Forum

GTHill Escribi?3:

Snaglpus00 Escribi?3:

I would go with Meru, especially if you don't have any RF people on staff. Meru Single Channel design works well.

I work for a VAR that sells Meru, so if you need info you can shoot me a PM.

~K

How does an SCA help if there is no Wi-Fi techs on staff?

Proper design from the beginning is more important than the vendor. Ease of management should be strongly considered for the system. When resources are slim, fewer AP's and better management is key.

GT

GT,

Why is SCA better in the long run? Lets say you design a working microcell deployment, works great for 6 months until something / anything changes. Such as you need more bandwidth, or put in a new wall, or lots of new students, etc. Pretty much any change that would require an AP move or AP addition. With a microcell you have to get out the RF maps and work out a new design to meet the "new" requirements. Right? Which is easy for RF people.

With a SCA, if you need more coverage...add a new AP with no planning. Need more density of users, add more AP's on a different channel that layers on top of the current and you just doubled your throughput capacity.

Also Meru has a GREAT monitoring and troubleshooting system which is easy for non-RF people to grasp so they can manage their network.

However, they are expensive!

But after you factor in the cost to RF surveys and designs the costs start to equal out. Especially when you need dense deployments where in microcell world you are turning down AP power to shrink cell size to support high density, Meru Ap's are all at full power and can support the same number of users as a Microcell.

~K

Snag,

Even with microcell deployments these days, most vendors have some sort of adaptive nature that automatically sets power and channel relative to the RF environment. If something changes, there is much the APs can do to channel and power around the movement of a wall, many new students, etc.

So for the most part, depending on things you wouldn't have to bust out the RF maps and work out a new design even if you weren't an "RF person."

For both architectures, microcell and SCA, if the change was great enough, it would require an AP move or an AP addition. Sure, with Meru's SCA you can just add it to the virtual cell. But with microcell architectures such as Aruba's ARM, you can just as easily add the AP and that's it. The adaptive nature of the APs bring the new AP into the fold and adjust power and channels accordingly. So SCA doesn't buy you much in the moves, adds and changes department.

Lastly, deploying APs at full power can have the effect of allowing you to deploy fewer APs in a given area, but you have little to no safety zone with regards to a failed AP or the introduction of a strong interference source. Any vendor can set their APs to max power, it's just that most don't for various reasons (including far-side EIRP differences..i.e. the client can transmit 30mW but the AP can transmit 100mW)...Meru isn't doing anything terribly special by setting APs to max power.

Double Snag.

Power control functions in AP's do just that, control power in AP's and does not control clients. You can have an AP set to 30mw but all the clients are at 100mw unless they are like Cisco and CCX extensions which allow power setting of clients.

As for AP failure in SCA design, it is standard to design the coverage area to "hear" 2 or more AP's at a minimum, so if one were to fail you have 2 as a back up.

Don't get me wrong, I still do Microcell deployments but those work better when you need just coverage and not a high user density. For example covering a office building that will not see more than 20 cleints per AP range.

SCA is better for large/dense deployments like 2500 students in a school where 30+ students in each classroom NEED access all at the same time, like during class.

There is always more than one solution, if the customer wants I'll be glad to put Meru up against Aruba!

~K

Snag,

If you're a Meru SE or partner, you of all people should know that not all client devices operate at 100mW. VoIP handsets are a perfect example. And while they are becoming less and less, there are in fact laptop client devices that don't do more than 30mW.

Client transmit power control will be very nice when it comes and is ubiquitous, but for now, we have to live with the lack of it in most clients.

Anyway, to the original poster's question: I would evaluate a few vendors for your uses....include Meru and Ruckus of course, but also add Aruba and Cisco--maybe even Trapeze if you can find a Belden rep.

bjwhite Escribi?3:

Snag,

If you're a Meru SE or partner, you of all people should know that not all client devices operate at 100mW. VoIP handsets are a perfect example. And while they are becoming less and less, there are in fact laptop client devices that don't do more than 30mW.

Client transmit power control will be very nice when it comes and is ubiquitous, but for now, we have to live with the lack of it in most clients.

Anyway, to the original poster's question: I would evaluate a few vendors for your uses....include Meru and Ruckus of course, but also add Aruba and Cisco--maybe even Trapeze if you can find a Belden rep.

If you wish to continue this discussion, lets start a new thread, this is not the place.

The original poster wanted opinions between Meru and Ruckus which have been offered with short explanation on why.

If the original poster is interested he can contact me via PM and I can supply contacts in universities and K12 schools that are using Meru.

~K

I am loving the passion here. :)

bjwhite Escribi?3:

Snag,

Even with microcell deployments these days, most vendors have some sort of adaptive nature that automatically sets power and channel relative to the RF environment. If something changes, there is much the APs can do to channel and power around the movement of a wall, many new students, etc.

So for the most part, depending on things you wouldn't have to bust out the RF maps and work out a new design even if you weren't an "RF person."

For both architectures, microcell and SCA, if the change was great enough, it would require an AP move or an AP addition. Sure, with Meru's SCA you can just add it to the virtual cell. But with microcell architectures such as Aruba's ARM, you can just as easily add the AP and that's it. The adaptive nature of the APs bring the new AP into the fold and adjust power and channels accordingly. So SCA doesn't buy you much in the moves, adds and changes department.

Lastly, deploying APs at full power can have the effect of allowing you to deploy fewer APs in a given area, but you have little to no safety zone with regards to a failed AP or the introduction of a strong interference source. Any vendor can set their APs to max power, it's just that most don't for various reasons (including far-side EIRP differences..i.e. the client can transmit 30mW but the AP can transmit 100mW)...Meru isn't doing anything terribly special by setting APs to max power.

1-When you deploy a network for 54Mbps/300Mbps, VOIP @-67dBm the fact that clients are @ 30mw and APs @100mw doesn't have an impact

2- changing dynamically channels and power is disruptive for many critical applications ie VOIP not even talking about the time it takes to converge.

3-

???¡é?¡é?????¡­?¡°but you have little to no safety zone with regards to a failed AP or the introduction of a strong interference source???¡é?¡é????????.

. Are you saying that a solution is more sensitive to strong interference when deployed @ 100mw then @ 30 /50 mw ? can you clarify ?

4-

"Any vendor can set their APs to max power", "Meru isn't doing anything terribly special by setting APs to max power"

I thought the reason for decreasing the power was to mitigate co-channel interference. Now you're saying that all the solutions can use APs @ max power. Can you clarify ? cause if it's true, if one AP fails and you increase the tx power of neighbor APs to compensate then you increase co-channel interference, I'm I correct ?

Roma,

You can't group all wireless network into the same design model. Wireless deployments will all be different. There are times where each different design will be a benefit over the other. Not every client will want the same level of wireless, nor do they want to pay for stuff they don't need.

A coffee shop,a stadium, a hotel, a Elementary school, a High School, a college, a apartment building, a high rise office - All have different needs and follow different designs.

Certain WLAN vendors have special technologies which work better for different applications, but that is not to say they can't work in the others...they just may not be as efficient.

This is why before we even quote customers on any wireless, we get a "wireless mission statement" from them. Pretty much it is what they "expect" the wireless to do. Based off that, we design a network. This also will CYA in the case where 12 months down the road the client has a "new" idea and tries to use the wireless in a way it was not designed for and calls to complain about how the wireless.

Example, design an office for 20 wireless users for email and light browsing over 10,000 sqft and works perfect...then they change that 20 to be 100 clients.....can you see what is going to happen?

~K

Does Meru AP's have the ability to do Mesh and Band Steering?

GT

GT,

Mesh yes, up to I think 6 AP's deep in a mesh tree.

Not sure if they do "band steering" in the way Aruba or others do, which is supposed to move clients to a less congested network / band. However since Meru networks are single channel that may not directly apply.

They do automatically do load balancing of users between AP's and between channel layers if you have more than 1 channel overlay in a given area which is used for more throughput or extreme densities. Which i think accomplishes the same goal.

~K