An article on The Register discusses why an alternative Internet architecture called SCION has seen some adoption in specific economies (like Switzerland) and industries (like banking and healthcare), without much further growth. We’ve discussed the architecture on this blog and APNIC’s PING podcast several times.
SCION manages bandwidth expectations using contracts between communicating parties, with a coordination model that imposes Public Key Infrastructure (PKI) -based certificates and price adjustments to ensure equity among participants.
There are contexts where it’s important to guarantee bandwidth and latency beyond standard Internet performance. Examples include remote surgery, where high-bandwidth, real-time video is needed to control robotic instruments, and financial trading systems, which require precise, predictable network delays to ensure fair execution of buy and sell orders.
The Register article goes beyond bandwidth management concerns to focus on SCION’s ‘security-centric’ model that allows communicating entities to establish routing trust among themselves.
Border Gateway Protocol (BGP) is an example of scaling a simple idea into a massive, world-scale system, where security has been added opportunistically as an afterthought. There are limits to what BGP security can prove and secure.
SCION approaches the trust problem differently, through a structured hierarchy of certification and certificate management, resulting in a different basis for what people can say about their networks, and what others can believe.
Beginning as a research project, SCION has strong roots in ETH Zurich and backing from the Swiss financial sector. The project has received newfound interest as concerns around ‘digital sovereignty’ have become more prominent.
While the Internet’s routing architecture is open, it’s seen by many as anchored in the interests of a particular state, reflecting a specific view of the role of the free market in networks and services that may not be universal.
SCION is perceived as being better aligned with European interests and could be suited to the pursuit of network certainty that can be found in the Asia Pacific region. While financial services are an especially strong use case for a deterministic, managed, structured network, they are perhaps not as central to the Internet in the Asia Pacific region as they are to an economy like Switzerland.
That said, the existence of an alternative routing system that governments can include in strategic planning for their own resilience is a good thing.
Currently, SCION looks a lot like an overlay network that depends on BGP as a routing fabric. To achieve adoption to the level of adoption that BGP has, SCION will have to be implemented over physical and link assets that people can use without concerns for BGP stability underneath.
While SCION has proponents and opponents, in the marketplace, it has to overcome an entrenched incumbent in the form of BGP, and decision-making driven by carrier costs, not national interest.
As long-haul traffic Internet traffic is displaced by private links between Content Delivery Networks (CDNs) that put services closer to users, there may be space for changes to how non-CDN services handle end-to-end traffic flows.
In that context, SCION is a contender with serious traction.
The views expressed by the authors of this blog are their own and do not necessarily reflect the views of APNIC. Please note a Code of Conduct applies to this blog.