Network Parameters and Classes

Modern Internet networks, as a rule, are based on 32-bit IP-addresses, consisting of two parts - the identifier of the network and the host. To determine which part of the address is the host and which part represents the network, two methods have been developed. Now, ISPs use the method of classless addressing, based on subnet masks. Classes of networks are the first, already outdated method, which was based on ranges.

The IP address of any object, be it a server or a regular computer, is closely related to the network name. Converts this name to a network address, a special DNS service that manages domain names. "Response" to the network name will only be the server registered in this service. The resources of such servers automatically become public, and you can use them through the Internet.

Having understood with IP-addresses, we pay attention to classes of networks. There are five in all, and each has its own characteristics. Classes A are used for huge global networks. Included in it and the Internet. The range of this class extends from zero to 127 and consists of 126 networks. One A-network accommodates more than sixteen million nodes. The network ID itself occupies only eight first bits, the remaining 24 bits are reserved for the host address.

The class of networks B consists of medium-sized networks that occupy the address range up to a value of 191. Here, the IP address is divided into the same 16-bit parts. One part is occupied by the identification number of the network, and the other is reserved for the host. The B-network unites 65534 nodes. As a rule, it is used in universities or large enterprises.

Class C supports small grids. They occupy a range of up to 223. The network number is the first 24 bits, and the remaining 8-bit space is provided to the host. The C-network can host no more than 256 nodes, two of which are reserved for IP mailings. It's worth adding that the addresses of these three classes are involved in routing and building subnetworks on a global network scale. That is why they are called "real", or "white".

The remaining classes of networks do not play such a significant role. D-networks are in the range up to 239. They do not implement access to nodes, but multicast IP-messages are broadcast. Networks of class E also do not contain nodes. Their range is up to 255, and they themselves are experimental.

All listed classes have address blocks reserved for private use. They are used exclusively in private LANs, so on the Internet these addresses are not routed and are called "gray", or "private". A NAT router is used to connect private LANs and exit through them to the World Wide Web.

Since the above classes of networks have a limited number of IP addresses, it is inconvenient to use them. The alternative is the way in which the number of bytes is not limited and the subnet mask is used. However, the old system is not forgotten. It is described in many textbooks, and addresses of class D and E are now used privately.