Thursday, October 26, 2023

Will 6G Try to Recreate Closed Networks?

Already, some are suggesting 6G will be different from 5G in a significant way: where 5G is still a connectivity mechanism, some tout 6G as a computing mechanism. It will “enable immersive, ubiquitous, and sensory digital experiences on a massive scale,” some argue. 


“Enable,” yes, in the same way that home broadband “enables” use of internet-delivered applications.  “Embed,” in the sense of the network itself being the supplier of the features, probably not, and for good reasons. 


Modern computing is based on layers. That is what allows us to innovate faster and avoid monolithic solutions because functions are compartmentalized; independent objects rather than integrated processes. 


Even if infrastructure suppliers want us to accept new forms of functions integration as a way of convincing us to buy their new platforms, we should resist the notion. We actually do want permissionless app creation, not “integrated” solutions. 


In other words, some are likely to argue for 6G standards that are more centralized and controlled than 5G networks. More “closed,” in other words. 


Some will argue this is necessary because 6G will need to support a wider range of complex and demanding applications, such as immersive virtual reality and real-time AI-powered services. 


We might want to resist that notion. It is a move in the direction of walled gardens, closed networks and app development controlled, to a larger extent, by the entities providing internet access. 


Is that going to be better? 


Some will argue for advantages such as enhanced security or privacy. But permissionless development enabled by the layered architecture  has worked well. It’s easy to see why some in the value chain would prefer more closed, centralized networks. 


It recreates the experience of the public switched telephone network, where telcos controlled “all” the apps running on the network. 


To use the network, you needed permission from the network operator. Is that going to be better?


Every mobile generation gets hyped. Each will, it is said, enable and revolutionize the experience. Improvements happen, yes. Latency is reduced; bandwidth is increased; energy efficiency gets better. 


6G should “enable” immersive experiences such as the metaverse by staying out of the way. Embedding such features into the fabric of the network--beyond measures to control latency and supply lots of bandwidth--will be a mistake.


No comments:

Have LLMs Hit an Improvement Wall, or Not?

Some might argue it is way too early to worry about a slowdown in large language model performance improvement rates . But some already voic...