Today, most radio spectrum is used ineffectively and its utilization varies by time and location. The unutilized parts of the spectrum results in ‘spectrum holes’ or ‘white spaces’. It has been proposed to allow the utilization of unused spectrum by other, non-licensed users. Spectrum sharing might be soon be feasible with new technology based on cognitive radios.

Cognitive radios are defined as radio systems that perform environment analysis, identify spectrum holes, and then operate in these unused holes. The most important feature in a cognitive radio is the ability to measure, sense, learn, and be aware of the parameters related to the radio channel characteristics.

Primary users can be defined as those who receive a higher priority to use a specific part of the available spectrum. Secondary users, who have a lower priority, might be able to exploit this spectrum but they must not cause interference to the primary users. Thus, the secondary users need to have cognitive radio capabilities, such as sensing current spectrum usage and adapting the radio parameters, in order to exploit the unused parts.

Following Suzan Bayhan’s presentation

[1], we can present different spectrum-sensing methods in terms of their accuracy and complexity.

Figure 1: A comparison of common sensing methods

Figure 1: A comparison of common sensing methods

We can also look at spectrum-sensing algorithms from other perspectives, including:

  • Proactive vs reactive
  • Local vs cooperative
  • Distributed vs centralized
  • In-band vs out-of-band
  • Synchronized vs asynchronous

Suzan Bayhan’s presentation [1] explains these in more detail.

Many R&D centers around the world are working on cognitive radio. Nutaq contributes to this field by providing software-defined radio (SDR) platforms and development tools that enable researchers to verify their algorithms in real-time.

CorteXlab [2] in Lyon, France is one of the most prominent examples of a large-scale cognitive radio laboratory. CorteXlab will foster significant scientific progress by allowing users to design, benchmark, and tune their cognitive radio protocols. It will enable the evaluation of cognitive radio designs in a real environment. Nutaq provided around 20 PicoSDRs to CorteXlab and also conducted a three-day training course. Recently, researchers from France published a remarkable paper about implementing multiply PHY layers of IEEE 802.15.4 on a PicoSDR.

 

CWC Oulu University in Finland is another leading R&D center working on cognitive radio, with a focus on spectrum-sensing algorithms. The researchers selected Nutaq’s Radio420S for use in its white space research as the RF front-end for its Xilinx ML605-based SDR research platform (described here).

 

Our PicoSDR, one of Nutaq’s flagship products, is designed for use as a cognitive radio research platform. With support for GNU Radio, you can offload processing from the PC and implement the most demanding or real-time parts of your algorithm, the PHY layer, in the FPGA and let the PC (using GNU Radio or C code) handle the less demanding operations (MAC and above). The onboard radio transceiver, the Radio420X, covers a frequency range from 300 MHz to 3.8 GHz with a software-configurable bandwidth of 1.5 to 28 MHz and includes software-selectable banks of baseband filters.

Figure 2: Radio420S block diagram

Figure 2: Radio420S block diagram

Providing powerful hardware is not enough for us. Our mission is to ensure our SDR platforms are easy for researchers to use. VHDL coding skills are not required. Figure 3 shows our software offering. The Board Support Development Kit (BSDK) is used when you program the FPGA in VHDL with Xilinx ISE/Platform Studio (XPS); the Model-Based Design Kit (MBDK) is for programming the FPGA from a Simulink flow graph with the Xilinx System Generator tools.

Figure 3. Nutaq's software development suite (includes IP cores, APIs, and I/O)

Figure 3. Nutaq’s software development suite (includes IP cores, APIs, and I/O)

Moreover, you do not have to start your design from scratch. Nutaq provides a QAM64 2×2 MIMO ODFM with AGC reference design for the PicoSDR. Based on this reference design, we also made a couple of demos, like the one presented at SDR-WinnComm in March (Nutaq’s cognitive OFDM demo) or this very short video: Cognitive OFDM.

References

[1] Overview of Cognitive Radio Basics and Spectrum Sensing, January 29th, 2013, Suzan Bayhan: http://www.hiit.fi/u/bayhan/uh/CN-S2013-SpectrumSensing.pdf

[2] CorteXlab website: http://www.cortexlab.fr/