Cross-Layer Architectures, Semantics and Strategies (CLASS)

While static parameter selections may be adequate for traditional layered network architectures, contemporary mobile ad hoc networks require cross-layer parameter optimizations to maintain desired performance. Though promising, the doctrine of cross-layer designs have limited proliferation due to unknown performance gains vs complexity trade-offs, ease of implementation, portability, scalability and security vulnerability. Phase I of this SBIR will provide a tool to analyze, test, evaluate and improve cross-layer designs. A thorough analysis of existing cross-layer techniques/algorithms and implementation strategies will be done. The SBIR proposes a high level design of the CLASS tool, which takes inputs from the user about performance objectives, cross-layer design techniques/protocols to be analyzed and the target deployment environment and/or mission scenario. The proposed tool will use a combination of analytical models and simulations to evaluate the existing design; make changes to the design to improve its performance and output the cross-layer design, including the expected performance goals. The analysis will include classification of these techniques according the expected performance gains from various designs and impact of cross-layer techniques/algorithms on network security. The Phase 1 design will provide the basis for a Phase 2 implementation effort to analyze cross-layered designs for large-scale networks with tens of thousands of components.