1.4.a [i] Define the tasks/goals for a preliminary site survey
1.4.a [ii] Conduct the site survey
1.4.a [iii] Determine AP quantity, placement and antenna type
Much like the previous section which focused on customer requirements and current lay of the land, defining the tasks and goals for the survey will follow much the same pattern through meetings between stakeholders to agree upon the scope of work. The key decision during those discussions will be whether to do a predictive, passive or active site survey with each containing its own pros and cons. Tied into any of these three would be the requirements previously gathered with the stakeholders.
The most basic (cheapest) of the three. Using tools like Ekahau Site Survey, Air Magnet, and – most importantly for the CCIEW – Cisco Prime Infrastructure, we are able to produce predictive site surveys of varying accuracy depending upon available information of the building environment. A very basic prediction would be to use a google map overlay of the site with perhaps a few walls drawn in and estimating building materials whilst a more complex prediction may take the up-to-date or as-built architectural drawings containing extensive building material data. Important to note: CPI 2.2 accepts PNG, JPG and GIF file formats. AutoCAD imports are also possible but only once converted to the appropriate file type. Once your building map is imported you are ready to open up the planning mode and hop to it.
The screenshot below shows parameters being manipulated such as Received Signal Strength Indicator (RSSI) and Channel width along with the many other options; which standards to support, anticipated AP model, angle of install (ceiling vs wall) and external antennas. Prime also allows you to export these surveys in PDF format which is quite useful for providing to customers with some proprietary blurbs on what they’re actually looking at.
Performing a predictive survey is fairly easy and realistically should only be performed to give a customer approximations of the potential hardware required to provide their desired coverage. Whilst they are also free to base the final design off the predictive survey it is not recommended and is often a case of you get what you pay for. It would not be surprising for further work to take place following an installation based on a predictive site survey.
The next one up on the price chart is a passive survey. Ideally using some site of the survey tools mentioned previously you will wander the halls of the building with a WiFi adapter or two configured to simulate your customers’ laptops, tablets and other devices in the pursuit of rogue APs and areas where coverage is not meeting requirements. This type of survey is popular given that a lot of information can be gathered before or after an installation without spending significant periods in each area validating active settings. It is key that the devices intended for use are actually simulated (or used) when performing the survey as you could run around tuning things for Macbook pros when your customer is only going to use some low powered tablets. Similarly you’ll want to match up the frequency, modulation rates, cell overlap etc when performing the survey.
From a Cisco standpoint the Cisco Spectrum Expert tool is available for analysing the layer 1 (physical) environment and checking for spectrum interference across the buildings. You’ll need to set this up on your chosen laptop with specific equipment and I’ll admit to never having done this! That doesn’t mean the tool is not usable however as you can turn any lightweight AP into “SE-Connect” mode and connect to it to use the tools.
I’d certainly expect (and have seen) questions on types of analysis in the exam with focus on:
- Standard WiFi usage
- DECT Phones
I think another post may be needed to cover explore this topic it a bit further!
The final survey type is an Active one which will yield the most data on a deployment. Typical devices (or adaptors configured to emulate them) are associated with the live network and a slower walk around takes place to gather more data on associated rates, SNR, RSSI and RF interference. Adaptors can also perform throughput testing and detect levels of packet loss and re-transmissions.
Throughout any of the surveys you are able to determine different levels of how APs will be placed. Predictive surveys are a very efficient way to plan your deployments and the Passive/Active surveys then help you validate this. If there are hard to reach places (lift shafts) or special requirements for directional coverage (shopping aisles) an external antenna may be needed along with an external AP supporting said antenna. See the reference guide link at the bottom for more info.
A lot of site survey work comes down to experience and learning from mistakes. There are a multitude of ways you can prepare and the main things Cisco want you to focus on are defining the scope of work properly with your customers, using the right tools to plan and validating post-install using good methodology. Referring to the vendor notes is always recommended to help with planning a deployment.
Cisco Spectrum Expert User Guide: http://www.cisco.com/c/dam/en/us/td/docs/wireless/spectrum/expert/users/guide/Cisco_Spectrum_Expert_User_Guide_Version_4-0.pdf
Cisco Spectrum Expert (WLC Config Guide 8.0): http://www.cisco.com/c/en/us/td/docs/wireless/controller/8-0/configuration-guide/b_cg80/b_cg80_chapter_010000111.html#ID1163
Cisco Aironet Antennas and Accessories Reference Guide: http://www.cisco.com/c/en/us/products/collateral/wireless/aironet-antennas-accessories/product_data_sheet09186a008008883b.html