Forum

  • This is a sore point for me. If people think they can truly deploy controller-based APs and not really plan out signal level propagation into an environment with some care we're going to have a bright future cleaning up after everybody's mess.

    Those of you with a fair amount of deployment experience in a variety of environments thoroughly realize that propagation varies widely depending on construction and what the environment is populated with. Tx power from the AP is only one side of the link. The uplink is often the biggest variable and headache in deployments due to many factors. 802.11h has some interesting implications here, but from what I've read I only see it as a value of turning down the Tx power of clients as most clients are already at full power most of the time. ...then we're talking about whether all developers for the client software actually take this aspect of 802.11h and implement it (not the bigger issue of radar avoidance).

    In my opinion deploying APs in an environment must be done using care and mapping signal levels based on the chosen antenna type using two-way communication measurements in order to understand where link budget problems occur. TPC doesn't really help this.

    Case in point...I'm guessing some genius marketing guy at our friends who make the controller products came up with the claim that if an AP goes down the surrounding APs will power up and fill in (assuming there's enough headroom in the amplifier and the network was designed at non-peak Tx power). Now, think about the uplink. Let's say you're on the phone and speaking into a speakerphone on the table in front of you and it drops. Let's say there's another speakerphone a couple of rooms away from you and it takes over your conversation and powers up its speaker to get the sound strong enough to you. Now, assuming you don't talk any louder (important concept here), how well do you think your voice is heard? You're at least uplink limited.

    I'm just skimming over the top here on the other issues involved, but in an effort to keep this simple, this one issue is a deal breaker.

    On another point, the other stated reasons for doing site surveys for spectrum analysis, etc. is spot on.

    Sorry for the passionate response :), but as I said, this is a sore point for me.

    Shawn

Page 1 of 1
  • 1