All Things 802.11ac - Question 9: What Does 802.11ac Mean for Applications?

  • 1
  • Question
  • Updated 5 years ago
  • Answered
Faster/better/cheaper has historically defined high tech, and WLANs are clearly no exception here. But what do the significant improvements in performance and other capabilities inherent in 802.11ac, that we’ve discussed in detail here today, mean for applications? Sure higher data throughput is always valuable, but are any new applications or application deployment strategies enabled or assisted by 802.11ac? And what impact will .11ac have on time-bounded traffic like voice, streaming video, and multicast?
Photo of Craig Mathias

Craig Mathias

  • 63 Posts
  • 0 Reply Likes

Posted 5 years ago

  • 1
Photo of Matthew Gast

Matthew Gast

  • 284 Posts
  • 63 Reply Likes
The last part of the question is the easiest one. The 802.11 MAC manages airtime utilization. Any time a new technology comes along that requires less airtime, things get better.

Or, more formally, the average latency on a network is related to the average queue depth. (You know this, of course, if you have to drive in rush hour. When traffic backs up on the entrance ramps and beyond, you know the highway isn't moving either.)

This is not to say that 802.11ac solves all the problems of time-bounded traffic, but anything that can be done to decrease queue length and reduce airtime utilization is a plus for time-bounded traffic.
Photo of Craig Mathias

Craig Mathias

  • 63 Posts
  • 0 Reply Likes
It would seem that most new smartphones are incorporating single-stream .11ac. I'm assuming this will be used for voice (I'm already doing with .11n on my iPhone 5, and it works great).
Photo of Matthew Gast

Matthew Gast

  • 284 Posts
  • 63 Reply Likes
I've used UMA for years on T-Mobile, which has been a great boon for me. I can't count the time I've spent using Wi-Fi for calling overseas at no extra charge. (The downside is that enables me to do tech support no matter where in the world I am!)
Photo of Craig Mathias

Craig Mathias

  • 63 Posts
  • 0 Reply Likes
I just got back from the Baltic area where I was testing a Wi-Fi-based mobile unified communications (MUC) solution we put together using free/open source components. Mostly worked great - I'm posting the details in my blog at Network World throughout this month (first entry tomorrow, in fact).
Photo of Craig Mathias

Craig Mathias

  • 63 Posts
  • 0 Reply Likes
One other thought - I wonder if the increased capacity inherent in .11ac will encourage more of a device-virtualziation approach to application deployment, with the client essentially dumb and all of the real action on the other side of the link. I personally like this approach, as it's easier both to manage security and to deploy the applications themselves.
Photo of Matthew Gast

Matthew Gast

  • 284 Posts
  • 63 Reply Likes
The cloud cluster down the street always has more power than your phone (even if the iPhone CPU is embarassingly powerful compared to the laptops I had when I first moved to Silicon Valley!).

We're starting to see more virtualization-type functions in devices, with storage containers on devices and management of individual containers. There's a question as to where the balance will fall -- is the device display-only, or does it get to store some data locally? But I agree that it's an interesting direction we're headed in.
Photo of Matthew Gast

Matthew Gast

  • 284 Posts
  • 63 Reply Likes
Multi-user MIMO has a potentially interesting effect on time-bounded traffic as well. Delivery-time sensitive traffic is often transmitted at slower data rates, either because the devices involved are battery-operated and implement fewer spatial streams, or they are using older chipsets. Within MU-MIMO, the access to the radio medium is controlled by high-priority traffic.

In the following figure (taken again from the book we're raffling off!), the green frame is a high-priority frame. Because the red and blue frames are lower priority, they can't seize the medium. However, if they are off in a "different" direction from the green receiver, they can still be transmitted at the same time!

Photo of Matthew Gast

Matthew Gast

  • 284 Posts
  • 63 Reply Likes
This is tangentially related, but still speaks to your question about applications. For most of my time in networking, we've always assumed that the IT staff knows exactly what's on the network. That was never true, though some network admins did better than others. With the increasing computing power now available in even what were traditionally embedded systems, there's much more awareness of what is happening out on the network. We're now at the point where network administrators can take action based on what they're being told by the network itself -- Palo Alto Networks was one of the first to do this, but that application visibility function is now able to run even at the edge of the network in an access point.
Photo of Craig Mathias

Craig Mathias

  • 63 Posts
  • 0 Reply Likes
True - but networks are getting more complex, and getting it right when it comes to both interpreting the data and taking action based on this can be very challenging indeed...