TDEB Top Bar

Trending
Semiconductor Industry News — The Digital Electronics Blog
Follow Us

TDEB Masthead

The Semiconductor Hub

Semiconductors.  Chips and Systems.  Innovation.

Indexed by leading Search Engines, AI and referenced by Engineers worldwide 500+ Articles 10M+ Pageviews 30+ Reports 50+ Citations

RIPE NCC Opposes China and Huawei's New IP Internet Plan

Murugavel Ganesan
by
0
Share Facebook X LinkedIn WhatsApp Email

News from April 2020. The RIPE NCC, the regional Internet registry that serves Europe, the Middle East and parts of Central Asia, publicly came out against "New IP", a proposal to redesign the Internet's core protocols that was backed by Huawei, Chinese telecom operators and the Chinese government. Its position, set out by Marco Hogewoning, was that today's Internet has technical challenges but does not need a whole new architecture to solve them.

RIPE stands for Réseaux IP Européens, French for "European IP Networks". RIPE itself is an open community forum interested in the technical development of the Internet; the RIPE NCC is the membership organization that supports that community and allocates IP addresses in its region. The statement covered here came from the RIPE NCC.

Key takeaways

  • New IP was presented to the International Telecommunication Union (ITU) in 2019 as a rework of the TCP/IP model for future networks.
  • Critics saw it as moving Internet design from open bodies such as the IETF to a state-led UN agency, with features that could make censorship and tracking easier.
  • In April 2020 the RIPE NCC said publicly that it did not believe New IP was needed.
  • The IETF also rejected the need for a top-down replacement of IP, and by December 2020 two ITU study groups had agreed not to take New IP on as new work.

What New IP proposed

As described in ZDNet's report of April 23, 2020, New IP had three main parts:

  • A revised version of the TCP/IP protocol stack, presented as necessary for future technologies such as holographic communication, very low latency services and large numbers of connected devices.
  • A "shutoff" mechanism that would let parts of the network be cut off when they were judged to be misbehaving.
  • A new "top-to-bottom" governance model that would place much more control in the hands of a small number of key network operators.

Huawei first presented the idea at an ITU meeting in September 2019. It came to wide public attention after a Financial Times report in March 2020. Supporters described it as a technical upgrade. Critics, including privacy advocates, read it as an attempt to build censorship features into the network itself, behind the language of a protocol redesign.

Why it was controversial

The Internet's protocols are developed in open, multistakeholder bodies, mainly the Internet Engineering Task Force (IETF), where anyone can take part and decisions come from technical consensus. The ITU is a United Nations agency where member states vote. Moving core Internet design into the ITU would shift influence from engineers and network operators toward governments.

The same ZDNet report described New IP as the Chinese government's attempt to export its model of a controlled Internet to the rest of the world's infrastructure, and reported that Russia, Iran and Saudi Arabia had expressed support for the proposal.

The RIPE NCC position

In a RIPE Labs article of April 22, 2020 titled "Do We Need a New IP?", Marco Hogewoning, then acting Manager Public Policy and Internet Governance at the RIPE NCC, answered his own question: "I don't think we do." His argument, in summary:

  • The current Internet does face technical challenges, but they can be solved by evolving the existing protocols, as has happened for decades.
  • A whole new architecture is not required to support new applications.
  • Decisions about the Internet's design should stay with the open, consensus-based processes that built it.

He also encouraged readers to talk to their government representatives, since ITU positions are set by member states. The RIPE NCC said it would keep following the topic, and Hogewoning presented "The New IP Proposal" to the community's Cooperation Working Group at RIPE 80 on May 13, 2020, according to the working group minutes.

Why this matters to the hardware industry

Network protocols are implemented in silicon: switch and router ASICs, network interface cards and programmable data planes all assume the IP packet formats and routing behavior defined by the IETF. A split into competing Internet architectures would mean separate chip and equipment ecosystems, much smaller volumes for each, and harder interoperability at borders. That is why the New IP debate mattered to chip and equipment vendors as well as to network operators.

Deep dive: the numbers and details

This part holds the technical arguments in more detail, the IETF's formal response and a dated record of what followed. Skip it if the overview is all you need.

The technical case in the RIPE Labs article

  • Little that is new. The proposal had few technical details, and much of it reused existing work from the IEEE, 3GPP and the IETF, so the novelty lay mainly in how and where it would be deployed.
  • Addressing. New IP's case for a new address space was weak, Hogewoning argued, because it too would start from 128-bit addresses, the same size as IPv6.
  • Latency and capacity. Limits on latency are set largely by physics, and capacity is better handled in IEEE and 3GPP work on the links themselves.
  • Security. Quantum-resistant encryption is a cryptography problem; existing protocols such as DNSSEC and TLS can adopt new algorithms without a new network architecture.
  • Governance. The core concern was that New IP would move control toward central authorities able to block data flows, in conflict with the Internet's open, user-centered model. The RIPE NCC used its ITU membership to file a formal response objecting to the proposal.

The IETF's formal response

On March 30, 2020 the IETF sent the ITU-T's advisory group TSAG a liaison statement, "Response to 'LS on New IP, Shaping Future Network'". Its main points:

  • The IETF saw no evidence that a new, top-down replacement for IP was needed; the IP stack has evolved for decades and extensions can meet the stated goals.
  • Many of the claimed gaps were already addressed: IoT work (for example RFC 8520), satellite transport, and low-latency work such as IntServ, DiffServ, active queue management, DetNet and L4S. It noted that sub-millisecond latency over distance is limited by the speed of light.
  • Separate, disjoint addressing systems would need translation between them, adding fragility, latency and state, while interoperability is essential.
  • Proposals to extend IETF technologies should be brought to the IETF first.

The statement concluded that "the creation of a top-down design effort to replace the existing IP protocol stack wholesale would be harmful."

A quick sense of scale: latency and the speed of light

The physics point in both responses is easy to check. Light in optical fiber travels at roughly two-thirds of its speed in vacuum, about 200,000 km per second, or 200 km per millisecond. A 1 ms one-way latency budget therefore limits the distance to about 200 km before any switching or queuing delay is counted, and a round trip within 1 ms limits it to about 100 km. No protocol redesign changes that arithmetic, which is the point both the RIPE NCC and the IETF made.

What happened next

The Internet Society's New IP FAQ (February 22, 2022) records the later steps:

  • July 2020: at ITU-T Study Group 11 and 13 meetings, the name was changed to "Future Vertical Communication Networks" (FVCN), with the substance unchanged.
  • December 2020: the SG13 and SG11 plenaries decided not to accept New IP related questions as new work items and to stop discussing New IP at least until the World Telecommunication Standardization Assembly in March 2022.
  • 2021: elements of the proposal kept appearing as separate work item proposals in ITU-T study groups, for example on "polymorphic networking" and immersive real-time communications in SG13, and on security for deterministic communication services in SG17.

The wider debate the proposal started, over whether Internet standards should be set by open technical bodies or by states, has continued at later ITU meetings.

Advertise on this space or blog. For more information, contact us at ads@digitalelectronics.co.in.

Murugavel Ganesan
About the Author
Murugavel Ganesan
Product Director | SerDes · IVN · Edge AI · Compute | Ex Intel · Infineon

A semiconductor product leader working across product architecture, definition, strategy, and silicon and software execution, from concept through volume production. He has worked with teams across Asia, Europe and the USA, including at Intel, Infineon and ArrayComm.

His work covers ASICs, SoCs, FPGAs and chiplets for wireless, wireline, automotive, IoT and AI, including PCIe, CXL and DDR interfaces alongside RF and mixed-signal design.

Share Facebook X LinkedIn WhatsApp Email

Post a Comment

0 Comments

Your comments will be moderated before it appears here.

Post a Comment (0)

#buttons=(Ok, Go it!) #days=(20)

Our website uses cookies to enhance your experience. Learn more
Ok, Go it!