Tuesday 26th July, 2005



Sports Arena
Business Guardian
Online Community
Death Notices
Classified Ads
Jobs in T&T
Contact Us
Privacy Policy


TSTT’s answers

Trevor Deane

By Mark LYndersay

In mid-July, TSTT’s GSM cellular network fell over again, tipped this time by a combination of circumstances that the communications provider made the pillar of a quirky apology to the nation’s cellular users.

I began calling for an interview with someone at the company on Monday of last week, but with the deadline for this column looming, I accepted the opportunity to have my questions answered via e-mail.

The official respondent to my e-mailed questions was Trevor Deane, project manager for TSTT’s Mobile Development Project, but given that his response came four hours after the questions were sent, readers should have no illusions that these statements are sanctioned official speak. Following each response is my own final word on the subject.

BD: How long and how widespread was the GSM collapse? The perception is of nationwide network loss throughout July 13 and 14.

Deane: First, it must be clarified that the GSM system did not collapse. The network congestion means the system was handling calls up to its design capacity and calls in excess of this had difficulty connecting. Calls being placed on the network during the period exceeded its capacity by a factor of four. The congestion lasted from July 14 through to July 17.

(FW: The company plans to accommodate a million users by the end of 2005, but if a factor of four kills the network, then 250,000 users will be the real capacity even after the upgrade. Or is it? TSTT isn’t saying.)

Isn’t this really a problem of network capacity? Has TSTT built a network around a strategy of a percentage of users being on the network simultaneously with an insufficient buffer for usage that probably came closer to 90 per cent during the outage, the kind of use that might be expected during, say, a hurricane watch?

Yes, it is a capacity problem but not an equipment failure problem. All the network components were working, albeit above their design threshold.

Earlier this year, TSTT announced that it was upgrading its network to serve almost 1,000,000 users. We are in the midst of that upgrade.

Before the end of the year we shall almost double the number of cellsites on the network, add a second mobile switch and introduce high-speed mobile data capabilities on the network.

No operator designs a system capable of supporting 100 per cent of users 100 per cent of the time, especially when there is a high concentration of the subscribers in a restricted location.

National emergencies and natural disasters tend to stimulate mass calling events which challenge services even in cities like London where there are multiple cell companies.

(Still no word on the capacity buffer. Adding high-speed mobile data implies a vast increase in theoretical bandwidth, but many of TSTT’s bandwidth promises remain vastly theoretical. I’m still getting speeds on my DSL connection that are far lower than I should from an alleged 256k worth of bandwidth.)

What happens in a real storm? How redundant is the network and how much of it goes down if 25 per cent of cell repeaters are damaged?

Because a significant part of communications infrastructure is above ground, cables and other plant are susceptible to flying debris and fallen trees and poles. If communications links to a cell tower get damaged, the rest of the network can continue to function.

If the hypothetical 25 per cent is in an area that has a low population, it means that most people will still have access to mobile services.

(Sounds like no redundancy in the system, particularly in high density areas and specifically in remote, low density rural areas where there is no ready landline access either.

The question during a storm isn’t whether the greatest number have connectivity but whether the people at greatest risk of being incommunicado can be reached. How about mobile repeaters that can temporarily extend the range in flooded or impassible areas?)

Is TSTT planning its network around a drop in its user base when competitors begin operating?

No, we foresee that the total customer base will grow and we are designing for beyond that growth.

(Wild optimism, I’d say, given that rather severe lack of enthusiasm people have for TSTT’s mobile services after their second major service failure this year.)

Isn’t it fundamentally naive to blame users for trying to take advantage, en masse, of a situation in which they might have access to free calls?

TSTT was not blaming users, the company was describing a situation in which customers could take conscious decisions to influence the service they were getting. You cannot blame customers in such circumstances in the same way you won’t blame them if they are caught in long lines at a grocery. We were merely giving customers information that they could act on.

(I’d really have liked to hear Trevor Deane say this. Out loud, with a straight face. Giving customers information they could act on might mean issuing an urgent news bulletin while the crisis was in progress, explaining the reasons for the system overload and asking customers to reduce their call placements.)

People are paying for periods of system availability and TSTT has promised to recompense people for lost access in the past. When is that going to happen? How long does it take to figure out a rebate on billings to customers or posting additional minutes of access to customer accounts?

Given the status of the network, TSTT has a responsibility to give customers compensation in a manner that will ensure that the network will withstand the additional demands placed on it.

It would be irresponsible of us, knowing how the market may respond to additional airtime, to implement a rebate that would trigger a mass calling event and deprive customers of properly enjoying the very benefit we gave them. Details of the compensation will be announced as soon as everything is in place to ensure a secure implementation.

(I have absolutely nothing printable to say to this. It’s evasion, pure and simple. If TSTT had begun crediting customers in alphabetical order on a rolling daily cycle, they would have been done with reparations to the first failure by now.)

My final, final word is actually something I’ve said to representatives of TSTT before. Frustrated about being repeatedly stymied by TSTT responses to a personal issue with my phone, I strongly, perhaps even rudely suggested that the company needed a customer ombudsman.

Someone who would stalk the corporate corridors with fists full of angry letters and furious phone messages who would bend the ears of the people responsible for its more appalling customer service failures.

Until then, TSTT will settle for full page advertisements full of earnest apologies and grand promises that fall on ears deafened by cruel experience.

©2004-2005 Trinidad Publishing Company Limited

Designed by: Randall Rajkumar-Maharaj · Updated daily by: Sheahan Farrell