Personal tools

Plug prize


Jump to: navigation, search

2010/11/18: Plug Prize Apritel


Cognitive Radio (Plug prize 2010)

Authors: Nuno Borges de Carvalho1, José Neto Vieira2, Arnaldo Oliveira1, Pedro Miguel Cruz1, Daniel Albuquerque2 and Nelson Silva1.

1 Institute of Telecommunications (IT), Department of Electronics, Telecommunications and Informatics – University of Aveiro.

2 Institute of Electronics and Telematics Engineering of Aveiro (IEETA), Department of Electronics, Telecommunications and Informatics – University of Aveiro.


Nowadays, we are facing the constant appearance of new telecommunications standards or new releases of older versions (e.g., Worldwide Interoperability for Microwave Access (WiMAX), 3GPP Long Term Evolution (LTE), Universal Mobile Telecommunications System (UMTS) – High Speed Downlink Packet Access (HSDPA), etc). So, in a scenario that allows the interoperability between the existing standards and also provides quality of service (QoS), the Software Defined Radio (SDR) technology could be a promising solution [1] to deal with the higher complexity of the radios.

The concept of SDR first appeared with the work of Mitola [2] in 1995. In this work, he proposed to create a radio that is fully adaptable by software, enabling the radio to adjust to several communication scenarios automatically. Thus, the ideal SDR will adapt to the transmission scenario by gathering information about all of the signals that are present in the air interface and so, leading to Cognitive Radio (CR) approaches. Even so, this will need the use of a system that can scan the spectrum from low to high frequencies using software. This concept has driven many researchers to study CR approaches, an idea also proposed by Mitola in [3], where the radio adapts itself to the air interface by optimizing the carrier frequency, modulation, and choice of radio standard to minimize interference and maintain communication in a given scenario.

One of the most promising applications of this technology is to increase the spectrum occupancy by the use of opportunistic radios, where the radio will make use of the spectrum that is free at a given moment. In order to be able to implement this ideal solution, the radio should “see” and be aware of the entire spectrum and of the communications being used at a precise time. This has motivated the scientific community to study different architectures for constructing radios that have the capability of detecting signals over a broad frequency band with high dynamic range. Although it is a very promising technology there are a lot of problems to be solved, for instance, several signals should be received simultaneously, which will bring up new problems not only due to bandwidth restrictions, but also due to high peak-to-average power ratios (PAPR) demands, or the co-existence of several RF interferers.

Building flexible multi mode/multi standard SDR and CR requires the digital processing of high frequency and wide band signals, which is challenging in terms of sampling rates, operating speeds, dynamic range and power consumption. For this reason, a pure software based implementation using only and Digital Signal Processors (DSPs) is not possible, being necessary the support of specialized hardware devices, such as the field programmable (FPGAs). The large integration capacity and features of modern FPGAs make them appealing for an increasing number of real-world applications. The reconfigurability makes them particularly suited for applications requiring fast design cycles, dynamic adaptation and/or field updates.

The New Front-End for SDR

The ideal SDR would be able to demodulate several signals simultaneously even if they have quite different frequency carriers. Moreover, if the SDR receiver has the ability to sense the free bands of the RF spectrum, multiband waveform symbols could be designed to transmit the same digital information signal. To this specific application the classical front-end approach for SDR with several down-converters working in parallel would be very restrictive.

One possible solution to this problem can be the introduction of an Hybrid Filter Bank [4,5] with an A/D converter at each output working in parallel at a lower sampling rate. This parallelized architecture would be much more flexible to answer the challenges raised by the concept of CR. Moreover, this front-end would be more robust to jamming, and will deal with PAPR [6] problem in a effective way. This late aspect is not a minor one. As the bandwidth of the signal received by a SDR increases, and also as the signals in different bands could present quite different relative power, the front would suffer from the blocking effect.

These problems are very similar to the ones faced by the Human ear. The hair cells inside the cochlea convert the vibrations to electrical impulses. However, these analog to digital converters are equivalent to low bandwidth and low dynamic range devices.

The evolution solved the problem by creating the cochlea, which is an acoustic transmission line that performs spectral analysis of the input signal. That way, the hair cells only sense a small bandwidth signal with a reduced PAPR.

Inspired on the human cochlea a new architecture for the radio frequency front-end was designed. The input signal is divided in several frequency bands and each one is converted to the digital domain using affordable ADCs. The DSP has to reconstruct the original signal from the sub-sampled signals.

A prototype of a RF cochlea was developed as a proof of concept with promising results [7-9].


This work has the support of the Portuguese Science Foundation under the project PTDC/EEA-TEL/099646/2008, TACCS – Cognitive Radios Adaptable Wireless Transceivers.


[1] SDR Forum, “Perspective and Views on Regulatory Aspects of Software Defined Radio”, Working Paper, 2002.

[2] J. Mitola, "The Software Radio Architecture", IEEE Communications Magazine, Vol.33, N.5, pp.26-38, May 1995.

[3] J. Mitola and G.Q. Maguire, “Cognitive Radio: Making Software Radios More Personal,” IEEE Personal Communications, vol. 6, no. 4, pp. 13-18, Aug. 1999.

[4] Christopher J. Galbraith, Robert D. White, Lei Cheng, Karl Grosh and Gabriel M. Rebeiz, "Cochlea-Based RF Channelizing Filters", IEEE Transactions on Circuits and Systems-I: Regular Papers, Vol.55, N.4, pp.969-979, May 2008.

[5] Davud Asemani, Jacques Oksman and Pierre Duhamel, "Subband Architecture for Hybrid Filter Bank A/D Converters", IEEE Journal on Selected Topics in Signal Processing, Vol.2, N.2, pp.191-201, April 2008.

[6] Hee Seung Han and Jae Hong Lee, "An Overview of Peak-to-Average Power Ratio Reduction Techniques for Multicarrier Transmission", IEEE Wireless Communications, Vol.12, N.2, pp.56-65, April 2005.

[7] Daniel F. Albuquerque, José M. N. Vieira, Nuno B. Carvalho and José R. Pereira, "Analog Filter Bank for Cochlear Radio", IMWS 2010, Aveiro, Portugal, January, 2010.

[8] Daniel Albuquerque, Sérgio Soldado, José M. N. Vieira and Nuno B. Carvalho, "Cochlear Radio", EuMW2010, Paris, France, October, 2010.

[9] Pedro Cruz, Nuno B. Carvalho and K. A. Remley, "Evaluation of Nonlinear Distortion in ADCs Using Multisines", International Microwave Symposium, IEEE, June 2008.