Doctoral research project

Person in charge of the project:
LAUWEREINS RUDY, member of research team Associated Section of ESAT - INSYS, Integrated Systems
Title:
Efficient Design of MIMO solutions for Software Defined Radio
Project summary:
Wireless communication standards are diversifying and evolving rapidly.It is often required that the first prototypes be available even beforethe standard is finalised. Software Defined Radio (SDR) is seen as foreseeable solution to reduce the design time cost. SDR baseband solutions employ flexible programmable architectures, so that the hardware can be reused across various wireless standards. The drawback is that the algorithm design to support various wireless standards efficiently on these architectures becomes very challenging. On top of that, the support for Multiple Input Multiple Output (MIMO) antenna schemes is increased with each new wireless standard to meet the high data rate requirements of users. Unfortunately, the high data rates offered by MIMO systems come at the expense of a high implementation complexity in the receiver. Althoughboth industry and academia have invested a lot of effort in MIMO-SDR receiver design, existing solutions implement relatively simple algorithmsthat cannot achieve optimal communication performance. In addition, thearea/energy efficiency of MIMO-SDR receivers is still significantly lower than the ASIC implementations.To tackle the aforementioned challenges, we apply a co-design approach to the MIMO-SDR receiver design.SDR platforms offer flexibility and programmability, which is usually the main cause of lower energy efficiency. However, this flexibility can also be used as an advantage to compensate for the energy inefficiency by implementing scalable algorithms that adapt to the dynamically varyingwireless channel. First, we use an architecture aware algorithm design approach, to enable run and design time scalable algorithms for the MIMOreceiver. The MIMO receiver algorithms proposed in this thesis are designed such that they can be implemented on a generic SDR baseband processor. Afterwards, a novel C-programmable ASIP template is proposed for MIMO detection. Specifically, the template is designed such that it can be re-configured to support various MIMO antenna configurations. The proposed scalable SDR solution for MIMO detection delivers a peak-throughput of 3.6 Gbps with 13.03 mW power consumption for a 4x4 MIMO system. In high performance mode it achieves a throughput of 400 Mbps with 16.85 mW power consumption with near Maximum Likelihood bit-error-rate (BER) performance.
ph.D student :
AHMAD UBAID UR REHMAN
Faculty of Engineering Science
Doctoral Programme in Engineering Science (Leuven)

ph.D defence : 01.12.2014
Full text ph.D