In order for data to be transferred over a network it has to pass through a number of "layers". Each of these layers functions are necessary to get the data through the network intact. The application layer, the only one visible to the user, prepares the data for transfer by converting it into data bits and attaching a header that identifies both the sender and receiver. This is the electronic envelope that we discussed earlier. The presentation layer translates the data bits into a generic language (usually ASCII, but it can be any number of network protocols) so that the receiving computer will be sure to be able to understand it. The presentation layer also handles compression and encryption tasks, if they are required, and adds data to the header specifying the language, compression and encryption types. The session layer opens communication and sets boundaries for the beginning and ending of the data being sent and writes these instructions in the session header. The next layer, called the transport layer, then takes over the data, protecting it by segmenting the data and creating checksum tests (mathematical data that can be used later to determine whether the data was corrupted) and making a backup copy of the data and writes the transport header in each packet with the checksum of each segment and its position in the data being transferred. The network layer forms these segments into packets counts them and selects a route for them to follow. The data-link layer supervises the transfer, confirming the checksum and once again copying the packet to a backup, which it holds until it receives confirmation that the packet was correctly transferred. The physical layer then encodes the packet into the form (analog or digital) it must take to travel over the medium and then sends it. As it travels the network medium intermediate nodes or stations verify each packet and pass it along. When the packets reach their destination the process starts all over again in reverse. It is within this layering that data can be filtered to reach its proper destination over the network.
Enterprise networking is the act of connecting multiple workgroup LANs within an enterprise. Inter-linking smaller LANs, as opposed to one giant LAN can provide better security and faster access. Properly implemented and administrated enterprise networks are much less restricted by such things as geography, the number of nodes supported, transfer rates, and packet traffic flow. A single LAN can only support so much data transfer. When traffic increases beyond its load capacity, stations attached to the LAN experience reduced throughput reducing productivity. Enterprise networking can effectively divide the load between multiple inter-networked LANs.
Enterprise networking makes use of three types of hardware in addition to the hubs, cabling and NICs we have previously discussed; bridges, routers, and switches.
Bridges and Routers
Routers originally designed to allow users to connect remote LANs across a wide area network, currently are the most widely used device in enterprise networking, but bridges are also be used for this purpose. Both Ethernet and Token Ring networks have specific limitations on maximum distances between workstations and hubs, hubs and hubs, and a maximum number of stations or nodes that can be connected to a single LAN. As you approach these limits performance begins to suffer. Bridges and routers were first used locally to extend these limitations by breaking up larger LANs into smaller segments and then linking these LAN segments together. This process was then extended through the use of telecommunications to cover large geographical distances.
In essence bridges and routers are filters, using information found in network protocol layers, so that only those packets that need to pass from one LAN to another are forwarded across the link. This keeps the packets sent between two stations on any one LAN from crossing over onto the other LANs and creating non-essential traffic. Since generally about 80 percent of the packets transmitted on a typical workgroup are destined for stations on that LAN. Both bridges and routers can be used to segment LANs.
Bridges filter packets between LANs by making a forward/don't forward decision on each packet they receive from any of the networks they are connected to. Filtering is done based on the destination address of the packet. If a packet's destination is a node on the same LAN segment it originated from, it is not forwarded. If it is destined for a node on another LAN, it is forwarded to the port that handles the destination;s LAN segment. What bridges lack in complexity they make up for in speed and cost-efficiency, which means they are very useful for high traffic networks.
Routers are significantly more complex network devices and, as such, are generally more expensive than bridges. They use network layer protocol information, discussed earlier, to route packets from one LAN to another. Routers must have the ability to recognize all of the different network layer protocols that may be used on the networks it is linking together. Routers also communicate with each other to determine the best route that links the sending and receiving nodes.
LAN switches are a relatively new type of enterprise networking hardware, and many believe that switched networks will be wave of the future. LAN switches and switching hubs are the first steps on a migration path towards asynchronous transfer mode (ATM) switching, an emerging technology that will be widely implemented in both LANs and wide area networks in the coming years.
Switches, like bridges and routers, are capable of linking several separate LANs using packet filtering to best route data transfers. A LAN switch is a device with multiple ports, each of which can support a single node or LAN. With a different LAN connected to each of the switch's ports, it can switch packets between LANs as needed. Essentially a high speed, multiport bridge; switches filter packets based on the destination address. Switches increase network performance by providing each port with dedicated bandwidth, and are quickly gaining popularity because no changes are required to any equipment, such as NICs, hubs, cabling, or bridges and routers currently in place. Using LAN switches allows designers to create several small network segments, meaning that fewer stations are competing for bandwidth.
ATM is a type of switching technology that switches small, fixed-length cells containing data. ATM networks can be run at ultra high speed, and most network techs believe that it won't be long until ATM is used to carry video, voice, and virtual reality data over both short and long distances. ATM is sure be one of the dominant enterprise networking technologies of the future, and many companies are beginning to develop strategies to incorporate ATM in their existing networks.