How many topologies are there

Network topology

A network topology is the typical arrangement and physical connection of devices in a network. Devices are hosts, such as clients and servers, that actively use the network. This also includes network components such as switches and routers, which have a distribution function and ensure that all network participants can establish a logical connection with one another. The network topology determines the components to be used and the access methods to the transmission medium.

The network topologies described below refer to packet-switching networks.

  • Point-to-Point (PtP) / point-to-point
  • Point-to-Multipoint (PtMP) / Point-to-Multipoint
  • Line / chain / line
  • bus
  • ring
  • Star / star
  • Tree
  • Mesh topology
  • Fully connected / fully meshed topology
  • Fabric / tissue

Point-to-Point (PtP) / point-to-point

With a point-to-point topology, there is only a simple and direct physical connection between two hosts or devices. The two devices can use these connections for mutual communication.
In an ad hoc environment, the hosts usually come together spontaneously and do not necessarily cooperate with one another over the long term.
In a network environment, the physical connection and the logical connection based on it usually exist permanently.
The point-to-point topology must not be confused with P2P or peer-to-peer architecture.

Point-to-Multipoint (PtMP) / Point-to-Multipoint

In a point-to-multipoint topology, several hosts are fed by a central system. All hosts within the topology must share a line to the central system. Each participant can have his own transmission medium. But only up to a branch point from which a line leads to the central system.

Line / chain / line topology (line technology)

With the line topology, several hosts are connected to one another. A line is laid from host to host. The two ends of the line are each terminated with a host.
The line topology is also called a daisy chain configuration, in which several hardware components are connected in series or series. This type of networking is often found in automation and safety technology.
The specialty of the line topology is that removing a host in the line leads to the line being interrupted and the network failing. You then have to bridge the gap or rewire at this point. As a consequence, this means that the connection logic that is based on it may have to be renegotiated between all participants (depending on the transmission system).
The line topology is often mixed up with the bus topology. There is often no distinction made in the physical cabling. With the transfer logic and access procedures based on it, yes.

Bus / bus topology

In the bus topology, all hosts are connected to one another via a common line. All hosts have access to the transmission medium and the signals that are transmitted over it. In order to prevent disturbances on the line and to improve the physical conditions, the two cable ends are provided with a terminating resistor.

There is no central network component that regulates the processes on the bus. An access procedure for the processes on the bus is responsible for this, and all hosts adhere to the rules of this procedure. The intelligence resides in the hosts and is usually specified by a protocol. The bus itself is only a passive transmission medium.

If the bus is to be expanded or if hosts are to be added or removed, the network can be down while the work is being carried out.

Ring / ring topology

The ring topology is a closed cable route in which the network participants are connected to one another with a continuous cable ring. This means that one cable arrives at each host and one cable goes out.
There is typically no need for any active network components in the ring topology. Control and access to the transmission medium are regulated by a protocol that all stations adhere to.
However, a ring manager makes sense. Because if the ring is interrupted at one point, it can switch to bus operation.
This means that the ring topology is redundant, which is important in production environments that rely on high availability.

Star / star topology

In the star topology there is a network component that maintains a physical connection to all hosts. Each host is connected to the central network component via its own line. It is usually a hub or a switch. The hub or switch takes on the distribution function for the data packets. For this purpose, the data packets are received and forwarded to the destination.
The data load on the central distributor can be very high, as all data and connections run through it. Hosts can be added or removed at any time. This has no effect on the operation of the network.

A network with a star-bus structure is a combination of star and bus topology. The stations are connected to a hub via a star structure. Several switches are connected to one another via a bus line.

Tree / tree topology

The tree topology is an extended star topology. Larger local networks have this structure. Especially when several topologies are combined with one another. Usually a higher-level network element, either a router or another topology, forms the root (shown here in reverse). From there a trunk with many branches and branches forms.

Mesh / mesh topology

The mesh topology or meshed topology is a decentralized network that does not have to be subject to any binding structures and in which all network nodes are somehow connected to one another.
A mesh network increases the range of the network, especially for the nodes on the edge. If a connection fails, there is usually always an alternative route to continue the data traffic without interruption. To do this, active network components must transmit the data packets within the network. For example through routing.

  • In a local area network (LAN), the mesh topology always occurs when a wild growth of different systems and several topologies emerges.
  • The Internet corresponds to the mesh topology because no central authority decides on the architecture of the network.

Fully connected

Fully Connected refers to a topology in which all hosts are connected to one another. This means that every host has its own physical connection to every host. The number of available interfaces on each host must be correspondingly large. At the same time, a transmission medium must also be available for each connection.
From a topology point of view, this is a perfect network, but it is only practiced to a limited extent. Fully Connected only makes sense if the full bandwidth of the physical connection is required for a logical connection between two network participants.

Fabric / braid topology (woven fabric)

A hierarchical structure and segmentation have long been the criteria for structuring a network. Because of the predominant client-server data stream, the star or tree topology with its centralized approach was the preferred cabling and networking architecture. Today, the data traffic has increased significantly due to dynamic content between the servers and the interaction of web, application and database servers. This means that the logical connections no longer only take place between clients and servers, but also between individual servers. In doing so, data streams occur that have very different characteristics. And that is why future network topologies and network architectures must be flexible and have more intelligence.

The concept of the fabric should meet all important requirements for the network in the data center:

  • high speed
  • Resilience
  • flexibility
  • easy management

A fabric has a star structure that does not have a central node, but connects the distributing components redundantly to form a structured, meshed topology. The fabric forms the basis for highly available, distributed systems.

The fabric is a network topology, the term and technical approaches of which come from the fiber channel world and have been used in storage networks for a very long time. They serve as a model for Ethernet in the data center. Here the fabric is a distributed architecture that, for example, combines several physical switches into one large logical switch.

For building networking, this means that additional core, floor and access switches must be used, which are connected to one another via additional lines. Here one often works with techniques from the area of ​​software defined networking.

Advantages and disadvantages of the basic topologies

Bus topology
  • easy to install
  • short lines
  • Network expansion limited
  • in the event of a cable break, the network fails
  • complex access methods
Ring topology
  • distributed control
  • large network expansion
  • costly troubleshooting
  • power failure in the event of malfunctions
  • high cabling effort
Star topology
  • simple networking
  • easy extension
  • high reliability
  • high cabling effort
  • Power failure if the hub fails or is overloaded
Mesh topology
  • decentralized control
  • infinite network expansion
  • high reliability
  • complex administration
  • expensive and high quality networking

Cabling expenditure in comparison

The high cabling effort of the bus and ring topology should not be underestimated. It may be that the cable route is shorter here than, for example, with the star topology. With the bus and ring topology, however, you often have to choose long and unusually winding cable routes because you have to route them from host to host. If you have to move a host, the cable pulling starts all over again.
It is much easier with the star topology. The cabling effort is a bit more complex in the first step. The lines of the star topology can be used flexibly for this. So you can operate different network technologies or other applications on the cable routes.

WLAN topology

A WLAN topology essentially consists of the wireless network subscribers, which are referred to as WLAN clients, and at least one WLAN base station, which is referred to as a wireless access point (WAP) or simply an access point (AP).

Other related topics:

Product recommendations

Everything you need to know about networks.

Network technology primer

The network technology primer is a book about the basics of network technology, transmission technology, TCP / IP, services, applications and network security.

I want that!

Everything you need to know about networks.

Network technology primer

The network technology primer is a book about the basics of network technology, transmission technology, TCP / IP, services, applications and network security.

I want that!