Unfortunately it will turn out that many of the

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: which the minimum-cost paths are ≤ H hops. Hence, the convergence time in the absence of packet losses of other is equal to the length (i.e., number of hops) of the longest minimum-cost path in the network. In the next lecture, when we augment the protocol to handle failures, we will calculate the bandwidth consumed by the protocol and discuss some of its shortcomings. In particular, we will discover that when link or node failures occur, this protocol behaves poorly. Unfortunately, it will turn out that many of the solutions to this problem are a two-edged sword: they will solve the problem, but do so in a way that does not work well as the size of the network grows. As a result, a distance vector protocol is limited to small networks. For these networks (tens of nodes), it is a good choice because of its relative simplicity. In practice, some examples of distance-vector protocols include RIP (Routing Information Protocol), the first distributed routing protocol ever developed for packet-switched networks; EIGRP, a proprietary protocol developed by Cisco; and a slew of wireless mesh network protocols (which are variants of the concepts described above) including some that are deployed in various places around the world. ￿ 18.5 A Simple Link-State Routing Protocol A link-state protocol may be viewed as a counter-point to distance-vector: whereas a node advertised only the best cost to each destination in the latter, in a link state protocol, a node advertises information about all its neighbors and the link costs to them in the advertisement step (note again: a node does not advertise information about its routes to various destinations). Moreover, upon receiving the advertisement, a node re-broadcasts the advertisement along all its links.4 This process is termed flooding. As a result of this flooding process, each node has a map of the entire network; this map 4 We’ll assume that the information is re-broadcast even along the link on which it came, for simplicity. CHAPTER 18. NETWORK ROUTING - I 10 WITHOUT ANY FAILURES Figure 18-5: Link-state advertis...
View Full Document

Ask a homework question - tutors are online