Investigation of Criteria for
Selection and Application of Software Defined
Inclusive Dates: 09/15/08 01/15/09
Background - With the increased popularity of software-defined radios (SDR) has come a wave of SDR standards (e.g., SCA, STRS, etc.). The selection of SDR standards is often done arbitrarily, leading to highly inefficient radio implementations that fall short of the reconfigurability goals promised by SDR standards. The objective of this project was to investigate the implications of implementing standard-compliant SDR frameworks to establish a set of guidelines for the optimized selection of SDR standards.
Approach - To accomplish this task, existing standards were analyzed to develop lists of characteristics of each standard. In addition, this effort experimented with selected framework implementations to learn the trade-offs in overhead associated with reconfiguration. This experimentation was conducted through a series of simulations in which the temporal effects of reconfiguration were assessed empirically.
Accomplishments - This research resulted in a list of criteria for SDR standards selection, of which radio reconfiguration criteria were found to be the largest of distinguishing factors. Middleware-based radio reconfiguration was found to be one of the most computationally burdensome requirements of compliance to many SDR standards. In many cases, fulfilling compliance to a small subset of the SDR standards' requirements while maintaining real-time performance was technically not feasible. Because of the large impact of some requirements for compliance, it was determined that selection of a particular SDR standard can often be just as important to system performance as selecting an appropriate radio technology.