Network topology
Diagram of
different network topologies.
Network topology is
the physical interconnections of the elements (links, nodes,
etc.) of a computer network. A local area network (LAN) is one example of a
network that exhibits both a physical topology and a logical
topology. Any given node in the LAN has one or more links to one or more
other nodes in the network and the mapping of these links and nodes in a graph
results in a geometrical shape that may be used to describe the physical
topology of the network. Likewise, the mapping of the data flows between the
nodes in the network determines the logical topology of the network. The
physical and logical topologies may or may not be identical in any particular
network.
Any particular network topology is determined only by the graphical mapping
of the configuration of physical and/or logical connections between nodes. The
study of network topology uses graph
theory. Distances between nodes, physical interconnections, transmission
rates, and/or signal types may differ in two networks and yet their topologies
may be identical.
Basic topology types
The study of network topology recognizes four basic topologies:
·
Bus
topology
·
Star
topology
·
Ring
topology
·
Tree
topology
Classification of network topologies
There are also three basic categories of network topologies:
·
physical
topologies
·
signal
topologies
·
logical
topologies
The terms signal topology and logical
topology are often used interchangeably, though there is a subtle
difference between the two.[citation needed]
Physical topologies
The mapping
of the nodes of a network and the physical connections between them – i.e., the
layout of wiring, cables, the locations
of nodes, and the interconnections between the nodes and the cabling or wiring
system[1].
Classification of physical topologies
Point-to-point
The simplest topology is a permanent link between two endpoints (the line
in the illustration above). Switched point-to-point topologies are
the basic model of conventional telephony. The value of a permanent point-to-point network
is the value of guaranteed, or nearly so, communications between the two
endpoints. The value of an on-demand point-to-point connection is proportional
to the number of potential pairs of subscribers, and has been expressed as Metcalfe's Law.
Permanent
(dedicated)
Easiest to
understand, of the variations of point-to-point topology, is a point-to-point communications channel that appears, to
the user, to be permanently associated with the two endpoints. Children's
"tin-can telephone" is one example, with a microphone to a single
public address speaker is another. These are examples of physical dedicated
channels.
Within many switched telecommunications systems, it is
possible to establish a permanent circuit. One example might be a telephone in
the lobby of a public building, which is programmed to ring only the number of
a telephone dispatcher. "Nailing down" a switched connection saves
the cost of running a physical circuit between the two points. The resources in
such a connection can be released when no longer needed, for example, a
television circuit from a parade route back to the studio.
Switched:
Using circuit-switching
or packet-switching
technologies, a point-to-point circuit can be set up dynamically, and dropped
when no longer needed. This is the basic mode of conventional telephony.
Bus
Bus network
topology
In local area
networks where bus topology is used, each machine is connected to a single
cable. Each computer or server is connected to the single bus cable through
some kind of connector. A terminator is required at each end of the bus cable
to prevent the signal from bouncing back and forth on the bus cable. A signal
from the source travels in both directions to all machines connected on the bus
cable until it finds the MAC address or IP address on the network that is the
intended recipient. If the machine address does not match the intended address
for the data, the machine ignores the data. Alternatively, if the data does
match the machine address, the data is accepted. Since the bus topology
consists of only one wire, it is rather inexpensive to implement when compared
to other topologies. However, the low cost of implementing the technology is
offset by the high cost of managing the network. Additionally, since only one
cable is utilized, it can be the single point of failure. If the network cable
breaks, the entire network will be down.
Linear bus
The type of
network topology in which all of the nodes of the network are connected to a
common transmission medium which has exactly two endpoints (this is the 'bus',
which is also commonly referred to as the backbone,
or trunk) – all data that is transmitted between
nodes in the network is transmitted over this common transmission medium and is
able to be received by all nodes in the network virtually
simultaneously (disregarding propagation
delays).
Note: The two endpoints of the common
transmission medium are normally terminated with a device called a terminator that exhibits the characteristic impedance of the transmission medium and which
dissipates or absorbs the energy that remains in the signal to prevent the
signal from being reflected or propagated back onto the transmission medium in
the opposite direction, which would cause interference with and degradation of
the signals on the transmission medium (See Electrical termination).
Distributed
bus
The type of
network topology in which all of the nodes of the network are connected to a
common transmission medium which has more than two endpoints that are created
by adding branches to the main section of the transmission medium – the
physical distributed bus topology functions in exactly the same fashion as the
physical linear bus topology (i.e., all nodes share a common transmission
medium).
Notes:
1.) All of the endpoints of the common transmission medium are normally
terminated with a device called a 'terminator'.
2.) The physical linear bus topology is sometimes considered to be a
special case of the physical distributed bus topology – i.e., a distributed bus
with no branching segments.
3.) The physical distributed bus topology is sometimes incorrectly referred
to as a physical tree topology – however, although the physical distributed bus
topology resembles the physical tree topology, it differs from the physical
tree topology in that there is no central node to which any other nodes are
connected, since this hierarchical functionality is replaced by the common bus.
Star
Star network
topology
In local area networks with a star topology, each network host is connected
to a central hub. In contrast to the bus topology, the star topology connects
each node to the hub with a point-to-point connection. All traffic that
transverses the network passes through the central hub. The hub acts as a
signal booster or repeater. The star topology is considered the easiest
topology to design and implement. An advantage of the star topology is the
simplicity of adding additional nodes. The primary disadvantage of the star
topology is that the hub represents a single point of failure.
Notes:
·
A
point-to-point link (described above) is sometimes categorized as a special
instance of the physical star topology – therefore, the simplest type of
network that is based upon the physical star topology would consist of one node
with a single point-to-point link to a second node, the choice of which node is
the 'hub' and which node is the 'spoke' being arbitrary[1].
·
After
the special case of the point-to-point link, as in note 1.) above, the next
simplest type of network that is based upon the physical star topology would
consist of one central node – the 'hub' – with two separate point-to-point
links to two peripheral nodes – the 'spokes'.
·
Although
most networks that are based upon the physical star topology are commonly
implemented using a special device such as a hub or switch
as the central node (i.e., the 'hub' of the star), it is also possible to
implement a network that is based upon the physical star topology using a
computer or even a simple common connection point as the 'hub' or central node
– however, since many illustrations of the physical star network topology
depict the central node as one of these special devices, some confusion is
possible, since this practice may lead to the misconception that a physical
star network requires the central node to be one of these special devices,
which is not true because a simple network consisting of three computers
connected as in note 2.) above also has the topology of the physical star.
·
Star
networks may also be described as either broadcast multi-access or nonbroadcast multi-access
(NBMA), depending on whether the technology of the network either automatically
propagates a signal at the hub to all spokes, or only addresses individual
spokes with each communication.
Extended star
A type of network topology in which a network that is based upon the
physical star topology has one or more repeaters between the central node (the
'hub' of the star) and the peripheral or 'spoke' nodes, the repeaters being
used to extend the maximum transmission distance of the point-to-point links
between the central node and the peripheral nodes beyond that which is
supported by the transmitter power of the central node or beyond that which is
supported by the standard upon which the physical layer of the physical star
network is based.
If the repeaters in a network that is based upon the physical extended star
topology are replaced with hubs or switches, then a hybrid network topology is
created that is referred to as a physical hierarchical star topology, although
some texts make no distinction between the two topologies.
Distributed Star
A type of network topology that is composed of individual networks that are
based upon the physical star topology connected together in a linear fashion –
i.e., 'daisy-chained' – with no central or top level connection point (e.g.,
two or more 'stacked' hubs, along with their associated star connected nodes or
'spokes').
Ring
Ring network
topology
In local area
networks where the ring topology is used, each computer is connected to the
network in a closed loop or ring. Each machine or computer has a unique address
that is used for identification purposes. The signal passes through each
machine or computer connected to the ring in one direction. Ring topologies
typically utilize a token passing scheme, used to control access to the
network. By utilizing this scheme, only one machine can transmit on the network
at a time. The machines or computers connected to the ring act as signal
boosters or repeaters which strengthen the signals that transverse the network.
The primary disadvantage of ring topology is the failure of one machine will
cause the entire network to fail.
Mesh
The value of fully meshed networks is proportional to the exponent of the
number of subscribers, assuming that communicating groups of any two endpoints,
up to and including all the endpoints, is approximated by Reed's Law.
Fully connected
mesh topology
Note: The physical fully connected mesh
topology is generally too costly and complex for practical networks, although
the topology is used when there are only a small number of nodes to be
interconnected.
Partially
connected mesh topology
Partially
connected
The type of
network topology in which some of the nodes of the network are connected to
more than one other node in the network with a point-to-point link – this makes
it possible to take advantage of some of the redundancy that is provided by a
physical fully connected mesh topology without the expense and complexity
required for a connection between every node in the network.
Note: In most practical networks that are based
upon the physical partially connected mesh topology, all of the data that is
transmitted between nodes in the network takes the shortest path (or an
approximation of the shortest path) between nodes, except in the case of a
failure or break in one of the links, in which case the data takes an
alternative path to the destination. This requires that the nodes of the
network possess some type of logical 'routing' algorithm to determine the
correct path to use at any particular time.
Tree
Tree network
topology
Also known as a hierarchical network.
The type of network topology in which a central 'root' node (the top level
of the hierarchy) is connected to one or more other nodes that are one level
lower in the hierarchy (i.e., the second level) with a point-to-point link between
each of the second level nodes and the top level central 'root' node, while
each of the second level nodes that are connected to the top level central
'root' node will also have one or more other nodes that are one level lower in
the hierarchy (i.e., the third level) connected to it, also with a
point-to-point link, the top level central 'root' node being the only node that
has no other node above it in the hierarchy (The hierarchy of the tree is
symmetrical.) Each node in the network having a specific fixed number, of nodes
connected to it at the next lower level in the hierarchy, the number, being
referred to as the 'branching factor' of the hierarchical tree.
1.) A network
that is based upon the physical hierarchical topology must have at least three
levels in the hierarchy of the tree, since a network with a central 'root' node
and only one hierarchical level below it would exhibit the physical topology of
a star.
2.) A network
that is based upon the physical hierarchical topology and with a branching
factor of 1 would be classified as a physical linear topology.
3.) The branching
factor, f, is independent of the total number of nodes in the network and,
therefore, if the nodes in the network require ports for connection to other
nodes the total number of ports per node may be kept low even though the total
number of nodes is large – this makes the effect of the cost of adding ports to
each node totally dependent upon the branching factor and may therefore be kept
as low as required without any effect upon the total number of nodes that are
possible.
4.) The total
number of point-to-point links in a network that is based upon the physical
hierarchical topology will be one less than the total number of nodes in the
network.
5.) If the nodes
in a network that is based upon the physical hierarchical topology are required
to perform any processing upon the data that is transmitted between nodes in
the network, the nodes that are at higher levels in the hierarchy will be
required to perform more processing operations on behalf of other nodes than
the nodes that are lower in the hierarchy. Such a type of network topology is
very useful and highly recommended.
Signal topology
The mapping of the actual connections between the nodes of a network, as
evidenced by the path that the signals take when propagating between the nodes.
Note: The term 'signal topology' is often used
synonymously with the term 'logical topology', however, some confusion may
result from this practice in certain situations since, by definition, the term
'logical topology' refers to the apparent path that the data takes between
nodes in a network while the term 'signal topology' generally refers to the
actual path that the signals (e.g., optical, electrical, electromagnetic, etc.)
take when propagating between nodes.
Example
Logical topology
The logical topology, in contrast to the "physical", is the way
that the signals act on the network media, or the way that the data passes
through the network from one device to the next without regard to the physical
interconnection of the devices. A network's logical topology is not necessarily
the same as its physical topology. For example, twisted pair Ethernet is a
logical bus topology in a physical star topology layout. While IBM's Token Ring
is a logical ring topology, it is physically set up in a star topology.
Classification of logical topologies
The logical classification of network topologies generally follows the same
classifications as those in the physical classifications of network topologies,
the path that the data takes between nodes being used to determine the
topology as opposed to the actual physical connections being used to
determine the gyutt
Notes:
1.) Logical
topologies are often closely associated with media access control (MAC) methods
and protocols.
2.) The logical
topologies are generally determined by network protocols as opposed to being
determined by the physical layout of cables, wires, and network devices or by
the flow of the electrical signals, although in many cases the paths that the
electrical signals take between nodes may closely match the logical flow of
data, hence the convention of using the terms 'logical topology' and 'signal
topology' interchangeably.
3.) Logical
topologies are able to be dynamically reconfigured by special types of
equipment such as routers
and switches.
Daisy chains
Except for star-based networks, the easiest way to add more computers into
a network is by daisy-chaining, or connecting each computer in series
to the next. If a message is intended for a computer partway down the line, each
system bounces it along in sequence until it reaches the destination. A
daisy-chained network can take two basic forms: linear and ring.
·
A linear
topology puts a two-way link between one computer and the next.
However, this was expensive in the early days of computing, since each computer
(except for the ones at each end) required two receivers and two transmitters.
·
By
connecting the computers at each end, a ring topology
can be formed. An advantage of the ring is that the number of transmitters and
receivers can be cut in half, since a message will eventually loop all of the
way around. When a node sends a message, the message is processed by
each computer in the ring. If a computer is not the destination node, it will
pass the message to the next node, until the message arrives at its
destination. If the message is not accepted by any node on the network, it will
travel around the entire ring and return to the sender. This potentially
results in a doubling of travel time for data.
Centralization
The star
topology reduces the probability of a network failure by
connecting all of the peripheral nodes (computers, etc.) to a central node.
When the physical star topology is applied to a logical bus network such as Ethernet, this
central node (traditionally a hub)
rebroadcasts all transmissions received from any peripheral node to all
peripheral nodes on the network, sometimes including the originating node. All
peripheral nodes may thus communicate with all others by transmitting to, and
receiving from, the central node only. The failure of a transmission
line linking any peripheral node to the central node will result in the
isolation of that peripheral node from all others, but the remaining peripheral
nodes will be unaffected. However, the disadvantage is that the failure of the
central node will cause the failure of all of the peripheral nodes also.
If the central node is passive, the originating node must be able to
tolerate the reception of an echo
of its own transmission, delayed by the two-way round trip transmission
time (i.e. to and from the central node) plus any delay generated in the
central node. An active star network has an active central node that
usually has the means to prevent echo-related problems.
A tree
topology (a.k.a. hierarchical topology) can be viewed as
a collection of star networks arranged in a hierarchy.
This tree has individual peripheral nodes (e.g.
leaves) which are required to transmit to and receive from one other node only
and are not required to act as repeaters or regenerators. Unlike the star
network, the functionality of the central node may be distributed.
As in the conventional star network, individual nodes may thus still be
isolated from the network by a single-point failure of a transmission path to
the node. If a link connecting a leaf fails, that leaf is isolated; if a
connection to a non-leaf node fails, an entire section of the network becomes
isolated from the rest.
In order to alleviate the amount of network traffic that comes from
broadcasting all signals to all nodes, more advanced central nodes were
developed that are able to keep track of the identities of the nodes that are
connected to the network. These network
switches will "learn" the layout of the network by
"listening" on each port during normal data transmission, examining
the data packets and
recording the address/identifier of each connected node and which port it's
connected to in a lookup table held in memory. This lookup table then
allows future transmissions to be forwarded to the intended destination only.
Decentralization
In a mesh topology (i.e., a partially
connected mesh topology), there are at least two nodes with two or more
paths between them to provide redundant paths to be used in case the link
providing one of the paths fails. This decentralization is often used to
advantage to compensate for the single-point-failure disadvantage that is
present when using a single device as a central node (e.g., in star and tree
networks). A special kind of mesh, limiting the number of hops between two
nodes, is a hypercube.
The number of arbitrary forks in mesh networks makes them more difficult to
design and implement, but their decentralized nature makes them very useful.
This is similar in some ways to a grid
network, where a linear or ring topology is used to connect systems in
multiple directions. A multi-dimensional ring has a toroidal topology,
for instance.
A fully connected network, complete topology or full
mesh topology is a network topology in which there is a direct link
between all pairs of nodes. In a fully connected network with n nodes, there are
n(n-1)/2 direct links. Networks designed with this topology are usually very
expensive to set up, but provide a high degree of reliability due to the
multiple paths for data that are provided by the large number of redundant
links between nodes. This topology is mostly seen in military
applications. However, it can also be seen in the file
sharing protocol BitTorrent in which users connect to other
users in the "swarm" by allowing each user sharing the file to
connect to other users also involved. Often in actual usage of BitTorrent any
given individual node is rarely connected to every single other node as in a
true fully connected network but the protocol does allow for the possibility
for any one node to connect to any other node when sharing files.
Hybrids
Hybrid networks use a combination of any two or more topologies in such a
way that the resulting network does not exhibit one of the standard topologies
(e.g., bus, star, ring, etc.). For example, a tree network connected to a tree
network is still a tree network, but two star networks connected together
exhibit a hybrid network topology. A hybrid topology is always produced when
two different basic network topologies are connected. Two common examples for
Hybrid network are: star ring network and star bus network
·
A Star
ring network consists of two or more star topologies connected using a multistation
access unit (MAU) as a centralized hub.
·
A
Star Bus network consists of two or more star topologies connected using a bus
trunk (the bus trunk serves as the network's backbone).
While grid networks have found popularity in high-performance computing applications,
some systems have used genetic algorithms to design custom networks that
have the fewest possible hops in between different nodes. Some of the resulting
layouts are nearly incomprehensible, although they function quite well.