Multiprotocol Label Switching Virtual Private Networks: Problems, Protocols, Possibilities

Multiprotocol Label Switching Virtual Private Networks: Problems, Protocols, Possibilities

Jan Schankin (Christchurch Polytechnic Institute of Technology, New Zealand) and Eduardo Correia (Christchurch Polytechnic Institute of Technology, New Zealand)
DOI: 10.4018/978-1-60960-732-6.ch020


Multiprotocol Label Switching (MPLS) was originally designed with the intention of improving the speed with which routers could forward packets in Internet Protocol (IP) networks. Due to significant improvements in packet forwarding, this is no longer an advantage, but the technology has found large-scale industry-wide acceptance because of its greatly widened scope and application. Multiple extensions and enhancements to the protocol make it capable of solving an array of current service provider and customer network requirements for a converged network in an IP dominated world. The chapter considers the use of MPLS for the provisioning of a virtual private network over a shared physical infrastructure, and discusses the logic and functionality of key protocols associated with MPLS. It then goes on to explore the problems, protocols, and possibilities of these technologies in current environments.
Chapter Preview


Service provider networks face the problem of connecting multiple customers at various locations across a shared physical infrastructure. Legacy networks followed a strict connection-oriented approach in which data is sent across a pre-established physical or logical path. This deterministic approach followed the method used by the original analogue public switched telephone network (PSTN) and led to the development of other circuit switched technologies such as Integrated Services Digital Network (ISDN) and in its original form Synchronous Optical Networking (SONET) or Synchronous Digital Hierarchy (SDH). Some packet switched technologies follow the connection-oriented approach by establishing virtual circuits between two or multiple points across a shared network, including the old but reliable X.25 and unreliable Frame Relay, as well as ATM and MPLS (Perros, 2005).

These technologies function at either Layer 1 (L1) or L2 but on a fundamental level, they actually serve the same purpose as the now ubiquitous IP protocol, which functions at L3. It seems strange, especially for a new generation of network engineers for whom networks are automatically IP and ATM or Frame Relay legacy protocols, to view IP as a protocol that simply connects multiple Ethernet-based networks. That, though, is the actual fundamental purpose of IP: connecting local area networks through a hierarchical addressing scheme. The great flexibility, ease of implementation and maintenance plus overall cost effectiveness have led to a near total dominance of IP in local area network sector with Ethernet as the protocol of choice for the Data Link layer.

Complete Chapter List

Search this Book: