9.1. Introduction

Routing refers to the process whereby a node finds one or more paths to possible destinations in a network. Routing consists of several components: neighbor discovery, topology discovery, and path selection. Neighbor discovery is typically the first step of routing whereby the nodes in a network discover the identity of their immediate neighbors and how they are connected to these neighbors. Topology discovery is the process by which a node discovers other nodes in the network and how they are connected to each other. Once the network topology is known, paths from a source to a destination can be computed using path computation algorithms.

Due to the tremendous popularity of the Internet, routing today has become synonymous with IP routing, or routing in the IP (network) layer [Stallings03]. Optical networks are different from IP networks in many ways. But IP routing protocols, with some adaptations and extensions, can be used to dynamically route connections in optical networks. Routing in optical networks is also related to routing in telephone networks and Asynchronous Transfer Mode (ATM) networks.

In this chapter, we review the routing architectures and protocols in different networks. Given the popularity of IP and the Internet, we discuss IP routing protocols in detail. We also briefly discuss routing in telephone networks and in ATM networks. In the chapters that follow, we discuss how the IP routing protocols can be extended and enhanced to handle connection routing in dynamic optical networks.

In this and the following chapters, we distinguish between routing within a network controlled by one administrative entity and routing across networks controlled by different administrative entities. The primary focus of the next chapter is routing within a network managed by a single administrative entity, also known as intradomain routing. Routing across networks managed by different administrative entities, known as interdomain routing, is the subject of Chapter 12.