Forum

  • I am currently working with a customer who operates an OEM'd Meru wireless network which was bought from Foundry before Brocade acquired them. The wireless network was designed by another company for data usage using 802.11 b/g. They are using AP-208s (non-diversity) with dipole antennas at 100mW spaced fairly far apart on the floors but the APs are stacked on top of one another from floor to floor. The building is 3 stories of newer construction and an AirMagnet passive survey reveals signal propagation between floors is very high, as I can see all three APs about the CCA value on any floor at each AP location. I have also found signal deficiencies inside of the patient rooms at 2.4 and 5.0 GHz.

    I am not very familiar with the SCA so my question is; how big of an impact do highly overlapping cells have on overall throughput? Would it be a worthwhile investment to relocate the APs in a staggered pattern so the floor to floor impact from neighboring APs isn't as drastic? Or maybe swap out the antennas for a more low profile pattern?

    Any advice would be appreciated.

  • Overlapping channels will surely have effect on the wlan. I suggest u start reading the cwts study guide at least. Followed by the cwna study guide. Do these readings and u won't go wrong.

  • I think you should contact Meru directly with your questions too.

  • A very interesting question. I always read that Meru's SCA defied the laws of interference because all APs are on the same channel. The claim that their APs were built so that you can put them anywhere you want without worrying about channel planning.

  • Meru's SCA is somewhat similar to trapeze(used to be with Belden) SCA. Both of them claimed no issues with their channel planning. SCA is good for voice deployment but not with data. MCA is still best for data and video streaming because of more aggregate tput compared to SCA.

  • Are the channels the same on all three floors?

  • Wlanman, Currently all access points are operating on the same channel within the same virtual cell. I would certainly entertain the thought of running multiple virtual cells on different channels (if that's possible), but I would concerned the floor to floor bleed through would also cause problems. What's to say a client on the 2nd floor wouldn't "roam" to an AP on the 3rd floor and then quickly have to "roam" back to the 2nd?

    In addition to web searches and combing through my ~4 gb worth of pdfs and whitepapers I have collected over the past 4 years, I have been browsing through the CWDP book for some additional information on the ability of SCA to handle drastically overlapping cells. One of the authors of the CWDP book (who may or may not monitor this forum :) alludes to the fact that this has been a heavily debated topic in the industry. He concludes that regardless of whether the controller is able to coordinate airtime from the access point's perspective, clients are out of their realm of their control. Clients are stupid. "Simply put, all of the same logic still applies to AP placement regardless of what type of technology you intend to deploy...?

    Only 6 more months until my cwna runs out. I guess I had better start doing more than browsing the cwdp book. Good stuff

  • If you ran a different virtual cell (channel) on each floor, the bleed through wouldn't be an issue. It would help with performance (because it is closer to a proper MCA). However, the client device wouldn't be able to roam from floor to floor so if you need that, no dice on the multiple cells.

    If you are having coverage issues, a higher gain (lower vert beamwidth) could help. I'd try a few experiments before committing to swapping all antennas though.

    GT

  • GT,

    Are you implying that users on different channels (w/ SCA) cannot be on the same subchannel?

  • Subchannel? Not sure what you mean there... if you meant SSID, then I'll take a stab at it.

    If the design was a different channel per floor (1,6,11, repeat as necessary) AND the SSID was all the same, that would not work right. Well, it would negate the benefit of the SCA because the STA could (and occasionally would) roam to an AP on a floor above or below.

    In order to solve that, each floor would need a different SSID which is kind of bunk too.

    GT

Page 1 of 2