Attenuation and Physical Structures and Obstructions

Unanswered Question
Feb 7th, 2008
User Badges:

Hi, I am looking for any information on structures and obstructions and how they impede and attenuate signals from a Cisco AP. I am a student and I need to put together a proposal for setting up wireless in our cinder block university. We presently have wireless and are using Cisco AP's, but the project is to redesign it or better yet start from scratch. Is this information or any other additional information listed here or anywhere else that might help me? Thanks...

  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
Loading.
scottmac Thu, 02/07/2008 - 08:56
User Badges:
  • Green, 3000 points or more

You're asking a pretty complex question, one that probably can't be thoroughly answered in the Cisco space provided.


About the best we can do is give you some "rules of thumb."


The reason is that there are just too many variables ... there are no easily predictable circumstances for any given scenario, aside from true line of site (If you can see it through an unobstructed path, you can PROBABLY (not always) connect to it).


"Not always" depends on things like interference (signal quality) and the transmit and receive characteristics of the AP/antenna(s) and the client/antenna(s) -- relating to signal strength.


If you search these Cisco wireless forums, the only consistent theme you'll find is that to cover an area properly, only a solid and comprehensive site survey can give you sufficient guidance.


Anything short of that is just a "good guess" and you'll either overspend and saturate the area (crappy network) or underspend and frustrate the end-users with weak or always-jumping-roaming connections (crappy system).


SO, given that you are a student, and you are tasked with improving the system with a "from scratch" implementation, ponder this approach:


1) Survey the existing system. Get current application and "typical" traffic flows; This is the baseline you need to improve against.


2) Query {whoever} and find out what the expansion plans are (application, type, traffic models, user density, etc). This should guide you to determine expansion options and designing a WLAN that can grow easily. You also need to determine the expectation of {whoever} in the WLAN performance and --> prioritize the needs and desires <--.



3) Commission a site survey (or figure out how you'd do one that meets your needs for extra credit). you ***ALWAYS*** need a site survey, certainly for a project of this size and scope.


4) Create you BOM (Bill of Materials) and price it out. Don't forget installation labor costs + fudge factor.


5) Compare that to your budget. This is where the "Prioritize the needs and desires" come in. Frequently the budget is not quite big enough for the wants / desires / needs. You have to work with {whoever} and find out if they want to expand the budget, or cut back on some of the wants / desires / needs.


6) Make sure the budget includes $$ for documentation. An undocumented network is a serious pain in the butt. It's hard to find, hard to fix, and makes everyone ... the users and the engineers that run it .. really frustrated.


7) make sure the budget includes post-install testing and needs verification, compared to the original project plan. You have to have some way of identifying to the "customer" {whoever} that you have accomplished their desires completely and within the agreed budget (or have documentation as to WHY you didn't).


This is where you can compare the new network against the baseline from #1 above. If it's not better, either the old network was *really good* or your new network "needs some additional adjustments."


You should also plan on several return trips to monitor the network traffic for utilization, dead spots, "enhancements" at the hands of students, etc.


It is rare that an WLAN install and implementation (at least big ones) happen perfectly at the first whack. Usually there will be "several" return trips for adjustment of one sort or another.


That's it for me


Good Luck


Scott

rmletcrmletc Thu, 02/07/2008 - 09:04
User Badges:

I was poking around and found a site survey article which I will definitely implement. My concerns were more for the quality of the walls. These being cinder block.


I guess I wanted to quantify the effects of walls being 5 feet apart. In any event all that you have written will definitely help. Thanks....

scottmac Thu, 02/07/2008 - 13:13
User Badges:
  • Green, 3000 points or more

Wall composition, the paint, the texture, and the angle of incidence for the RF are all factors.


Secondary are the type of the antenna, the radiation pattern, altitude/distance relative to ALL of the surfaces, RF absorbent things in the path (like plants, people or other water-bearing things, including fog & high humidity.


There's some other stuff in addition to the above.


At some point, you'll definitely get the feeling that instead of 10 hours of calculations and computations and modeling, it's easier to drag out an AP and client and just "play a little".


Good Luck

Scott



2.4GHz doesn't penetrate much, but it bounces really well.


You can't quantify unless you have all the variables (or at least, most of them).

Actions

This Discussion