In the modern age, saturated with a large number of computers and equipment, the services of designing computer networks are particularly relevant. The stability and long operation of all devices will depend on how the layout is carefully planned. As a basis, you can take the manual on the design of a computer network of an energy company.
Introduction
A computer network is a computer system that combines computers and the necessary peripheral equipment located in a certain limited area. In most cases, this is one office or building, or one branch of the company. Today, a computer network has become one of the main attributes of computing systems in which there is more than one computer.
The main advantages that can be obtained from the local network are the ability to share some common resources, exchange data, have centralized storage, use printers or access the Internet.
An equally important aspect is to obtain a fault-tolerant system that can continue to function if some of its parts fail. This is achieved through the use of so-called redundancy and duplication in the design of computer networks.
As a rule, the construction of a reliable system that will meet all the stated requirements and have the lowest cost, begins with the development of the plan. It takes into account various characteristics, selects the necessary topology, software and hardware.
Description of the required network
When designing a computer network, you need to consider that it will have special characteristics and parameters. In particular, it should enable the information department of the company to maintain the operability of all hardware and software, allow it to administer the necessary network resources and create backup copies of the database.
To ensure all clients of the network, it is necessary that it has access to high-speed Internet. It is also necessary to organize a large repository for all stored documentation. The network considered in the article will be designed for 40 jobs. They will be divided into two segments, each containing 20 cars. There will also be one shared server. The location of the wires connecting all the machines is assumed in special channels, which, in turn, are placed under an artificial floor covering. This will save on the length of all cables.
Network topology
Any computer network is primarily divided into computers and data transmission networks. They can be divided into several categories:
In this case, a local area network will be used. This means that the machines located in it will be located within either one room or a building. Such networks have several basic advantages:
at a high data transfer rate, within 100 megabits per second;
not a big chance of error during data transfer;
short cable length i .
The word "topology" means the physical location of computers on the same network. This is one of the standard terms used by specialists to describe the placement of elements. The following characteristics very often influence the choice of a particular topology:
with the equipment left to be used in it;
Characteristics of this equipment;
in the ability and ability of the network to expand;
with the administration and management.
Basically, there are 5 types of different topologies: bus, star, ring, cellular, honeycomb or concentric.
Bus topology
It is often called a linear bus. It is used in the simplest and most common computing systems. Only one cable is used, which is a trunk. In turn, all computers of this network are already connected to this highway. Data transmission in this type of topology occurs due to the addressing of data to a particular node sent via cable through electrical signals. At a certain point in time, only one machine can transmit. And this means that the performance of this topology very much depends on the number of computers participating in this system. The larger their number, the slower the exchange will occur.
Naturally, this factor is not the only one that affects the data transfer rate. There are also general characteristics of the system hardware, the frequency of computers accessing it, the category of network applications running on machines, the type of cable, the distance between computers. At its core, a bus is a passive topology. This means that machines can only listen to data that is transmitted over the network. That is, they cannot participate in the direct transfer of this information to another computer.
One of the features of the network is the fact that data moving along it can be reflected when it reaches the end of the cable. This may interfere with normal operation. Therefore, it is necessary to somehow suppress these signals when they are received by the addressee. For this, special terminators are used.
If the integrity of the cable is violated, the network completely stops functioning. At the same time, machines can continue to continue independent work. These features must be taken into account when designing computer networks.
Ring topology
As you can guess from the name, the segments of this network are arranged in a circle. That is, in a closed ring. With such a design system, one of the advantages is sending a request immediately to all workstations of a computer network.
The main problem of this topology is the fact that each station should be directly involved in the transmission of information. And this means that the failure of one participant will cause a violation of the integrity of the entire system. Also, accepting a new member into the ranks will require opening.
Star topology
This view was developed for large computers. In such a system, there is one host machine that processes data coming from peripheral devices or other terminals. The basic principle of such a topology is that all information between two workstations should always pass through a common node of the entire computer network.
The throughput of such a topology is characterized by the processing power of the nodes and is described for each workstation. And since this system uses a common center, it means that you need to connect every workplace with it. This means that the cost of laying cables will be much higher than in other topologies. But the star is one of the fastest, as data is transmitted between two workstations through one common node. From here comes a direct correlation of the performance of the entire network on the performance of this node.
In general, a star topology has the following advantages:
If one workstation fails, the network will continue to operate;
n allows you to easily scale the system by connecting new jobs and stations;
Easily detect a malfunction or break;
It's very easy to administer.
Along with this, there are several disadvantages:
If the main node or hub fails , the network will be inaccessible;
Organization requires a large amount of cable.
It is this computer network design system that is ideal for solving the current problem.
A bit about OSI models
The methodology for designing computer networks involves a large number of operations that ensure the transfer of information from one node to another. At the same time, the user who is working on the current machine does not suspect how the process of data exchange occurs. In fact, many interesting processes are happening behind the scenes. Firstly, the information is divided into special blocks, each of which carries a special service information. The resulting blocks are formed in the form of network packets. They can be encoded, encrypted, and then transmitted through electrical or light signals. It is the network models that precisely serve to streamline all levels of procedures that occur during transmission and reception of packets.
In the modern world, the OSI reference system is most widely used. It was approved in 1984. Since then, it has been used as a basis by almost all equipment manufacturers to take into account all the features of computer network design.
In short, the model divides the entire transmission of information into seven major tasks or levels. The 2 lowest of them are characterized by hardware implementation. The top five implement the transfer through software.
However, this model is still theoretical, therefore, during the design of computer networks of the enterprise there are always some deviations or changes. In general, of all they are particularly interested in the TCP / IP network protocol .
TCP / IP Protocol Stack
TCP / IP is a tool for exchanging data between machines in a single network.
The main reason why this protocol is very common is the fact that it can allow computer network design systems to connect to the Internet or create new segments. All information transmitted through this protocol is IP packets. Recognition of the addressee in it occurs by issuing an IP address.
Choosing software and hardware
Among the main stages of designing computer networks, the choice of tools that will ensure the interaction of nodes stands out. You can start with cable groups. In any computer network, they can be present up to 3. This is coaxial, twisted pair and optical fiber. The first, due to its high cost, is not widespread. On the other two, you can stay in more detail.
Cable
Twisted pair is a type of communication cable that consists of several pairs of insulated conductors twisted together. It is inexpensive, easy to install, so it is widely used in the installation and design of computer networks. Connection is made using an RJ-45 standard connector.
Fiber optic provides one of the fastest types of data transmission. But since it is quite expensive to maintain and connect, it is mainly used when connecting the Internet to a building, where further the network is built using twisted pair cable.
Equipment
As switching equipment, several different elements can be used. Repeaters are used to extend the network connection by repeating the electrical signal. Hubs, or hubs, combine several nodes into one common segment. When a signal is received, the hub sends it to all available ports. But due to the large number of collisions, hubs are now being used less and less. They were replaced by switches, or switch. They are much more productive than hubs and are able to transmit data to a specific destination. Routers help reduce network load by splitting into multiple domains, as well as packet filtering.
Workstation selection
For the design and administration of computer networks, jobs are selected in accordance with available financial resources and preferences. The server, in most cases, has much greater resources and increased computational characteristics. In fact, it will be he who will be the main “thinking” node in the network.
Conclusion
The article gave an example of the theoretical foundations of designing computer networks. In fact, this process is quite lengthy and complicated. At each stage, you will need to solve a lot of problems. Therefore, the implementation of the laying and installation of the network is usually carried out by experienced specialists who possess both theoretical knowledge and practical experience in this area.