Professor of Signal Processing
If you purchase a wireless router for your home, there are usually several antennae on it. Sometimes they are visible, and sometimes they are hidden under the shell. With the help of these antennae, the radio receiver can listen in directions that contain interesting information, and avoid listening in directions that contain the most interference. In the same way, the information can be sent to the desired recipient, and be suppressed in directions where interferences are not wanted. This is an example of how signal processing can be used to improve the performance of wireless communications systems.
Mats Bengtsson performs research on how one can use available radio frequencies more efficiently by optimising the reuse of radio resources in time, frequency and space. This is done by using multiple antennae, and collaborating between different base stations. The results are often based on optimisation methods. In these, one searches for solutions that are decentralised and robust against uncertainty, both in the mathematical models of the hardware and in the measured properties of the radio propagation and ambient interferences. Statistical modelling, convex and non-convex optimisation, as well as game theory, are some of the tools used in this research.
The research group has participated in several large EU projects that have developed the bases of 4G/LTE, and the upcoming fifth generation mobile system.