Download presentation
Presentation is loading. Please wait.
1
IPng: Paths not taken worth revisiting
Jon Crowcroft,
2
IPng is > 20 years old, why now
IPv6 is sort of out there But not really addressing core challenges But the IPng requirements were broader More recent Future Internet programs havn’t gained a lot of traction Perhaps because smart money was elsewhere Cloud, data center, apps, smart phones etc
3
IPng alternatives nack then….
SIP v. PIP Simple & FTIFs Nimrod (& BGP) Maps v. Path Vector Catnip Pluralisms (viz Plutarch)
4
Circumstances not what they were
Scale – yes, roughly as predicted Mobile/wireless – no No real ip-level seamless handover Cloud, virtualisation Not at all predicted Computation, Storage Somewhat under-estimated
5
IPv6 v. Loose Source Route/soft circuit
One of the age old debates Brittleness v. predictability Maybe with slices/vpn, can re-visit But then what for? Route cache/fwd table lookup opt? Maybe when 100% of planet covered Three times over Can start to see long term stability? Including diurnal/timezone variance etc
6
Maps v. Link State or Path Vector
Nimrod (+to some extent PNNI routing) Based on exchange of maps Rather than vectors or LSAs Maybe people were afraid of scale Or revelation of internal details To competitors or attackers Especially if combined with source route But can mitigate with accountable IP And maybe other tricks….see later
7
Multi-protocol diversity
Virtualised switches Many functions Fast instantiation Per flow or customer etc Partitioned… Programmable NICs Offload v. bypass (extreme case rdma fpga brainwave )
8
Low Latency and Resilience
Need cooperation between Aps At spectrum, wireless, coding, packet level But we’re competing in these levels too “Coopetition” Looks a bit like problem BGP solves But fine grain, faster And BGP doesn’t converge or operate fast enough or describe right resources Exchange (encrypted) resource maps?
9
Exchange Encrypted Maps
And cooperating nodes have keys Non-coop nodes use MPC to compute paths MPC run on coop nodes or use homomorphic crypto perhaps Paths could be braided for resilience multipath Braiding can be generalised Resource pool spectrum, access lines, cell towers backhaul, etc
10
Many apps don’t need paths
Generalise virtual path to virtual tree Hose model+braiding Virtual braided tree from net>customer For media/cast Virtual braided tree from customer>net For IoT, collector nets
11
Virtualised switches&NICs
Can run resource pool algorithms Over heterogeneous links, virtual paths i.e. recursively over layers too (c.f. John Day) Braiding roughly like bundles of pathlets Virtual path/braided treelets
12
Re-combination of Ipng ideas
From Nimrod & Catnip updated to use virtualisation & crypto To get isolation and cooperation
13
How can we build this from now?
First steps design performant crypto Needed anyhow for commercial reasons Get buy-in from others requires openness By definition, this doesn’t work unless inclusive Advantage for Huawei? Leadership! First mover etc
Similar presentations
© 2025 SlidePlayer.com Inc.
All rights reserved.