What is the future of WiFi (from a network security standpoint)
cloudi-fi.com
cloudi-fi.com
I've now taken my hindsight glasses off and recognize that it's hard to imagine that assholes and criminals would so easily ruin it for the rest of us. 50+ years later and ~30 years later of the interwebs and now it's an "but of course they would". The sad thing is that there should no longer be a blank sheet of paper for a startup, but every "new" sheet should already have defensive strategies on it. It's just not sexy, and hard to get to MVP so it's easy to drop/ignore/delay. I'm guilty too, but maybe I'm worse because I'm well aware that I'm doing it????
You can still argue about the endpoint security, but at least the network participants and their messages can be secured, even if an endpoint is compromised.
Put captive portal in place and force the users to validate through their phone numbers, social media accounts or emails. You can even integrate your existing IdP to the captive portal and even ask your employees to connect through the captive portal.
OWE, which it mentions at the end is genuinely useful, reducing the drawbacks open networks had in the past.
Last I knew: Due to the nature of the medium (radio waves), wifi was a 'circuit-switched' network - any two devices communicating monopolized the entire network. That is, unlike a switched wired network, only two devices can communicate at once (usually an endpoint and a base station).
More devices using the network, which of course was usually necessary, required some sort of time-division multiplexing - the network needed to slice up circuit access by time. The only other solutions were multiple networks, which were created with multiple radios on different frequencies, and [edit: beam forming], which created different networks to different physical locations - if that actually worked in practice.
Has anything changed? Are there new solutions to the circuit-switching limitation? Has someone managed to turn wireless into a packet-switched network?
Also, modern wireless systems can even multiplex by location, using phased arrays, it will beam a spotlight of radio to your direction, and can noisy data from receiving phase information as well.
The neat thing about phased arrays is that they are solid state, aka no moving parts, so they have a much faster response.
Here is a really good presentation on how they work and how to build one.
Beamforming is achieved using phased arrays. I'm not sure if all beam forming uses phased arrays, but one of the terms is just talking about the wave itself (beamforming), while phased array is talking about how the beamforming is achieved.
That being said, I don't want to presuppose that you can't beamform form without phased arrays.
Wi-Fi 6 adds OFDMA, which splits channels into sub-channels so multiple clients can transmit simultaneously, and MU-MIMO allows simultaneous streams to different clients. Beamforming helps too with directional signals.
tl;dr; it is a shared medium but with lots of tech on top of it that mitigate much of the issues!
> MU-MIMO allows simultaneous streams to different clients
That's great, but the question is the same: how? How does it solve what I called the circuit-switching problem (i.e., each tranmission utilizing the channel 100%)?
as I understand it, a channel can be effectively divided into K separate sub-channels, and then in a crowded space with N active users, your time-division only needs to contend with N/K other users.
this dovetails well with Wifi 6E expanding to the 6ghz range, and Wifi 7 (among other things) expanding channel bandwidth to as large as 320mhz. (compare this to classic 2.4ghz having only 3 non-overlapping channels of 20mhz width each)
0: https://en.wikipedia.org/wiki/Wi-Fi_6
1: https://en.wikipedia.org/wiki/Orthogonal_frequency-division_...
> starting with 802.11ax ("Wifi 6") [0] they've started doing frequency-division
How does that differ from multiple radios running multiple connections? Is it that there are many more frequencies, closer together, dynamically allocated?
Also, some beam-steering tricks that likely won't be available in the US for another 5 years... =3