IoT - The end of the internet

William Herrin bill at
Thu Aug 11 05:38:18 UTC 2022

On Wed, Aug 10, 2022 at 3:29 PM Christopher Wolff
<chris at> wrote:
> Use Case 1:  Augmented Reality/Virtual Reality.  It is stated that round trip latency must be <4ms with 100mbit full duplex at the cell edge to prevent nausea and dizziness while wearing goggles for a long term.

Hi Christopher,

Not really IOT. Call it Netflix Part Two. Look for the discussions
around Netflix's impact on the Internet. I recall some folks
calculating what it would take for every household to stream their
television via the Internet and fretting over it. Needless to say,
folks are streaming their TV and the Internet hasn't collapsed.

> Use Case 2:  A little closer to “IoT”. An autonomous vehicle under remote control requires 100 feet to stop with LTE vs 20 feet with 5G.
> Use Case 3:  A Lidar near-miss sensor at an intersection requires 1ms from the traffic operations center.

These sound like absolutely terrible designs. IoT does NOT imply that
all the compute is located remotely or that local comms can or should
be relayed through a central system. So your near-miss sensor sends
its packets out local radio, cryptographically authenticated with a
key the vehicle collected from central while it was still two blocks
away. No latency difference.

Think of these devices like the Mars Rovers. The Mars Rovers aren't
operated by a dude with a screen and a joystick. They receive
directions from central but follow them autonomously. If in the course
of following the directions they exceed any of dozens of safety
parameters, they stop and wait for new instructions.

> Will tomorrow’s applications require a re-thinking of “The Internet” and protocols that are low latency compliant?

No, because speed of light constraints will continue to cause us to
implement the latency-critical components close to the user. It's
basic physics man.

Bill Herrin

For hire.

More information about the NANOG mailing list