K. Lukac (Switzerland), Z. Lukac, and M. Tkalic (Croatia)
WDM, RWA, learning automata
To build and scale new optical networks wavelength division multiplexing (WDM) promises to be one of the best solutions for both reliable wavelenghth transport and wavelength-level traffic engineering. However, the routing and wavelength assignement problem (RWA) arises. We attempt to minimize connection blocking using a heuristic approach based on the concept of learning automata which apply the idea of reinforcement learning. The blocking performance of the proposed model has been studied and compared with those of the fixed alternate routing assignment. The new approach is shown to be a viable solution to the RWA problem. It achieves lower blocking rates than fixed-alternate routing over all available wavelengths.
Important Links:
Go Back
