TY - JOUR
T1 - Parallel QoS scheduling for WDM optical interconnection system using a new ranked hopfield neural network
AU - Tien, Po-Lung
AU - Ke, Bo Yu
PY - 2011
Y1 - 2011
N2 - In this paper, we propose a parallel QoS scheduler for a WDM optical interconnection system (WOPIS), using a new ranked Hopfield neural-network (RHNN). The WOPIS contains a set of Clos-like optical switches and a handful of output FDL-based optical buffers. The RHNN scheduler determines an optimal set of neurons (I/O paths) to be enabled, achieving maximal system throughput and priority differentiation subject to the switch-and buffer-contention-free constraints. Structured with ranked neurons, the RHNN allows higher-rank neurons (higher-priority and/or lower-delay paths) to disable lower-rank neurons that have been enabled during previous iterations. Ranking the neurons unfortunately gives rise to a convergence problem. We present two theorems that give the sufficient conditions for the RHNN scheduler to converge to the optimal solution. We demonstrate via simulation results that, with the computation time within one system slot time, the RHNN scheduler achieves near 100% throughput and multi-level prioritized scheduling.
AB - In this paper, we propose a parallel QoS scheduler for a WDM optical interconnection system (WOPIS), using a new ranked Hopfield neural-network (RHNN). The WOPIS contains a set of Clos-like optical switches and a handful of output FDL-based optical buffers. The RHNN scheduler determines an optimal set of neurons (I/O paths) to be enabled, achieving maximal system throughput and priority differentiation subject to the switch-and buffer-contention-free constraints. Structured with ranked neurons, the RHNN allows higher-rank neurons (higher-priority and/or lower-delay paths) to disable lower-rank neurons that have been enabled during previous iterations. Ranking the neurons unfortunately gives rise to a convergence problem. We present two theorems that give the sufficient conditions for the RHNN scheduler to converge to the optimal solution. We demonstrate via simulation results that, with the computation time within one system slot time, the RHNN scheduler achieves near 100% throughput and multi-level prioritized scheduling.
KW - Hopfield neural networks
KW - Quality of service (QoS)
KW - optical interconnect
KW - parallel scheduling
UR - http://www.scopus.com/inward/record.url?scp=80051540968&partnerID=8YFLogxK
U2 - 10.1109/JLT.2011.2159578
DO - 10.1109/JLT.2011.2159578
M3 - Article
AN - SCOPUS:80051540968
SN - 0733-8724
VL - 29
SP - 2436
EP - 2446
JO - Journal of Lightwave Technology
JF - Journal of Lightwave Technology
IS - 16
M1 - 5876292
ER -