This thesis addresses the paucity of research on network mobility with a taxonomy and a
quantitative comparison on a real test bed of existing and forthcoming network mobility solutions. Broadly, network mobility solutions can be divided into two broad categories, intra-domain and inter-domain solutions. The two are distinct enough to warrant separate solutions. Network mobility solutions can also be divided into four catogories according to their implementation, (i) Mobile IP-based solutions, (ii) Mobile IP Mobile Routing, (iii) intra-domain routing protocols, and (iv) Mobile IP Foreign Agent-based solutions. This latter division by implementation serves as a useful starting point for a discussion and study of these protocols. The qualitative and quantitive comparisons presented here yield two "winning" solutions, Mobile IP with Network Address Translation (NAT) and Mobile IP Mobile Routing, both based on Mobile IP, the draft standard for Internet mobility. The two were chosen because they are the only solutions fit for production networks, not necessarily because they are the best performing solutions. Indeed, nearly all of the other solutions are promising alternatives and some, such as mobile ad hoc network (MANET) routing protocols and the solutions based on the next generation Internet (IPv6) version of Mobile IP, are particularly so. One of the solutions uses a novel extension for Mobile IP, the Multiple Home Address extension, developed and specified in this thesis, that turns a Mobile IP mobile node into a mobile router able to support dynamically sizing mobile networks while using the existing Mobile IP infrastructure.
