shown in figure 1-2. Only one cable is required from the
central computer to each PCs network interface card to
tie that workstation to the LAN. The star is one of the
earliest types of network topologies. It uses the same
approach to sending and receiving messages as our
phone system. Just as a telephone call from one person
to another is handled by a central switching station, all
messages must go through the central computer or
network server that controls the flow of data. You can
easily add new workstations to the network without
interrupting other nodes. This is one of the advantages
of the star topology.
Another advantage of star topology is that the
network administrator can give selected nodes a higher
priority status than others. The central computer looks
for signals from these higher priority workstations
before recognizing other nodes. Also, the star topology
permits centralized diagnostics (troubleshooting) of all
functions. It can do this because all messages must first
go through the central computer. This can prove
invaluable for ensuring network security has not been
breached. So much for the good news; now for the bad
news, or the disadvantages of the star network. Of all
the topologies, the star is the least reliable because it has
a single point of failure. The network relies mainly on
the central computer for all functions. If it fails, all
nodes also stop functioning, resulting in failure of the
entire network. This is precisely the same weakness
multi-user computer systems have that rely on a central
The Bus Network
The bus topology is like a data highway. That is, all
components or nodes are connected to the same cable,
Figure 1-2.A star network topology.
and the far ends of this cable never meet (see figure 1-3).
Bus LANs are best suited to applications involving
relatively low usage of the bus coupled with the need to
pass relatively short messages from one node to another.
In many such networks, the workstations check whether
a message is coming down the highway before sending
their messages. Since all nodes share the bus, all
messages must pass through the other workstations on
the way to their destinations. Each node checks the
address attached to the message to see if it matches its
own address. Bus topologies allow individual nodes to
be out of service or to be moved to new locations
without disrupting service to the remaining nodes.
Unlike the star topology, where dozens of cables
come together at the central computer causing logistical
problems, bus cabling is simple. The bus topology is
very reliable, because if any node on the bus network
fails, the bus itself is NOT affected, and the remaining
nodes can continue to operate without interruption.
Many of the low-cost LANs use a bus topology and
twisted-pair wire cabling.
A disadvantage of the bus topology is that generally
there must be a minimum distance between
workstations to avoid signal interference. Another
disadvantage is that nodes must contend with each other
for the use of the bus. Simultaneous transmissions by
more than one node are NOT permitted. This problem,
however, can be solved by using one of several types of
systems designed to control access to the bus. They are
collision detection, collision avoidance, and token
passing, which we will discuss shortly. Also, there is no
easy way for the network administrator to run
diagnostics on the entire network. Finally, the bus
network can be easily compromised by an unauthorized
network user, since all messages are sent along a
common data highway. For this reason, it is difficult to
maintain network security.
Figure 1-3.A bus network topology.