Energy-efficient framework for time and wavelength division multiplexed passive optical networks

For the first time, to the best of our knowledge, an energy-efficient framework that optimizes the number of active wavelengths and uses sleep/doze mode to improve the energy savings of a delay-constrained time and wavelength division multiplexed passive optical network is presented in this work. In the proposed framework, the optical network units (ONUs) operate under the maximum polling cycle time that satisfies a given delay constraint to achieve maximum possible energy savings at the ONUs. When the average bandwidth requested by an ONU exceeds the maximumallowable bandwidth, a new wavelength is introduced to the network to satisfy delay constraints. Conversely, when the bandwidth requested by an ONU is small, idle wavelengths are switched off to increase energy savings at the optical line terminal (OLT). The proposed framework is analyzed using online and offline just-in-time dynamic wavelength and bandwidth allocation (DWBA) algorithms. The performance of both algorithms, under the proposed framework, is evaluated in terms of average delay, percentage of energy savings at the ONU, and percentage of energy savings at the OLT. Our analytical and simulation results indicate that under the proposed framework, the average delay of the network remains below a specified maximum allowable delay, in both algorithms. The online DWBA algorithm, however, results in improved energy savings, both at the OLT and at the ONUs, compared to the offline DWBA algorithm.