« PreviousContinue »
METHOD AND SYSTEM FOR MOBILE IP NODES IN HETEROGENEOUS NETWORKS
 This invention relates to a method for mobile IP nodes in heterogeneous networks in which a home agent dynamically assigns a temporary IP care-of address to a static IP home address when the mobile node is moved in the heterogeneous networks, the dynamic care-of address indicating the topologically current network location of the mobile node and IP data packets having the IP home address of the mobile node as destination address being redirected to the care-of address of the mobile node. In particular, the invention relates to a method for mobile nodes in heterogeneous networks with real-time applications.
 In the last few years, the number of Internet users worldwide and thus the information offered there has increased exponentially. Although the Internet offers access to information worldwide, we normally have no access to it, however, until we have arrived at a particular network access point, such as e.g. an office, school, university or at home. The growing range of IP-capable mobile devices, such as e.g. PDAs, mobile radio telephones and laptops, is beginning to change our concept of the Internet. An analogous transition from fixed nodes in networks to flexible requirements through increased mobility has only just begun. In mobile telephony, for example, this tendency has also manifested itself inter alia in new standards such as WAP, GPRS or UMTS. To understand the difference between the current reality and the IP linking possibilities of the future, one can call to mind, as a comparison, the development of telephony in the direction of mobility in the last twenty years.
 Mobile computer use must not be confused with computer use and network capability as we know them today. With mobile network use, an existing IP access to applications on the mobile node should not be interrupted when the user changes his location in the network. On the contrary, all changes of link and interface e.g. during a change into different networks (Ethernet, mobile radio network, WLAN, Bluetooth, etc.) should be able to take place automatically and not interactively, so that the user does not even need to know about them. This also applies to a change of interface e.g. during the use of real-time applications. Real mobile IP computing has many advantages based on a stable access to the Internet at all times. With such an access, work can be done freely and independently of a desk. The requirements for mobile nodes in networks differ in many ways, however, from the mentioned development in mobile radio technology. The end points in mobile radio communication are usually people. With mobile nodes, however, computer applications can carry out interactions between different network participants without any human assistance or intervention. Enough examples therefor are to be found in airplanes, on ships or in automobiles. Thus mobile computing in particular with Internet access together with other applications, such as e.g. in combination with position determining devices, such as the satellite-based GPS (Global Positioning System), can be useful.
 One of the problems with mobile network access via Internet Protocol (IP) is that the IP used for routing the data packets from the source address to the destination address in the network uses so-called IP addresses (IP: Internet Protocol). These addresses are assigned to a fixed location in the network, similar to how the telephone num
bers of the fixed network are assigned to a physical jack. When the destination address of the data packets is a mobile node this means that a new IP network address must be assigned with each change of network location, which makes impossible transparent, mobile access. These problems were solved by the mobile IP standard (IETF RFC 2002, October 1996) of the Internet Engineering Task Force (IETF) in that mobile IP allows the mobile node to use two IP addresses. One of these addresses is the normal, static IP address (home address), which indicates the location of the home network, whereas the second is a dynamic IP care-of address, which designates the current location of the mobile node in the network. The assignment of the two addresses allows the IP data packets to be rerouted to the correct, momentary address of the mobile node.
 Not all the problems of mobile network use are solved with the mobile IP of the IETF, however. If, for instance, a user would like to switch between two different network interfaces while an IP application is running, the IP connection is interrupted at the moment when he leaves the old network link. This connection is interrupted at least until at the mobile node the new link to the network has been made and until the new location, i.e. the new care-of address, is known and has been registered at the so-called home agent. The home agent is normally a fixed network node, which administers the two addresses of the mobile node (home address and care-of address) and reroutes or routes the corresponding data packets. If the interruption time for the change exceeds the time-out delays specified e.g. in the TCP (Transfer Control Protocol) for dead times, the IP connection is interrupted of course anyway. Even when the interruption time lies within the time-out delays specified in the TCP, however, the IP applications are not able to maintain the connection if a physical network interface is not permanently available. Examples of this are the change of the network card in a mobile node (e.g. a portable PC) having only one available card plug-in for the physical network interfaces. In the case of such a change of physical network interface, the IP applications or respectively the kernel receive the message that no physical network device can be assigned any longer to the IP data tunnel, and cut off the connection. This leads to the IP applications having to be restarted normally after change of network card in order to be able to access a particular IP data tunnel. Another problem is that, on the side of the mobile node, the data packets get lost in downtime between the connections since no physical network device is assigned anymore. Not only does this result in a loss of data, but it also causes the transmission rate of the IP packets through the IP applications to be slowed down corresponding to the duration of the downtime. As soon as the new connection has been made, the transmission rate is increased, at first only step-by-step, however. This slows down the IP application unnecessarily with every change of interface or location.
 Network interfaces are traditionally divided up into different layers. Of interest for the present invention are the lowermost layers. A distinction is made between layer 1 (LI), which corresponds to the physical network interface (e.g. the network interface card NIC), layer 2 (L2) at which an initial recognition and identification of the interface through software is made possible, and layer 3 (L3) as IP layer (IP: Internet Protocol), at which level the distinction between different IP network links for software applications of the system takes place as well as the connection of the IP