You are here

Driver of Change: Networks, Standards and Interoperability

The global, general purpose, and interoperable nature of the Internet, which is fuelled by open and voluntary standards, has been at the core of the Internet’s success to date.   However a number of participants in our work have expressed concern that the future of key internetworking principles and open standards is increasingly uncertain. In particular, the underlying architectural principles of the Internet may be threatened by choices that lead to technical fragmentation of the Internet.

At one extreme, some are concerned that the Internet of the future may look and behave in a fundamentally different manner than we’ve come to expect. As one participant commented, “the nature of the end-to-end, global Internet is fading away and may fundamentally shift into something unrecognizable.” Others wonder whether the Internet of today will be fit for future applications and challenges of tomorrow’s applications. “Will the “best effort” Internet be able to keep up with technical innovations,” a participant asked, or will we see “the growth of networks and systems aimed at greater control and predictability of Internet services?”

At the same time, however, others note that there have been similar concerns about the Internet in the past, including that it would not keep up with explosive growth and new innovations. Yet the Internet has always been able to adapt and keep up with new technical demands, changes in use, and challenges to its underlying principles.

The extent to which intelligence and computing power will be concentrated in the future into the core of network operations rather than at the edges of the network has also been raised. One commenter noted, for example, that the Internet is “moving away from dumb at the core and smart at the end.” Yet others see a potential resurgence of the end-to-end model, as applications such as the Internet of Things increasingly require low latency. 

Furthermore, the uneven adoption of key standards, particularly the IPv6 protocol, may lead to “island networks” in the future that are unable to communicate with others over the global Internet. Questions were also raised regarding the future support for open standards and their development, with some seeing the possibility that fewer entities will invest in them in the future. Combined with the potential growth of proprietary Internet of Things implementations, eroding incentives for open standard use and development could fundamentally change the nature of Internet in the future.

The role of regulation will also impact the evolution of architecture and standardization, with some concerned that governments in the future may be tempted to mandate Internet technology for political and economic reasons.

Key Challenges and Uncertainties  Raised by our Community 

•    Potential movement away from and lack of support for open standards and interoperability, increasing walled gardens.
•    Centralization of Internet intelligence and operational functions towards the middle of the Internet, rather than at the edges.
•    Government mandates of protocols and standards, including those that conflict with open Internet standards.
•    New uses, applications, and challenges to the Internet growing more quickly than standards processes can adapt.
•    Possible limits to the design and functionality of the Internet Protocol / ability of the "public Internet" to keep up with current and future technology demands.
•    Slow adoption of current standards (such as IPv6) leading to technical fragmentation and islands.
•    Potential impact of the Internet of Things on future of Internet requirements, architecture, and design.

Questions from our Community:

  • To what extent will traffic traverse private networks and rely on the end-to-end "global" Internet only for select services?
  • To what extent will Internet and technology companies embrace the use of open standards versus closed, proprietary standards?
  • Is intelligence moving more towards the centre of the network rather than being distributed at the edges?  If so, what will the consequences be on the Internet experience of the future?