An IPv6 Distributed Client Mobility Management approach using - - PowerPoint PPT Presentation

an ipv6 distributed client mobility management approach
SMART_READER_LITE
LIVE PREVIEW

An IPv6 Distributed Client Mobility Management approach using - - PowerPoint PPT Presentation

An IPv6 Distributed Client Mobility Management approach using existing mechanisms draft-bernardos-mext-dmm-cmip-00 Carlos J. Bernardos Universidad Carlos III de Madrid Antonio de la Oliva Universidad Carlos III de Madrid Fabio Giust


slide-1
SLIDE 1

80th IETF, Prague draft-bernardos-mext-dmm-cmip-00

An IPv6 Distributed Client Mobility Management approach using existing mechanisms draft-bernardos-mext-dmm-cmip-00

Carlos J. Bernardos – Universidad Carlos III de Madrid Antonio de la Oliva – Universidad Carlos III de Madrid Fabio Giust – Institute IMDEA Networks & Universidad Carlos III de Madrid Prague, MEXT WG, 2011-04-01

MEXT WG, 2011-04-01

slide-2
SLIDE 2

Motivation

  • Current IP mobility approaches rely on a central

anchor point (either HA or LMA)

  • Issues:
  • Sub-optimal routing
  • Reliability
  • Scalability
  • Lack of granularity (mobility is offered on a per-

mobile basis)

  • Signaling overhead

80th IETF, Prague draft-bernardos-mext-dmm-cmip-00 MEXT WG, 2011-04-01

slide-3
SLIDE 3
  • FAMA. Basic principles (I)
  • Flat Access and Mobility Architecture (FAMA)[1]
  • DMM approach for Client MIP, using existing

approaches

  • Mobile IPv6 : RFC 3775
  • Authorizing MIPv6 BU with CGAs: draft-laganier-mext-cga
  • The HA is moved to the edge
  • Distributed Anchor Router (DAR)
  • Deployed in the MN’s default gateway (first hop router)
  • Each time an MN attaches to a DAR, it gets a

topologically valid address

[1] ¡F. ¡Giust, ¡A. ¡de ¡la ¡Oliva, ¡C. ¡J. ¡Bernardos, ¡“Flat ¡Access ¡and ¡Mobility ¡Architecture: ¡an ¡IPv6 ¡Distributed ¡Client ¡

¡ ¡ ¡ ¡ ¡Mobility ¡Management ¡solu=on”, ¡accepted ¡in ¡Mobiworld ¡2011, ¡co-­‑located ¡with ¡IEEE ¡INFOCOM ¡2011

slide-4
SLIDE 4
  • FAMA. Basic principles (II)
  • While attached to a particular DAR, the MN can

send/receive traffic using the address from that DAR

  • Every time the MN moves, it obtains a new address
  • The MN can preserve the reachability of IPv6 addresses
  • btained at previous DARs, by sending a BU to the DARs
  • How this dynamic decision is taken is out-of-scope of the draft (for

example, it can be done on an application-basis)

  • DARs play the role of the HA for those addresses

that the MN want to keep reachability,

  • and only for the period of time decided by the MN
  • MNs simultaneously handle several IPv6 addresses
  • Each of them anchored at a different DAR

80th IETF, Prague draft-bernardos-mext-dmm-cmip-00 MEXT WG, 2011-04-01

slide-5
SLIDE 5
  • FAMA. Basic principles (III)

80th IETF, Prague draft-bernardos-mext-dmm-cmip-00 MEXT WG, 2011-04-01

slide-6
SLIDE 6
  • FAMA. MBIP BU auth with CGAs
  • With a DMM approach like FAMA, many IPsec

SAs would be required to follow RFC4877 security

  • We adopt the use of CGAs to provide

authentication between the DAR and the MNs

  • As introduced in draft-laganier-mext-cga

80th IETF, Prague draft-bernardos-mext-dmm-cmip-00 MEXT WG, 2011-04-01

slide-7
SLIDE 7
  • FAMA. Signaling

80th IETF, Prague draft-bernardos-mext-dmm-cmip-00 MEXT WG, 2011-04-01

MN DAR

CGA config PHKT caching BU + CGA param + signature MN auth

(first handoff)

BA + PHKT

MN DAR

BU(PHKT auth) MN auth

(subsequent signaling)

BA PHKT refresh, next handoffs, de-reg