CONTENTS
1. Abstract
2. Introduction
3. Technology behind Optical Networks
3.1. Taxonomy of Optical Networks
3.2. Broadcast and Select Networks
3.3. Wavelength Routing Networks
3.4. WDM Link Networks
3.5. Access Networks
3.6. Optical TDM based networks
4. Optical Components
4.1. Transmitters
4.2. Receivers
5. The Media Access Protocols
5.1. Overview
5.2.
6. Current Market Trends
7. Conclusion
8. Reference
Abstract - This paper provides a comprehensive review of Medium access protocols (MAC) for
optical networks. First we compares the different underlying network topologies used by WDM
based LANs then we investigate suitable media access protocols for these networks.
INTRODUCTION
All-Optical networks, where the entire path between end nodes is optical, represent the
third generation in the use of fiber in communication systems. In this generation, fiber is
used for its unique properties instead of merely as a replacement for copper.
Wavelength division multiplexing(WDM) can be used to provide multiple channels on a
single fiber, each operating at a macimum data rate set by electronic technology. This
technique avoids the electronic bottlenck due to the network interface, and networks
with an aggregate bandwidth approaching a terabit per second can be realized. It is
important to note that with transmission rates of 1 Gb/s or higher, the ratio of
propagation delay to transmission time increases to the point that there can be a large
number of packets “in the pipe”[5]”. AS a result, carrier sensing and token passing
protocols suddenly become impratical.
IN recent years there has been a wave of research toward the development of
wavelength-division multiplexing (WDM)-based local area networks (LAN’s) [1]–[10].
Most of the proposed protocols and architectures are based on a broadcast-select star
network architecture. Some of the protocols are based on random access and
consequently result in low throughput due to contention [3], [4]. Other protocols
attempting to minimize contention, through the use of some form of reservations,
require that the system be synchronized and slotted, and many of
these protocols require multiple transceivers per node [5]–[8]. Despite the added
complexity of these systems, most still fail to achieve high levels of utilization due to
inefficient scheduling scheme that fail to deal with receiver contention or ignore the
effects of propagation delays. A comprehensive survey of WDM multi-access protocols
and their properties is presented in [1],[2].
BACKGROUND AND DEFINITION
Potential areas of applications for this class of photonic networks are expanding rapidly,
due to the significant advances in wavelength tunable laser diode transmitters and
wavelength tunable receivers. Tunable receivers can be achieved via wavelength
tunable electro-optic or acousto-optic filters with direct detection, or coherent receivers
[10]. Tunable filters with direct detection and coherent detection can be viewed as
complementary approaches: coherent receivers have higher sensitivity, selectivity and
cost; whereas direct detection has reduced performance levels but reduced complexity
and cost. This expands their region of application in the performance-cost requirement
spectrum. Characteristics such as increased fanout, very large bandwidth, high
reliability, low power requirements, reduced crosstalk, and immunity to EMI (ElectroMagnetic Interference) make optical networks highly desirable. Multiple optical channels
can be formed on a single optical fiber through wavelength multiplexing to form
Wavelength Division Multiple Access (WDMA) channels. This is an approach to
circumvent the speed mismatch between the optics and the interface electronics:
multiple channels are created on a single fiber rather than creating a single (very) fast
channel. The WDM channels form a set that can be individually switched and routed.
However, tunable transmitters and receivers are required to achieve wavelength
division multiplexing. The following sections describe the
tunable optical devices, the star-coupled configuration, and the network architecture
considered in this paper.
NETWORK ARCHITECTURES
This section briefly outlines the current architectures of optical networks. Optical
networks can be divided according to the taxonomy tree in the Figure 1. The main
distinction between the various network types is based on the multiplexing scheme:
whether it is done in the frequency domain (WDM) or in the Time domain as in optical
time division multiplexing (OTDM). WDM networks may be future split in to Point-toPoint links – in which both ends of the link have identical equipment to transmit and
receive the channels. Access Networks – in which one side of the link gets split among
different locations (homes) and requires simpler equipment. Broadcast-and-select
systems – in which the signal is broadcast to multiple endpoints rather than a single
endpoint. This may further be classified those networks that have a single hop and
multiple hops networks.
Optical Networks
WDM
Broadcast and Select
WDM Link
TDM Based WDM
OTDM
Wavelength routing
Access Networks (PON)
Figure 1: Taxonomy of Optical Networks
WDM links
WDM technology is based on the ability to transmit several light signals on a single
fiber, using different wavelengths. It turns out that such different light wavelengths do
not interfere with each other, and thus they can be split apart at the other end of the
fiber to form separate channels, as shown in the Figure 2. The WDM link essentially
comprised of the following elements:
(i)
Different interfaces per port, to enable different protocols to communicate over
the link
(ii)
an elector-optical converter, which includes a laser per channel at different
wavelengths,
(iii)
an optical multiplexer, typically a piece of glass called a grating,
(iv)
(v)
(vi)
(vii)
Due to attenuation amplifiers may be need along the fiber or at the endpoints,
When the signal gets to the other end of the fiber it is split by an optical
demultiplexer, which acts like a prism, to separate wavelength-specific optical
signals (the same grating could be used again),
A wavelength insensitive receiver converts the signal into electrical form,
The signal is output via the specific interface of the channel’s port.
Broadcast and Select networks
Broadcast and select networks are based on a passive star coupler device, connected
to several stations in a star topology. This device is a piece of glass that splits the signal
it receives on any of its ports to all the ports. As a result it offers an optical equivalent of
radio system: each transmitter broadcasts its signal on a different wavelength and the
receivers can tune to receive the desired signal. See Figure 3 for a schematic drawing
of such a system.
The networking challenge in such networks pertains to the coordination of a pair of
stations in order to agree and tune their systems to transmit and receive on the same
wavelength. One design issue that must be determined before deciding on these
protocols is the tunable part of the system. It is either possible to have the transmitters
fixed on a different wavelength each and have tunable receivers, have fixed receivers
and unable transmitters, or have tuning abilities in both components. It has been shown
in [CDR90] that it is more advantageous to have tunable receivers and fixed
transmitters than the other way round.
The advantages of these networks are in their simplicity and the natural multicasting
capability. However they have severe limitations, since they do not enable reuse of
wavelengths and are thus not scalable beyond the number of supported wavelengths.
Another factor, which hinders scalability of these solutions, and disables it from
spanning long distances, is the splitting of the transmitted energy to all the ports. For
these reasons the relatively high costs of WDM transmitters and receivers, compared to
the low costs that other technologies provide – ATM and switched Ethernet, for example
– do not enable broadcast –and-select networks to be competitive in this arena
currently. The few niches, which appear to be appropriate for such networks, are
broadcast studios and supercomputer centers.
Wavelength routing networks
A scalable optical network can be constructed by taking several WDM links and
connecting them at a node by a switching subsystem. Using such nodes (also called
wavelength routers) interconnected by fibers, diverse networks with complex and large
topologies may be devised.
Each wavelength router makes its routing decision based on the input port and
wavelength of a connection going through it. Thus, if a light signal of wavelength 1
enters a router at a port x it is switched to some output port y. At the other end of fiber,
attached to y, the signal enters another router in which a similar routing decision is
made. This process continues until the signal is switched to an output port of the
system. Another optical signal coming into the same router on a different wavelength 1
will be routed differently. Such an end-to-end connection is called a lightpath, and it
provides a high speed transparent pipe to its end-users. At the same time, another
ligthpath can reuse the same wavelength in some other part of the network, as long as
both lightpaths do not use it on the same fiber. Since such “spatial reuse” of
wavelengths is supported by wavelength routing networks, they are much more scalable
than broadcast-and-select networks. Another important characteristics, which enables
these networks to span long distances in that the energy invested in a lightpath is not
split to irrelevant destinations.
There is a large diversity of capabilities that a wavelength router can provide,
depending on the components in use and design of the node. Most notably, may
provide” configurable lightpaths versus fixed routing, full wavelength conversion versus
limited conversion versus no conversion at all, fault tolerance in the optical layer versus
reliance on higher layers. Nodes may also vary in their scalability to increasing the
number of local or network ports.
As for the design of the node itself, current commercial technology enables one of the
following first two designs. The third design relies on large optical switches and
wavelength converters, a technology that is far from being commercially available, and
is therefor a longer-term option.
i.
Elector-optical node: Converts the optical signal into the electrical, perform
the switching in this domain, and regenerate the optical signal at the outputs
(Fig –5X). this design enables wavelength conversion easily and maintains a
high quality signal for multiple hops. On the other hand it does not support
transparency. This design is mentioned to represent evolutionary phase
towards all-optical networks.
ii.
Simple-all-optical node: separates the different wavelengths from each input,
sends all channels of wavelength i to the same switch, which optically
switches them to the output ports (fig- 5b). This design does not allow
wavelength conversion, thereby restricting the reuse of wavelengths in the
system. This may prove to be a cost-effective solution, as it does not require
a (costly) transceiver per channel per node.
iii.
Full conversion all-optical node: Such a node enables each wavelength to be
converted to any other wavelength. It is based on a large optical switch, which
takes a channel and switches it to any other channel (on any fiber). Before
being multiplexed into the fiber, each channel is converted to the appropriate
wavelength, by fixed wavelength converters ( see Figure 5c).
Passive optical access networks (PONs)
Passive optical access networks (PONs) enable bi-directional communication between
a server (such as a cable TV provider) and a set of customers. The main challenge in
these networks is to design such a system based on WDM technology, in which the
equipment at the customers’ side is as cheap, simple and durable as possible, but
above all – identical for all the customers (or else it will create impossible management
overheads, since it will be necessary to ensure that no two customers have identical
wavelength-specific devices).
Alternate Technologies
Optical TDM: Optical TDM attempts to copy conventional TDM ideas and realize them
optically, thereby achieving much higher speeds. Since this technology is in its very first
steps, where the most elementary estimate that it will not be a potential player in the
commercial arena in the next 10 years.
Optical Components
WDM networks require tunable transmitters and/or receivers to switch between the
multiple channels created on the single optical fiber. This section briefly outlines the
functional characteristics of the tunable devices.
Transmitters: Tunable lasers can be achieved through thermal, mechanical, injectioncurrent, and acousto-optic means [18, 19]. Thermal and mechanical approaches
achieve slow tunable devices (ms- ). Injection-current techniques are capable of tuning
speeds of a few ns. Dense WDM networks have become possible through the
significant advances in narrow line width Distributed Feedback (DFB) and Distributed
Bragg Re ector (DBR) tunable lasers and filters, and low cost star couplers [20]. The
range of a DFB laser has a limit of 10-15 nm, due to heating and non radiative
recombination [21]. A DBR laserdiode was also demonstrated in [21] to achieve 50
separate channels with a switching time between channels of 15 ns. Spectral slicing is a
low cost alternative to tunable laser diodes that has been recently introduced for this
environment
___
Receivers:
Star Coupler:
MEDIA ACCESS PROTOCOLS
Broadcast and Select network MAC protocols.
Media access control protocols developed for photonic star-coupled WDM networks
may be broadly classified into reservation (sometimes reffered as Pretransmission
Coordination) and pre-allocation strategies [9]. Reservation techniques may designate
one wavelength channel as the control channel that is used to reserve access on the
remaining channels (designated as data channels) for data packet transmission. The
control channel is used transmit control information and reserve access on the data
channels. Media access control protocols are required to provide arbitration on both the
data and control channels.
Pre-allocation techniques pre-assign the
Broadcast and Select
channels to the nodes, where each node has
a
home channel that it uses either for all data
Reservation
Pre-Allocation
packet transmissions or all data packet
receptions. This eliminates the requirement
that a
Token
node possess both a tunable transmitter and a
tunable receiver. Pre-allocation may be
Static
approached by either specifying the channel a
node
Random
will use to transmit (requires a tunable
receiver and a fixed transmitter) or receive
Hybrid
(requires a tunable transmitter and a fixed
receiver). The proposed approach achieves further reduction in system complexity by
eliminating the requirement of a control channel. Reservation protocols are often more
complex than pre-allocation protocols since the transfer is based on two stages:
reservation and transmission [5]. Depending on the implemented protocol, collisions
may occur during control and/or data packet transmission, which require a
retransmission of both. Pre-allocation approaches appear to be very promising due to
their low implementational and operational complexity.
Protocols based on Pretrasmission Coordination
Reservation is generally required when Tunable Receivers are used. The receiver must
be informed of the wavelength of each incoming transmission. All of the protocols based
on reservation employ one or more control channels, possibly embedded on the data
channels, for accomplishing this purpose. This coordination is not required if the
transmission schedule is fixed or pre-determined in a distributed manner at each
station. With the fixed receivers no pertransmission coordination is required since each
receiver is at a fixed wavelength for reception of incoming traffic.
The protocols requiring reservation are summarized in table 1 using the following
metrics:
 Equipment : This lists the components required at each station. Note that C is
defined to be the number of channels.



Channels: the first number of the pair is the number of control channels required
by the protocol. The second number is the number of data channels. N is defined
to be the number of stations in the network.
Tell and go: This feature allows a station to inform the destination it is
transmitting a packet and then transmit it without waiting. In reservation protocols
a station must wait for at least one rout-trip delay before transmitting.
Throughput: this is a relative measure of the maximum achievable throughput.
Collisions and/or destination conflicts limit the throughput as the offered load
increases. In general dynamic reservation schemes that avoid collisions and
destination conflicts can achieve the higest throughput.

Conclusion
MAC protocols that provide high throughput, low delay, simplicity, robustness, and
support for priorities and different traffic classes are good candidates for future
research. Continued development of fast tunable, wide range lasers and filters is
needed to implement the tunable transmitters and receivers required for the above
protocols. Another key to the widespread deployment of all-optical networks is cost
reduction for optical components. A final important issue is the development of an
efficient protocol stack that supports current and future heterogeneous network traffic.
As research in optical technology progresses, decisions will have to be made on the
optimal protocol layering for high-speed WDM networks.