IEEE Consumer Electronics Magazine - March/April 2023 - 6

Security
efforts. " If the network does not know the intent
of the packet, it cannot provide services such as
reliable and timely delivery. Retransmission, for
example, requires arbitrary buffering in the network,
which frustrates streaming.
More to the point, it cannot know which bits
are good bits and which bits are bad bits-the
network equivalent of Maxwell's Demon. If it
claimed to know, it would be second-guessing
the intent of the applications causing perverse
failures and complexity.
Only the applications know enough to manage
their trust relationships and " favors " inside
the network are a problem because they do not
have the necessary knowledge and become a
source of difficult-to-diagnose failures.
There is not a universal answer to trust, but
we can provide better tools. One approach I
wrote about is to create communities of thingsc
sharing a secret. That does not build in the limits
of physical topologies, and devices can be part
of multiple communities. It is commonly done
within manufacturers' products, but we need
such approaches as a successor to managing
keys.
To get a sense of the complexity, consider
the problem of trusted devices being in error
even with the best intentions. There is also the
problem of bitrot-when the software and
assumptions remain valid, but the conditions
around it change. You might trust " childrens.
candy, " but when the DNS entry for that name
expires, it might be repurposed by a bad actor.
That is a reason for the relationships to be
defined using abstract identifiers and to not
depend on building semantics into the plumbing-a
topic I plan to write about in a future
column.
Things get even more complicated as the protocols
are woven into the fabric of the physical
world. When you have smart devices in the
home, how do you implement policies without
requiring people to log in to each light switch
and without having to be explicit about ambiguous
social policies? Do your children have the
ability to turn on your bedroom light at 3 AM?
Those whose kids tell Alexa to play loud, annoying
music as they leave the house have experienced
the problem.
LEARNING
As in the early days of Windows, many devices
assume they are behind a security perimeter.
As imperfect as the approach is, it has given
us breathing room to learn better approaches.
Unfortunately, many have used it as an excuse
to declare victory rather than trying to take
advantage of the opportunity for learning.
Just as we learn to safely navigate the physical
world, we need to develop new best practices for a
connected world. That will not happen without
tools for managing communities of devices. Even
then, we will get failures. But we can limit the harm
if we are wary about trusting other devices and do
not naively rely on firewalls as security boundaries.
Wetend to focus on the use cases that work with
our current approaches. This complacency too
often frustrates innovation and limits our future.
Bob Frankston is best known for writing VisiCalc-the
first electronic spreadsheet. While at
Microsoft, he was instrumental in enabling home
networking. Today, he is addressing the issues
associated with coming to terms with a world
being transformed by software. Contact him at
ieeecemagazine@bob.ma.
c
[Online]. Available: https://rmf.vc/IEEECommunitiesOfThings
6
IEEE Consumer Electronics Magazine
https://rmf.vc/IEEECommunitiesOfThings
IEEE Consumer Electronics Magazine - March/April 2023

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - March/April 2023
