In a network of computers, the manner in which multiple nodes are physically interconnected is referred to as its topology. The communication media that is used for interconnection can be broadly categorized into two types: point-to-point and broadcast.
In case of point-to-point channel, communication media of different types such as twisted-pair, co-axial cable or optical fiber are used to connect any pair of nodes in the network. In case of broadcast communication, a single communication channel such as a common bus or air is shared by all the nodes in the network. Information sent by any machine is received by all others.
Here we will discuss common topologies implemented using these two types of communication media.
One straight-forward approach is to use a fully connected or Mesh topology. In this topology, if there are n nodes or computers in the network, each node has (n-1) ports to connect to the remaining (n-1) nodes using point-to-point links. We require n(n-1)/2 full duplex links to realize the fully connected topology.
In this topology the cost of cabling and I/O hardware grows at the rate of square of the number of nodes. As adding a new node involves modification of the entire cabling and adding one port to each node, the topology is inflexible. As a consequence, this topology is not commonly used.
In a star topology, all the nodes are connected to a central node or bus as shown in the figure below.
For example, the central node can be a switching element such as a PBX telephone exchange, which is used to establish dedicated path between any pair of nodes using circuit switching techniques. The central node can also be an active node, such as a computer. In this case, the central node performs all routing of messages between any pair of outer nodes or to a remote node from any one of the outer nodes. Since there is a central control point, the control of the network is easy and priority can be given to selected nodes. Moreover, star topology facilitates easy maintenance and testing of the network. However, the reliability of the entire network depends on the reliability of the central node and there is a heavy burden on the central node.
In this topology, the nodes are interconnected in the form of a ring with each node connected by unidirectional point-to-point links to its adjacent nodes on either side as shown in the figure below.
This topology is commonly realized by interfacing each node to a repeater, which regenerates and re-transmits each received bit. This involves one bit delay. Each received bit is re-transmitted with a small delay, required to allow the repeater to perform required functions. Ideally the delay should be of the order of one bit time i.e. the time it takes for a repeater to transmit one complete bit onto the outgoing line.
The repeater is also provided with the capability of data insertion, data reception and data removal. Any node can insert data into the ring in the form of a packet, which contains the address of the destination node. After a packet is inserted, it moves from one repeater to the next repeater in a particular direction. As the data passes through the repeater of a particular node, the address field is copied and is compared with its own address. If there is a match, the entire data packet is copied. This operation is known as data reception. The packet is usually removed from the ring by the transmitting node after it has made one trip around the loop. Apart from the data removal operation, this approach facilitates automatic acknowledgement and multicast addressing, i.e. sending data to multiple nodes.
In ring topology no separate action is needed for routing data. It is however prone to single node or single link failure. Reliability can be improved by using bypass relays to bypass faulty nodes or links. Bypass relay also allows forwarding of data without any delay.
In a tree topology nodes are arranged in a hierarchical order having several levels as shown in the figure below. This type of topology usually results from the cascaded connection of networking equipment, such as hubs and switches for interconnecting computers. This topology helps to grow the network in an incremental manner without disturbing the existing setup.
In a bus topology all the nodes are connected to a common transmission medium as shown in the figure below. As a consequence, only one node can transmit at a time.
The data packet sent out by a node propagates throughout the medium and is received by all other nodes. As each node has a unique address in the network, a node copies a packet only when the destination address of the packet matches with the self address. This is how data are received by a node. To transmit a packet, a node has to gain control of the medium. For this purpose usually distributed medium access control technique are used.
As the bus is a broadcast medium, no special technique is needed for routing or forwarding a message to the destination node. The lack of routing and lack of centralized medium access control gives substantial reliability to this topology.
Another advantage of this topology is its flexibility. Nodes can be added or removed very easily and the failure of a node does not usually affect the operation of other nodes.
Some of the disadvantages of this topology are; restriction on the length of the cable due to delay, signal unbalancing due to multi-point access.
All the topologies discussed so far are symmetric and constrained by well defined interconnection pattern. However, sometimes no definite pattern is followed and nodes are interconnected in an arbitrary manner using point-to-point links as shown in the figure below.
Unconstrained topology allows a lot of configuration flexibility but suffers from the complex routing problem. Complex routing involves unwanted overhead and delay.
The choice of transmission media and topology cannot be made independently. They together decide the reliability, expandability and performance of the network. The topology used for different media are given by the following table.
|Twisted-pair||Bus, Ring, Star|
|Base-band Coaxial||Bus, Ring, Star|
|Broadband Coaxial||Bus, Tree|
|Optical Fibre||Ring, Tree|
Rajeev Kumar is the primary author of How2Lab. He is a B.Tech. from IIT Kanpur with several years of experience in IT education and Software development. He has taught a wide spectrum of people including fresh young talents, students of premier engineering colleges & management institutes, and IT professionals.
Rajeev has founded Computer Solutions & Web Services Worldwide. He has hands-on experience of building variety of websites and business applications, that include - SaaS based erp & e-commerce systems, and cloud deployed operations management software for health-care, manufacturing and other industries.