach. There is no need to engage in vituperation. He can do his thing; I can do mine. One or both will work. The consumers will benefit. The proof of the pudding is in the eating; the superiority of the solution is in the results.

    Mr KKG was a member of the committee that made recommendations on restructuring the industry in 1982. He was Director of Telecom for most of the 1980s. He supervised the giving of connections to firms in the Katunayake FTZ in the early 1980s and the efforts to prevent people from using automatic diallers, because they had been given phones but not dialtone.

    He set the interconnection terms for the first new entrants–Celltel in particular (which was the model for subsequent interconnection arrangements). He was responsible for the 1991 Act that corporatized the Department of Telecom and made him the first Director General of Telecom. I came into the picture much later, in 1998. Whose actions yielded the better results?

    My suggestion is that Mr KKG use his considerable energy and intelligence to solve some problem: LLU or broadband QOS, or whatever. No point in venting all this anger on me. I am not standing in his way. I’d much rather do something useful than engage in these endless polemics.

  13. Why this guy Gunawardene or somebody brings BB prices in Japan to discussion? Do we normally compare Sri Lanka with Japan? Do we compare Micro cars with Toyotas? Or Shinkansen with Udarata Menike?

    Not only BB prices, but even phone charges are very low in Japan. It is logical. Japan has an enormous market potential for telecom services and companies have made huge investments in information infrastructure knowing the returns. So they can offer speeds up to not 8 Mbps but even 100 Mbps at low rates. When the volumes are high, production cost of services too will come down and quality improves.

    Given the minute market size in Sri Lanka for BB (What is our PC penetration any way?) who will make such investments? Forget the volatility of the market and inflation.

    Gunawardene, compare apples with apples, not with oranges.

  14. Mr K.K.Gunawardena’s letter addressed to Dr Rohan Samarajiva in its substance and form clearly indicates how and to what extent anger, personal animosity, professional jealuosy and intelectual bankrupsy can drive a those days professional to mental and physical deteriorationis. Hope he himself understand his predicament and avoid more public engagements.
    There is a day every one of us to retire from public life.

  15. The traditional view has been that facilities based competition that involves new entrants building their own local loops is costly and places an undue burden on new entrants. LLU forces incumbents to resell their lines to new entrants and significantly lowers barriers to entry and stimulates competition. There is significant evidence of how mandated unbundling in the EU since 2001 has stimulated DSL growth and competition in the broadband market. EU countries that require LLU have higher broadband penetration compared to countries that don’t require LLU.

    The above is the traditional view. Things have changed significantly since 2001. Local loop is not as significant a bottle-neck as it used to be with the availability of new wireless technologies.

    For example, it is precisely because of WLL that new entrants in Sri Lanka have been able to compete with the incumbent. Lower cost per line of CDMA versus copper lines has lowered barriers to entry. I doubt if SLT invests in copper lines anymore. Fixed-mobile substitution has also resulted in consumers giving up their fixed lines in large numbers in the USA as in India.

    So fixed lines aren’t what they used to be in recent past. Furthermore, in developing countries such as Sri lanka, LLU was meaningless when penetration was in the single digits. What lines are you gonna unbundle? From policy perspective, it may make more sense to promote facilities based competition when large parts of the country didn’t have any communication infrastructure to speak off.

    Coming to broadband and LLU. With wireless technologies like WiFi, Wimax, HSPA, LTE it is not only possible to bypass the local loop of the incumbent but also to provide broadband services on the move to consumers. The convenience of mobility is huge and fixed lines are at a great disadvantage. In this context, if someone still insists on LLU I would think that person is stuck in the past. The competitive landscape has changed considerably and policy prescriptions must also change.


  16. I suggest, as a reputed professional organization, Institution of Engineers (Sri Lanka) to seriously consider revoking the membership of any member, irrespective of the class and seniority, who openly shows derogatory behavior not only by making petty personal attacks but also using their status in doing so. Otherwise they will never learn.

  17. As a representative of ITU, an inter-governmental organisation dedicated to the development of international standards, I would like to add that I’m happy to see this issue being debated with such passion and gusto. It is certainly important that Broadband QoS testing is carried out in a way which makes international comparisons possible, using objective criteria. However is also worth remembering that broadband speeds, worldwide, are doubling roughly every 12-18 months, and therefore QoS measurement is very much a moving target. Thus, while QoS measurement provides useful data on performance for customers and service managers, it should not be perceived as a tool for micro-managing the development of the industry.

    In this context, I think a useful lesson is offered by the Singapore regulator, IDA (see, for instance, and Regulation/20060425150155.aspx). The Singapore approach might be summarised as follows:

    — report just a small number of variables (availability, international latency, national latency and complaints), but do so regularly (monthly) so that performance problems can be quickly recognised;

    — distinguish clearly between international and national bandwidth as this relates to issues that can can can’t be controlled;

    — include the number of complaints as a consumer-oriented measure to supplement technical measure;

    — be flexible about measures, dropping those that become less useful over time.

    I hope this contribution to the debate proves useful.

  18. Seminar-on QoS Benchmarking Project & the Sidetracked Local Loop Unbundling (LLU) the vital prerequisite to Boost Market Competitiveness of Broadband Connectivity

    1. Comments Of Lirneasia and its stakeholders published at its website

    This refers to my email dated 25/03/2008 on the subject and the comments of Lirneasia thereto published at its blog. Most of these comments except in a few instances are of flippant nature and not pertinent to the Broadband debacle (BBD) that is inflicting harshly the sector efficiency. These comments pertinent to the issue endeavor to confuse the public and conceal the adverse impact of the EGO licence, and the stalling of the LLU that is yet thriving unabated This is also self evident from obstacles encountered by Airtel, the recently licensed operator that was also affirmed by the TRC

    To dispel any confusion and doubts that may arise from these comments the root cause of the BBD, its adverse outcome, benefits deprived to the market, the revenue losses to the SLT and the Government and solutions thereto are dealt with at paragraphs 2 to 6. Also given at paragraphs 2.4 of the attachment is the subtle linkage of the EGO licence with the BBD that is not readily revealing and has been and is exploited to confuse decision makers . For details of the causes and solutions in particular the interconnection problem, LLU the long overdue national directory and number portability you may also kindly study the proposal given to the TRC’s emissary dispatched to the International Telecommunications Union (ITU) following the failure of the dismal EGO licence forewarned of by the ITU

    2. The Adverse Impact of Weakened Broadband Connectivity on the Goals of e Sri Lanka and the Mahinda Chintana

    It is rather baffling to note that baring facts to alert the Telecom Sector’s vision confused and dimmed ( by misleading expertise who have made telecom acts their heaven of employment ) to focus on the most vital few crucial to boost its enfeebled market competitiveness, has perturbed the Lirneasia. As stated in my email to Dr Samarajiva (DrS) the broadband debacle is inhibiting migration to NGN technology that is of immense importance for realizing the goals of e Sri Lanka and the Mahinda Chintana .The NGN technology could on the overall cut down cost of information services dramatically and also offer a wide range of feature abound applications.

    2.1. Telecommuting services To Improve National Productivity and Easing the Strained Transport Sector

    One of such important feature abound application is telecommuting. It has the capability to immensely improve the national productivity, in particular that of the strained transport sector: telecommuting could help to spread out the diurnal peak hour passenger loads uniformly over the day thereby improving utilization of the available off peak idling capacity , improve commuting time and workforce productivity, cut down significantly diurnal migratory work force involved in information processing commuting to cities, reduce the escalating transport sector fuel cost ; mitigate considerably the harmful impact of the strained transport sector on the economy and the environment etc. and thereby help to curb the rising cost of living

    2.2. Interconnection of Disparate Networks on Demand

    Another application of the NGN technology is its capability to readily connect on demand disparate networks of any operator current or new entrant. To satisfy this requirement it is vital to comply with the latent public good character of interconnection as practiced by the other sectors of the economy cited in my previous email.

    2.3. ADSL Monopoly squelching sector regulation and competition

    One may argue that it is far more important to privatize quickly; one can deal with the issues of competition and regulation later. But the danger here is that once a vested interest has been created, it has an incentive, and the money, to maintain its monopoly portion. squelching regulation and competition and distorting the political process along the way. Following the consultancies of 2002 engaged by an entity called the Public Interest Program Unit (PIPU) it is precisely the current status of Sri Lanka’s Telecom sector This is self evident from the dismal failure of the External Gateway Licence (EGO) that barred new entrants from operations , the resulting high international call tariffs as compared with that of the global market; the emergence of international call smuggling ; the avalanching of the broadband debacle, the obstacles encountered by the recently licensed Airtel and several other critical issues such as the high domestic tariffs that PIPU failed to address in 2002, not withstanding the trend of near zero tariff predicted by the World bank as far back in 1995 .

    2.4. The unavailing of Sri Lanka’s Proven Lead Role n the Telecomm sector Digitalization Liberalization in the Region & The Subtle Linkage to the Broad Band Debacle

    Had these consultancies of 2002 not been engaged and TRC’s efforts reported at 4.1 not been foiled then apart from saving the exorbitant consultancy fees, just as much Sri Lanka, in the Region, was the first to digitalize the national network and also the first to liberalize/privatize the telecom sector it could have also led in the region the Local Loop Un-bundling . These consultancies engaged by the PIPU not only inveigled the then Government to enforce the recommendation of these consultancies that stalled the entry of new operators (nearly 25 ) to the market and, the LLU, it also stifled the growth of the IP based backbone capacity these operators would have invested to provide low cost VoIP international telephone services. The diurnally varying idling capacities of these backbones could have been easily availed of to carry data traffic appropriately mixed with voice at near zero marginal cost Likewise it also blocked growth of users’ access to broadband services This in turn deprived colossal revenues
    to the SLT and the Government of Sri Lanka. The revenue would have surged with upfront cash flow and annual revenues arising from activation of the idling latent bandwidth of the copper loop and the resulting rapid growth of broadband access services. In other words had the PIPU not inveighed the Government to enforce the recommendations of these consultancies the broad band debacle would have completely averted and also international call charges would have significantly dropped

    2.5. The Lirneasia’ s Narrow Perspective Squelching the Goals the White Paper on Sector Reform

    Alerting the Sector’s vision dimmed by the misleading expertise to focus on the vital few essential to solve the Broadband debacle seemed to have been viewed narrowly by the Lirneasia as an infringement of and a threat to the interest of its stake holders. On the contrary these facts are bared to address the squelched precautionary measures essential to protect interest of the nation and the general public spelled out in the White Paper that started the telecom sector reform process The co authors of this white paper were Professor K.KY.W Perera Dr Mohan Munasinge and Professor J.A Gunawardana Were it not for their unswerving and dedicated efforts SLT also would have ended up like the CEB The asymmetries created by the numerous consultancies engaged in 2002 have squelched these precautionary measures endangering the interest of the public. The EGO licence is just but one of these numerous costly and unproductive consultancies

    3. The driver of the Broad Debacle – The Harshly Asymmetric Interconnection Conditions imposed in the EGO licence

    Transparency and market competitiveness are cardinal perquisites for effective regulation of a liberalized market and accountability It is for this reason that Dr Samarajiva (Dr S) was requested to provide his invaluable observations in respect of these asymmetric Interconnection conditions of the EGO licence and the recommendation to block LLU that triggered the broadband debacle As he has neither published the attachment to my email nor made his observations on the said licence, quoted below is an excerpt of these conditions for the information of the public

    The ferocity of these conditions extremely hostile to new entrants was compounded by the recommendation made by the same consultant to stall unbundling the local loop. Its aim was not only to block competition in the wired broadband access services but also to block access to cheap international calls available on the internet Its subtle purpose was to strengthen the international telephone service oligopoly created by the EGO licence The aforementioned subtle linkage of these asymmetries with the current broadband debacle, squelching regulation , competition and distorting the political process explained previously is unlikely to be readily revealing to many let alone even some professionals on the subject

    3.1. QoS Bench Marking a Prescription to Prolong the Agony of the Broadband Debacle

    Had the new entrants been able to operate and, LLU was not stalled not only the international call charges would have dropped significantly there would been as aforementioned abundance of backbone capacity on IP availed of for voice services and as well as broad band access at the end users premises The efficiency of the backbone would have improved through the use of optimal mix of real and non real-time applications thereby further aiding to not only improve the QOS but also helping to bring down the cost to the user It may be now clear that QoS bench marking as justified by Lirneasia is not reflective of limiting factors of the network and hence is not a remedy but a prescription to prolong the agony of the debacle

    4. The Lirneasia’s Vacillating Views on LLU

    What is intriguing is that the Lirneasia’s stance on LLU is same that of consultant engaged by the PIPU but modulated in time by varying convoluted views devoid of understanding the basics : on one instance Dr S says that problem is in the Telecom Act but fails point out the specifics thereof; on another instance to the press he comments “ The most cost-effective broadband technology for ordinary people and small businesses is ADSL. .only one service provider in the industry is permitted to lay copper….. This should be stopped and a level playing field created for the industry…”. At yet another he says the “Concepts that work in developed-market settings cannot be imported wholesale to countries that do not have the necessary conditions. …that these conditions do not exist in Sri Lanka, and most recently he cited that India has abandoned the LLU implying that Sri should follow like wise

    4.1. The TRC’s Efforts Foiled by the PIPU

    The TRC officials at the time deserve a special commendation for referring the EGO license’s asymmetries to the ITU the UN’s specialized agency in telecommunications, for its opinion. As mentioned previously The TRC was promptly forewarned of the imminent failure of EGO licence , advised not to proceed with it and offered expertise free of charge to redo the EGO license. The ITU also offered assistance free of charge to address the more important issues such as price caps and unbundling The TRC said that the PIPU has turned down the offer on the grounds it has awarded the consultancy on the approval of the PM

    Subsequent to the dismal failure of the EGO licence TRC sent an emissary to the ITU The emissary was apprised of the solution and thereafter followed up with all the documentation of the solution It is understood that solution proposed to the emissary sent to the TRC have also been squelched by the same vested interests

    5. The Threats of the Lirneasia’s Defense Force A Reminisce of Dr Stiglitz’s Publication on Globalization and Its Discontent

    To the defense force of Lirneasia its General HS, the very familiar apple orange nursery story teller , the cowardly engineer of non disclosed identity purporting to be a member of the IESL, its stake holders and others on its payroll alike whose comments though not pertinent to the issue, I apologize if I have infringed on your interest commercial or any other. I reaffirm my concerns are in respect of Public Interest unaddressed by the Lirneasia My colleagues in this context say that Lirneasia’s ADSL blog was scathingly criticized by several including expatriate Sri Lankans and academics accusing it to be far worse than the notorious “ Great Impostor” and its activities far worse than that of Wolfervitz the ousted president of the World Bank But at the time the Lirneasia’s defense force has been passive and silent Their comments reminisce of the threats received by those who dare to challenge doers of the globalization’s ugly business reported in the “ Globalization and Discontents ” by Dr Stglitz’s the Nobel laureate and at one time the Chief Economist of the World Bank

    6. TRC, IEE IESL forum to address the Broadband Debacle

    I do not wish to engage in any more unproductive squabbles with Lirneasia . The solutions to these pending problems have been provided as far back in 2003.In 2005 another report at the request of Hon Mr T Witarana was submitted to the ICTA advisory committee of the Ministry of Science and Technology. It is best that in order to help TRC to solve the long overdue Broadband debacle that TRC IESL and IET jointly convene a forum of the expertise on the subject to identify the cause of and the speedy solution for the broadband access debacle to quash its exploitation of the end user and inflicting immense harm to the national economy

    Kind Regards


  19. To the regular readers of LIRNEasia blog Mr. K. K. Gunawardana’s long and shallow response brings reminiscences of the Sinhala Unicode chapter. Then we had Mr. Donald Gaminitilake leading a lone but stern battle to push his CAT. Now we have Mr. K. K. Gunawardana doing the same for LLU.

    We are happy that they selected our blog to carry out the debate, but unfortunately broken gramophones make only noise, not music.

    There are also few lessons Mr. K. K. Gunawardana can learn from Mr. Gaminitilake.

    Mr. Donald Gaminitilake was always open about his agenda. All he wanted was ICTA to give him some money to ‘correct’ the mistake. Mr. K. K. Gunawardana is not so honest and only passing hints to TRC. So it is left for readers to decide. Anyway, Mr. Gunawardana need not be shy. We all work for money.

    Mr. Donald Gaminitilake was also more disciplined. He never deviated from the topic and always avoided personal affronts. Sadly we do not see such discipline from Mr. K. K. Gunawardana. Apart from verbal diarrhea, there is little information in his response. Pity. I thought with the experience he claims to have, we could learn something from him. I was mistaken.

    Both me and Divakar have raised questions about the possible effect of (or lack of) LLU on the broadband prices in the Sri Lankan context but sadly Mr. K. K. Gunawardana conveniently evade them, leaving us nothing to respond.

    We like to engage in a public debate on all issues preventing quality broadband services in Sri Lanka (not just LLU) but that will be of public interest only if confined to the topic. General public would not attend a seminar to see ‘experts’ slinging mud at each other.

    Finally Mr. Gunawardana seems to have mistake that all those who respond here work for LIRNEasia. No, certainly not. This is an open blog and anybody is free to comment.

  20. Isay Gunawardene, whoever you are please stop talking nonsense. You seem to know nothing about BB market in Japan. That is still you try to compare Japan to Sri Lanka and call others nursery story tellers. Isay if there is any nursery story teller it is you.

    If you do not know learn now. Do you know when Japan started offering BB and what was the price then. Even around 2001, a monthly DSL package cost about $ 4,000 and that was for 3 Mbps. It was with the massive expansion that now Japan can offer 100 Mbps BB at such low prices. It is no surprise. Any country with such a high volume can offer low prices. Number of BB subscribers in Japan is more than the population in Sri Lanka. Are you still comparing such a large market with Sri Lanka? Man, you know neither about Japan nor Sri Lanka.

  21. Thanks BB, your comment provides some food for thought for me.

    To unbundle or not how many local loops we have in Sri Lanka? Not more than 700k or even less. The number of ‘fixed phones’ might be about 2 mil but that include a large number of CDMA and before that WLL phones. SLT is no more bothered providing Cu loops to homes.

    So are we really taking about ‘unbundling’ these 700k loops? What difference it can make? Mind you, about 80% of these lines might be in Colombo.

    For example a village like Mahavilachchiya still there is not a single Cu loop to a home. So what competition can we make by ‘unbundling’ ?

    I agree in ‘wired’ countries like Japan, Korea or UK the case is different.

    I have seen in a Sunday newspaper article Dr. Gihan Dias advising if you do not already have a Cu loop (ie a fixed phone line) to go for WiMax to get broadband. I agree. Why worry about ADSL when there are so many more options?

    But on the other hand, unbundling mere 700k loops too can be a lucrative consultation assignment for somebody.

  22. Pardon me for barging in the middle of this discussion for something bit off topic, but thought of doing so after seeing the names of Prof Joseph E Stiglitz and Wolfowitz mentioned.

    I do not know how common the name Gunawardana in Sri Lanka, but if this is the same Gunawardana who was the key consultant in a WB funded telecom reform project in Bangladesh few years ago with N. H.Choudhury, one of the well known retired corrupted public servant as its project manager, his comments are really funny.

    Because we found what Choudhury and Gunawardana did was just a pure waste of WB money without any benefit to Bangladesh at all. These two were not too different to Wolfowitz.

    I would be grateful if Gunawardana can answer what this project did to telecom reforms in Bangladesh? (If it is the same individual. Otherwise please accept my apologies)

    Was n’t achievements of this project was limited to providing computer facilities to BTRC and also PCs to some unknown individuals who reconginsied them as consumer organisation representatives? What happened to those PCs? Were they handed over to WB after the project period? Gunawardana can also elaborate on the size of this project and why such a large amount was needed just to lay a computer LAN and to buy some PCs.

    I understand Gunawardana and other consultants like Dr Eun-Ju-Kim, Raimondo Giuliani and R.K. Bhatnagar were paid for conducting workshops (Gunawardana can tell the amounts) I will be glad to know what he taught at these workshops and for whom and what was the outcome? Did these workshops train even a single staff member of BTRC? I am glad to know the answer from Gunawardana.

    Finally I would like to know from Gunawardana whether he ever did anything he suggests doing in Sri Lanka, in Bangladesh? If he did not do that when given enough opportunity why not?

  23. We appreciate if all our readers stay within the topic. That is the only way to make the discussion productive.

    In case you have forgotten what the topic of this thread is please see above. We still wait for someone to give a constructive feedback on our ‘AshokaTissa’ broadband benchmarking methodology.

  24. Dhathika Sirimanne

    Please don’t smack me for saying this but Mr. K. K. Gunawardana is right. Unbundling is an essential condition to expand the DSL services in any country. Please see the graphs of DSL development. They tell enough.

    Even you recognise the importance of unbundling in your own research.


    One major policy stumbling block in the growth of broadband services is the issue of Local Loop Unbundling (LLU). TRAI had suggested, in April 2004, non-discriminatory LLU so that the access networks of the basic service operators (BSOs) could be shared with internet service providers and other competing operators in order to improve broadband penetration.

    However, BSNL and MTNL view LLU as infringements on their property rights and have hence prevented the government from implementing LLU. TRAI, in its letter6 to the Department of Telecommunication (DoT), pointed out that:

    While recommending for Broadband Policy, TRAI had made many significant recommendations, a few critical ones which were not accepted which are mentioned below:

    a) Recommendations pertaining to Local Loop Unbundling
    b) Other Fiscal measures like tax concessions for

    Broadband equipments & services

    While framing the recommendation for Local Loop Unbundling, TRAI had detailed discussions with the incumbents (BSNL & MTNL) as well as other service providers and considered the best international practices. It was suggested in the recommendations that Local Loop owners should be given an opportunity to decide the areas where they would make their own investments for providing the broadband services and also to decide on the type of unbundling, depending on their commercial objectives. Also the local loops which were installed 5 years back only were recommended to be unbundled and those installed in last 5 years were left to be utilized by the owner themselves.

    It was also suggested that, in case the broadband target is not achieved in the first year, review of above specified arrangement should be conducted. It can be observed from the current trend that the incumbents (BSNL & MTNL) are not able to make full utilization of their infrastructure themselves and neither through franchisee option provided to them through Broadband policy. Because of this, the existing local loops could not be utilized by incumbents as well as by the private operators.

    Although the success rate of LLU in other countries is still under debate, in a competitive basic service market, LLU as a regulatory intervention is an instrument to discipline market power, reduce monopolistic bottlenecks and potentially to provide the way for innovative service offerings such as broadband connectivity. Most of the current broadband connections are in large cities, where private BSOs and cable operators also provide service through their own local loop. However, penetration is poor in smaller cities and outside urban areas where only government operators are providing service.

    This is where LLU will have a positive impact. TRAI recommended LLU only for lines installed five years or older. LLU will not be successful if competing operators can share only older lines with longer loop lengths and poor line conditions, as these are not suitable for broadband connectivity. Pricing of unbundled local loops as set by the regulator is crucial. If the price is set too high, loop-sharing may not be attractive for competing operators. If it is too low, investment incentives are destroyed. Ideally, prices should reflect their long‐run incremental cost plus a mark‐up, to ensure that costs which are common to the line and other services of the incumbent carrier can be recovered.

    LLU gives BSOs, especially BSNL, the opportunity to leverage on its infrastructure of 50 million copper cables. However, it can be successful only if the incumbents do not view it as a threat and competing operators sense opportunity for broadband services. In India besides BSNL/MTNL, GAIL, Rail Telecom & Information Technology & PGCL provide the broadband, optical fibre –based transmission infrastructure which may be leased by service providers at competitive prices. Licensed service providers may put up their national and international gateways and connect them with every other network.

    Issues of infrastructure sharing do not lend themselves to any cookie-cutter solutions and the regulator indeed has to tread carefully in these matters. Many trade offs are involved and the final solution has to preserve the incentive structure that the regulations will produce. Our view is that it all depends on the details and on the presence of alternative infrastructures (such as cable, for example). In the US, mandatory unbundling (UNEP) did not work as the regulators did “too much”, so entrants could just sit on the incumbent’s network. The European perspective is that some unbundling is necessary to create competition; however, at some stage, entrants must have their own facilities. This is sometimes referred to as the “ladder of investments”, so unbundling can help new entrants move up the ladder. Unbundling is one of several tools in the policy makers arsenal of pro-entry policies that should finally result in facilities-based infrastructure competition (on either a wholesale or vertically-integrated basis). Once this demand is realised and facilities-based competition exists, in that case, mandatory asymmetrical unbundling should, in theory, no longer prove necessary. Precautions should be taken that unbundling should not be used to create a static incumbent-centric perpetual resale model, where everybody purchases their primary input from a single monopoly provider.

    Unbundling can be viewed as a two-stage process. In the first stage, unbundling should be used to stimulate new alternative non-incumbent demand. In the second stage, new facilities-based entry should be encouraged to serve this consolidated demand. Unbundling in the Indian context requires to be seen in the light of its implications on last mile connectivity sharing and opening to the many Internet service providers.


    If it is good for India, why unbundling bad for Sri Lanka?

    Mr. K. K. Gunawardana has always maintained his consistency on this issue. I wish TRC will listen to him and process unbundling in Sri Lanka. Otherwise we can never bring the ADSL prices down in Sri Lanka.

    Do you think SLT will ever reduce prices unless there is unbundling?

    Please don’t delete this post.

    Dhathika Sirimanne

  25. Dear Mr. K. K. Gunawardna sir,

    I am a young engineer fortunate enough to be benefitted by your vast amount of knowledge of telecommunication engineering knowledge you have been so generous to share with us at the IEE sessions.

    Sir, why you do no more take part in IEE activities? Please sir we wait for you to provide us guidence in our work sir.

  26. Response to Dhathika.

    We never delete comments.

    The issue is not ADSL development; it is broadband development. Let’s not get hung up on a specific technology. As Chanuka pointed out, the theoretical maximum number of ADSL lines would be in the range of 800k, This can never be reached, because wireline connections include many government connections, pensioners, etc. who may not want broadband. We need to think bigger than that.

    I have no problems with LLU. I have said that it is theoretically optimal but difficult to implement under the inferior regulatory conditions that exist in countries like Sri Lanka.

    There is no reason to think of these solutions as mutually exclusive.

    I believe removing the restrictions on anyone other than SLT laying copper wire should be removed. I’ll continue to advocate this.

    Mr KKG and you believe that LLU is needed. You should advocate for that. All strength to your efforts. No conflict with what I propose.

    More people will be connected through broadband wireless than wireline. We need to remove the barriers standing in the way of wireless broadband. We’ll advocate for that too. Again, no conflict with your preferred course of action.

    We’ll publish benchmark data on broadband quality and price, irrespective of technology. Should be wind for the sails of all the ships.

  27. May I have Mr. K. K. Gunawardene’s email address please?