ADONIS project: Algorithms for Dynamic Optical Networks based on Internet Solutions

Welcome

ADONIS - Algorithms for Dynamic Optical Networks based on Internet Solutions is a jointly project among Universita' di Trento (DIT), Politecnico di Torino (TNG) and Politecnico di Milano (TNG).

The project aims at proposing an evolutionary methodology for the design and management of optical networks in order to support the migration from the current static assignment and routing techniques to adaptive, dynamic techniques based on IP-centric control as soon as new generation equipment becomes available.

The specific objectives of our research project are:

to study resource allocation and contention resolution algorithms suited for the optical technology and compatible with the dominant IP paradigm. In particular, history-sensitive intelligent heuristics will be considered for NP-hard optimization problems.
to suggest a unified evolutionary design approach, which supports the current static network structure, but is aware of the dynamic future requirements. By a proper design of the static network, these techniques will allow a seamless migration to dynamic management as soon as equipment and protocols become available.
to develop efficient algorithms and protocols for establishing lightpaths in wavelength-routed networks with dynamic traffic demands. These protocols will consider traffic engineering capabilities and guarantee effective protection/restoration mechanisms
to propose and evaluate new switching architectures to support the dynamic traffic patterns
to implement the algorithms through a careful design of supporting data structures, by considering both sequential and distributed realizations, and to study the proposed methods through theoretical analysis and statistically sound comparisons.

The resulting static design methodology will produce scalable architectures that will adapt to larger numbers of wavelengths and enable migration towards an efficient dynamic design and management of optical networks. Traffic engineering schemes to deal with fine granularity at sub-wavelength resolution will be developed mainly exploiting OTDM. The network design will also consider the resilience property of the topology and traffic uncertainty.

New techniques will be proposed for the design and management of dynamic optical networks based on IP-centric control, in particular dealing with fast dynamic lightpath establishment, protection and fault restoration. Distributed "intelligent" algorithms will be considered, in which some parameters will be adapted in an online manner through feedback from the previous history. Finally, new switching architectures will be evaluated, capable of realizing the proposed protocols.