Thesis etd-06162010-212140 |
Link copiato negli appunti
Thesis type
Tesi di laurea specialistica
Author
MONTI, MASSIMO
URN
etd-06162010-212140
Thesis title
A Signal Processing Approach to the Analysis of Chemical Networking Protocols
Department
INGEGNERIA
Course of study
INGEGNERIA DELLE TELECOMUNICAZIONI
Supervisors
tutor Meyer, Thomas
relatore Prof. Luise, Marco
relatore Prof. Giannetti, Filippo
relatore Dott. Tschudin, Christian
relatore Prof. Luise, Marco
relatore Prof. Giannetti, Filippo
relatore Dott. Tschudin, Christian
Keywords
- chemical networking protocols
- dynamics analysis
- frequency response
- metabolic control analysis
Graduation session start date
19/07/2010
Availability
Withheld
Release date
19/07/2050
Summary
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.
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.
File
Nome file | Dimensione |
---|---|
The thesis is not available. |