LDI-backed Anti-voip Move – Now More Evil

There is nothing bad in an Internet world than to fear new technologies. In fact, fearing new technologies in general is a bad idea itself.

Pakistan – like the rest of the developing countries – stands strong as a potential beneficiary from the worldwide ICT related businesses provided a liberal Internet policy is adopted at all levels.

When PTCL was a monopoly in telecommunications in Pakistan and in came the cellular operators, positive things happened for the consumers – services improved, rates declined and availability got almost ubiquitous. Of course, the financial success that the cellcos met in Pakistan is an over-stated fact.

The left-behinds in all the high-water mark events mentioned above were the LDI operators who, while started off with great zeal and put in a lot of money too, found the telecoms environment too diluted and with much of undercutting going on. The rates (for termination inside Pakistan) went as low as 1 cent per min for wholesale carriers. A number of LDI operators burnt out in these conditions never to be seen again.

Later, around the beginning of this year (2008), sanity finally prevailed and the LDI operators and the authority decided that it was time to end the next to free rates to Pakistan termination traffic and raised the nominal tariff to 10 cents per minute. This move, which was essentially about LDIs putting their own house in order, alone injected a new life in their business. Salaries for the staff started coming in time and new equipment started being ordered by the operators.

Despite some undercutting, the rates in the international market for wholesale termination inside Pakistan are still above the 7 cents per minute mark and this leaves a considerable room for the operators to keep and take their business forward. According to rough industry estimates, only the recorded business is worth in access of 700 million minutes (or $35 million @ 5c/min) per month (grey traffic not included).

The LDI operators, in an effort to pump more out of this new found oil well, requested en mass the PTA to go after the grey traffic operators and invested in equipment that claims to detect and mitigate voip traffic in real time. This equipment has been on the international exit points in Pakistan. Suspected IP traffic was detected and investigated to see if it grey, or belongs to an un-intimated call center. This stuff was reportedly being done manually so far.

So far, the business-saving and law-enforcing arrangements by the LDIs and PTA appear to be logical and permissible. No one likes grey operators – the steal legal traffic from licensed players, do not pay taxes and do not help when LEA wants their help in tracing crimes and these guys are generally of, well, grey character themselves.

But when you see this item in today’s The News, it starts getting really uncomfortable:

PTA to start automated blocking of IPs
By By our correspondent
ISLAMABAD: Pakistan Telecommunication Authority (PTA) will start automated blocking of Internet Protocol Addresses (IPs), involved in illegal termination/origination of international traffic, in a bid to check grey traffic flowing into the country. The facility will be operational within the next few days.

This was announced by PTA Chairman Dr Muhammad Yaseen in a meeting with the CEOs of major Internet Service Providers (ISPs) of the country held at the PTA headquarters on Tuesday.

He requested all the ISPs to declare their IP addresses along-with the antecedents of their customers so that illegal telecom traffic could be monitored. It was emphasised that the operators should oversee their customers to make sure they are not involved in grey traffic termination. He sought the operators’ cooperation to stem the menace of grey telephony.

The ISPs appreciated the recent steps taken by the PTA including an announcement in the press wherein call centres were asked to provide their IP addresses to the PSEB to ensure uninterrupted services.

Since the inception of technical facility in May 2008 at the PTA, the IP addresses found to be involved in illegal activities were being blocked manually and in the process, over 14 million minutes (worth around Rs100 million) have been saved on monthly basis. Now these would be automatically blocked if any IP, not authorised to carry voice, is found doing so. Under the current policy, only LDIs and international call centres are authorised to carry voice across national boundaries.

The problems with the above arrangements are many:

  1. There ‘real-time’ nature of the voip blocking apparently takes away the manual process and the sanity that can possibly accompany these efforts
  2. The regulator will now essentially be ‘peeking’ inside the contents of the traffic. True, they are looking for voice packets but one is justified to think, ‘what’s next’?
  3. PC-to-PC voip traffic – which unfortunately has been neatly wrapped in clouds of uncertainty by the regulator despite the industry literally begging for clearer guidelines on this topic – might get disrupted. The optimistic take here is that the Naurus gear would be intelligent enough to have thresholds that can distinguish between an occasional PC-to-PC voip caller and a bulk grey operator.
  4. For the network applications administrators, troubleshooting will now get more complex as the high-end IP transit operators are no more acting transparently and the traiff is getting actively peeked inside and the Naurus system would be fiddling with it if it passes the defined litmus tests of being grey voice.

The LDI operators are best advised to invest more into the reliability and reach of their network instead of lobbying the regulator to take effects that could be counter ICT development. The industry is clearly against illegal activities but at the same time, a liberal Internet regime is one of the prime enabler of a saner IT featuring future of Pakistan.

All are requested to keep an eye on their network performances with the possibility of network issues cropping up due to this imminent implementation of automatic voip traffic suppression.

And let us hope the equipment vendor is not making us a guinea pigs for their new software releases!


Internet-based LL Services

Fellow blogger and friend Babar Bhatti has complained about the problems he is facing in using Braintel’s local loop services (probably from out there in US). A reader has commented on this citing a PTA ban on such operations by the local loop operators in Pakistan.

The main issue here is the ability of the LL operator to replace the traditional copper for its end user or the expensive 1900 Mhz WLL frequencies for its end user with the ‘Internet cloud’. It is technically possible but the use of Internet for voice problem had its own fair share of cloudiness. At the center of this debate is the use of the word ‘long distance’ – whether it is to be taken physically or network-wise.

When last LDI/LL licenses were issued, the LL license itself was a cheap affair – though the spectrum (both for 1900 Mhz and 3.5 Ghz) was auctioned at high prices. At that time, a number of Internet savvy people – Brain included – took the LL license only banking on new VoIP technologies to come to their rescue later.

A few VoIP technology companies (a cross between operators and technology vendors if you will) had been constantly chasing the smaller LL operators (as described above) evangelizing the use of VoIP in local loop operations where the ‘affinity to local numbers’ is the actual ‘good sold’ and the profit comes in from volumes of such ‘numbers’ beings sold worldwide at fixed monthly rates (but not actually always being used).

These VoIP technology companies met some success during the past couple of years with a number of LL operators signing up them either as partners or just technology vendors. At that time, cellular industry was priming and people, operators and the regulator somehow had little time to attend to this possibly controversial issue.

Now that the market is nearing a tele-density of over 50% and market consolidation has started, these by-issues will get more attention (and probably get more debated).

PTA’s clarification on use of VoIP (available from PTA’s website and discussed on some related online forums) is of little help as it leaves room for guessing by the readers.

My personal position here is that given the non-deterministic nature of public Internet (specially when the bandwidth is not directly coming from a T1 operator) as a transport mechanism for real time traffic, such ‘Internet-glued’ LL services should be allowed – they will always be placed at number 2 in terms of voice quality and deterministic performance. Such services could be declared a new class of service with clear requirement of informing end consumers about emergency services not working on them as well as possible degraded voice performance due to third party packet networks that happen to lie between the end networks. Given a considerable population of expatriate Pakistanis around the world, these services stand a fairly good chance of catching on in popularity and can earn some part of the much needed forex for country by the local companies offering such services worldwide.

A Newage Inspiration

We all know Internet killed geography long time ago. Just like TiVo took the live TV out from the clutches of time and Slingbox freed your TV from your living room, there is one company that thought about enabling any phone company on earth worth its name to let their ‘phone numbers’ set free anywhere on the Internet and make money along the way.

It is not always that a Pakistani origin new-age Internet company makes a wonderful product that not only proves to be a financial success but is also an excellent example of innovative thinking and visionary excellence.

Meet DIDX – DID Exchange, the star product of Super Technologies. Without exaggeration, an attempt to describe Super Technologies, its various products and its contribution to the telecommunication regulation politics in Pakistan will consume a full blog on its own. The company that started its journey from the (then) innovative idea of Internet faxing back in 1987, morphed into a VoIP service company with a Pakistani network, changed to a hosted VOIP platform and finally, the team hit the sweet spot (incidentally as a by-product) of becoming a marketplace for an unusual telecommunication resource that nobody had not spotted before in the same context.

So what is DIDX and what do they do?

One of the core forces behind DIDX is Rehan Ahmed Allahwala. Rehan works alongside his American mates who front the company worldwide in commercial and marketing domains. Rehan leads the team that works on new ideas and commercial gymnastics. Suzanne Bowen is the CEO of the company (and out of her interest in the Pakistani Telecommunication market which, as a by-product, has given him a partner like Rehan – she’s a lurker at TGP too!) The company has a number of developers working OUT of Pakistan. So while it would be technically wrong to call this a ‘totally Pakistani venture’ because of the mix of fine people from so many places, the strong Pakistani contents in their HR is a matter of pride for all of us back at home.

DIDX is about selling and buying an unusual, untapped asset that a lot of telecos are sitting upon – the E.164 numbers they give out to ‘their’ subscribers who take one (or more) service from them. Most of the time, the teleco focuses on the service it is providing and tends to overlook the ‘affinity’ that those numbers carry and the possibility of that affinity being marketed and traded – anywhere on the Internet.

DIDX is the opposite of traditional call-termination market places. In these call-termination market places, participants compete to terminate your call at the cheapest of the rates possible. At DIDX, participants compete to sell their ‘DID numbers’ to those who value them because of their ‘callers’ having a certain affinity or proximity with those numbers. DIDX participants generate revenue because either the number is used from their very own network (for which they get paid anyway by the caller) or from foreign networks (which, again, pay them the termination charges). The only hidden cost of the DIDX participant here is the Internet bandwidth that is utilized for each of the call that originates as a VOIP call from their SIP box to the wild Internet and end up at the original user’s MSN, Google Talk or FWD soft phone (it could be a hard one as well).

Explained in even simpler terms, DIDX enables any telephone company to procure some cheap Internet bandwidth, host a box on that bandwidth, and start selling their telephone numbers to worldwide buyers. Since these numbers would be used by individuals seeking to ‘receive’ calls from the old telephony world to their voiped-up digital universe, there is a very high over-subscription that is possible because not everyone would be using the service all the time. Selling a million odd numbers via DIDX, for example, for $10 ea could give a juicy business figure to a lot many telecommunication companies especially outside North America.

On the other side of the eco-system, DIDX suddenly enables voip islands, like Skype and MSN a steel bridge back to the 4 billion people on earth that still are stuck in the older telephony universe. It takes little time to appreciate the impact that DID has created on VoIP worldwide.

Ever since DIDX went out in the wild, its participants are growing. While the list is long, Skype, Google putting them up as their partners and (just today), Jeff Pulver the grand daddy of voice over IP announcing the partnership of FWD with them are rock solid proofs of the brilliance of the idea and the company behind it.

So best wishes for DIDX, Rehan and Suzanne, and we hope you keep innovating the voip space!

More Details on PTCL IPTV Plans

More details are emerging on the new broadband infrastructure that PTCL has been working on.  Daily Times says in a report that the project cost for the Optical Fiber Access Network was $58 million (Rs 3.5 billion) and that PTCL plans to offer triple play services on the new infrastructure with as many as 250 IPTV channels, voice lines and faster Internet access all for Rs 1,500 ($25) per month. [Let us hope they put all this together before stuff like Joost takes away most of the juice out of regular TV and channels!]

The network will continue to be build beyond the three key cities of Karachi, Lahore and Islambad:

PTCL plans to expand the service in phases across the country. During the first phase, the company will provide the service in Lahore, Karachi and Islamabad. The required network will be installed across the country within a year. Two Chinese companies have installed more than 350 ONUs from exchanges to telephone cabinets by replacing old copper cables with optic fibres. One optic network unit costs Rs 10 million.

PTCL’s new management apparently knows it is easy to put the nuts and bolts together but difficult to build up customers’ confidence in their service attitude that is very badly tarnished due to the people in its ranks that are used to of decades of monopolistic environment. Let’s hope they put together a better support team they are currently seeking people for.

Provided the network is properly rolled out in time (with adequate customer support backing the higher speeds it promises), it will open up opportunities of more BPO businesses by existing or new SMB outfits and other virtual transactions.

A key challenge that PTCL will face in the roll out of this network would be provisioning of real IPv4 addresses to its envisaged 100,000 always-on customers. Since this would require close to the equivalent of 400 class C blocks and given the shortage of IPv4 addresses worldwide, it is most probable that PTCL will resort to network address translation (NAT) and put its customers behind fewer real IPv4 addresses. While browsing and download speeds will increase with the new network, a lot of fun would go away because of the absence of real IPv4 addresses. It would be logical to expect that there would be an exception to this for the ‘gold plan’ customers on the network. It would also be good to see PTCL experimenting with IPv6 deployment along side this network roll out though shortage of applications for IPv6 will remain a tough challenge to surpass.

Wateen Bulk IP BW: Sales Calls Begin

Wateen Telecom has started making inquiry calls to potential bulk transit Internet bandwidth customers. While price points were not available with the calling representative of the company, they are planning to offer a single rate for IP bandwidth across Pakistan which will be major impact news.

Whether IP bandwidth would be subject to VoIP service restrictions or not is not known yet. However, unless canabilization is not a problem for Wateen which is also actively pursuing domestic players for nationwide transmission services, most of the tier two players would want to opt for the flat rate Internet across Pakistan and use the medium to save their transmission costs.

Currently, PTCL offers transit Internet bandwidth to Internet Service Providers at a flat rate across Pakistan. However, the same bandwidth could not be used for transporting voice traffic by any player as major VoIP services are disabled on this bandwidth. (PTCL has  traditionally been using a so called ‘Network Security & Surveillance Wing’ to actively detect voice traffic on PTCL provided links to ISP and other enterprise customers).

Backbone Interconnections: Wateen would initially be buying bulk capacity from SMW4 consortium members.

Internet Exchange and Private Peering: Busy in their network roll-outs and customer acquisitions, and want of realized volumes, none of the players (TW, PTCL, Wateen, Mobilink) have so far announced any private peering arrangements. Similarly, the Internet Exchange plan also remains a dream because the country lacks agile nationwide ISPs that own a lot of IP users and hence would potentially benefit from an Internet Exchange.

Death by CLI!

The hoax took the nation by storm. And it actually became a social virus much to the delight of whoever must have released it in the first place. And true to the traditions of the ‘Jinn in London Mosque‘ and ‘Bomb Blast at Grumander’ news, this was aired on the TVs only to be later declared as a hoax in the slides.

We saw a number of interesting things happening out there. As if the constant calls from friends and knocks at the office room from colleagues was not enough, my doctor wife wore the most serious and somber of the faces I have seen on her in the past seven years when she asked me: ‘What will happen now?’ – taking the troublesome news item as a ground reality in the first place. And every tech savvy person in my circle had similar stories to tell about from their respective surroundings.

On the face of it, it seems a bad situation that we had the hoax getting so successful in the public. However, this very event once again proved the fact that the nation has truly gone cellular in the past few years with everyone – okay ‘almost’ everyone – having a piece of this convenience. We cannot blame the masses for the lack of technical acumen to separate a hoax from a real trouble – this is how the general public behaves the world over. The other silver lining very much visible was the voluntary squad of techies that stood up and fought the hoax in a distributed manner. This again is a sign of the youngsters getting a hang on the technology and its possibility brackets.

Since the majority of the hoax message talked about ‘an unknown or weird’ number being the culprit, I am tempted to ponder on the possibility of this being an attempt to curb illegal voice termination where it is given that the origination number would be hidden behind a dummy number which generally is also weird and distinct from regular numbers that we receive during the day. Of course, such an measure to counter illegal call termination would be silly for it will have a production cycle of days if not hours but it seems probable that for as long as the hoax caused the greatest panic, illegal calls were probably not terminated into Pakistan 🙂

Another thought that quickly hit after we saw the panic waves around the network was whether the industry has now reached a state of stability and harmony to do a structured search from the SMS records of the six cellular operators and make an attempt to track where the stuff started from in the first place. Even if we are too late, or the records are too many or the stuff originated out of Pakistan, an effort in this direction will sure teach us a thing or two. Internet is even more complex and multi-partied and yet virus and malware writers do get identified when efforts are put in. This would be more productive than the pacifying announcement by PTA on the same topic.

And finally, since a lot of friends were debating whether there is a possibility of this being crafted by the cellular operators in the first place for increased SMS traffic between their users, I can only confirm that this is a long established practice in the operators to ‘seed’ interesting and most-likely-to-be-forwarded messages to small number of customers every now and then only to see a ripple effect of the messages moving around the network, brining in revenue.

Bandwidth Tariffs – PTA’s Fresh Determination

PTA has issued Determination on 5th October 2006 on Bandwidth Tariffs offered by PTCL. You can grab your copy of the determination (53 pages, ~1 MB) from here. PTA’s website has a nice summary of the same here.

This determination is the conclusive part of the industry consultation process that PTA started in early September 2006.


  1. PTCL has been made to reduce prices of IPLCs further. Tariffs for ISPs/Data Operators (read customers) and Voice / LDI operators (read competitors) are still different except the difference has narrowed down. Cost of half-circuits for high capacity IPLCs will, for the first time, fall under the $100K figure. Under $52K, an Voice or an ISP operator can now get an STM-1 (155 MB). This is high impact item. If even a few providers opt for leased international STM1 capacities, Pakistan will get some much-needed network-layer diversification. Yes, we will still be travelling on the same media (SMW3 / SMW4) but we should fare much better against DoS attacks and network config problems with this diversification.
  2. IP Transit services are still differentiated based on usage (something most of the people except PTCL disagreed with). However, the new reduced rates are welcome. For example, an ISP could now get STM1 capacity (155 MB) for around $47K as against $60K previously.
  3. PTCL has been asked to provide contention based (1:2) transit Internet services @ $1,000 per E1. This is in accordance with the point 6 of the recommendation I gave (full recommendation here). With lots of OFAN customers getting on-board, this is one enforced-decision that PTCL should actually be happy about. A good decision that PTCL was not taking on its own has now been forced on it. This is again, high impact. At these rates, lots of smaller ISPs that serve business customers will now be faced with a competitive price of $500 per MB while the going rate in the market is around $800 per MB.
  4. It will still cost a Voice operator $238K (down from $361K) to subscribe to 155 MB of transit Internet in Islamabad that could be used for VoIP.
  5. DPLC (Domestic Leased Private Circuits) costs have been slashed down by 20% to 30%.

What to Expect:

  1. PTCL crying foul once again with the Ministry of IT & T and/or moving legal courts against PTA’s determination.
  2. PTCL’s IP gear enforcing voice/non-voice usages on high-capacity circuits to face hickups and even getting tripped when large number of STM1 circuits get loaded. And, in the longer run,
  3. Better DSL and cable modem experiences