Personal tools



Jump to: navigation, search

Title Cognitive Radios Adaptable Wireless Transceivers
Reference PTDC/EEA-TEL/099646/2008
PI Nuno Borges de Carvalho
Participants José M. N. Vieira, Arnaldo Oliveira, Teófilo Monteiro
Funded by FCT
Global funding (€) 153510€ (UA 67204€)
RU funding (€)
Starts 2010/01/01
Ends 2012/12/31

Software Defined Radio has already start its way to conquer traditional radio configurations, by moving base band (de)modulators completely to the digital world and thus implementing it by software. Nevertheless, it is expected that Cognitive Radio, CR, will be the next frontier in terms of radio architecture design, in fact the new paradigm that CR can bring to traditional ways to think on radio architectures, is so huge, as it was the transition from Analog to Digital systems. CR will allow the optimization of radio spectra occupancy, but also the implementation of the supreme universal radio, that is radios that talk to each other’s no matter the original implementation. But in order to achieve such a new deployment, the radio should be constructed with some form of processing capabilities, which it is not possible with traditional schemes, that processing capabilities arises from some kind of processor that treats the transmitted/received signal in real time, and makes decisions over it. That processing capabilities should be done at the digital counterpart, but also, and most of the time neglected, at the analog part. In fact the main conclusion of the last SDR forum meeting was that the analog part, usually called agile radios, should have a huge research effort in order to implement most of the CR concepts. This project aims exactly to study most of the physical layers of CR, starting from agile RF front ends, but also considering digital implementations, either FPGA based, but also DSP based. The project will try to give an answer to software engineers, by carefully studying the last piece of the system that is avoiding the complete implementation of CR. The main difficulty of implementing these CR architectures is the non-ideal behavior of the analog components, due mainly to bandwidth limitations, and to signal amplitude limitations, dynamic range, by creating nonlinear distortion that severely degrades SNR. Thus the correct identification of the system components behavior, that is nonlinear dynamic models, the correct understanding of the new waveform behaviors, which will be created in the CR approaches, will allow the correct design of the transmitted signal. This signal should be pre-distorted in order to be transmitted with perfect spectrum masks, but the received signal should also be compensated either by overcoming the non-idealities of the receiver components but also by reducing interference distortion due to the existence of jammers outside the band. Thus the now proposed work will make new contributions on (1) behavior models for non-ideal analog RF components; (2) new signal identification methods, either for evaluation of the analog needs at transmission or at receivers, but also for spectrum sensing; (3) new waveform generation for CR applications, in order to maximize the SNR, but also for maximize the transmitted bit rate; (4) new design of reconfigurable architectures based on FPGA, for efficient CR implementation; (5) improved receivers for spectrum sensing sub-systems by maximizing the gathering of large bandwidths; (6) new improved receivers for maximization of dynamic range; and finally (7) new transmitters for linear and high efficient signal communications. The proponent team is composed of a number of different knowledgeable elements, coordinated by a senior researcher of IT with previous international and recognized work on nonlinear distortion modelling, RF circuit/system design and SDR system implementation. For signal processing capabilities the team has two researchers with very important achievements in the area of Sampling Theory and Signal Reconstruction, plus a researcher with knowledge on FPGA-based reconfigurable architectures and hardware accelerators, digital systems modeling, simulation and synthesis. The team is complemented by the work of three PhD students that are working in these fields, and we hope to have the help of two “bolseiros” that will assist in the simulation, laboratorial measurements, and design of most of the new proposed CR architectures. Moreover there is no other Portuguese research team working in this new emergent and future area of communication circuits and systems applied to SDR and CR RF front ends. These team combined with the new ideas will allow the proposed project to give a step further in the state of art in the area of agile RF front ends for CR implementations.