Markku Renfors - On-going and recent projects

Hydrid Analog-Digital Signal Processing for Communications Tranceivers (HYBRA):
This project is focusing on the physical layer of wireless communication networks, where one of the main challenges is the development of flexible radio transceiver implementations supporting the utilization of an increasing number of radio access technologies in multi-mode terminals. Analog receiver front-end processing is facing new challenges and provides also new opportunities. In low-cost implementations, nano-scale VLSI processing optimized for digital circuits are to be used also for analog RF. Reduced supply voltage leads to reduced linear range of amplifiers, and there are great difficulties in implementing receivers based on the conventional superheterodyne or direct-conversion principles. However, in recent few years, it has been demonstrated that receiver structures based on sampling directly at RF frequencies are better suited to the VLSI technology trends. These so-called digital radio processor designs utilize decimating discrete-time analog processing after the sampler, yet before the actual ADC that is operating at baseband. An important aspect in recent RF sampling architectures is that certain kind of embedded filtering can be combined with the sampling process. It is clear that there are still great challenges in making such designs feasible for the recent and emerging wideband wireless standards. Also, the solutions presented so far do not answer to the needs of multi-mode radios on one hand and field-programmable radios on the other.The scope of this project proposal includes analog and digital signal processing circuits and algorithms for advanced radio receiver front-end solutions to support different radio interfaces in a flexible manner. Methodologies for flexible and optimized hybrid analog-digital signal processing solutions are the core area of this project. The main application is to optimize the overall receiver filtering chain, including the embedded filtering in sampling, decimating analog discrete-time filtering stages, and digital filtering stages after the ADC. In this development, the characteristics of charge-domain sampling and processing with nano-scale CMOS processes are taken into account.

Physical Layer for Dynamic Spectrum Access and Cognitive Radio (PHYDYAS):
Physical layer best suited to the new concepts of dynamic access spectrum management and cognitive radio is needed for future efficient wireless and mobile radio networks. The requirements of high data rates and flexible spectrum allocation are met by multicarrier techniques, which can approach the theoretical capacity limits in communications. The scheme used so far, Orthogonal Frequency Division Multiplexing (OFDM), is a block processing technique, which lacks flexibility and has poor spectral resolution. In contrast, a filter bank-based multicarrier (FBMC) technique offers high spectrum resolution and can provide independent sub-channels, while maintaining or enhancing the high data rate capability. In the project, a short term objective was to develop and demonstrate algorithms for single and multiantenna terminals, scalability and adaptivity, and multiple access. In the longer term, the impact on cognitive radio will be investigated. The success of the proposed physical layer will contribute to the dissemination and exploitation of the new radio concepts and to the advent of improved environment friendly services to the users. PHYDYAS was a STREP type project in EU FP7 programme during years 2008-10, with a strong academic participation. TUT contributed the project in the field of filter bank optimization, channel estimation, equalization and synchronization in single antenna and multiantenna configurations, as well is in spectrum sensing algorithms. TUT had a major role in defining the signal processing algorithms for the project demonstrator. More information is available at the project's own web-site.