In a typical wireless LAN, a wireless client connects to an AP, and communication takes place between the client and the access point (AP) as the user browses the Internet, sends and receives e-mail, or talks to someone on a wireless IP phone. This communication includes an uplink path from the client to the AP and a downlink path from the AP to the client. For example, when a user opens their browser, the client device sends a URL page request through the uplink path to the AP, then the Web pages are sent through the AP to the client over the downlink path. Another example is when a person -- let's call her Mary -- is talking on a wireless IP phone to Bob. In this case, Mary's voice packets flow over the uplink path from her IP phone to the AP, and Bob's voice packets eventually travel over the downlink path from the AP to Mary's phone.
At least, this is what we hope occurs; otherwise, none of these applications will work properly.
Traditional Testing Techniques
APs periodically send beacons, which travel over the downlink path from the AP to the client devices. Most Wi-Fi site survey tools receive these beacons and display the signal strength and SNR (signal-to-noise ratio) associated with the beacon signals. A person performing a radio frequency (RF) site survey determines the optimum installation location for APs by utilizing the beacon signal strength to determine where adequate coverage is provided.
For example, you might define the target range boundary to be 20dB SNR. You then use a test tool to ensure that there is at least 20dB or better SNR throughout the covered area.
Keep in mind, however, that this is only in relation to the downlink path. It doesn't take into consideration the uplink signals from the client devices.
Limited Uplink Reduces Range
Something to consider: most APs have a significantly higher EIRP (equivalent isotropically radiated power) -- the transmit power plus antenna gain -- compared to wireless clients.
APs, for example, are typically set to their highest transmit power, usually around 100 milliwatts (mW). This is done to maximize the signal propagation and coverage from each AP, which minimizes the number and costs of APs. Wireless clients, though, tend to have a much lower EIRP due to smaller, lower-gain antennas and the desire to conserve battery power.
In this situation, the downlink signal strength will be relatively high, and the uplink signals much weaker. Thus the effective range between the AP and the client device is governed by the uplink signal strength. As a result, the use of only AP beacons (i.e., downlink signals) for determining coverage leads to much better coverage than what may be available when the clients interact with the AP. Nothing is wrong with the downlink communications, but weaker uplink signals limit the effective range and likely disrupt communications when client devices move into areas where uplink signal strength is not good enough to support communications.
Tips to Consider
Identify the client devices. Start by determining which client device has the lowest EIRP. You can do this by looking at the specifications. Be certain to take into consideration the transmit power that you'll be using (it may be possible to turn the transmit power up higher than the default value, which will help) and the antenna gain.
If the specifications don't define the EIRP values, do some testing. Associate each client to an AP, and log into the AP to view the association signal strength of each device. Do this test with all the client devices located at the same distance from the AP. Identify which has the lowest association signal strength. Make sure that the client you're planning to support will ensure two-way communications in all covered areas for the client device having the lowest EIRP or weakest uplink signal when associated with the AP.
Base range measurements on the weaker uplink signal. If the EIRP of the client device is lower than the AP that you plan to use, or the signal strength of the client device at the AP (viewed by logging into the AP) is lower than the signal strength of the AP beacons at the client device (measured by the client device or signal measurement tool located next to the client device), take the uplink signal into consideration when performing signal coverage testing. Be sure to use the weaker wireless client if the network will support multiple clients.
Perform signal testing by measuring the uplink signals. To do this, measure a client's uplink signal strength as you walk with the device throughout the covered area. You'll need to log into the AP, probably with a laptop (through a Web browser), to view the signal values.
In addition, you'll need to periodically have the client device send something to the AP and refresh the AP display to see signal updates as you move the client device about the facility. In some cases, the weaker client device and the laptop will be the same device. This can be cumbersome to perform if you need to carry a client device and a laptop while interacting with the AP's management screen. Use a cart with wheels. Or a friend.
Perform signal testing by measuring the downlink beacon signals with a calibration applied to the measurements. In order to use test tools made for only measuring the downlink signals while performing an RF survey, figure out a calibration factor first. You can do this by finding the EIRP specs for the AP and weakest client device, and subtract the AP EIRP from the client device EIRP. The resulting value is the calibration factor.
If the calibration factor is positive, then add it to the cutoff that you're using to define the range boundary. This will reduce the measured effective range sufficiently to compensate for the weaker uplink signal strengths.
For example, if you find that the calibration factor is +10dB and you're using 15dB SNR as the range boundary cutoff, then the calibrated range boundary cutoff is 25dB. As a result, you'll need to ensure that there is 25dB SNR in order to have enough coverage to compensate for the weaker uplink signals of the client device.
If you find a negative calibration factor, downlink beacon signal values are weaker than the uplink. With this condition, which is rare unless you purposely turn down the AP transmit power, you don't need a calibration factor. Since the downlink signals are weaker, range boundary measurements based solely on the downlink signals will indicate sufficient coverage for the client devices.
Consider turning down the transmit power of the access points. If you want to maximize performance, consider adjusting the transmit power of the AP to balance the uplink and downlink signal strengths. Turn the AP transmit power down by a value equivalent to the calibration factor. For example, if the calibration factor is +3dB, turn down the transmit power of the AP by 3dB. This will make the downlink equal the uplink signal strength. You'll have to increase the density of APs and increase cost due to a larger number of APs required to cover an area -- fewer users will associate with each AP -- but it may improve performance because of the higher capacity. If you pursue this approach, just measure the downlink signal strength of the AP beacons without any calibration factor.
No matter how you deploy a wireless LAN, be sure to take into account the differences between the uplink and downlink signal strengths. By ignoring the weaker client device signals, you may end up with a network that doesn't provide required coverage, and users won't be happy.
Jim Geier is the principal of Wireless-Nets, Ltd., an independent consulting firm assisting companies and cities with the implementation of wireless network solutions and training. He is the author of the books Deploying Voice over WLANs (Cisco Press), Wireless LANs (SAMS) and Wireless Networks - First Step (Cisco Press).
Article courtesy of Wi-Fi Planet