Demonstrate the working principle of distance vector routing algorithm (1)

Demonstrate the working principle of distance vector routing algorithm (1)

Modern computer networks usually use dynamic routing algorithms because these algorithms can adapt to network topology and traffic changes, the two most popular dynamic routing algorithms are the distance vector routing algorithm and the Link State routing algorithm ".

Distance Vector Routing (DV) is the earliest Routing Algorithm Used in the ARPANET network. It is also called the Bellman-Ford Routing algorithm and the Ford-Fulkerson Algorithm, it is mainly used in the RIP (Route Information Protocol) Protocol. Cisco's IGRP and OSPF routing protocols also adopt the DV routing algorithm.

The basic idea of "distance vector routing algorithm" is as follows: each vro maintains a distance vector (usually a delay variable) table, then, the distance vector table is updated through the distance vector announcement between neighboring routers. Each distance vector table includes two parts: the optimal output line to reach the destination node, and the time or distance required to reach the destination node. Each Other router in the Communication Subnet occupies one table item in the table, and used as the index of the table item. Each time, the router sends a distance table to all neighboring nodes, and also receives the distance table sent from each neighbor node. In this way, after a period of time, the Distance Vector information obtained by each vro in the network can be unified on each vro, in this way, each vro only needs to view the distance vector table to find an optimal route for different source groups.

Assume that latency is used as the distance measurement. For example, 7-37. Assume that at some time, vroy Y receives the distance vector of its neighbor router X, where m is the latency that Y estimates to reach vrox X. If the Router Y knows that its delay to neighbor Z is n, it can know that it takes m + n to reach X through Y. If the Z router has other adjacent routers, the distance vector received from each other's neighbors is calculated similarly, finally, select the least time-consuming route as the best route for Z to X, update its route table, and notify its neighbor router.

Simple Example of Distance Vector Routing Algorithm

Here, we use an example 7-38 to describe how to determine the route in the distance vector algorithm. The delay of each link is marked in the figure. Five routers A, B, C, D, and E represent the following route table: A → B → C → D → E (this is related to the sequence of vro startup ). The specific process below.

(1) In the initial state, each vro only collects the delay information of the directly connected link. Each vro node obtains its own initial vector table 7-39. Because route information is not exchanged between nodes, their initial route tables are also like their vector tables.

Figure 7-38 routing confirmation example of Distance Vector Algorithm

Vector table of each node in the initial state

(2) Now vroa A sends its route table to vrob B. In this case, the route table sent from router A and its own initial route table are updated to A new vector table, as shown in the 7-40 left figure (deep color in the final vector table ). It can be seen that there are two paths from Node B to node E, one is direct, and the other is through node. In addition, the overhead of the two lines is different. The overhead of arriving at the E node through node A (7) is lower than that of the direct line (8). Therefore, in the route table, change the line arriving at node E to the line passing through node A, as shown in the figure 7-40 on the right.

New vector table and route table of Node B

(3) B sends the final route table to router C. Similarly, router C should combine its original route table with the route table sent from router B to form a new vector table, figure 7-41 on the left (deep color in the final vector table ). In the new vector table, in addition to the vector between the B and D nodes directly connected at the beginning, the vector information that reaches the and E nodes is also collected. Because node C is not directly connected to node A and node E, the routing information does not reach the two nodes in the initial routing table. Therefore, only the routing table sent from router B is used, the path from Node B to node A and node E.