Optimising wireless bandwidth

2 min read

With the number of mobile internet users growing and accessing increasingly rich multimedia and interactive content wirelessly, there is a problem of spectrum scarcity.

To help get over this Dr Sangarapillai Lambotharan of Loughborough University's communications department is developing a technique to make better use of the available spectrum, increase coverage and provide a better service to people surfing on the move.

One method of optimising bandwidth is spatial diversity, which uses multiple antennas at the base station transmitter allowing it to send signals to many people at the same time in the same frequency band, and sometimes at the receiver end too. This is known as MIMO — multiple input, multiple output.

'This way we can increase the data rate, in other words the speed of transmission,' said Lambotharan. 'If the current wireless local area network (LAN) could support up to 100Mbit/s, we can increase this and the coverage distance from the base station too, simply with more antennas.'

When transmitting signals through multiple antennas, a key issue is that to increase the data rate, serve multiple users at the same time, or increase the coverage, different signal processing algorithms have to be applied. Lambotharan's research aims to develop algorithms that can achieve them all simultaneously.

'If you want to get the benefit of MIMO, you need to have feedback from your mobile terminals to the base station,' he said. 'Using multiple antennas, you can steer narrow electronic beams only toward where the user is. For that to work, the user needs to feed back some information — this is called channel feedback.'

The user's location is transmitted back to the base station so the signal can be transmitted towards that direction. This minimises interference and improves signal quality, as one user will not pick up signals transmitted to others.

However, by the time the feedback arrives at the base station, the user could be in a different place, so there would be a difference between the feedback and the actual position. This data is a condition known as 'uncertainty in the feedback information'.

This Loughborough project will use advanced mathematical techniques known as convex optimisation effectively to introduce robustness to this uncertainty. Rather than transmit a very narrow signal towards a specific direction, the signal radius will be made wider as and when required, meaning the if the user moves, there is still consistent, high data rate coverage.

Other project collaborators are world-leading expert in robust signal processing techniques, Prof Alex Gershman of Germany's Darmstadt University of Technology, who will concentrate on wireless communication. Prof Malcolm McLeod at Qinetiq will direct the team to apply their algorithms to actual wireless communication scenarios and provide real data for testing.

The team's aim is to publish new algorithmic signal processing techniques and to work with academic collaborators to use them in prototype wireless system development.

Berenice Baker