“We are working with visions, not with science fiction. Our research is visionary and is placed just right,” says Erik G Larsson, professor of communications systems at Linköping University.
He and his research group are leading technological development in multiple-antenna systems, a technique that is super “hot” in the rapid development towards 5G.
“We are pioneers in what is known as Massive MIMO. The concept was invented by Bell Laboratories in the United States at the end of the last decade. Most people in the world of research were sceptical, but we quickly identified the potential. Now there are many others who have put a lot of energy into following in our wake,” says Professor Larsson.
Massive MIMO
MIMO stands for Multiple Input, Multiple Output; briefly, the technology consists of replacing the large rotating antennae – the oblong 50 kilo slabs that we see most everywhere around us today – with hundreds of small antennae of perhaps 10mW each. They are set together in something that looks like a standard flat screen TV. Thanks to intelligent signal processing, the electromagnetic field from each little antenna can be directed in a very narrow beam towards where the receiver, sensor or cell phone is located. This produces a combination of low output, high energy efficiency and superior capacity.
“This will mean a paradigm shift in how we build base stations. The technology exists, and it works in theory. Our research colleagues in Lund have built a prototype and the concept has been tested. We have the laws of physics on our side,” Professor Larsson states.
But there is a lot of work left before the technology can be used commercially.“We need to develop smart algorithms for signal processing and new protocols for networks, with so many antennae. We do not currently know how best to build and manage systems that have contact with 30-40 terminals simultaneously. We also need to find methods to deal with non-perfect hardware,” Professor Larsson says.
Emil Björnson, reader in communications systems and expert specifically in Massive MIMO, is convinced of the benefits.
“In theory we can transfer 100 times more bits per second with this technology. A massive MIMO base station would, for example, have the capacity to supply all the visitors to a festival with high speed internet at once. But now we have to develop a demonstration model and show how great the benefits in practice will be,” he says.
The multiple antenna system is one way to go in order to reach the target of 1,000 times greater data capacity. Another alternative is to raise the frequencies, thereby increasing bandwidth. Currently 1-3 GHz is used as the carrier frequency for wireless networks, but there is greater free space around and above 10 GHz, which would then place significantly greater demands on the electronics.
A third way is to build denser infrastructure – building base stations closer together, which is the approach used most at the moment.The need for all three alternatives in the future is something that Professor Larsson and his colleague Atila Alvandpour, professor of electronic circuits and systems, are in complete agreement on. Because when it comes to the hardware as well – the actual electronics themselves – the expertise is here at LiU too, just a few corridors down from Professor Larsson and his research group.
Professor Alvandpour and his group have their core competence specifically in the construction of fast, efficient, low-energy circuits. 
Foto: 20 JPEG
Rapid broadband circuits
“In conjunction with Ericsson, we are already developing rapid broadband circuits used in radio amplifiers, where high demands are placed on the electronics. We’re also working on the hardware that makes the antennae work together. But there is a need for further research into systems and hardware architecture. It is a multidisciplinary challenge and our teams need to work together,” Professor Alvandpour states.There is also research concerning the third alternative – building base stations closer together – at LiU. This is the area where Professor Di Yuan and his mobile telecommunications group are the experts.
So will these high transfer speeds be achievable by 2020? It’s only five years away.Mr Björnson laughs:
“We’re not alone in the world; guaranteed, Samsung will claim that they have 5G for the Olympics in South Korea in 2018.”
Professor Alvandpour fills in the details:
“I believe that the improvements will be gradually implemented up through 2020, but it may perhaps take a little more time to get the next generation out into general use.”
From 1G to 5G
Ever since the first mobile network, NMT was launched in 1981, a new generation has appeared roughly every ten years. 1G - NMT; 2G - GSM, 1992; 3G - W-CDMA, 2001; when the development of 4G began at full power. The next generation, 5G, is therefore expected to be out on the market by around 2020.
According to the prognoses we will then have 500 billion interconnected components communicating with each other, streaming video in the car on the way up to the mountains; in the environment all around us there will be countless sensors reading various parameters in indoor and outdoor environments.
The fifth generation will therefore allow 1,000 times greater data traffic data, speeds of hundreds of Mbits/second almost everywhere and very short response times in milliseconds.
Ericsson’s 5G for Sweden
The project, launched last week, is a research and innovation programme supported by Vinnova. The programme will increase the competitiveness of the industry by, for example, developing and integrating various 5G-based communications solutions in a number of products and services. So 5G for Sweden deals more with use than technological development. Here, Ericsson is working together with partners in research and in industry; Linköping University, the Royal Institute of Technology, Chalmers, Lund University and Swedish ICT are all taking part. Partners in industry include Scania and Volvo CE.
