Integrating IP-based Micro-Mobility in Future Cellular Multimedia Networks

1. Integrating IP-based Micro-Mobility in Future Cellular Multimedia Networks Youssef Khouaja, Emmanuel Coelho-Alves, Philippe Bertin

2. Plan • Introduction • The requirements for a new micro-mobility protocol. • Existing proposals. • Common base principles of Cellular IP and HAWAII. • Are the listed requirements being verified and how? • The limits of these new micro-mobility protocols. • The movement detection. • The support of Paging. • The semi-soft handover of Cellular IP. • Conclusion.

3. Introduction • Retain everywhere seamless access to a rich set of information. • A global approach for providing IP-based mobility management over various access technologies. • Mobile IP offers a flexible mechanism but is non optimized to micro-mobility.

4. The requirements for a new micro-mobility protocol (1/5) • Micro-mobility is characterized by frequent and fast movements. • The use of Mobile IP implicates: • An overload of the network core. • An important delay in the diffusion of the new location. • A long interruption of communications. • An enormous loss of packets. • A degradation of the quality of services. • Non-support of real time applications. • The necessity of a new protocol.

5. The requirements for a new micro-mobility protocol (2/5) • This new micro-mobility will have to: • 1 Manage local movements without informing the core network. • Send update messages outside the access network only at the time of inter-network movements. • 2 Decrease the update traffic of the new location. • Not transmit update messages independently of the movement nature. • 3 Limit the diffusion of update messages. • Reduce the diffusion time of the new location.

6. The requirements for a new micro-mobility protocol (3/5) • 4 Minimize the delay in the new location update. • Not emit a classical Mobile IP registration request at each movement. • 5 Eliminate the packet losses during handovers. • Minimize the delay in the update diffusion. • 6 Provide superior QoS and support real time services. • Maintain QoS during all communication. • Facilitate the support of real time applications.

7. The requirements for a new micro-mobility protocol (4/5) • 7 Define an optimal use of radio resources. • Reduction of update traffic initiated by the mobile node. • Minimize the battery power consumption by the introduction of a stand-by mode as defined in GPRS. • 8 Support paging. • Capability to inform the mobile nodes in stand-by mode that they'll be receiving packets. • 9 Interact with Mobile IP. • Define a good interaction between the two protocols.

8. The requirements for a new micro-mobility protocol (5/5) • 10 Be independent of the radio technology. • Support the maximum of public and private mobile/wireless networks. • 11 Insure the robustness. • The tolerance to failures. • 12 Be scalable. • Not complicate the mobility management mechanism.

9. Existing proposals (1/11) • Cellular IP from the University of Columbia and HAWAII from Lucent Bell Labs. • These two protocols are based on the same base principles: • Network architecture on two levels. • The router linking the two levels masks the local movements. • Routing entries mapping the mobile node address to the neighbor from which the update packet arrived. • Entries updated at each movement and refreshed periodically.

10. Existing proposals (2/11) • Distinguish between a mobile node in active mode with a mobile node in stand-by mode. • Stand-by mobile node needs neither to update nor to refresh its routing entries. • Define paging entries able to find the mobile with expired routing entries. • Mobile node passes in active mode thanks to the emission of an update packet of routing entries.

11. Existing proposals (3/11) • Manage local movements without informing the core network. • HAWAII domains and Cellular IP networks. • HAWAII domain root router and gateway Cellular IP. • In a visited Cellular IP network. • Mobile addressed by home address. • The gateway play the foreign agent role. • Packets decapsulated by the gateway Cellular IP. • In a visited HAWAII domain. • Mobile addressed by a dynamically configured address. • Packets decapsulated by the mobile node.

12. Existing proposals (4/11) • Decrease the update traffic of the new location. • Others kinds of messages necessary to update and refresh routing entries. • In Cellular IP, • Route-update packets from mobile to the gateway. • Refreshment by uplink data packets and route-update packets. • In HAWAII, • Path setup power-up messages from base station to domain root router. • Path setup update messages from new base station to old one. • Refreshment by specific messages. • The new traffic remains lower than that of Mobile IP.

13. Existing proposals (5/11) • Limit the diffusion of update messages. • At the Cellular IP gateway or the HAWAII cross-over router. • Minimize the delay in the new location update. • Cross-over router. • Eliminate the packet losses during handovers. • The semi-soft Cellular IP handover. • Two update schemes in HAWAII.

14. DRR Gateway Cellular IP A A R0 R0 B C B C A A A A R1 R2 R1 R2 C B C B C B C B A A A A BS1 BS2 BS1 BS2 B B B B M M Existing proposals (6/11) HAWAII o. @ mobile  B n. @ mobile  C o. @ mobile  C n. @ mobile  A n. @ mobile  B o. @ mobile  B n. @ mobile  A n. @ mobile  B Cellular IP o,n. @ mobile  R0 o. @ mobile  B n. @ mobile  C o. @ mobile  C n. @ mobile  B o. @ mobile  B n. @ mobile  B

15. Existing proposals (7/11) • Provide superior QoS and support real time services. • Reduction of packet losses. • Cellular IP foresees delaying packets at the cross-over router. • HAWAII avoids starting again all resources reservation after a handover. • Define an optimal use of radio resources. • The diminution of signal traffic. • Cellular IP defines active mode and idle mode. • HAWAII defines active mode, stand-by mode and null mode. • The entries are updated and refreshed only in active mode.

16. Existing proposals (8/11) • Support paging. • In Cellular IP, • Paging entries implanted in paging caches. • Caches placement defines paging areas. • Route-update and paging-update packets update paging entries. • Packets destined to an idle mobile node are routed by the paging entries. • Diffused towards all neighbors at the nodes without paging caches. • Broadcasted in the paging area. • Update of routing entries.

17. Existing proposals (9/11) • In HAWAII, • Paging entries located in routing tables. • Dynamically created on the path between mobile and domain root router. • Entries updated by path setup messages and paging update messages. • Paging areas defined by Multicast groups addresses. • Packets destined to a stand-by mobile node are buffered. • A paging request is initiated. • Multicasted in paging area. • Response to paging initiator. HAWAII

18. Existing proposals (10/11) • Interact with Mobile IP. • In HAWAII, • Mobile sends registration request. • Base station verifies movement nature. • Generates a path setup message or relays the request to the home agent. • Receives acknowledgement. • Initiates registration response. • In Cellular IP, • Mobile sends a router-update or paging-update packet towards the gateway. • Gateway realizes admission control and transmits response. • Mobile emits registration request.

19. Existing proposals (11/11) • Be independent of the radio technology. • Act mainly at layer 3. • Some assumptions can question the independence. • Paging support on level 2 in HAWAII. • Quick switch between two base stations in Cellular IP. • Insure the robustness. • Home agent break down. • Periodic refreshment. • Be scalable. • The new micro-mobility traffic remains less important.

20. The limits (1/5) • The movement detection. • Accomplished at layer 3. • With Base stations agent advertisement. • Easy to integrate and independent of the radio technology. • A considerable delay between the real change and the detection. • A loss of packets and non support of real time services. • Made in layer 2. • With the most powerful signal. • Fast. • Manage some parameters and thresholds of change. • Network management? .

21. The limits (2/5) • The paging support. • In HAWAII. • A paging request. • All routers can initiate paging. • Entries dynamically created. • Layer two of the radio link supports the paging. • Not verified in IEEE802.11 and HIPERLAN. • Routers must support multicasting.

22. The limits (3/5) • In Cellular IP. • No paging packet. • Without paging caches  Broadcast to all neighbors. • Placement of paging caches is very complex. • Compromise with the number of nodes where they are placed. • Data packets overload the access network.

23. The limits (4/5) • The Cellular IP semi-soft handover. • During update procedure, packets are delivered to the old base station and they are lost. • During a semi-soft handover: • Mobile executes a fast handover. • Sends route-update-packet. • Returns to listen old base. • At the cross-over router, a new entry is added without erasing the old one. • Packets will be delivered to the two stations. • Mobile makes the real move with minimal packet loss. • Sends a route-update packet with S bit cleared.

24. The limits (5/5) • Radio handover isn't instantly made. • Cellular IP introduced the possibility to delay data packets and to buffer them at the cross-over router. • Quickly switch between two base stations. • Handover time faster than the route-update diffusion.

25. Conclusion • Cellular IP and HAWAII offer very interesting solutions. • Some mechanisms remain dependent of radio technology. • Better define the movement detection procedure and the handover execution decision. Acknowledgements: This work has been performed in the framework of the IST project IST-1999-10050 BRAIN, which is partly funded by the European Union. The authors would like to acknowledge the contributions of their colleagues, although the views expressed are those of the authors and do not necessarily represent the project.