Friday, April 23, 2010

IT experts say on the Uganda fibre optic cable

By Esther Nakkazi

The Uganda National Backbone Initiative project (NBI) for the past months has been facing so much criticism for its sloppy and shoddy work.

I have been speaking to experts and following the debate on i-network mailing list. Here are some issues that have come up:

Some experts are calling to uproot the already laid cable, something experts are talking about with mixed feelings after the time and funds invested in the $110 million Chinese funded project.

In the meantime, NITA has asked the Chinese-Huawei Technologies, to stop work on the project and to undertake a forensic audit and fix mistakes at the firms cost.

But Huawei officials said the dig-ups in the city by utility companies are responsible for the damages on the cable. They insisted that each party comes up with a team to investigate works on the cable.

Huawei officials like some experts in the IT sector and on I-Network mailing list insist that the G652 is a standard widely used cable.

Dr. Ham Mulira pioneer ICT minister and also an IT expert says he does not understand why they say the cable can not work. It is a standard cable the most widely used, it is sad when people make wild statements.

“The cost of the project is not only the fibre, there are civil works, compensation for the right of way, you just do not divide cost by distance,” said Mulira of the cost of the Ugandan cable that is said to be more expensive than Rwanda.

As a showcase and reference for a cable laid in Africa, Huawei would never do shoddy work. Many African countries were in Uganda to study the project including Malawi, ICT ministry official said.

But Samuel Besweli on the I-Network mailing list says the G652 cable limits capacity to 2.5Gbps-10 Gbps and it will require additional repeaters or DCF/SEQs to compensate for the long haul applications, which could simply become more expensive than initially laying the G655 cable.

Experts say the technology to solve these problems is in place and at advanced stages so it may not necessarily be costly after 15 years when the cable needs upgrading. Although the question would then be why not set up one that does not need upgrading initially.

Here are some excerpts from Mark Tinka, the chief Network Architect, Global Transmit Communications, Malaysia and others on the I-Network Mailing list about the cable: the average life span of a terrestrial or submarine cable system (other external factors notwithstanding) is between 15 - 20 years.

G652 is capable of pumping 10Gbps, and if the cable is fairly new with minimal to no splicing or fibre cuts, it could manage 40Gbps within 10 years of its useful lifecycle so the capacity issue is a non-issue.

Fibre is cheap, and with the ducting already in place, sending G655 through will be, easy, moreover, upgrading the whole network can be in phases. While those running 40Gbps or less are still on G652 you can install the G652.

This is an operational issue, not a technical one even if the fibre was G652, G653 or G655. We cannot predict that the next revision of G655 or G656 or whatever future cable will be much better.

Stakeholders in the Industry are also worried that the cable was not laid deep enough, which will create cuts and down time.

Safety in terms of depth remains to be seen when it is in operation, the trade off is between cost, security and timelines. Pulling out fibre and replacing it for no major technical or commercial reason is counter-intuitive.

The significance of being able to pump 880Gbps up to 1.6Tbps or 160 wavelengths @ 10Gbps each, worth of bandwidth over a single core of newer, low attenuation G652 fibre should not be underestimated.

I would suggest going ahead to light what has been laid, continue to lay the remaining G652 until it runs out, and then purchase and lay G655 for the remainder of the
project. However, all current purchases of G652 must halt!, advised Tinka.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.