Research

Introduction

Spectrum is the lifeblood of RF Communications. To avoid interference, radio communication is subject to regulations, which assign dedicated parts of the radio frequency spectrum to dedicated services, such as FM-radio, TV broadcasts, GSM, etc. Measurements all over the world, such as the one made in Bern by Swisscom (figure 1), show that large parts of the spectrum are not or rarely used, while other parts are overcrowded.

Figure 1: Spectrum measurement performed by Swisscom in Bern

To better exploit the scarce available spectrum, new radio devices could monitor the locally available spectrum, and use unused or "white space", but "step aside" if traditional radio services enter the picture. This requires a smart agile radio transceiver, referred to as a Cognitive Radio (CR). The term 'cognitive' stands for awareness of the surrounding environment; in our project we limit this awareness to availability of the spectrum.

Traditional radio transreceiver circuits are narrowband in order to be robust against interfering signals in the spectrum. Because the frequency band is fixed, external (expensive) RF-filters can be used for filtering. In CR, the frequency bands are not fixed and may even vary during a single session, so a lot of flexibility is required. This flexibility may also include flexibility in modulation schemes, because of regulations by governing authorities or simply to achieve a high spectral efficiency under heavily varying channel conditions.

CR builds on top of Software Defined Radio: a hardware system that is flexible enough such that it can be controlled by software. At the same time, it must be cheap to produce, have a low power consumption to guarantee a long battery life, have a high performance to quickly respond to varying conditions and give a high throughput.

CMOS is cheap and allows both analog and digital electronics to reside on one-chip, which makes it suitable for integration in many systems, such as mobile phones. The goal of the project is to design a fully functional CR chip in CMOS (raw bits in, raw bits out). We do not focus on higher level communication, unless this is absolutely critical information for the lower network levels.

Spectrum Sensing

An important task of a CR is spectrum sensing: scanning the spectrum to look for unused frequency bands. With respect to the OSI-layer model, spectrum sensing is a cross-layer problem, which makes it very interesting but also very challenging. Important questions for the physical and data-link layer are:

• What techniques can be used for spectrum sensing?
• Do they require knowledge of the modulation of the signal?
• What is their performance in terms of mathematics and in terms of system requirements / parameters?
• What will be the influence of analog imperfections, such as flicker noise, I/Q mismatch and frequency offset of oscillators, on the performance of spectrum sensing?
• Can we intelligently reuse analog and/or digital building blocks that we need for other functionalities of the CR?
• Can we use digital techniques to reduce analog imperfections?

Figure 2: Feature detection is one of the techniques for spectrum sensing, but is limited by oscillator frequency offset [Cabric, PhD Thesis, 2007]

On higher OSI-levels (which is not the main focus of this project) spectrum sensing can also be important:

• How can multiple CRs cooperate to detect weak signals (to prevent false negatives)?
• How can CRs know what type of modulation is used in that band (local/global database, ...), because more knowledge allows a better chance of detection?
• How can local regulations be obeyed?
• Can a CR learn from previous experience; can they learn from each other? 

People

This research is shared by two projects: CMOS Cognitive Radio (CMOS CR) and Ad-hoc Dynamic Radio-spectrum Exploitation via Multi-phase Radio (AD-REM). The following people are involved in the research:

• Eric Klumperink (CMOS CR, AD-REM)
• André Kokkeler (AD-REM)
• Amir Ghaffari (AD-REM)
• Dlovan Mahrof (AD-REM)
• Mark Oude Alink (AD-REM)
• Saqib Subhan (CMOS CR)

BSc / MSc assignments

Students intrigued by the above problem statements can always ask for more information, but the following assignments are directly available:

• MSc-project: Suppression of out-of-band power spectral density (More info)
• MSc-project: Comparing FX- and XF-correlators (More info)
• BSc-project: Cross-correlator testbed (More info)

Currently, the following students are working (partly) on my part of the project:

• Chiel Hakkenberg & Matthijs Dam: exploration of a software-defined radio platform (BSc project)

The following projects are finished:

• Wim Korevaar (MSc project, June 2010): A Time-Frequency Localized Signal Basis for Multi-Carrier Communication
• Boris Hupkens van der Elst (BSc project, August 2010): Analyzing mixed-signal frequency mixers that use polyphase harmonic product cancellation
• Arnout Smeenge (MSc project, October 2010): Improving Cross-Correlation Spectrum Sensing Using Two Antennas
• Arjan van Heusden (MSc project, August 2011): Automatic Gain Control ADC based on signal statistics for a cognitive radio cross-correlation spectrum analyzer
• Pepijn Bicker (MSc project, October 2011): Analysis and design of an ADC for a spectrum analyzer
• Peter Prins (MSc project, January 2012): Signal detection and noise correlation with two antennas
• Marijn Ufkes (BSc project, January 2012): Towards the implementation of an XF correlator in the USRP2 FPGA

Educational Activities

• Elektronische Basisschakelingen en Functies (AT, 121173)
• Elektronische Basisschakelingen (EL, 121175)
• Elektronische Functies (EL, 121177)
• Integrated Circuits and Systems for Mixed Signals (EE, 121087)
• Functioneel Programmeren (INF, 211205)
• Embedded Computer Architectures 1 (CS, 213024)
• Instrumentation for Embedded Systems (CS, 121000)

Biography

Mark Oude Alink was born in Hengelo, the Netherlands, in 1984. He received the MSc degree in electrical engineering and the MSc degree in computer science (both with honors) from the University of Twente, the Netherlands, in November 2008, for which he received the University Thesis Prize. During his studies he worked as student-assistant for numerous courses and for five months as a trainee at BlueShift Technologies, a developer of modular wafer handling systems.

From November 2008 until the present day, he works as a Ph.D. student at the same university at the Integrated Circuit Design group headed by Bram Nauta and at the Computer Architectures for Embedded Systems group headed by Gerard Smit. His research includes system level design, modulation and digital processing of a Cognitive Radio system.

Publications

[1] M.S. Oude Alink, A.B.J. Kokkeler, E.A.M. Klumperink, K.C. Rovers, G.J.M. Smit and B. Nauta, "Spurious-Free Dynamic Range of a Uniform Quantizer," IEEE Trans. Circuits Syst. II, vol. 56, no. 6, pp. 434-438, June 2009.

[2] M.S. Oude Alink, E.A.M. Klumperink, M.C.M. Soer, A.B.J. Kokkeler, B. Nauta, "A 50MHz-to-1.5GHz Cross-Correlation CMOS Spectrum Analyzer for Cognitive Radio with 89dB SFDR in 1MHz RBW," Proc. 4th IEEE Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN), Singapore, April 6-9, 2010, ISBN 978-1-4244-5188-3

[3] M.S. Oude Alink, A.R. Smeenge, A.B.J. Kokkeler, E.A.M. Klumperink, G.J.M. Smit, B. Nauta, "Exploring the Use of Two Antennas for Crosscorrelation Spectrum Sensing," Vehicular Technology Conference, San Francisco, September 5-8, 2011

[4] M.S. Oude Alink, A.B.J. Kokkeler, E.A.M. Klumperink, G.J.M. Smit, B. Nauta, "Lowering the SNR-Wall for Energy Detection Using Crosscorrelation," IEEE Trans. Veh. Technol., vol. 60, no. 8, pp. 3748-3757, 2011

[5] M.S. Oude Alink, E.A.M. Klumperink, A.B.J. Kokkeler, M.C.M. Soer, G.J.M. Smit, B. Nauta, "A CMOS-Compatible Spectrum Analyzer for Cognitive Radio Exploiting Crosscorrelation to Improve Linearity and Noise Performance," IEEE Trans. Circuits Syst. I, vol. 59, no. 3, pp. 479-492, March 2012.

[6] W. Cheng, M.S. Oude Alink, A.J. Annema, G.J.M. Wienk, B. Nauta, "A Wideband IM3 Cancellation Technique for CMOS Attenuators," in Proc. IEEE Int. Solid-State Circuits Conf. - Dig. Tech. Papers, Feb 2012, pp. 78-79

[7] W. Cheng, M.S. Oude Alink, A.J. Annema, J.A. Croon, B. Nauta, "RF Circuit Linearity Optimization Using a General Weak Nonlinearity Model," IEEE Trans. Circuits Syst. I (accepted)

[8] M.S. Oude Alink, E.A.M. Klumperink, A.B.J. Kokkeler, W. Cheng, Z. Ru, A. Ghaffari, G.J.M. Wienk, B. Nauta, "A CMOS Spectrum Analyzer Frontend for Cognitive Radio Achieving +25dBm IIP3 and -169 dBm/Hz DANL," RFIC2012 (accepted)

[9] M.S. Oude Alink, A.B.J. Kokkeler, E.A.M. Klumperink, Z. Ru, W. Cheng, B. Nauta, "Improving Harmonic Rejection for Spectrum Sensing using Crosscorrelation," ESSCIRC2012 (accepted)

Other

[A] M.S. Oude Alink, "Cognitive Radio," Invited talk at the 16th URSI Forum of the Benelux, May 18, 2010, Brussels, Belgium

[B] M.S. Oude Alink, "A Crosscorrelation CMOS Spectrum Analyzer with Improved SFDR for Cognitive Radio," Invited presentation on ProRISC 2010, 20th Annual Workshop on Circuits Systems and Signal Processing, November 18-19, 2010, Veldhoven, The Netherlands.

[C] M.S. Oude Alink, "Cognitive Radio Research at the University of Twente," Invited Talk at the 17th Annual Symposium on Communications and Vehicular Technology in the Benelux, November 24, 2010, University of Twente, Enschede, The Netherlands.

[D] M.S. Oude Alink, A.R. Smeenge, A.B.J. Kokkeler, E.A.M. Klumperink, G.J.M. Smit, B. Nauta, "Crosscorrelation Spectrum Sensing Exploiting Two Antennas," ICT.Open 2011, 21th Annual Workshop on Circuits Systems and Signal Processing, November 14-15, 2011, Veldhoven, The Netherlands.

[E] W. Cheng, M.S. Oude Alink, A.J. Annema, G.J.M. Wienk, B. Nauta, "A Wideband IM3 Cancellation Technique for CMOS Attenuators," RFIC and Technology Workshop, February 29, Nijmegen, The Netherlands