The FCC released its National Broadband Plan citing the exponentially growing demand for mobile data services and the critical need to better utilize the radio spectrum. Beyond retargeting certain frequency bands, the FCC is considering a paradigm shift of dynamic spectrum access (DSA) technology and policy. DSA allows 'secondary' radios to transmit in underutilized 'white space' provided they create minimal interference to 'primary' radios. Significant white space, as much as 85%, has been observed across time, frequency, and location. However, even in the TV bands, the challenge of inconspicuous utilization of white spaces has not yet been achieved by the two leading solutions, distributed spectrum sensing and centralized emitter databases.
This project focuses on a comprehensive approach to DSA based upon spectrum sensing that combines the following in a feedback loop: novel modeling of primary and secondary transmissions; design of optimal spectrum sensing algorithms and secondary access protocols based upon these models; and experimental validation within a testbed of software-defined radios. Preliminary results develop primary Markov models and optimal sequence detection algorithms for spectrum sensing that build upon the well-known Viterbi and forward-backward algorithms and expose fundamental limitations due to primary mismatch for the commonly used energy detector. The connections to trellis- and graph-based algorithms from the channel coding community should introduce a sizable new toolbox to DSA researchers. Secondary access protocols that take advantage of Markov process models for the primary users similarly exhibit improved performance tradeoffs between primary interference and secondary throughput. Collaborators in industry and regulatory bodies will be kept informed of the research results with an aim toward impacting DSA technology and policy development at national and international levels.
