In indoor environments such as homes and offices, wireless communication technology is spreading rapidly. To enable increased throughput, this technology will increasingly exploit multiple antennas at both the transmitting and receiving ends of the communication system. This approach is often referred to as MIMO (Multi-Input Multi-Output). The first WLAN (wireles local-area network) products using MIMO technology have recently entered the market. As yet these products are relatively costly and power-hungry, and have limited resistance to (ever increasing) interference. For MIMO technology to attain mass-market appeal, these drawbacks need to be overcome. The present project aims to make an important step in this direction by joint optimization of the radio-frequency (RF) and baseband (BB) parts of the transceiver. These parts are normally designed largely independently, and a joint optimization is expected to enable substantially lower power dissipation, lower cost, and higher resilience to interference. Possible approaches include:
a) Novel transceiver topologies, with e.g. fewer RF front-ends than antennas, and a redefinition of the RF/BB borderline.
b) Dealing with RF imperfactions in the BB part of the transceiver, rather than preventing them at RF. This approach is sometimes called `dirty RF'.
c) Inclusion in the RF part of calibrations (of e.g. components or bias currents), based on information extracted at baseband. In this way the RF front-end can accept considerably higher intrinsic tolerances, potentially resulting in significant reductions of complexity/cost and power Dissipation.
d) Dynamic adaptation of the RF part of the transceiver in response to conditions of the communication channel as estimated at baseband. This would, for example, require a high power dissipation only during the small fraction of the time when reception conditions are poorest.
Exploitation of these possibilities requires a collaborative effort between experts in RF transceiver technology and in baseband signal-processing. For this reason the project involves two Ph.D. positions, one focusing on RF technology and embedded in the mixed-signal microelectronics (MSM) group at TU/e, and the other one focusing on baseband signal processing, embedded in the signal processing systems (SPS) group. The project is based partly at TU/e and partly at the Connectivity Solutions group at Philips Research Laboratories. It involves experts from TU/e and Philips and forms part of a broader effort, together with the Universities of Delft and Twente, towards a demonstrator of a `mass-market' MIMO system. This demonstrator should incorporate some of the approaches developed within the project.