Filter Bank Approach to Cognitive Radio

This research effort will develop a Cognitive Radio (CR) architecture to include spectral sensing and adaptive radio communications. Spectral sensing algorithms use a filter bank approach with wavelet-based spectral band edge detection to detect existing communication channels and spectral white space. Cyclostationary techniques are used for modulation recognition of existing communication channels. Adaptive communication techniques leverage the results of spectral sensing to exploit existing communication channels using burst QPSK waveforms and spectral white space using direct filter bank synthesis enabling robust communications even in dynamic spectral environments that include interference and multipath. Algorithms will be benchmarked in simulated communications environments using MATLAB, and prototyped in the REDHAWK Software Communications Architecture (SCA) framework. Prototyping in REDHAWK enables a straightforward migration path to COTS Software Defined Radio (SDR) platforms including our own wideband multi-channel SDR products. The use of multi-channel radios enables the exploitation of spatial diversity for improved performance in both spectral sensing and adaptive MIMO communications. This research has immediate direct impact to current tactical communication needs, but also provides a solution for many commercial market needs including TV white space exploitation, higher utilization of cellular bands, and as an adaptive wireless communication bridge (cross-banding between various wireless protocols).