@ Those who know wireless

Neeko

I'm trying to clarify some things related to wireless design and deployment and can't find much out there that doesn't just talk about physical data rates, modulation and the more theoretical side.

1) If I was to deploy APs throughout a multi-story building, with floor space of 200x200 meters, roughly how many APs would be needed? Assuming a generic building with little in the form of obstacles. 802.11n APs for example, have a maximum range of around 90m right? I'm guessing, in order to overlap cells and acheive high data rates throughout, there would need to be quite a few APs... 9 or so?

2) Is the best bet to fix APs on the ceiling to cover multiple floors? So those 9 (or whatever the required actually is) could cover 2 floors, and a 4 floor building would therefore require 18 APs. My numbers may be way off mind.

3) In terms of implementing a typical layered switch design for redundancy and scalability, what is the best approach when wireless is involved? A switch only has so many trunk ports, how do you go about trunking that many APs from a lower number of access switches? What is the recommended number of switches per every 10 APs, for example?

If there are any online sources that anyone knows of that discuss WLAN design then I'm prepared to read, it just seems the majority of wireless texts discuss the intracasies of modulation, encoding MAC efficiency etc rather than design.

GT-Rob

Are you looking to run VoWLAN? That will effect your requirements, especially with overlap.


We run APs on each floor, with omni-antennas. You have to be careful about channel-overlap when you start going between floors, and im not sure what coverage would be like.


Are you dunning these on off a WLC?


You can connect the APs to access ports, they dont need to be trunked. Think of the AP as just a bridge between wired and the network card on the client.

dynamik

1. Keep in mind that each WAP is a shared medium, so you need to take usage into account as well. You'll only get about half of the maximum in actual throughput, and then you need to divide that by the number of users. i.e. 54/2 = 27 / 10 users = 2.7mbps per user. You may want to have more APs and tone down the signal strength for better performance and less overlap/interference.

2. I don't think attaching to the ceiling is going to make a significant difference. You're removing what, 8ft of emptiness? I'd say whatever is in between the floors is going to affect the signal a lot more. It also depends on what antennas you use, power settings, etc.

3. You're really going to want to look at a WLC (WLAN Controller) for something of that scale. Cisco and numerous others offer those.

Neeko

GT-Rob wrote: »

Are you looking to run VoWLAN? That will effect your requirements, especially with overlap.


We run APs on each floor, with omni-antennas. You have to be careful about channel-overlap when you start going between floors, and im not sure what coverage would be like.


Are you dunning these on off a WLC?


You can connect the APs to access ports, they dont need to be trunked. Think of the AP as just a bridge between wired and the network card on the client.

In order to implement VLANs though, I think they would need to be trunked. I've read how VLANs can be implemented on APs by SSIDs and a trunk to a switch.

WLC isn't something I know about, what does that do in terms of implementing wireless to act like a layered wired LAN would?

Voice is not an immediate requirement, although that could be incorporated. Wouldn't overlap for clients provide the physical scope suitable for cordless VOIP phones anyway? I know there are some standards for QoS and fast roaming that are ratified, so not sure what implicatons you speak of there.

So if you run APs on each floor, I assume you would reduce signal strength in order to avoid channel overlap like dynamik said, which will achieve higher data rates?

What about number of APs then for the size of floor I mentioned? Based on what you guys have said seems like even more would be needed than I anticipated.

darkerosxx

Let me google that for you

Neeko

darkerosxx wrote: »

Let me google that for you

lol thanks, its not like I was doing that in the meantime

dynamik

No offense dude, but it seems like there is an enormous gap between your current level of knowledge and what is necessary for this project. You should go through the CWNA and CCNA:W books before even attempting this. You're going to want to do some serious site survey work, possibly even buying expensive hardware and/or software. While their may be no noticeable interference, you may find something else, such as microwaves, cordless phones, etc. that really play havoc with your signal. Maybe you need to go to 802.11A instead of B/G. This is an enormous undertaking.

Neeko

dynamik wrote: »

No offense dude, but it seems like there is an enormous gap between your current level of knowledge and what is necessary for this project. You should go through the CWNA and CCNA:W books before even attempting this. You're going to want to do some serious site survey work, possibly even buying expensive hardware and/or software. While their may be no noticeable interference, you may find something else, such as microwaves, cordless phones, etc. that really play havoc with your signal. Maybe you need to go to 802.11A instead of B/G. This is an enormous undertaking.

None taken, you're right that I don't know enough but its not a real scenario, its for coursework. I thought as soon as the brief was released that it was unfeasible for undergrads on a course that has rarely mentioned wireless to design such a network that is actually realistic and could be implemented, but given how the professor has written such a ridiculous brief, I doubt he'd know any different no matter what I include. The average level of knowledge on my course is much lower than mine, so I have no idea what other people are considering for this. To cover 4000 square meters of floor space on multiple floors with wireless is not anywhere near as simple as is being implied.

Turgon

Neeko wrote: »

I'm trying to clarify some things related to wireless design and deployment and can't find much out there that doesn't just talk about physical data rates, modulation and the more theoretical side.

1) If I was to deploy APs throughout a multi-story building, with floor space of 200x200 meters, roughly how many APs would be needed? Assuming a generic building with little in the form of obstacles. 802.11n APs for example, have a maximum range of around 90m right? I'm guessing, in order to overlap cells and acheive high data rates throughout, there would need to be quite a few APs... 9 or so?

2) Is the best bet to fix APs on the ceiling to cover multiple floors? So those 9 (or whatever the required actually is) could cover 2 floors, and a 4 floor building would therefore require 18 APs. My numbers may be way off mind.

3) In terms of implementing a typical layered switch design for redundancy and scalability, what is the best approach when wireless is involved? A switch only has so many trunk ports, how do you go about trunking that many APs from a lower number of access switches? What is the recommended number of switches per every 10 APs, for example?

If there are any online sources that anyone knows of that discuss WLAN design then I'm prepared to read, it just seems the majority of wireless texts discuss the intracasies of modulation, encoding MAC efficiency etc rather than design.

Can't advise you myself but I will be interested to see how you get on with all this. Wireless to me is the devil and edge work. That said I expect to be looking at it again next year. Some good consultancy roles are available if you really know what you are doing with wireless. Going by the amount of jobs asking for wireless skills it seems there is something of a shortage.

GT-Rob

I will echo Dynamik, and say that this is a big project, trust me. We have around 300 APs over 5 offices, and about 10 controllers. We also have a large hanger type building that has a trade show 2 times a year, with 300 users on wireless. Oh, and we need complete VoWLAN roaming.

Needless to say, wireless takes up a large chunk of my week some weeks lol.

Its not like a switch you install, configure, and come back when it dies. Its a very dynamic, changing, and touchy technology. You will learn a lot and slowly become an expert as long your company understands that its not something you can read about then put in. It will be an ongoing project until you die. Take the time to do one section at a time, and learn from the behavior you see.


Trust me, you want a WLC.


Also, buy the new CCDP (arch) book. There are 2 chapters there on wireless design and its exactly what you need to read right now.

Neeko

GT-Rob wrote: »

I will echo Dynamik, and say that this is a big project, trust me. We have around 300 APs over 5 offices, and about 10 controllers. We also have a large hanger type building that has a trade show 2 times a year, with 300 users on wireless. Oh, and we need complete VoWLAN roaming.

Needless to say, wireless takes up a large chunk of my week some weeks lol.

Its not like a switch you install, configure, and come back when it dies. Its a very dynamic, changing, and touchy technology. You will learn a lot and slowly become an expert as long your company understands that its not something you can read about then put in. It will be an ongoing project until you die. Take the time to do one section at a time, and learn from the behavior you see.


Trust me, you want a WLC.


Also, buy the new CCDP (arch) book. There are 2 chapters there on wireless design and its exactly what you need to read right now.

I recognise the complexities of such a project, but reiterate that it is purely theoretical and not for a real project on a live network (thankfully ).

You say 300 APs for 5 offices. How many APs do you think would be necessary for a 6 story building with the measurements I stated?

miller811

Typical steps in an RF site survey process include the following:
Step 1. Define customer requirements.
Step 2. Identify coverage areas and user density.
Step 3. Determine preliminary AP locations.
Step 4. Perform the actual RF site survey.