Towards a highly flexible spectrum sensing platform for cognitive radio networks

Abstract

Multi-nodal spectrum sensing is a promising approach to robust cognitive radio networks, which provides increased adaptability to database-anchored sharing approaches. Yet its commercially viable implementation requires compact, low-cost, flexible sensing receivers. This paper describes an implementation approach relying on a fully differential 0.25 mym SiGe:C BiCMOS low-noise down-converter RFIC in the front-end, which covers 0.3 - 6 GHz. Its quadrature outputs are processed in a commercially available baseband processor that combines dual channel 14-bit, 250 MS/s analog to digital converters (ADC) with a Virtex 6 FPGA. The design and implementation status of the system will be described, as well as its validation under realistic noise and interference conditions, sensing white spaces in the UHF TV band.