Future High Capacity Backbone Networks

This thesis - Future High Capacity Backbone Networks - deals with the energy efficiency problems associated with the development of future optical networks. In the first half of the thesis, novel approaches for using multiple/single alternative energy sources for improving energy efficiency are proposed.

The work focuses on energy efficient routing algorithms in a dynamic optical core network environment, with Generalized MultiProtocol Label Switching (GMPLS) as the control plane. Energy ef- ficient routing algorithms for energy savings and CO2 savings are proposed, and their performance is studied in details with dynamic network simulations using OPNET.

Dynamic routing optimization methods are proposed. The influences of re-routing and load-balancing factors on the algorithm are evaluated with a focus on different re-routing thresholds. Results from dynamic network simulations show that re-routing strategies can further lower CO2 emissions compared to basic energy source routing scheme, and that a lower re-routing threshold achieves more savings.

The increased blocking probability brought by using rerouting schemes can be compensated by applying load balancing criteria. A trade-off between blocking probability and obtained CO2 savings is studied. Specifically, the use of solar energy as an alternative energy source is also studied with the assumptions of bundled links usage.

However, due to the incoherence between the solar generation level and the traffic variation of the day, the algorithms proposed aiming for reducing the dynamic part of the energy consumption of the network may increase the fixed part of the energy consumption meanwhile. In the second half of the thesis, the conflict between energy efficiency and Quality of Service (QoS) is addressed by introducing a novel software defined integrated control plane.

The programmable control platform enables exchange of information between different network domains, and traffic flow concepts are introduced to replace the traditional routing/forwarding mechanisms. The functional design is defined, which introduces new possibilities to the routing methods and the control over QoS.

In the presented case, the integrated control plane collects the network energy related information and the QoS requirements of different types of traffic. This information is used to define the routing behavior for a specific class of service. Due to the flexibility of the routing structure, results show that the energy efficiency of the network can be improved without compromising QoS parameters such as delay or blocking probability.

A multi-objective evolutionary algorithm is employed to further improve the performance from the dynamic network simulations under the context of the integrated control plane structure. Results show improvements of energy efficiency over three types of traffic, while still keeping acceptable QoS levels for high priority traffic.