RF and IF channelizers for wide-band sensing in cognitive/Software-Defined-Radio receivers

This paper proposes an original architecture for a Software-Defined-Radio receiver based on an Hybrid Filter Bank. The key issue for its proper operation is the channelization of the incoming signal into multiple sub-bands, that is performed at two stages: radio-frequency (RF) and intermediate-frequency (IF) levels. Thus, low-cost analog-to-digital converters (ADCs) can be employed for the sampling of these sub-bands to be subsequently processed at the digital level for the overall signal reconstruction. Special attention is given to the RF and IF channelizers, for which two suitable designs consisting of an eight-channel double-star-junction and a cochlea-like multiplexer, respectively, are devised. For experimental verification, two prototypes in microstrip and lumped-element technologies are developed and characterized.