DYNAMIC RECONFIGURATION OF OPTICALLY INTERCONNECTED NETWORKS WITH TIME-DIVISION MULTIPLEXING

Citation
Cm. Qiao et al., DYNAMIC RECONFIGURATION OF OPTICALLY INTERCONNECTED NETWORKS WITH TIME-DIVISION MULTIPLEXING, Journal of parallel and distributed computing, 22(2), 1994, pp. 268-278
Citations number
22
Categorie Soggetti
Computer Sciences","Computer Science Theory & Methods
ISSN journal
07437315
Volume
22
Issue
2
Year of publication
1994
Pages
268 - 278
Database
ISI
SICI code
0743-7315(1994)22:2<268:DROOIN>2.0.ZU;2-5
Abstract
Routing performance of optical interconnection networks is limited by both switch complexity and network connectivity. One way to overcome t hese limitations is to allocate the network bandwidth in a time-divisi on multiplexed (TDM) fashion. With this technique, an appropriate subs et of input-to-output connections can be established during a time slo t and all possible connections can be established over several time sl ots. Emulating a fully connected network, however, requires a large mu ltiplexing degree, and thus introduces latencies which may be prohibit ive. As a solution, we propose a technique called reconfiguration with time-division multiplexing (RTDM). With RTDM, only a subset, as requi red by applications, of all possible connections needs to be multiplex ed in the network by letting the network go through a sequence of conf igurations. Network reconfiguration with TDM can be done either static ally or dynamically. Static RTDM is applied when communication require ments of an application are known a priori. This paper deals mainly wi th dynamic RTDM, which requires run time control to accommodate dynami c connection requests. We show that reconfiguration overhead can be am ortized over a sequence of configurations. In particular, we describe how the complexity of a dynamic reconfiguration control algorithm can be reduced through pipelined processing of requests. Our simulations s how that dynamic RTDM allows for fair and fast allocation of network r esources to connection requests. As a result, network service time dec reases and communication efficiency increases. (C) 1994 Academic Press , Inc.