Imagine you have two cellphones, both embedded in a single handset. Now imagine this phone handles both CDMA and GSM calls. As technology progresses and such ‘dual band’ phones leave the boardrooms and enter middle class homes, expect the network to be jammed — even more jammed than it is today.
And then there is the likelihood of new problems. Another flood the size of Mumbai’s could paralyse subscriber databases, making it impossible for people to reach each other for weeks.
Like spam has taken over the Internet, burgeoning teledensity could make handsets bulge with lost calls, unreachable networks and unsent messages. In short, network management and disaster planning are likely be the biggest headaches for mobile operators.
But there seems to be a solution. Scientists at Bell Labs’ India wing in Bangalore are on the edge of delivering an artificial intelligence (AI) that will fix all bugs in mobile networks, and never need human interference.
This Indian end of Bell Labs was set up a year ago — in September 2004 — and focussed on network management from the start. Its only goal: to find a way for mobile networks to manage themselves, eliminating human errors or interference.
‘‘During the Mumbai floods, electricity wires in some places touched water, threatening entire substations. Since substations are linked, if one went down, all can fall. AI will create built-in mechanisms to alert such faults. In future, it will also decide the course of action during disasters—AI will ensure the network is always on,’’ says Dr Eshwar Pittampalli, CTO, Lucent India.
So think of Reliance, Tata, Bharti, Hutch, Idea and the others, spreading their infrastructure all over the country. Each has centres from where they manage the local and regional infrastructure. And all, to a large extent, still depend on human alertness to identify errors.
‘‘AI is essential to telecom tomorrow. Let’s say that all new mobiles are going to be dual band. Let us assume that they also have the old mobile embedded within. Obviously, with mobile infrastructure heavily loaded, some intelligence then needs to direct the signals from one mobile technology to the other, automatically. If GSM is overloaded, let CDMA take over and so on,’’ says Pittampalli.
Bell Labs’ innovation is yet to reach the proof of concept stage, but in 15-20 months, Pittampalli expects to prove the feasibility of this new technology.
‘‘We are positive, we are sure that the fruits of this research will be visible in a very short period,’’ he says. Obviously, the way mobile communications are taking off, there is urgent need for AI to take over where human interference left off.