You’re reading it here first: The Indian government is holding a limited non-public consultation on a Chinese-led effort to revamp the internet’s architecture. New Internet Protocol, or IP, is a proposal mainly pushed by Chinese tech giant Huawei that seeks to fundamentally alter the standards on which the internet has been based for decades. MediaNama received a copy of the consultation, put out as a white paper by the Department of Telecommunications, from multiple sources. The DoT gave Indian industry associations a mere five days to respond.
Inputs from this private consultation will be used to formulate India’s response to Huawei’s New IP standard at the international level, the DoT said.
Internet Protocol is a term used to describe standards which devices on the internet rely on to communicate with each other. Huawei’s proposal at the International Telecommunication Union seeks to make major changes to internet standards, raising concerns (from nonprofits like ISOC and Mozilla) around the centralisation of networks and government control of the internet, as well as around whether such radical changes are needed in the first place.
The proposal’s future will be clearer at the World Telecommunication Standardization Assembly to be held in Hyderabad next February.
What the DoT asks in its white paper
On the Huawei proposal, the DoT has raised the following issues for consultation:
- what the shortcomings of IPv4 and IPv6 are;
- how these existing standards support the goals that New IP sets out to achieve;
- what standardisation bodies have developed so far for concerns around supporting heterogeneous networks, deterministic forwarding (which enables data to be sent with less delay and low data loss), intrinsic security, ultra-high throughput (i.e., a large amount of data transmission), and “user-defined customized request for network services”, which New IP is built around;
- whether autonomous subnetworks — which New IP proponents argue are something that their proposal will boost the efficiency of — are something the internet in its current form is incapable of handling;
- what it might cost to migrate to a new architecture (like New IP);
- for Network 2030, an ITU working group under which New IP was proposed, whether existing technology can be scaled, or if a new technology like New IP is required;
- how interoperability issues can be solved in the present regime;
- how flexible address spaces (New IP suggests that smaller devices in local networks should be able to do with less metadata to conserve power and computing resources), as proposed in New IP, can address interoperability challenges;
- “How guaranteed delivery of information over a network within certain parameters may be ensured with current IP regime”;
- how the current IP architecture can deal with issues surrounding “data traversing distance”.
“It is understood that that the proposal is still under naive stage,” the white paper concluded. “However, GoI [the Government of India] wishes to timely gear up [sic] to respond.”
BIF pushes for more time to discuss, questions need for New IP
In a submission filed by the Broadband India Forum that MediaNama obtained from a source, the industry body pushed for the involvement of a “larger group of stakeholders which includes industry, tech and research institutions, academia, civil societies, etc” for this consultation. BIF also added that more time was needed for this consultation. We have reached out to BIF to verify this filing and comment on it, and will update this post if we receive a response.
BIF argued that New IP as proposed by Huawei “effectively goes against […] openness and multi-stakeholder participation in Internet governance”. It also questioned the need for New IP, making the following arguments:
- IP has worked just fine: BIF argues that IP has been able to deal with several changes in the internet over the years, including massive increases in data volumes and number of devices. It argues that the challenges Huawei mentions only apply to “intranets” and not the global internet. “The IP layer itself does not introduce the constraints mentioned (e.g. inability of deterministic forwarding or lack of trust and security) – these capabilities can be built at the upper layers of the architectural stack, providing adequate solutions to specific requirements,” BIF said. The industry body said new types of networks like Content Delivery Networks (CDNs) were able to emerge because of this versatility.
- The internet has always supported heterogeneous networks: The internet and its underlying architecture “have demonstrated to be flexible enough to support many different applications,” BIF argued (emphasis theirs). For applications that need specialised requirements, BIF argued, like factory arms, being connected to the internet is not advisable in the first place and different technologies should be used instead. BIF said that work going on in organisations like the Institute of Electrical and Electronics Engineers and the Internet Engineering Task Force to address emerging issues were sufficient to deal with the challenges of future networks.
“Part of the flexibility [of the internet] also arises from the “bottom up” approach followed so far, which reflects multi-stakeholder collaboration and collaborative governance practices. In contrast, New IP establishes a “top down” system completely controlled by a single chokepoint. This could, in the long run, end up compromising on the flexibility of the system in addition to risks of partisan interests dominating regulation of the internet. It also gives rise to significant concerns on exercising mala fide powers to exercise a switch-off function, discriminate between users, and enable surveillance.” (emphasis ours)
- Costs will be high: BIF said that New IP would cost a lot, as installing new network equipment at several levels may be required. It also said that backward compatibility generally harms quality of service, as “network protocol interworking through gateway functions” can add complexity to the system. It also cited Indian service providers’ investments in IPv6, and argued that such investments would be “jeopardiz[ed]” by New IP.
Regulatory issues: “We have to carefully review any proposal which changes the nature of the network to assess its compatibility with regulatory aspects of net neutrality, law enforcement access etc.” (emphasis ours)
- New address spaces may be self-defeating: New IP suggests variable address spaces — which are metadata that accompany a piece of data — should be variable, so that devices in smaller networks can save power. “We do not agree with [the] premise” that existing IP standards are incapable of dealing with these problems, BIF said, and citing an IETF statement, added that changing address space specifications like this might add complexity that achieves the opposite of what New IP sets out to do: higher latency and worse quality of service.
- It’s impossible to solve distance-related constraints: In response to a question on how existing systems could deal with the issue of “data traversing distance”, BIF had this response:
Physical limitations such as speed of light and limits of spectral capacity are limits exposed by physics, they cannot be solved by engineering efforts – for any system. This sets some natural limits on communication. For instance, in glass fibre, light will take 1 ms to traverse approximately 200km. Making sub ms responses (two way) becomes impossible around 100km distance. (emphasis theirs)
Read BIF’s full filing here.
Who was asked to comment
In addition to some of its own offices, the DoT requested inputs from the National Security Council, the Ministry of Electronics and Information Technology, the Ministry of Home Affairs, the Telecom Engineering Centre, the Telecommunications Standards Development Society of India. The DoT official who signed off on the letter did not respond to a request by MediaNama to confirm the letter. We have filed an RTI with the DoT for a copy of all comments received in the consultation, as this consultation and responses to it should be in the public domain.
The DoT also asked for inputs from the following industry associations: the Cellular Operators Association of India, the Internet Service Providers Association of India, the GSMA, the Telecom Equipment Manufacturers Association of India, the Telecom Equipment & Services Export Promotion Council, Broadband India Forum, the Association of Competitive Telecom Operators, and the Internet and Mobile Association of India. None of the associations returned our request for comment on their response to the white paper, except for GSMA, which declined to comment.