Validated Reference Design Guides

Very High Density 802.11ac Networks Validated Reference Design

MVP
MVP

Very high-density WLANs are defined as RF coverage zones with a large number of wireless clients and APs in a single physical space. For purposes of this reference design, a VHD WLAN is one that is designed to serve at least 100 devices per cell. A VHD WLAN may serve as many as 500 devices per cell. With the proliferation of wireless-enabled personal and enterprise mobile devices, a surprisingly diverse range of facilities need VHD WLAN connectivity:

 

  • Large meeting rooms
  • Lecture halls and auditoriums
  • Convention center meeting halls
  • Hotel ballrooms
  • Stadiums, arenas, and ballparks
  • Concert halls and amphitheaters
  • Casinos
  • Airport concourses
  • Passenger aircraft and cruise ships
  • Places of worship

To help our customer and partner engineers succeed in meeting these new requirements, we have written this new validated reference design (VRD) exclusively about very high-density (VHD) WLANs. The guide captures the best practices of our field engineering teams. Aruba also built a dedicated VHD test facility with 300 of the latest 802.11ac devices to produce updated performance data. We are releasing this data to the public to assist in capacity planning

 

PLAN OF THE VRD

Aruba is introducing a new and modular approach with this VRD. We recognize that the people who are interested in this topic have a wide range of knowledge levels and interest levels. To that end, Aruba is publishing this VRD as a series of three guides, plus two individual scenarios that deal with specific deployment use cases. These guides can be mixed and matched as desired based on each reader’s experience and interests. This organization is shown below.

 

Screen Shot 2015-03-19 at 12.13.44 PM.png

VRD Volume

Click on Desired Format to Download

Planning Guide

PDF

HTML

Engineering & Configuration Guide

PDF

HTML

Theory Guide

PDF

HTML

Scenario 1: Adjacent Large Auditoriums

PDF

HTML

Scenario 2: Large Indoor Arena

PDF

HTML

Whole VRD

ZIP HTML

Frame Time Calculator

XLSX  
Version history
Revision #:
23 of 23
Last update:
‎04-23-2015 05:09 AM
Updated by:
 
Contributors
Comments

Thank you @Chuck_Lukaszewski and your team for this amazing VRD!

ishanmindika

Thanks a lot for the VRD's lot of useful info.

DResseguie

This is one of the best documents I have seen to date on the topic of High Density Wireless! This does a great job of showing all the factors to be considered including Theory, Planning, Engineering and Scenario Plans. This material is very useful regardless of the wireless vendor you are currently using. You do not have to be an Aruba customer to access this information. GREAT JOB Aruba Networks for helping improve the industry for everyone. - Douglas Resseguie

Thanks @DResseguie The team appriciates it! @Chuck_Lukaszewski

pcasanave

Is this being updated for 8.x?  I know much of the theory behind these practices have not changed, but the user device population, 802.11 technologies, and methods for supporting them have evolved in some senses.  It would be interesting to see an update that takes into account the additional tools Aruba has provided via 8.x ArubaOS code and Mobility Master.  

 

Cheers,

 

clukas
pcasanave,
 
It makes more sense to look at any possible revision in conjunction with 802.11ax next year.  802.11ax includes a variety of multi-user features including OFDMA and UL-MU-MIMO, as well as improved versions of the DL-MU-MIMO that was delivered in 802.11ac Wave 2.
 
None of the capacity methodology is dependent on ArubaOS version.  While the clustering architecture does get more interesting in 8.X, this is already addressed in the User Guide and is really orthogonal to the purpose of the VHD guide.  The vast majority of VHD config for SSIDs and radio profiles is unchanged in 8.X.  We're studying AirMatch performance in VHD environments, and will factor any guidance there in 11ax timeframe.
 
 
pcasanave

clukas,

 

Thanks for the reply.  I am guessing that mobile adoption of 802.11ax will occur at the same speed as adoption of 802.11ac Wave 2 happened.  Based on that, I am assuming that the clear majority of mobile devices, specifically Apple ones, are 5+ years from 802.11ax unless their hands are forced.  That being the case, any gains associated with 802.11ax won't be realized for many years to come, in the same way the 802.11ac Wave 2 still hasn't proven to be fruitful in large mobile device dominated venues at this time. 

 

Much has changed in the video world since 2015, which is the major reason this guide is out of date.  Streaming video providers have updated mobile profiles to move away from a 480p "mobile HD" standard to 720p as the new mobile HD profile.  Add to that, there is a huge prevalence of 60fps video hitting the market.  These are more than just available qualities, they are the new expectation of the audience.  This causes this guides per user bandwidth and cycle impacts to be off greatly from when this document was written.  This should be the greatest driving force for an update, not 802.11ax which is years from relevance.  So, the majority of us paying attention to arena and venue networking, won't be considering 802.11ax for two to three more years.  Please don't hold off in updating this until then and don't loose vision of real impacts of sports 60fps HD video, which is domintant service being consumed in these venues.  I am not saying 802.11ax is not important, but it is the least of tomorrow's concerns when looking at modern designs from a realistic perspective.

 

Cheers,

P

clukas
pcasanave,
 
Feel free to contact me directly chuck.lukaszewski@hpe.com to discuss further offline.
 
We're acutely aware of improvements in mobile device screen resolution, which are of course offset by improvements in video encoders.  There is tremendous R&D being done in this area, Netflix and Google among others maintain public blogs talking about the encoding ladder algorithms they use, the typical encoding gain over time for a given resolution, etc.  In case you are not aware, many of these services dynamically adjust to the screen resolution and CPU capacity of a device, which means that a Galaxy S9 watching a 4K stream may not be receiving the same stream at all as a laptop with a 4K display.  We've taken some measurements along this line and there is extreme variation that is observable. 
 
In terms of the VRD, all of this is to some extent irrelevant.  One of the core messages of the VRD is that the Total System Throughput (TST) of a VHD area is a fixed block of capacity that is independent of the device count and offered load.  That capacity is then subdivided to Active Device Throughput (ADT) as explained in Chapter EC-2.   But because the TST is largely determined by the RF design and facility type, it is the case that one must adjust the video encoding rates to fit within the availabale ADT.  Not the other way around.  Since most large venues are designed for 1Mbps per-seat peak rate, this is what you have to work with in peak conditions.  During non-peak conditions you will get more ADT per device.  The video encoders typically adapt to channel capacity, and so should uprate/downrate accordingly.
 
By way of example, the best performing stadiums >70,000 seats are generating ~8Gbps sustained load on 20,000 concurrent devices, which implies an average of just 400Kbps per device.  These "peak" scenarios tend to be dominated by uplink traffic (concert streaming, big event/penalty on the sporting field).  So for the downlink video streaming case you cite typically loads are far smaller and the "effective instantaneous ADT" would be higher.
 
To drive ADTs higher for a given TST, we need higher prevalence of MU technologies like DL-MU-MIMO and OFDMA. 
 
FYI, we are anticipating a far more aggressive rollout of 11ax radios in smartphone class devices.  Because the phone manufacturers want Target Wait Time and other features in 11ax that reduce energy consumption.   Won't be overnight of course, but by Spring of 2020 I wouldn't be surprised if WLAN dashboards were reporting 30% or higher 11ax populations.  And every 11ax device running an MU mode leaves more airtime for the legacy SU clients.
 
pcasanave

Chuck,

 

My company works in the streaming video distribution side also, so extensively familiar with gains make on the encoding side, and the limitations of modern mobile devices when it comes to things like HVEC.  I am also aware of dynamic bit rates based on DASH, HLS, and other.  That being said, the modern millennial user expectations for streaming sports video quality on mobile devices does not allow for downgraded quality due to wireless capacity or deployment.  The new expectation is 5-10Mbps streaming availability as needed, meaning of the 10,000's of devices, some 10-30% are going to be looking for that rate at peak.  This bandwidth requirement is driven by the combined lack of H.265 support on many mobile devices, coupled with 60fps HD streams.  This percentage can even be higher at breaks in play, i.e. end of a period or half time.  I hope you are right, and 802.11ax adoption occurs much faster than wave 2 did.  Peers are seeing as high as 80% take rates in stadium student sections, with nearly all being demanding connections during extended breaks; so, 802.11ax will be great once we pass it being the outlier device type.

 

Thanks for your responsiveness and look forward to updates on this and the rest of the portfolio. 

 

Cheers,

P

Search Airheads
cancel
Showing results for 
Search instead for 
Did you mean: