Design And Simulated Implementation Of Energy Saving Algorithm For Multicast Traffic In EPON

With the rapid development of society and economy, global warming, energy crisis and other world-wide problems become more and more serious. Due to the large energy consumption of existed network systems, it has become essential for network system to save energy. Access network which occupies a large proportion of energy consumption has become an important object of energy saving technologies. In recent years, Ethernet Passive Optical Network (EPON) technology gains rapid development and is widely deployed, so it has already been thought of the best choice of achieving the "last mile" access network by the industrical community. However, the popularity of multimedia service accelerates the increse of multicast traffic in EPON, and thus brings about more energy consumption.Therefore, how to reduce the extra energy consumption of multicast traffic has become one of the urgent problems for the development and advance of EPON in future.As one of the typical energy-consumming devices in EPON, Optical Network Unit becomes a major part of the total energy consumption of EPON due to its massive deployment. This thsis studies the problem of energy saving for multicast traffic in EPON. An energy-saving algorithm of ONU sleeping is proposed based on Travelling Salesman Problem (TSP). Research works in this algorithm includes three parts:First, a model for ONU energy consumption is constructed. The energy consumption minimization problem is formulated mathematically by introducing the constraints on transmission of multicast traffic. Second, a comprehensive analysis on the relathionship between multicast traffic groups scheduling and time slots. It is derived that the matching between multicast goups and time slots is critical for the multicast traffic scheduling. Finally, the problem of minimizing ONU energy consumption is transformed into a TSP instance. The adjacency array approach is employed to construct a complete dericted graph of state transitions between different multicast groups. The energy of each multicast group and the state transition between diffrent groups are stored in the array. The recursive method is used to calculate the total energy consumption of different scheduling sequences. The schedunling sequence with the lowest energy consumption is determerind to support the transmission of multicast traffic.The professional software OPNET for network simulation is employed to establish the simulated EPON system. The performce of proposed algorithm is analyzed in the simulated EPON system. It’s showed that the proposed energy-saving algorithm is effective in reducing the energy consumption of multicast traffic in EPON.