Routers and routing algorithms

telecommunications networks. data - is. routers. (routers), called in English vellum and routers.

I think our readers understand how the general functioning routers, how to manage the network and allow traffikomi find the best route for data transmission (we 'll talk about the packets of information). In this paper we consider routing algorithms - how to find these optimal routes.

basics.

To find the shortest network path that is the way with a minimum of transportation between the routers to the time the inevitable delay in switching, routing algorithms use some.

Based on how routers gather information about the surrounding network, we can distinguish two major routing algorithms: centralized (global routing algorithms), and ( distributed ), decentralized (decentralized routing algorithms). As part of a distributed algorithm, each router has information on routers with which it is directly connected, he does not know about all routers in the network. This algorithm is also known as DV (distance vector). In a global, centralized routing, each device has information on all routers in the network and the state of the network data transfer. Another name for this algorithm - LS (link state).

LS- algorithm ( centralized routing).

By using this LS- algorithm, each router must sovresh sleduyuyuschie steps:.

a. Identify devices that are physically connected with it and get their IP- addresses. When the router is a network, it first sends a sort of 'Hello'- packet in the network and each router that receives such a message, sends a response containing the IP-address.

2. Measure the time delay (or average data rate ) for the neighboring devices. In order to do that, routers send echo- packets into the network, which receive the appropriate response. Dividing the time of the request on the road for 2, one can calculate the time delay, and at this time will be laid down as the period of delay in packet processing by routers.

3. Broadcast the information about the structure of its network segment and receive relevant information from. other routers. At this stage, routers determine the best routes for each node using a known algorithm for finding the shortest path Dijkstra (Dijkstra). Under this method, the router is based on information gathered builds a network graph showing the relative positions of routers and networks between. Each vertex of this graph is marked with numbers, called weights. It is an indicator that reflects the time delay, and sometimes the number of transitions between nodes. Thus, to establish a connection, the router uses the path with the lowest weight assigned to the nodes in the path.

Pros and cons.
ease of implementation.
- complexity of the topology change and load.
- there is no exchange of routing data between two routers.

DV- algorithm ( decentralized or distributed routing).

Algortitm distributed routing, also known as Bellman- Ford algorithm (Bellman-Ford) and Ford - Fulkerson (Ford-Fulkerson).

Also known as Distributed Bellman-Ford Routing or Ford Fulkerson Algorithm. It can be metaphorically described as: ... Each router maintains a routing table with one entry for each node in the subnet. The table is a vector containing two components: the selected line and the distance. A node estimates the distance ( number of hops between nodes, delay or queue length ) to each neighboring device and sends it to his neighbors, who repeat these steps. As a result of the information collected, each. router. again generates a routing table.

That is, the router performs the following actions:.

a. Assesses the weight of links directly related devices.

2. At certain times of sending this information to the neighboring router it gets relevant information from a neighboring router.

3. Based on this information builds or updates its routing table.

Pros and cons.
Self-Organization.
the relative ease of implementation.
low - convergence, or ...
- expansion of the network due to some difficulties.

The hierarchical (mixed ) routing.

The distributed and centralized routing is inevitably hampered the growth of network nodes. To solve this problem is designed hierarchical division of the network.

In hierarchical routing, the routers are divided into groups. Each router only contains information about the routers in the group and has no information about the routers and the other groups. Moreover, groups can be nested, allowing you to distribute routing information. Such a scheme can be used multiple routing protocols within the group and is widely distributed on the Internet.

Source:. http://computer. howstuffworks. com /.