Mobile IP is the IETF proposed standard solution for handling terminal mobility among IP subnets and was designed to allow a host to change its point of attachment transparently to an IP network. Mobile IP works at the network layer (layer 3), influencing the routing of datagrams, and can easily handle mobility among different media (LAN, WLAN, dial-up links, wireless channels, etc.).
The generic problem with IP mobility is that when an IP node moves to a new subnet, it either has to change its IP address to reflect the new point of attachment, or the routers must have host specific routes for the mobile node. Both these alternatives have their drawbacks. Host-specific routes in general cannot be scaled up for Internet-wide use. Changing the IP address seen by the transport and the application layers every time a MN (Mobile Node) moves to a new network may be a solution to infrequent roaming, but not to mobility in general. This is because the transport layer (e.g. TCP) uses the IP address as an identifier, correlating IP packets to transport sessions. If this IP address is changed, then the correlation is lost and the sessions need to be restarted.
The generic problem with IP mobility is that when an IP node moves to a new subnet, it either has to change its IP address to reflect the new point of attachment, or the routers must have host specific routes for the mobile node. Both these alternatives have their drawbacks. Host-specific routes in general cannot be scaled up for Internet-wide use. Changing the IP address seen by the transport and the application layers every time a MN (Mobile Node) moves to a new network may be a solution to infrequent roaming, but not to mobility in general. This is because the transport layer (e.g. TCP) uses the IP address as an identifier, correlating IP packets to transport sessions. If this IP address is changed, then the correlation is lost and the sessions need to be restarted.
MOBILE IP with IPV6
Mobile IP solves the mobility problem by managing the correlation between a changing IP address (care-of address) and the static home address. The transport and application layers keep using the home address, allowing them to remain ignorant of any mobility-taking place. The home address is naturally routed to the Home Agent (HA), which maintains the mapping (“binding”) from the home address to the current (primary) care-of address (CoA). The HA will tunnel packets to the MN at its current point of attachment via the CoA. In Mobile IPv4 the care-of address can be either hosted by a Foreign Agent (FA in Figure 1) or co-located with the mobile node itself. The visited network always assigns the CoA, so that the routing of the packets to the mobile node will remain transparent to the routers in transit. The packets from the MN to the correspondent node (CN) will be routed naturally without going through the home agent. As the MN moves from one subnet to another, and its CoA changes, it will inform the HA of the new binding.
Mobile IP was originally defined for IPv4 (IETF RFC 2002). This definition has suffered from the fact that mobility support for IPv4 is an add-on, and the vast majority of IPv4 nodes do not support Mobile IP. For IPv6, the mobility support has been on the list of required features from the beginning. The Mobile IPv6 specification is on its way to becoming a standard, so it is expected that virtually all IPv6 deployments will include at least the minimal mobile IP support (i.e. the correspondent node functions).
Mobile IP was originally defined for IPv4 (IETF RFC 2002). This definition has suffered from the fact that mobility support for IPv4 is an add-on, and the vast majority of IPv4 nodes do not support Mobile IP. For IPv6, the mobility support has been on the list of required features from the beginning. The Mobile IPv6 specification is on its way to becoming a standard, so it is expected that virtually all IPv6 deployments will include at least the minimal mobile IP support (i.e. the correspondent node functions).
No comments:
Post a Comment