Virtual and physical WLAN site surveys
Many designers using radio frequency (RF) have experience in the very high frequency (VHF) spectrum (30 to 300 Mhz). VHF behaves much differently than RF in the C and S bands, in which Wi-Fi operates. Wi-Fi is considered microwave radiation, which propagates best in line-of-sight applications. RF in this spectrum is of very short wavelengths: at 2.4 GHz, the wavelength is about 2.5-in. long, while in the 5 GHz spectrum the wavelength is about 2-in. long. It is a characteristic of RF that lower frequencies (longer wavelengths) have greater range and "punch"-lower frequencies travel further per unit time/power and can penetrate dense materials easier than higher frequencies under the same time/power. While RF follows basic rules of propagation behavior, it does modify based on frequency.
VHF is effective in long-distance communication for the reasons stated above. In order to determine proper antenna and transmitter configuration, a "path study" is performed to determine the propagation characteristics of the shortest propagation path between communicating devices. If a facility has widely separated facilities that must be monitored or controlled, it is essential that the RF path be analyzed for any interference or Fresnel Zone obstructions in order to ensure reliable communication. It is virtually impossible to guarantee a reliable radio link without a path study. Most requests for proposals (RFPs) will list the path study as a required deliverable.
In the Wi-Fi spectrum, a greatly modified and specialized path study called a "site survey" is performed. A site survey differs from a path study in that the survey is localized to a much smaller area—typically within a facility. A path study involves distances, typically in excess of several miles. Within this smaller area, several factors come into play: building materials, other WLANs, microwave ovens, people, and reflective surfaces. All of these are sources of attenuation and interference and must be identified. Two types of surveys are available to the designer: virtual surveys and physical surveys.
A virtual, or predictive, survey allows the designer to predict the propagation and coverage of a network without ever stepping foot into the facility. This type of survey uses sophisticated software to simulate propagation throughout a facility based on known building material characteristics. Most packages allow the import of a floor plan as an image file (JPEG, BMP), PDF, or DXF file. The drawing is then populated with material data identifying walls, desks, elevator shafts, outside walls, etc. The material data introduces the known attenuation characteristics of that particular material of construction. After the drawing is fully annotated, the access points (APs) are placed on the drawing in the desired locations. The simulation is then run, and the expected propagation at the desired power level can be observed. Below is an example of a predictive survey for a water treatment plant using two APs.
The RF energy drops off in a predictable fashion in Figure 1 and is shadowed by the structures that are present. This plot is for the 2.4 GHz ISM spectrum and shows more than half of the plant process area having adequate coverage. If our intention was to cover the entire plant area, then it is apparent from this survey that additional APs are required, or, as it appears, the existing APs need to be repositioned. Instead, let’s add three more APs:
This image shows the effect of installing five APs on the same footprint. Unfortunately, this predictive software package does not allow for the antenna propagation to be simulated so the prediction is somewhat skewed. Using directional antennas to "sculpt" the coverage could eliminate the need for at least one AP while still providing RF coverage to the entire plant area. It should be noted that this survey was done with all APs transmitting (simulated) at 100 mW (20 dbm). The coverage areas shown can also be modified and their area diminished by lowering the power. In conjunction with directional antennas, RF coverage can almost be precisely placed within the process area.
Physical, or site surveys, require that the technician or engineer be on-site to measure the RF propagation in real time. These surveys are often done to prove the predictions of the virtual survey. While the predictive survey is used to get a preliminary picture of the expected propagation patterns of the system, the physical survey measures the actual RF propagation patterns that exist on the site. This requires that certain equipment be acquired and used for the survey. Many of the same predictive packages can be used to input real-time data as observed and measured in the field. The physical survey can become very sophisticated depending upon the customer requirements or application. For applications involving inventory control or custody transfer, the failure of a radio link could result in losses and costly litigation. Your reputation is not worth skimping on equipment and software; without solid, observed data, the risk of an unreliable wireless network increases exponentially.
To do a physical survey, the designer needs to have at least one AP with adjustable power output (two would be better); a spectrum analyzer to detect the presence and strength of any other RF source, as well as the signal strength (return signal strength indicator (RSSI)) of the test APs. Several Wi-Fi utilities are also available free or at little cost that allow the designer to plot the coverage and RSSI of the test APs, producing a coverage map in the process.
It is important that adequate signal levels be maintained at the limits of the AP coverage area—this is the actual limiting factor. Adjoining AP coverage areas should overlap one another to ensure adequate signal strength and signal-to-noise ratio (SNR) to allow for seamless roaming between APs. Roaming happens when the client device disassociates from the first AP and associates with the second as the client moves from one area of coverage to another. This is triggered by the SNR—when the SNR drops below a characteristic value, the transfer occurs.
As far as possible, try to conduct a physical survey during normal operating hours, particularly in a carpeted environment where many bodies are present. Recall that water is a very efficient attenuator of RF; the human body contains between 55% to 75% water. One well-placed body can reduce the signal by as much as 3 dB. A group of people in an auditorium or classroom can significantly attenuate RF. Proper placement and quantity of APs can mitigate this problem by placing the APs at the ceiling in each corner of the room rather than directly blocking the devices if placed on a desk or wall.
In an industrial setting, conducting your physical survey while the day shift process machinery and personnel are active will allow the designer to determine the amount of electromagnetic interference (EMI) present as well as attenuation due to moving equipment and workers. Motors are very noisy devices, and if variable frequency drives (VFDs) are being used they can wreak havoc on a wireless network (any network, really) if not isolated properly. Overhead cranes and heavy mobile equipment also present unique challenges to the designer; these moving obstructions are sometimes impossible to design around owing to their mobility. One possible answer is to provide for several alternate propagation paths. The bottom line is that every installation is unique and needs to be physically confirmed to ensure proper operation—not unlike any other system design.
It is possible, however, to forego a predictive survey and design the WLAN based upon empirical field data alone. Using the same equipment as listed above, a designer or technician can place APs based upon what is observed in the field. One time-honored technique is to place an AP in a corner of the facility or room and move away from the AP while observing an RSSI; when the RSSI reaches -85 dbm, for instance, that is where the first AP should be placed (not in the corner). The next AP can be placed at the point where the first AP signal reaches -75 dbm to allow for overlap and away from the first AP. Also, each AP should be assigned to a different channel to avoid interference.
Depending upon how much risk you want to assume, or due to budget constraints, the AP power level should not be set at maximum during this operation (or during a survey). A setting of 75% of maximum should be used to allow some room for movement later on if the configuration changes for any reason, but could require additional APs to effect total coverage. Understand this is a "quick and dirty" method of network design that is fraught with possible failure. It is mentioned here only to illustrate how survey methods can be used to get a network up and running quickly if need be. It is not a reliable design method and cannot be considered to produce a robust and available network. One can simply put an AP in the center of the room and hope for the best—but do so at your peril.
Performing a full predictive and follow-up physical survey is a sure way to design a reliable and robust network, definitively identify all sources of interference and attenuation, and ensure adequate coverage and capacity to the entire facility. The data collected from these procedures can be used to justify design changes or enhancements and will often save money by showing the most efficient design configuration. The objective is to accurately predict the patterns of RF propagation within a given facility for maximum effective coverage, which translates into adequate capacity, while minimizing spillover into neighboring spaces and causing possible interference.
– Daniel E. Capano, owner and president, Diversified Technical Services Inc. of Stamford, Conn., is a certified wireless network administrator (CWNA); email@example.com. Edited by Chris Vavra, production editor, CFE Media, Control Engineering, firstname.lastname@example.org.
www.controleng.com/blogs has other wireless tutorials from Capano on the following topics:
WLAN design preparation and needs analysis
WLAN design basics and wireless network considerations
IEEE 802.11ah, energy efficiency, extended battery life for wireless devices
www.controleng.com/webcasts has wireless webcasts, some for PDH credit.
Control Engineering has a wireless page.