A Signal Processing Approach to the Analysis of Chemical Networking Protocols

Chemical Networking Protocols (CNPs) are communication protocols, whose design is based on chemical models: distributed networks of reactions and molecule species. The benefits arise from the dynamics analysis of chemical models and thus from the prediction of CNP behaviors. Chemistry already supplies tools for the analysis of reaction network dynamics: the Chemical Master Equation (CME) and Differential Rate Equations (DREs). However, both procedures often lead to complicated solutions. We propose another deterministic approximation of dynamics, like DREs, but based upon the frequency characterization of chemical models. We used the signal processing background, adapting it into this new scenario and integrating it with analysis methods of other fields(e.g. the Metabolic Control Analysis (MCA)). In linear reaction networks, we identified and frequency characterized elementary building blocks which constitute chemical models. By linking these building blocks with series, parallel and feedback interconnections, we could replace chemical networks with schematics composed by transfer function blocks only. In nonlinear networks analysis, we applied the MCA to linearize the model. We showed dynamics of some existing CNPs and gave recommendations for the CNP design (i.e. for possible congestion control CNPs). We also considered chemical links with delays