Sparse Regeneration in Translucent Wavelength-Routed Optical
Networks: Architecture, Network Design and Wavelength
Information Sciences Institute, University of Southern California, USA; Department of Computer Science and Engineering,
University of Nebraska-Lincoln, Lincoln, NE, USA
Department of Computer Science and Engineering, University of Nebraska-Lincoln, Lincoln, NE, 68588-0115 USA
Received April 1, 2004; Revised February 28, 2005; Accepted March 3, 2005
Abstract. In this paper we study an alternate network architecture, called translucent network, to the fully transparent and fully
opaque network architectures. In a translucent wavelength-routed optical network, a technique called sparse regeneration is used to
overcome the severe lightpath blocking due to signal quality degradation and wavelength contention in a fully transparent network
while using much less regenerators than in a fully opaque network. In this paper, we present a node model and a network model that
perform sparse regeneration. We address the problem of translucent network design by proposing several regenerator placement
algorithms based on diﬀerent knowledge of future network traﬃc patterns. We also address the problem of wavelength routing under
sparse regeneration by incorporating two regenerator allocation strategies with heuristic wavelength routing algorithms. We compare
the performance of diﬀerent regenerator placement algorithms and wavelength routing schemes through simulation experiments. The
beneﬁt of sparse regeneration is quantitatively measured under diﬀerent network settings.
Keywords: translucent optical network, network design, sparse regeneration, physical impairments, signal quality, bit error rate (BER),
regenerator placement, regenerator allocation, routing and wavelength assignment (RWA), wavelength-division multiplexing (WDM)
Due to the advent of new applications, such as the
real-time data, voice and video, e-Business, and
multimedia, today’s computer and telecommuni-
cation networks, particularly the Internet, will
continue to see an ever-increasing demand for
network bandwidth. Wavelength division multi-
plexing (WDM) is a promising technique to
accommodate such a demand by utilizing the huge
bandwidth of optical ﬁbers. With the advances in
optical switching technologies, WDM wavelength
routed networks (WRN) emerge as promising
candidates for future Internet backbones. WRN
diﬀers from conventional networks in that its
traﬃc has a coarser granularity, i.e., at a wave-
length level, and that it uses optical signals for
both transmission and switching.
In previous studies, a WRN is mostly referred to
as an all-optical network, where a lightpath is
routed from the source node to destination node
without undergoing optical-electrical-optical (O/
E/O) conversion at any intermediate node [1,2].
Although there are many practical reasons sup-
porting the deployment of all-optical networks, in
the current phase these networks must face the
technical diﬃculties in overcoming the physical
impairments introduced by long-haul ﬁbers and
Corresponding author. This work was carried out while the author was a graduate student at the University of Nebraska-Lincoln
*This work was supported by NSF grants (ANI-0074121 and EPS-0091900).
Portions of this work have appeared in the Proceedings of the OSA Optical Fiber Communications (OFC 1999) Conference  and the
Proceedings of the IEEE Global Telecommunications (GLOBECOM 2001) Conference .
Photonic Network Communications, 10:1, 39–53, 2005
2005 Springer ScienceþBusiness Media, Inc., Manufactured in The Netherlands.