This Small Business Innovation Research (SBIR) Phase I project will perform research toward developing new behavior characterization methodologies that allow adaptive/cognitive radios to be rapidly introduced into the market improving spectrum utilization and therefore broadband access for all. Expected results include a closed loop test fixture design that dynamically creates realistic RF environments in response to measured adaptive system behavior in a laboratory setting, controlled by an efficient adaptive search algorithm. The complexities involved in simulating realistic dynamic RF environments, sufficient to characterize the behavior of adaptive systems, drive the number of testing cases into the hundreds of millions and test time to many years. The adaptive search problem will be solved using globally optimized experimental alternatives and multi-dimensional splitting levels yielding high confidence results and reasonable test times. The project will combine this efficient search adaptive, rare-event approach with the company's unique dynamic RF environment emulation capabilities to create virtual RF environments that support comprehensive hardware-in-the-loop testing. The intellectual merit of the proposed research lies in combining and improving these unique capabilities in the critical application area of adaptive/cognitive radio testing yielding improved spectrum access.
The broader impact/commercial potential of this project is acceleration of the deployment of adaptive/cognitive radios, offering the prospect of ubiquitous broadband access to the nation's underserved. Lack of universal broadband access, increasingly essential to daily life, is hurting the U.S. The 2012 PCAST report recommends using adaptive/cognitive radio as a solution to meeting these increased demands with existing spectrum, but acknowledges the lack of robust testing capabilities to ensure non-interfering coexistence with incumbent systems. The challenge of deploying adaptive radio technology is that it cannot be fielded without comprehensive testing, and cannot be tested in a live, densely populated RF system environment for fear of interfering with existing spectrum users. This project will develop a system to efficiently validate adaptive radios in the laboratory with field level fidelity. Efficient, high quality testing offers the potential for quicker deployments at significantly reduced costs. The potential uses for adaptive/cognitive radios include commercial fixed and mobile wireless, ad-hoc network, satellite, aviation, DoD, Intel and public safety with potential revenues of hundreds of millions of dollars. This project may benefit millions of underserved individuals and enhance scientific and technological understanding of adaptive, rare-event search techniques as applied to adaptive/cognitive radios.
Under this NSF Phase I SBIR award, Echo Ridge and George Mason University (GMU) conducted research toward the development of a prototype test platform which can be used to efficiently and effectively determine adaptive / cognitive radios behavior. There are major segments of our population that do not have sufficient access to broadband coverage, and additional available spectrum is a key part of the solution. Cognitive radio technology allows sharing of the spectrum on a non-interfering basis and is a critical technology needed to bring ubiquitous broadband access to all segments of the population. As part of this NSF grant, we are developing a prototype test platform which, when commercialized, will provide technology solutions to address two spectrum sharing challenges. The first technology solution provides a flexible, adaptive emulation environment enabling stakeholders to increase the trust level and demystify cognitive radio technology. This solution will be targeted at independent third party organizations that are trusted by policy makers to objectively validate cognitive radio performance. The second technology solution is an interactive tool allowing spectrum facilitators, who manage secondary spectrum use, to optimize the use of available spectrum in shared spectrum environments. During Phase I, we achieved the three defined program objectives. We successfully demonstrated the feasibility of characterizing adaptive / cognitive radio behavior in a test platform. We completed development of a prototype test platform and ran an emulation scenario that showed an increase in spectrum sharing area in a TV whitespace (TVWS) scenario for a given basestation (BS) location over existing legacy approaches. Through analysis, we showed a significant reduction in emulation execution time for complex cognitive radio test scenarios and an increase in results confidence for a fixed run time. The prototype test platform is based on our DYSE (DYnamic Spectrum Environment emulator), which allows the behavior of the spectrum users in the sharing scenario to be characterized by dynamically creating realistic RF environments in a laboratory setting in response to measured spectrum user behavior (including adaptive/cognitive radios). The DYSE architecture consists of three major components: the Adaptive Test Scenario Scheduler (ATSS) which manages the test execution, Virtual RF Environment (VRE) which emulates the RF environment and the software-defined, or virtual spectrum users (VSUs) (including adaptive) which replicate the behavior of physical Radio Frequency (RF) devices in the context of our test system. ATSS is the system controller and contains the search algorithm that determines the best sequence of environmental conditions to stimulate the VSUs to most efficiently (as measured by the fewest iterations) find the area of high information (where the generated RF environment provides insight into the cognitive radio behavior) in the fewest number of iterations. The algorithm intelligently and adaptively characterizes the behavior of cognitive radios in a controllable, repeatable and realistic RF operating environment. This platform provides for high-fidelity, scalable, and repeatable testing in a lab with near field quality characteristics. The complexities involved in simulating realistic dynamic RF environments, sufficient to characterize the behavior of adaptive systems, can drive the number of testing cases into the hundreds of millions and test time to many years. The adaptive search problem is solved using globally optimized experimental alternatives and multi-dimensional splitting levels due to the stochastic nature of the data. The algorithm yields high confidence results and reasonable test times by minimizing the total number of test runs. Echo Ridge combined this efficient search approach with the company’s unique dynamic RF environment emulation capabilities to create virtual RF environments that support comprehensive hardware-in-the-loop testing (HiTL). The intellectual merit of the proposed research lies in combining and improving these unique capabilities in the critical application area of cognitive radio testing yielding improved spectrum access. During Phase II, we will mature and demonstrate/validate our prototype platform toward usable/ reliable products for characterizing cognitive radio behavior. We will improve usability, improve search algorithm methodologies, increase fidelity and performance, add new virtual device models and add external interfaces. As part of system demonstration/validation, we will validate the core platform in spectrum sharing scenarios, validate spectrum sharing channel determinations and validate the platform as a cognitive radio test bed.
