Telecoms lessons learnt in the flood

There are as many lessons to take from the Queensland, NSW and Victoria floods as there are days in the year. When it comes to communications, the widespread service outages of recent weeks have provided invaluable insight into just how and where modern fibre-optic, copper-based and wireless telecommunications systems — none of which existed in their current state the last time floods reached these levels, back in 1974 — are vulnerable.

That there has been an outpouring of camaraderie to get the facilities back online, is commendable. But that hasn't eased the task ahead of the engineers dispatched to pump water out of more than 200 Telstra exchanges, Optus facilities, AAPT's offline Brisbane datacentre, Vodafone's base station sites and even an electricity network so badly damaged that authorities are already looking to how its renaissance could actually speed the shift towards the smart grid.

Indeed, an analyst I recently spoke to half-jokingly referred to flooded Queensland as "the world's biggest greenfields site" — and, not to minimise the seriousness of the devastation, he's right. Taking stock of the current situation, there are unprecedented opportunities to test new technologies, and write off old ones that have been pushed over the brink by these extreme conditions. I'm not sure how much of Telstra's controversial South Brisbane copper-for-fibre replacement project has been completed so far, for example, but the fact that much of that area was underwater would certainly have given Telstra a hands-on lesson in managing fibre-to-the-node infrastructure under duress.

Ditto Nextgen Networks, which only recently learned how a fibre-optic backbone cable buried 1.5 metres underground could have been severed by flood debris; buried as it is, fibre normally keeps on trucking without any dramas. But when flood waters washed away the dirt road under which the cable was buried, the cable became vulnerable to outside elements such as cascading debris, and it's only a matter of time until something breaks.

Thankfully, fibre networks are redundant and IP is reroutable, which is helping maintain services to nearly one-fifth of Australia's population while the services are fixed. But one cannot help but notice that the star performers throughout this crisis were the wireless networks: Vodafone, which was already suffering availability issues before the floods, was able to quickly get back online as its lack of fixed infrastructure means it has far fewer assets to fix. And successful efforts to avert the flooding of a 3 datacentre in Brisbane apparently kept that part of VHA's network online.

Meanwhile, Telstra and Optus engineers are furiously manning the pumps. And, throughout the crisis, it seems the most useful form of communications in flood-affected areas was the mobile phone — which continued working long after home landlines were underwater. Does this reinforce the argument that wireless next-generation services make more sense than fixed ones? On one hand, yes: wireless networks are clearly more flexible and resilient in times like these than their fixed counterparts. The battery life of on-site backups notwithstanding, it's quite possible for carriers to set up temporary base stations to supplement coverage for residents and emergency workers alike throughout crisis areas; it would seemingly have been relatively straightforward to anchor a barge-mounted base station, complete with generators and an ample supply of fuel, in the river or on a more slowly flowing floodplain to maintain services if necessary.

The fact that so many sites have suffered water damage and outages does rightfully raise questions.

When your telecommunications exchange is filled with water, there's not a lot you can do until that water is gone. But the fact that so many sites have suffered water damage and outages does rightfully raise questions — such as why crucial equipment wasn't located on high floors of local exchanges, safe above peak water lines. And why these exchanges are located in low-lying areas of flood-prone areas, rather than having been placed somewhere nearby but, say, up a hill. And why the ducts carrying Telstra's copper network, to which NBN Co will soon lease access for a tidy sum, are currently little more than underground rivers that will need to be pumped and tested in their thousands of kilometres.

There will be write-offs, of course, and it's hard to imagine that an event of this magnitude wouldn't exacerbate shortcomings in Telstra's already hit-or-miss local loop network. Telstra is racing to restore services, but now reckons it will be three months before services are back to normal in south-east Queensland. And one can say with great certainty that NBN Co engineers will be watching closely to consider how their own network designs might need to be modified to ensure maximum uptime and resilience during natural disasters. The company is, among other things, engineering 120 points of interconnection for which the disaster plans will likely need to be adjusted based on the recent events. And it will need to consider extra protective measures for network equipment in flood-prone areas; even assuming that wireless networks continue playing their important role in disaster recovery, they need backhaul to plug into.

NBN critics have also used the floods to raise questions for NBN Co's battery strategy, which has been identified as a potential disaster issue as residents with optical network terminal (ONT) installed batteries will theoretically lose services faster than those with centrally-powered copper phone connections. Yet as company representatives have pointed out, in such serious flooding you've probably already left the house; and, in any event, the limited battery life of cordless phones would take its own toll.

There are no easy answers to the flood crisis, and there will be many more difficult choices as the clean-up continues. Indeed, beyond all the technical minutiae will be an even more important question: does the previously unanticipated need for what is sure to be billions in disaster relief funding, mean we should slow down major projects like the NBN roll-out to reallocate that money and get Queensland's economy back on its feet?

There are already rumblings to this effect, and Julia Gillard will ultimately need to decide whether changing circumstances make it important to reconsider funding for the NBN, or other major projects, to preserve any hope of the promised surplus budget. For now, it's safe to assume she will not — and with $1.6 billion in new contracts assigned this week, big new commitments certainly need to be honoured. But as the mud and debris are cleared, her spending decisions will ultimately help shape the larger picture of our future communications infrastructure — and, potentially, accelerate our shift towards the NBN.

What do you think? Does the demonstrated importance of wireless strengthen the case for a wireless NBN? And will the flooding delay the NBN, or further justify it?