Forum

I have a question on channel allocation in an 802.11b network where channel interference is unavoidable. A situation continues to come up in my site survey process that causes me to wonder if using channels in addition to 1, 6 and 11 would be a benefit.

The situation is this. I do a lot of site surveys in hospitals and one thing they seem to have in common is that the Access Points must be placed in the hallways outside of patient rooms. Each patient room is densely populated with machines, furniture and people; also each has a bathroom which for sanitary proposes is usually ceramic tile. All these things contribute to signal attenuation that requires the power on the AP to be cranked up to maximum. To do otherwise would cause so many APs to be used as to be cost prohibitive.

Since the APs are located in the hallways and radiating maximum power, the signal is hardly attenuated as it travels in the free space of the halls. I can use channels 1, 6 and 11 and keep channel interference to a minimum on long straight hallways. The problem comes where hallways intersect. A lot of the hospitals I have worked in are built in the shape of a Cross which causes channel interference where the hallways intersect. There is no way to have less than 4 APs “visible” at these points. It is also very common to have nurse stations at these intersecting points, so the RF network has to be very robust here.

I use Ekahau site survey 2.0 software on an Acer tablet PC with very good results. The software allows me to adjust the channels on placed APs on the fly, allowing me to play what if scenarios. By adjusting the channels on the APs at these intersecting points to include channels 3, 4, 8 and 9 the software shows less interference than when only 1, 6 and 11 are used. Now the big question, just because the software shows less interference does it translate into less interference in the real world? Is it beneficial to have a few channels interfering over half of their bandwidth instead of a couple of APs on the same channel?

Your thoughts and experience would be appreciated.

Hi:

Balancing signal interference with manual settings may have a place in the future of automatically managed AP channel and power settings. I personally do not imagine it being a worthwhile design for static settings.

What keeps you from putting one AP in the intersection and staging non-interfering APs down each of the four halls?

Thanks. /criss

Hi NovellRed,

I've visited this situation more than once! :-) The way we normally fix this situation is to get very familiar with non-omni antennas. Most site surveyors, for some odd reason, think that omni antennas are the only ones that exist. :-) What we used to do is to put a patch antenna at the end of one of the hallways, let the backlobe cover the nurse's station, and then have enough gain (perhaps 6 or 9 dBi) to shoot down the hallway. If it won't reach the other end, then put another one at the other end on a different channel facing directly toward the first. The things to watch out for in this situation:

1. That the rooms are covered all the way to the windows. Make sure your antenna has AT LEAST a 60 degree horizontal beamwidth in a situation like this.
2. Make sure the rooms on both sides of each patch have good coverage. The closer the rooms are to the patch antennas (on the sides), the worse the coverage).
3. Check your vertical beamwidths. If this is a 5-6 story hospital with stacked identical floors, you might be able to get away with only doing this scenario on floor 2(which covers 1, 2, & 3), floor 5 (which covers 4, 5, 6), etc.
4. When you are shooting beams down the hallways that cross the first set of hallways, make sure to minimize co- or adjacent channel interference as much as you can...it's usually not an impossibly difficult task.

Hope this helps.
Devinator

Hi guys. Thanks for the input. It helps widen my vision. Sometimes I get so used to doing things “the way we always do” that I don’t even think of different approaches. Kind of like a horse with blinders on. That’s what makes this forum so valuable to me.

Placing one access point at the nurse’s station and using non interfering channels down each hall works good for coverage, but my customers usually stress capacity and overlapping coverage at the nursing stations. Historically this has been the gathering place for nurses to come and update their patient charts, the charge nurse and nursing supervisor also do their work here. The customer usually loves to see my Ekahau report showing nurse station coverage from 3 or 4 access points. It gives them a warm fuzzy that they wouldn’t be called in at 2 am because a single AP failure. I can see a case to be made for education my customers a bit.

I am now thinking of modifications that can be made to my site survey rig to make the placement of patch antennas easier and more accurate. At first it seems counter intuitive to use directional antennas for shooting down hallways, since hallway coverage isn’t the problem, getting the signal to the outer edges of the patient rooms is. I can now see the benefit of being able to direct the coverage to reduce channel interference. I always have several different patch antennas with differing gains with me. I will build some extra time into my next site survey and experiment with them.

Have you really been able to get coverage a floor above and a floor below? Maybe construction is heavier here, but I can’t get much signal through floors/ceilings. A foot or more of concrete between floors is pretty standard around here. Add all the rebar in the concrete, a layer of corrugated steel, the ventilation ductwork and assorted plumbing and it pretty much stops signals from traveling between floors. I am curious as to what I get using directional antennas, maybe focusing the energy will help here too.

Thanks again guys, I appreciate your taking the time to share your thoughts/experience. I will update the forum after the next site survey to let you know how things work out using the techniques you offer.

Hi NovellRed,

I did a site survey in KY once that was as big as a city. We were able to cover the floor above and below with a single AP...guess it was built like a house of cards, eh? ;) Let us know what you find with your next survey - i'm interested to hear.

devinator

I've had MANY experiences with similar issues. I have had to get very creative at times to get good coverage while keeping overlapping channels apart. I have had pretty good success from experience as well as good old common sense. (of course this is with understanding the principles of RF)

For example, I NEVER use omni antennas - (unless we need to light up a small cafe or something with 1-2 AP's max) I use highly directionalized antennas - and depending on the building structure, I will actually use a poured concrete wall (or similar) to act as a "RF Wall" to keep the backplane to a minimum so I can use additional directional antennas without compromising my signal.

Another trick is to turn down the transmit power. I have found that lower power often works better and provides greater coverage - especially if there are other radios in the facility to interfere.

Finally, get to know the equipment you are installing - every manufacturers equipment varies in performance, even though RF is RF... Once you know what the typical results you should see in a survey are the easier it is for you to determine what issues you are having.

Good luck
Scott