Photonic Network Communications, 1:3, 207±218 (1999)
# 1999 Kluwer Academic Publishers, Boston. Manufactured in The Netherlands.
Distributed Control Protocols for Wavelength Reservation and their
Department of Computer Science, Florida State University
R. Melhem*, R. Gupta
Department of Computer Science, The University of Pittsburgh
Y. Mei, C. Qiao
Departments of CSE and EE, University at Buffalo (SUNY)
Received April 28, 1999; Revised June 25, 1999
Abstract. This paper describes distributed (or decentralized) protocols for establishing wavelength paths in point-to-point WDM networks.
Distributed control improves reliability as well as scalability and reduces implementation cost of a network, but also presents major challenges
in managing/allocating wavelengths ef®ciently. Two types of distributed wavelength reservation protocols are proposed and evaluated, namely
forward and backward. The bene®t of wavelength conversion is assessed ®rst based on the evaluation of the forward reservation protocols, and it
is found that wavelength conversion can result in a performance advantage in a distributed environment. For networks without wavelength
conversion, a class of backward reservation protocols is studied and shown to generally perform better than their forward counterparts.
Keywords: wavelength division multiplexing, path setup, dynamic connections establishment, decentralized wavelenght selection, all-optical
networks, distributed control
All-optical networks, especially wavelength-division
multiplexed (WDM) networks, have received an
enormous amount of attention, as exempli®ed by
several projects including AON , MONET ,
TBONE  and various European projects (e.g.
ACTS-OPEN). In such networks, a lightpath is
usually established before data is transferred between
two communicating nodes, thus eliminating the need
for buffering and electronic-to-optical or optical-to-
electronic conversions during the data transfer at the
intermediate nodes. Most of these lightpaths may be
semi-permanent when used to provide virtual topol-
ogies to higher layers (e.g. ATM or SONET/SDH) but
some applications such as those supporting bursty
traf®c directly on top of WDM or fast protection and
restoration at the WDM layer may require frequent
recon®guration of the lightpaths.
Two basic approaches, namely path multiplexing
(PM) and link multiplexing (LM) [13,16], can be used
to establish lightpaths in all-optical WDM (or TDM)
networks. In PM, a lightpath is established by using
the same wavelength that is available on all the links
along the path. In LM, different wavelengths that are
available on different links along the path can be used
provided that all-optical wavelength converters are
available at each node [11,24].
The network control (or signaling) required for
establishing a connection may be either centralized or
This work was supported in part by NSF award MIP 9409864 and ANIR 98-01778 to University at Buffalo (SUNY), and by NSF awards CCR-
9157371 and MIP 96-33729 to the University of Pittsburgh.
* Preferred Address: Rami Melhem Department of Computer Science (219 MIB), The University of Pittsburgh, Pittsburgh,
PA 15260. Phone: 412-624-8426. Fax: 412-624-5249. email: email@example.com