Active Intra-ONU Scheduling with Cooperative Prediction Mechanism in EPONs

Dynamic bandwidth allocation in EPONs can be generally separated into inter-ONU scheduling and intra-ONU scheduling. In our previous work, the active intra-ONU scheduling (AS) utilizes multiple queue reports (QRs) in each report message to cooperate with the inter-ONU scheduling and makes the granted bandwidth fully utilized without leaving unused slot remainder (USR). This scheme successfully solves the USR problem originating from the inseparability of Ethernet frame. However, without proper setting of threshold value in AS, the number of QRs constrained by the IEEE 802.3ah standard is not enough, especially in the unbalanced traffic environment. This limitation may be solved by enlarging the threshold value. The large threshold implies the large gap between the adjacent QRs, thus resulting in the large difference between the best granted bandwidth and the real granted bandwidth. In this paper, we integrate AS with a cooperative prediction mechanism and distribute multiple QRs to reduce the penalty brought by the prediction error. Furthermore, to improve the QoS and save the usage of queue reports, the highest priority (EF) traffic which comes during the waiting time is granted automatically by OLT and is not considered in the requested bandwidth of ONU. The simulation results show that the proposed scheme has better performance metrics in terms of bandwidth utilization and average delay for different classes of packets. Keywords—EPON, Inter-ONU and Intra-ONU scheduling, Prediction, Unused slot remainder

[1]  Glen Kramer,et al.  Ethernet passive optical network (EPON): building a next-generation optical access network , 2002, IEEE Commun. Mag..

[2]  Tohru Inoue,et al.  QoS-aware dynamic bandwidth allocation scheme in Gigabit-Ethernet passive optical networks , 2004, 2004 IEEE International Conference on Communications (IEEE Cat. No.04CH37577).

[3]  Dessislava Nikolova,et al.  Dynamic bandwidth allocation algorithms in EPON: a simulation study , 2003, OptiComm: Optical Networking and Communications Conference.

[4]  Hung-Yuan Chen,et al.  Active scheduling algorithm for intra-ONU Dynamic Bandwidth Allocation in ethernet passive optical networks , 2009, 2009 First International Conference on Ubiquitous and Future Networks.

[5]  Martin Reisslein,et al.  Ethernet passive optical network architectures and dynamic bandwidth allocation algorithms , 2008, IEEE Communications Surveys & Tutorials.

[6]  Jong-Tae Lim,et al.  Dynamic bandwidth allocation algorithm in Ethernet passive optical networks , 2003 .

[7]  Walter Willinger,et al.  Proof of a fundamental result in self-similar traffic modeling , 1997, CCRV.

[8]  H. T. Mouftah,et al.  Media access control for Ethernet passive optical networks: an overview , 2005, IEEE Communications Magazine.

[9]  Walter Willinger,et al.  Self-similarity through high-variability: statistical analysis of Ethernet LAN traffic at the source level , 1997, TNET.

[10]  Chadi Assi,et al.  On the fairness of dynamic bandwidth allocation schemes in Ethernet passive optical networks , 2006, Comput. Commun..

[11]  Nirwan Ansari,et al.  Limited sharing with traffic prediction for dynamic bandwidth allocation and QoS provisioning over Ethernet passive optical networks , 2005 .

[12]  Zheng Wang,et al.  An Architecture for Differentiated Services , 1998, RFC.