Network layer mobility: an architecture and survey

Mobile end-systems frequently change their point of attachment to the network. In such an environment, in order for mobile devices to run without disruption, an inter-networking infrastructure is needed. In addition, a common networking protocol is required which can support network-wide mobility. Mobile devices also need to communicate with the existing pool of information servers and le servers, which means that internetworking solutions for connecting stationary and mobile systems are also required. Unfortunately, the Internet Protocol (IP), which forms the fabric of the current world-wide data communication network, falls short of meeting this demand. The current Internet suite of protocols (TCP/IP) were designed under the assumption that end-systems are stationary. If during an active network session one end of the connection moves, the network session breaks. Naturally, all networking services layered on top of TCP/IP are also disrupted when end-systems become mobile. There are two approaches for solving this problem. One is to completely redesign internetworking protocols with the speci c goal of supporting mobile end systems. The other approach is to provide additional services at the network layer in a backward compatible manner which make mobile internetworking possible. The rst approach, though an interesting possibility from a research viewpoint, is infeasible since it would require radical changes to the currently deployed networking infrastructure. It is the latter approach that is the focus of our investigation. To ensure inter-operability with the existing infrastructure, the handling of mobility should be completely transparent to the protocols and applications running on stationary hosts. In other words, from a stationary end-system's perspective, a mobile host should appear like any other stationary host connected to the Internet. This means the same naming and addressing conventions, those originally developed for stationary hosts, must apply to mobile hosts. In addition, any changes in a mobile's network attachment point should be completely hidden from the protocols and applications running on stationary hosts. In this paper we explore various network layer concepts that play a crucial role in the design of mobile networking systems. We show that mobility is essentially an address translation problem and is best resolved at the network layer. We have identi ed the fundamental services that must be supported at the network layer to carry out the task of address translation. Using these service primitives as building blocks, we describe a network layer architecture which enables smooth integration of mobile end systems within the existing Internet. In the second half of this paper, we