Second of a multipart series.
China’s high-technology communications and networking industries are proposing a host of future capabilities to come if vendors cast their lot with companies such as Huawei and ZTE. But these new technologies, once ensconced, would lead their users down a path closed to others and open to Chinese government control, say Internet experts.
This strategy is accompanied by attempts to influence international standards organizations, both local and global as they shape the evolution of Internet protocol (IP). At the heart of this effort is China’s “New IP.” The technologies that would enable it are complex and not interoperable with the existing Internet, and that complexity ultimately could force customers to choose between the open interoperable Internet and a closed version that locks its users into a control regimen over which they have little to no control.
Vint Cerf, vice president and chief Internet evangelist at Google, offers that several objectives seem to be in play. One is increased influence in the technology used by the rest of the world. That provides economic benefits to China if the technology comes from Chinese companies such as ZTE and Huawei.
He continues that China probably is pursuing a multi-pronged strategy. It is working to maximize its presence in the markets and place as much of its own equipment in the global networking realm as possible, which is the goal of every manufacturer, he notes. Simultaneously, China is striving to be more influential in the standards community. These efforts might be more coincidental than coordinated, he offers.
The two main thrusts of the telecommunications takeover are the evolution of 5G and the introduction of the New IP. Cerf’s biggest issue with China’s 5G approach is the architecture of the 5G control plane. It offers a wide range of functionality, which would be attractive to providers who want to charge differentially for various types of service. The problem with this approach is that it bleeds control away from the Internet layer of protocol and instills it in the hands of the 5G service providers.
“You can’t even get access to those functionalities of 5G without having physical access to that control plane,” he points out. “So, it offers more control of service to the communications and telecoms service providers while at the same time leeching control away from the Internet layer.”
An even greater problem applies to the New IP, which promises capabilities such as holographic imagery and virtual reality. Achieving these will require sacrificing independence in favor of a system that controls its bandwidth allotments.
Ted Hardie, Internet boffin at Google, warns that the technological makeup of the New IP will force a split in the Internet that he refers to as a fork. This fork will establish conditions that will create two classes of haves and have-nots: the free and the controlled.
Hardie notes that, a few years back, China proposed the Next Generation Protocol to the European Telecommunications Standards Institute (ETSI). It was led in part by groups from the European operator community accompanied by Huawei officials. These diverse organizations never were able to agree on the path forward, Hardie relates, so they closed that working group and let ETSI go forward with a related proposal from New IP that takes a circuit-switched approach for tight latency characteristics.
“They have established a set of use cases that are relatively small, and their aim is to get those use cases handled by this New IP approach, and then see what other use cases might be handled,” Hardie says. This opens the door to follow-on niche approaches related to the original ones.
But another approach concerns Hardie more. To accommodate advances such as a haptic communication feedback system, including holographic communications, low-earth-orbit satellite links and virtual reality, the network would need to be rebuilt. Uniting all these capabilities would be resource reservation, in which system capacity and low latency are always available for these high-demand applications.
Doing resource reservation will prevent the use of differential services, Hardie points out. Services would be designated for specific approaches and would receive specific treatment by the network. This new Internet would lose some of the multiplexing capabilities inherent in the current Internet. Users would not be able interact smoothly if they use different means of Internet access, such as fiber, WiFi or Bluetooth. Also, some of the reserved capacity would be idle at certain times.
Above all, this configuration records a contract. The format described in offerings displays a new set of IP headers, a contract between the network and the packet sender, and the user payload. The network and the individual requesting the traffic have a new relationship unlike that existing today. “The idea that you could go into a coffee shop and browse, and the network wouldn’t know who you were, kind of goes away,” Hardie states. “If you go into a coffee shop and set up a contract with a network, then of course at that point they would know who’s the signatory to each side of the contract. That is a really very fundamental change in the social dynamics of how a network works.”
Similarly, someone could be deemed socially ineligible for a network contract. That person would have no ability to get on the network at all. “We’ve reached about half the world in terms of IP networking, and we really want to reach the other half,” Hardie says. “Creating systems that make it much more difficult to do that is problematic for reaching that goal.
“What they’re building is an Ultranet,” Hardie charges. “The result of that could be that we lose interoperability between the parts of the network that follow along this [current] path, and those that follow along the evolution that is sort of managed by the IRTF [Internet Research Task Force] and processes like those.
“Losing interoperability is losing the Internet itself,” he declares.
Planners in the ITU-Standards (IUT-T) group must carefully consider the consequences of any splintering of the Internet, Hardie says. “The ITU-T must think carefully about what the interoperability requirements are on any work that it takes on in this space.”
Avoiding this fork will require efforts at all levels. Hardie says that people need to insist on interoperability with their vendors, who must be made to understand what is at stake with a single Internet. This message is even more important in the developing world, which may not realize the potential damage to its economic growth if the Internet forks, he emphasizes.
National security arguments are likely to fall on deaf ears, Cerf offers. The parts of the world that may give Huawei a significant part of their market don’t care about concerns expressed by the United States. Instead, the way to preempt Huawei’s market domination may be competition.
That approach is not without precedence, Cerf notes. Years ago, when a large U.S. high-technology company discovered that foreign competition was selling a product similar to its own at a retail price equal to the U.S. company’s manufacturing cost, the U.S. firm met the challenge head-on. It re-engineered its processes and reduced internal costs to become competitive again in the global marketplace. “It may very well be that the incentive of the U.S. manufacturers is, in fact, driven hard by driving costs down,” he says. Hardie suggests that customer insistence on interoperability can keep costs down and improve competition.
In Part III: The driving force behind China’s telecommunications efforts.