The Government is to fund research into LED-based Gbit/s free-space optical networks.
The heavyweight academic team includes the Universities of Cambridge, Edinburgh, Oxford and St Andrews, and is lead by the University of Strathclyde. Spin-outs mLED (Strathclyde) and pureVLC (Edinburgh) are also involved.
Over four years, the project will use data-modulated LED lighting arrays to side-step RF spectrum limits within room, inside equipment, and outdoors.
To prove the concept, 124Mbit/s has been sent from an off-the-shelf LED light bulb.
“Imagine an LED array beside a motorway helping to light the road, displaying the latest traffic updates and transmitting internet information wirelessly to passengers’ laptops, netbooks and smart phones," said Strathclyde professor Martin Dawson, overall project. "This is the kind of parallelism that we believe our technology could deliver.”
"Eventually, it could be possible for the LEDs to incorporate sensing capabilities too. For example, your mobile phone could be equipped with a flash that you point at a shop display where everything has been given an electronic price tag, and the price of all the items and other information about them would show up on your phone’s display," said the Engineering and Physical Sciences Research Council (EPSRC), which is funding the four year project.
Professor Martin Dawson from the University of Strathclyde leads the research
The project will push optical and modulation science to see how far it will go in the real world and, in view of the number of unknowns, the EPSRC is allowing more flexibility than is usual.
"We are trying to explore. The nature of the grant is there is enough room to reposition, unlike normal grants which specify all," said Dawson.
Building blocks include: micron-scale LEDs whose low capacitance allows GHz bandwidths (up to 1.5Gbit/s at the moment, according to Dawson); modulation techniques that spread data across time, frequency and space; lenses and non-imaging optical antennas; and fast p-i-n diode-on-CMOS optical receivers.
Professor Harald Haas of the University of Edinburgh, who coined 'Li-Fi', is working on modulation: developing optical equivalents of the radio MIMO (multiple input multiple output) techniques used in later Wi-Fi standards.
Instead of using many antennas to increase data throughput, he uses multiple LEDs and multiple photodiodes.
Just as with radio MIMO, signal processing can remove cross-talk - so photo-diodes and LEDs need not be paired one-to-one. And modulation schemes can be developed that incorporate the spatial diversity of an LED array to add another dimension to modulation space.
According to Haas, optical MIMO can be done with no lenses at all.
"We are trying to find out if we can avoid imaging optics and we have some evidence this can be achieved with spatial modulation," he told Electronics Weekly. "There could be multiple receivers pointed in different directions each receives different reflections from multiple LEDs. We have seen 10 sources working. I know we can get higher."
In a lighting application, a standard 1-2mm2 power LED might be replaced by an array of micro-LEDs, possibly reconfigured depending on link demands.
"You can run every micro-LED with same signal, or you can group them into sub-groups, or you can also modulate them as individuals," said Haas.
In his concept, data would reach a ceiling-mounted LED array via power-line communications in an existing building or, in new buildings, at higher data rates on power-over-Ethernet.
Modulation techniques include OFDM, the processing intensive but bandwidth-efficient scheme used in DAB radio, ADSL and Wi-Fi.
Professor Harald Haas, from the University of Edinburgh studies optical MIMO-OFDM, and coined the term 'Li-Fi'
Haas foresees receivers with enough computational power to demodulate 1,024 beams at the upper end.
One of the characteristics of OFDM is that the resulting signal in space has large magnitude variations that have to be received faithfully, according to Haas. "There are big variations in amplitude, which are a problem with RF, and not a problem in light," he said.
Over in Oxford, Professor Dominic O'Brien also looks into spatial optical communications.
"We specialise in a different flavour of MIMO from Harald," said O'Brien. "We start at the sharp image end, and he is at the no image [lens-less] end. There is a continuum. The maths is complex and it is hard to see where the sweet-spot is. It is one of the questions we will be asking."
Oxford will also look at close-up optical communication between portable devices like phones, but its main mission in the project is system engineering: to pull together developments at the other universities.
"Our job is to look at components required to build a complete communications system; how best to put a micro-LED transmitter from Martin Dawson with a receiver from Edinburgh and Robert Hendy's pin-diodes on CMOS - he is trying to push the technology forward, seeing if he can get really high-speed communications from CMOS," said O'Brien. "We will be trying to work out the balance of the system to get best overall performance, and then working out how to build the components physically."
What would O'Brien like to have in four years?
"There are a lot of open questions," he said. "For me, I would like a really good understanding of the technology, so if someone asks for something we will know how to make it. And I would like some really high data rate demonstrations showing saleability, and to have some vision of what the future might hold."
The EPSRC has put an audio slide show on YouTube - 'Li-Fi multi-tasking micro-lights could spark a communications revolution'.
