The term "topology" has many meanings, one of which is used in the computer world to describe networks. What is the topology of computer networks, and will be discussed further. But, running a little ahead, in the simplest case, this concept can be considered as a description of the configuration (location) of computers connected to the network. In other words, it all comes down to understanding not even the connections themselves, but the geometric shapes that correspond to each type of terminal arrangement.
What do you mean by LAN topology?
As it's already clear, computers connected to a single network are not connected to them randomly, but in a strictly defined order. To describe this scheme, the understanding of topology was introduced.
In fact, what is topology? Map, scheme, chart, map. The descriptive process, as is already clear, is somewhat akin to elementary knowledge of geometry. However, this term cannot be considered only from a purely geometric point of view. Since it is notonly about connections, but also about the transfer of information, in this regard, this factor should also be taken into account.
Main types of networks and their topologies
In general, there is no single concept of computer topology. It is generally accepted that there can be several types of topologies, collectively describing a particular network organization. Actually, networks can be completely different.
For example, the simplest form of organizing the connection of several computer terminals into a single whole can be called a local area network. There are also intermediate types of networks (urban, regional, etc.).
Finally, the largest are the global networks, which affect large geographic regions and include all other types of networks, as well as computers and telecommunications equipment.
But what is meant by the topology of a local network, as one of the simplest forms of organizing the connection of several computers to each other, in this case?
Based on the described processes and structures, they are divided into several types:
- physical - a description of the actual structure of the location of computers and network nodes, taking into account the connections between them;
- logical - description of the signal passing through the network;
- informational - description of the movement, direction and redirection of data within the network;
- management of the exchange - a description of the principle of using or transferring rights to use the network.
Network topology: types
Now a few words about the generally accepted classification of topology types by links. In the context of what topology is, it is worth noting separately another type of classification that describes only the way a computer connects to a network or the principle of its interaction with other terminals or main nodes. In this case, the concepts of fully connected and non-fully connected topologies become relevant.
A fully connected structure (and this is recognized all over the world) is extremely cumbersome due to the fact that each single terminal included in a single network structure is connected to all the others. The inconvenience in this case lies in the fact that for each computer it is necessary to install additional communication equipment, and the terminal itself must be equipped with a sufficiently large number of communication ports. And as a rule, such structures, if used, are extremely rare.
A non-fully connected topology in this regard looks much more preferable, since each individual terminal does not connect to all other computers, but receives or transmits information through certain network nodes or accesses directly to a central hub or hub. A striking example of this is the star network topology.
Since we are talking about the main methods of connecting terminals into a single whole (network), we should dwell on the main topologies of all main types, among which the main ones are "bus", "star" and "ring", althoughthere are also some mixed types.
Bus network topology
This type of networking of terminals is quite popular, although it has very serious drawbacks.
Consider what a "bus" topology is by a simple example. Imagine a cable with multiple taps on either side. At the end of each such branch is a computer terminal. They are not directly connected to each other, and information is received and transmitted through a single trunk, at both ends of which special terminators are installed that prevent signal reflection. This is a standard linear network topology.
The advantage of such a connection is that the length of the main trunk is significantly reduced, and the failure of a single terminal does not affect the operation of the network as a whole. The main disadvantage is that in case of violations in the operation of the backbone itself, the entire network turns out to be inoperable. In addition, the "bus" topology is limited in the number of connected workstations and has a rather low performance due to the distribution of resources between all terminals in the network. Distribution may be uniform or uneven.
Star topology (star)
The topology of the "star" network in a sense resembles a "bus", with the only difference being that all terminals are connected not to a single backbone, but to a central distribution device (hub).
Just through the hub all computers can interact with each other. Information is transmitted from the hub to all devices, but is received only by those for which it is intended. The advantages of such a connection include the possibility of centralized management of all network terminals, as well as the connection of new ones. However, as with the bus, the failure of the central switching device is fraught with consequences for the entire network.
Topology "ring" (ring)
Finally, we have one more type of connection - the ring topology of the network. As is probably already clear from the name, computers are connected sequentially from one to another through intermediate nodes, as a result of which a vicious circle is formed (naturally, a circle in this case is a conditional concept).
During transmission, information from the starting point passes through all terminals that are in front of the final recipient. But the recognition of the ultimate beneficiary is based on token access. That is, only the terminal marked in the information flow receives information. Such a scheme is practically never used due to the fact that the failure of one computer automatically entails a disruption in the operation of the entire network.
Mesh and mixed topology
This type of connections can be obtained by removing some connections from the above connections or adding them additionally. In most cases, this scheme is used in large networks.
In this regard, several mainderivatives. The most common are schemes such as "double ring", "tree", "lattice", "snowflake", "Close network", etc. As you can see even from the names, all these are variations on the theme of the main types of connections, which are taken for the base.
There is also a mixed type of topology, which can combine several others (subnets), grouped according to some characteristic features.
Conclusion
Now it's probably clear what topology is. If we make some general result, this concept is a description of how computers are connected on a network and how they interact. How this is done depends solely on the method of combining the terminals into one. And to say that today it is impossible to single out any one universal connection option. In each specific case and depending on the needs, one or another type of connection can be used. But in local networks, if we talk specifically about them, the most common is the "star" scheme, although the "bus" is still widely used.
It remains to be added that in the topology of networks one can also meet the concepts of centralization and decentralization, but they are mostly associated not with connections, but with the system for managing network terminals and exercising control over them. Centralization is clearly expressed in star connections, but decentralization is also applicable for this type, providing the introduction of additional elements in order to increase network reliability when the central switch fails. A fairly effective development in this regard isthe "hypercube" scheme, however, it is very difficult to develop.
