Forum

  • In preparation for the CWDP, I though it would be good to get some feedback from you all on your biggest questions and desires for a deeper dive on WLAN design.

    If you have anything you'd like to see in the industry's best, vendor neutral design book (yeah, I am a psychic and can see the future) I want to hear it. While we are well underway on the text, it is never too late to solicit feedback.

    If you don't want your comments public, please feel free to PM me.

  • I think its such a vast area.

    From the basics of cell overlap for classic designs using three non overlapping channels etc on 2.4GHz to the different requirements for voice, data and RFID.

    I would also like to see a comparrison of the different architectural deployments.

    Autonomous
    Centralised controllers
    Decentralised hive (Aerohive)
    Single channel blankets (which is a query I am going to post about)

    The future of wireless 802.11ac and perhaps the impact of virtualisation.

  • By (Deleted User)

    I have to think about it but there are a few things that come to mind:

    1. Define the Core, Distribution, and Access Layers as it applies to a Wireless Design and Deployment.

    - This seems to be shaky so far in the CWNA book I'm reading. I'd suggest a more concrete definition of which elements are required versus desired and when/why.

    2. Define the Process of creating a Wireless Deployment Project - There should probably be different models:

    - SMB
    - Campus
    - Enterprise
    - Public / Free
    - Service Provider (commercial)
    - Other?

    3. The Project Management issues can be lost in the cloud.

    4. Designing with diagrams is really not a possibility - although very few design guides cover this in any degree of depth - perhaps it is ASS-U-ME(D). I bring this up since I recall a conversation from some newly minted CCIE's who questioned if they needed to know Visio at all. Hmm? Some things get lost in translation.

    5. RF Design Aspects

    6. Antenna Selection - Why?

    7. Wireless Security Elements - including Radius, PKI, WIPS, IPSec, etc.

    Let me think on it a bit.

  • The most common design issue I'm running into lately is trying to provide 801.11na services that don't drop because of DFS. So a few pages on design considerations surrounding DFS would be useful.

    Also as the UK has 13 2.4GHz channels what about a discussion on the feasability of a 4 channel layout. I've read whitepapers with conflicting advice and have never got round to doing some experimentation.

  • Paul

    You are right. Experimentation is the key for a particular RF environment. We always see those beautiful diagrams of channels 1, 6 and 11 all at the exact same amplitude. That almost never happens due to a whole bunch of factors [ Tx power, AP location etc ]. One of the biggest problems comes from the sidelobes caused by the sinx/x distribution pattern. I'd definitely experiment with trying to stick another channel "up at the top" and see what happens. Obviously there are a zillion other factors involved [ noise reliability etc ], but it would be interesting to see what happens on a test or empirical basis.

    Dave

  • I have tried using 4 channels as I am based in the UK, unfortunately I haven't done it for several years, the problem always came down to clients simply not wanting to connect on any channel above 11, I assumed right or wrong at the time as everything practically came from the US it was only really tested on channels upto 11. However there are still major issues with clients so it would be interesting to see.

    Just as an aside Japan uses 14 channels with channel 14 being having a centre frequency 12MHz higher than channel 13, as oposed to the normal channel seperation of 5 MHz so they often deploy 4 channel models.

    Addition info I would like to see in the design book would be IDS/IPS and guest architectures.

  • By (Deleted User)

    From this side of the equation, here's what I see and hear are the challenges that those who are implementing, or planning, or selling, or re-implementing, or troubleshooting WLANs come across...

    [list=1]How to implement VoWi-Fi (or VoWLAN or VoFi or whatever they're calling it these days)
    How to build from the ground up for data, voice, video and perhaps RTLS
    How to design a mission critical WLAN in a warehouse environment
    How to design a mission critical WLAN in a carpeted environment
    How to design for 802.11n for capacity, speed, resiliency, and certain apps
    How to design for high capacity campus environments, in which you'll have hundreds or even thousands of users hitting the network simultaneously for short bursts, e.g., between classes[/list]

    Just to name a few...

  • I would like to see more design. I have the CWNA and CWSP books and they cover the theories and security. I would like to see real world installation best practices. Maybe even include a CD where you can perform a simulated site survey and design a complete wireless network for the customer. Those are the skills I think would be the best taught.

  • By (Deleted User)

    How about a best practices for integrating A/B/G/N in real world networks?

    I work in a medical network with a lot of clinical applications. Some of those apps are wireless and work a lot of different frequencies, so there is always a need to design with a Spectrum Analyzer as well as the typical site survey.

    I'm sure you've allotted for these kinds of scenarios, but just in case.

  • Well I sat the ACMX (yeah I know its vendor specific) and it was pretty much all about design. The whole concept of the the exam was to give you a customer scenario and have you sit down and design a wireless solution around that and then explain why you choose what you did and _then_ implement it. My list would be something like this;
    1) Site survey - critical, explain what you approach is, whether manual, predictive or something in between and be able to explain why you choose that approach and what the tradeoffs are.
    2) Network and information gathering - If you don't know what the requirements for the design are, you'll never be able to meet them! You especially need to know how the solution is going to integrate with the customer's wired network and what applications its intended to be used for.
    3) Bill of materials - I've been in a lot of situations where someone has just gone and purchased equipment and thrown it in without relating it to some of my previous items. Its critical to get with the customer before he starts buying stuff so that a) they have all the parts they need, b) they know what they brought is going to meet their needs and especially c) they don't buy stuff on the basis of some fancy tech they don't understand and will never actually use!
    4) Integration - Part of the design is knowing what the integration points are with the existing network. For example, how many people have considered that WIPS information can be integrated into a SEIM to produce some pretty amazing information on what's happening in the WLAN?
    5) Project planning - This can be something as simple as a sheet with what you are aiming to do each day to a large MS Project with many intricate pieces. I have done quite a few large implementations now and I can say without a doubt having a good project manager is critical to success. Even though its more implementation than design I really thing it should be one of the basic design elements.

    I know there are a lot of extra parts I missed, like how to pick the right antennas, how to figure out what the Fresnel zone is, etc but these are some of the basic things I go through as part of every implementation I've been involved in. Hope that helps!

Page 1 of 3