A constraint satisfaction language for intelligent self-correction of software-defined radio configurations

This paper presents an embedded Artificial Intelligence (AI) engine based on Constraint Satisfaction Graphs (CSG) that enables rapid analysis and modification of networked Software Defined Radio (SDR) configurations. We developed a Constraint Satisfaction Language (CSL), an extensible AI framework for solving CSGs, and a set of unique strategy modules, and combined them to provide a scalable configuration troubleshooting tool for large complex SDR networks. These elements are combined into a tool called CINCH (Cognitive Interactive Network Configuration Helper). CINCH runs on an Android tablet, and can analyze hundreds of SDR configuration files, each with large numbers of parameters, in less than a minute. The extensible framework allows rapid support of additional SDR waveforms. In addition, the integration of this solution into a Cognitive Radio (CR) framework, to enable real-time modification of CR settings, is feasible with minimal effort.