Physical site surveys are the traditional method for planning wireless deployments of all kinds. Prior to technology advancements such as the Aruba ARM, APs were planned to use static power and channel settings as determined by the physical site survey. Though many of the same principals in planning apply today, ARM eliminates the need to carefully select exact channel and power settings. The three types of physical surveys are passive, active, and spectrum clearing.
The typical wireless passive survey methodology utilizes professional survey software utilities from companies such as AirMagnet and Ekahau to measure existing signal propagation within the designated coverage areas. Due to the many varieties of building structures, designs, and materials that can impact the RF signal, the survey tool effectively captures the “actual” RF signals that originate from the APs.
Figure 33 Passive survey with Ekahau Site Survey Professional Version 4.4
To perform a passive survey:
1. Obtain the current electronic floor plan of the facility.
2. Using the site survey software application, walk through the coverage area and sample the RF path every few feet.
3. Analyze the data to produce heat maps of the existing coverage and to look for sources of external interference.
4. Have an experienced WLAN engineer use the passive survey data to validate the choice of AP locations.
An active AP survey uses operational test APs for many purposes:
• To determine the best placement of the APs
• To determine the AP density necessary for a given building construction
• To ascertain any pre-existing RF conditions that may influence the outcome of the implementation
Follow these steps for each test area:
1. Obtain the current electronic floor plan of the facility, and mark the locations at which active tests are to be performed.
2. Using an Aruba Instant AP, provision the AP with ARM minimum power that will be used in the deployment.
3. Mount the AP to a portable tripod, speaker mount, or other stable, movable platform. If available, you can also mount the AP directly to the ceiling rails.
4. Position the AP at a test location and connect it to a power source and data link.
5. Use a professional site survey application to complete a passive survey of the area that immediately surrounds the test AP.
6. Repeat steps 4 and 5 for all identified test locations.
7. Have an experienced WLAN engineer analyze the active survey data to determine the proper AP density for the coverage area.
The analysis software makes hundreds of RF measurements throughout the test, which are then visualized by superimposing their values in color over the relevant facility map. (See Figure 34.) This section presents sample heat maps at 2.4 GHz and 5 GHz generated with AirMagnet during a retail survey in a grocery store. In this case, five APs were set up in the locations shown on the heat map. Because voice communications are being used, the customer requirement is for 48 Mb/s cell edge data rate (equivalent to a signal-to-noise ratio [SNR] of 20 dBm or -65 dBm minimum signal strength in the 2.4 GHz and 5 GHz bands). Configure AirMagnet with a -65 dBm display filter. As a result, areas that fall below this threshold appear gray and areas that exceed it are in color. Because almost the entire floor is in color at 2.4 GHz, the survey in the figure shows that coverage meets the requirement in that band.
Figure 34 2.4 GHz active survey with AirMagnet Survey 6.0
In the 5 GHz band, a gray area appears in the middle of the store, which indicates that higher AP density is required. (See Figure 35.) This result occurs because Free Space Path Loss (FSPL) increases proportionally with frequency, so radio signals in the 5 GHz band travel approximately half as far as 2.4 GHz signals, assuming constant power. In addition, this part of the grocery store contains freezers, which significantly attenuate the signal. This example shows how the AP density that is appropriate for 2.4 GHz is inappropriate for 5 GHz.
Figure 35 5 GHz active survey with AirMagnet Survey 6.0
By its very nature, the unlicensed 2.4 GHz and 5 GHz spectrum is shared by a multitude of devices that operate in the same frequency space and create interference for one another. This situation can result in poor 802.11 network performance. Common examples of such devices include APs in neighboring stores or warehouses, cordless phones, analog and digital video cameras, Bluetooth devices, and microwave ovens in break areas. When you design a wireless network, it is important to understand the overall RF environment typical of the facility types where the network will be deployed to mitigate any interference problems. Spectrum clearing refers to the use of a portable spectrum analyzer to discover and pinpoint interference sources before the network is deployed. After the interferers have been identified, you should remove or migrate any devices you can to lower the interference effects.
The two methods to monitor the spectrum are dedicated RF software or using Aruba APs and the RFProtect® license to scan the environment.
Figure 36 2.4 GHz spectrum analyzer display
The spectrum clearing process involves these steps in a software solution:
1. Configure the spectrum analyzer to record peak, average, and maximum hold for the 2.4 GHz and 5 GHz bands. If supported, also enable a swept spectrogram for both bands. If the analyzer includes omni and directional antennas, begin by using the omni antenna.
2. Walk a carefully planned route for each selected location and look for active devices. (An active device is any electrical equipment that broadcasts or radiates in the same frequency bands as the proposed Aruba network.)
3. If strong interfering signals are observed, pause in that location and record a spectrum trace for 60-90 seconds.
4. If interferers are found, pinpoint them using these steps:
a. Attach a directional RF antenna to the spectrum analyzer.
b. Slowly rotate the antenna until you see an interfering source of RF energy in the 2.5 or 5 GHz band.
c. Attempt to determine the RF channel number of the interference and whether or not it impacts your proposed network coverage.
d. If it does impact your coverage, move the antenna closer to or farther away from the source of the signal.
e. Using this dynamically changing signal, identify the offending device and determine its exact location.
f. Decide what to do about the interferer (remove it, shield it, or replace it for example).
Figure 37 shows results from a spectrum analyzer that shows the presence of DECT cordless phones in the 2.4 GHz band.
Figure 37 Display that shows 2.4 GHz interference from DECT phones
Figure 38 shows the significant interference effect of a microwave oven in the area.
Figure 38 Display that shows microwave oven interference
If an Aruba Network is in place, many of the same charts are directly available from the AP through the mobility controller. (See Figure 39.)
Figure 39 Aruba spectrum analysis
A set of tools that are very similar to the software solution is available in the Aruba Mobility Controller. Different Aruba AP models have different capabilities with regards to spectrum output and hybrid spectrum capabilities. For a longer discussion of the capabilities of the Aruba Mobility Controller, see the Managing and Optimizing RF Spectrum for Aruba WLANs Application Note available at http://www.arubanetworks.com/vrd.
For this planning exercise, an active site survey was conducted at the Aruba PoC Lab test facility. For details of that facility, see Chapter 5: Example Facility. We will use the plan developed in Case Study: Planning the PoC Lab Facility Using VisualRF Plan on page 37 as the basis for AP placement and as a comparison for the results. We will use AirMagnet Site Survey Pro to gather test data.
To perform the survey, first we must have actual APs for the software to visualize. In our case, our building had been previously occupied and wiring that is still in the cubicles leads back to a common datacenter. In this case, we could easily mount a controller, but because our lab had a working DHCP server in place. Instead we chose to use Aruba Instant IAP-105s for the testing. An Aruba S3500S provided power-over-Ethernet+ (PoE+) for the IAP. We also tested using the IAP, a power injector, and a laptop running a DHCP server to run the setup, which made the design more contained and less infrastructure dependent. Finally we clipped the AP to the ceiling rails, so we had a realistic test of the AP in the position it would eventually be mounted in.
The IAP broadcast an SSID in 5 GHz only using WPA2-PSK. This setup was chosen as the most self-contained method to complete our comparison testing. We also set our IAP to keep power at the minimum level that ARM would use later when assigning power levels to the campus APs. This setting allowed us to confirm that at minimum power levels we would have a solid coverage plan for 5 GHz. More details are covered in Appendix A: Using Aruba Instant in a Physical Site Survey.
We created a new building and planned our methodology for performing the survey. We had special requirements in this particular lab due to the way the lab operates. For instance, the large meeting room at the front of the building is typically used for specific, large demos and runs its own AP set. So we knew that space would have existing coverage. We also have RF cages at the back of the building that will cause shadows, so we did not spend a lot of time in those areas.
When using AirMagnet, we picked a path to walk around each AP and paid attention to signal strength and where the next AP placement will be. As we were looking for voice quality, we watched for the signal to dip below -65 dBm on the AirMagnet display as shown in Figure 40.
Figure 40 AirMagnet signal strength indicator
With each AP placement, we started a new survey path and we saved at the end of each pass. This results in multiple passes, which can be viewed individually or merged into a single display. Figure 41 shows the pattern walked around the location of the first AP.
Figure 41 First pass walking pattern
After the pattern has been walked, we can change the display mode and show the signal strength measured as seen in Figure 42.
Figure 42 First pass heat map
After we complete the first pass, the AP is moved to its next location. As the testing continues, each survey path results in a similar pattern and heat map. When testing is complete, the different paths are merged into a single view. In the merged view in Figure 43 we see the path that the team walked across all the survey passes. Some areas were hit multiple times to account for the different AP placements. Some areas did not receive coverage because the building has been reconfigured since the architect originally drew the floor plan.
Figure 43 Merged survey path
The other thing that can be done is to display the heat map that is generated from the survey data. This map is also presented in a merged form in Figure 44.
Figure 44 Merged signal strength
After the survey has been completed and the data merged, you now must determine if the AP placement meets the requirements of the deployment. In AirMagnet, move the slider representing signal strength to the minimum acceptable level and look for areas of the map that do not meet that level. In Figure 45, we have reduced the signal view to -50 dBm to show how the tool displays coverage. If we had a requirement, such as -65 dBm, we can set the slider to that level and the areas that do not support that coverage level will be grayed out.
Figure 45 Areas that do not have acceptable signal strength
Site surveys are time-consuming projects. The survey of this building took approximately four hours to complete with a two-person team. You should expect that surveys will take many days or weeks depending on the size of the buildings, number of buildings, mounting options, and building materials.