Tuesday, February 11, 2014

WLPC Conference Day 1 Highlights

I'm here at the WLAN Professionals Conference (#WLPC if you're following on Twitter). This is the first of what hopefully will turn into an annual conference dedicated to the Wi-Fi industry. But this conference is a bit different than what you might think a typical conference is. First and foremost, it's got a grassroots, peer-to-peer focus. It's engineers talking about Wi-Fi and gathering for discussion. It's not overly promoted by vendors or full of presentations with marketing drivel. Instead, it's just people who are passionate about this technology coming together to share their knowledge and experiences with each other to better everyone! What a great concept!

There are over 100+ attendees, many of whom are also presenters. I hear there was more demand than seats available, so next year Keith Parsons, organizer of the event, should have a solid baseline to grow the conference and allow more of you (the community) to attend and get involved. What's also great is that many of the presentations have been interactive, with great questions and quality discussion fostering the entire group to share information. The focus on the technology instead of the marketing that so often surrounds the technology and products in this industry. That's refreshing!

Day one of the conference was full of great content. Since there are two tracks of presentations, I can't cover all of the great presentations that occurred, but all sessions are being recorded so I plan on going back and watching the ones that I missed. Here are the highlights that stuck out from day one for me.

First up, Matthew Gast presented on 802.11ac. In typical Matthew Gast fashion, "minds were blown!" Matthew mixes the technical geeky details along with practical implications of the technology on real-world networks and the motivations that IEEE standard developers considered when drafting the protocol amendment. At one point Matthew also entered The Matrix (no not that virtual world in which machines rule mankind, rather the steering matrix for RF beamforming), but luckily spared the audience by dumbing down the mathematics for normal engineers :) Attendees also received a copy of his 802.11ac book.

Matthew Gast presents on 802.11ac

After a much needed coffee (and brain) break, Chuck Lukaszewski presented on high-density WLAN design. Chuck's presentation highlighted the method he uses to gather a Rough Order of Magnitude (ROM) to scope customer expectations and frame the design and budget early on in the process. This included understanding the associated user capacity, active user capacity, AP layout requirements, infrastructure dimensioning, and developing the ROM quote. This serves as a great starting point to ensure all parties are on the same page early on in the project and to focus more detailed activities that will follow with on-site visits such as site surveying and design.

High Density Rough Order of Magnitude (ROM) Process

Chuck then turned to RF coverage design in stadiums, detailing the mounting options available and the preferred methods his team uses in different situations to minimize co-channel interference. One point that he highlighted was music to my ears... the fact that CCA Busy is triggered by the physical preamble and PLCP header (and NOT by the ability for a receiver to decode the MAC header and read the Duration/NAV value). This impacts the distance at which an AP or client causes CCI because the preamble and PLCP header is encoded at the minimum PHY Basic rate (e.g. 1, 2, or 6 Mbps) and can be decoded at great distances! I've been explaining this to anyone who will listen for several years and it has a HUGE impact on RF network design.

Interference (CCI) goes MUCH farther than you think!

Brad Crump from CWNP discussed certifications and your career. This was one of the most engaging and useful discussions that I've had at the conference so far. Brad posed several questions to the audience about learning methods (live class, online, self study, boot camp) which prompted some passionate debate in the room. Several members of the audience were current or former instructors and had some very good information to share about how they've been able to effectively train students. Additional discussion on vendor-neutral and vendor-specific training was lively as well. Brad framed the discussion by explaining that there is a trade-off between acquiring knowledge and attaining a certification that HR managers are looking for in employee candidates. In short, HR managers are looking for "Expertise" and they often recognize vendor brand certifications like Cisco more than vendor-neutral. But Brad and team (including Julia Baldini, marketing whiz) are working to build the CWNP into a globally recognized brand! Woohoo, I wish them great success in this endeavor. I'm a big fan of CWNP content and certifications. Go get some CWNP people!

What is expertise?

Charlie Clemmer talked in the afternoon about RF in warehouse environments. Since I used to be directly responsible for managing over 3 dozen large warehouses with varying sizes and product inventory, this is a topic near and dear to my heart. Charlie shared is insights on why warehouses are not as easy as most people think. Sure, the clients typically require low bandwidth for telnet/SSH and warehouse management applications. But the environment can be extremely challenging due to legacy client compatibility issues, very high ceilings, unique freezer environments, high availability requirements, and some interesting IDF and wired network restrictions. Once again, great discussion ensued with the audience. Several seasoned wireless engineers who deal with warehouses shared their experiences and how to solve for some of the unique challenges that can be encountered. One topic of large discussion was how to cost-effectively design WLANs by performing a hybrid site survey with predictive modeling that is grounded with real-world data from on-site surveying in select sample areas of the warehouse. Otherwise, performing a full site-survey for warehouses that can be thousands or millions of square feet in size is too time consuming and expensive.

Charlie Clemmer presents on RF in warehouse environments

Finally, I'd like to give a quick shout out to all of the old friends that I've seen this week and new friends that I've met in person for the first time. These types of events are absolutely great to meet fellow peers in the industry! Also, a shout out to Prime Image Video who are recording all of the sessions and working their magic to make this content available for everyone online who couldn't attend in person. Ben and Andrea are awesome people and they are outstanding media professionals! If I haven't met you in person at the conference yet, don't be shy, come up and say 'Hi'!

Cheers,
Andrew von Nagy


P.S. If you're at the conference, I'll be speaking about capacity planning for every WLAN on Wednesday morning. I hope you'll swing on in!

4 comments:

  1. Andrew--thanks for your detailed Day One recap--first written recap I have seen (wish I could have been there...will do it next year).
    Do you have a link yet for where the video sessions will be located (or is Keith doing that....)

    Thanks!
    Glenn Cate
    Duke Energy
    gcatewifi.wordpress.com

    ReplyDelete
    Replies
    1. Hi Glenn,
      Keith will be posting the videos from all of the sessions. I plan on writing another blog post about the conference and including a link the video archives once they are online.

      Cheers,
      Andrew

      Delete
  2. Great post Andrew. I want to confirm my understanding of this: "This impacts the distance at which an AP or client causes CCI because the preamble and PLCP header is encoded at the minimum PHY Basic rate (e.g. 1, 2, or 6 Mbps)" My interpretation being the Minimum 'basic' rate, as set by the AP/infrastructure, is what is used to encode the Preamble & PLCP Header. Example being: I have 24 set as Basic (what Cisco WLC's call "mandatory") and 12 & 18 set to "supported". If a client is operating at 12, will it encode the Preamble & PLCP Header @ 12, however if it is operating at 54 it will Preamble @ not 24.

    This is a huge consideration when shrinking cell sizes from the AP's point of view. Client CCI is as much (or more) a CCI issue.

    Mike

    ReplyDelete
    Replies
    1. Hi Mike,
      Great question. To clarify, each 802.11 PHY (b, g, a, n, ac) specifies a hard-coded data rate for the PLCP header. It is not configurable or affected by the data rates set as Basic or Mandatory by an administrator.

      The data rates used for PLCP are:
      802.11-1997 = 1 Mbps
      802.11b long preamble = 1 Mbps
      802.11b short preamble = 2 Mbps
      802.11g = 6 Mbps
      802.11a = 6 Mbps
      802.11n and 802.11ac use the respective 802.11a/b/g data rate depending on the band

      A hard coded data rate is necessary to ensure the receiver can lock onto the incoming preamble, acquire the signal, and decode the PLCP header to determine key information for the incoming L2 MAC frame (such as higher data rate used and length of the frame). It has to be hard coded otherwise the receiver wouldn't know how to properly decode the very initial part of the signal. Only the L2 MAC (MPDU) is at a variable data rate.

      Cheers,
      Andrew

      Delete