Types of Site Surveys

Among network engineers, RF knowledge has been something that is typically not understood deeply. As WLANs have evolved, there has been a push to simplify deployment, making it easier for non-RF engineers to install and manage an RF WLAN network. Part of this push has been in deployment tools, or site survey tools. Today there are three general categories for such tools, as described in the following sections:

Manual
Automated and assisted
Theoretical

Manual

The manual site survey has been the main method for determining how to deploy a WLAN. This type of survey requires physically being at the location and taking actual RF readings throughout the site. Using logical methods for working through the site, an AP is placed in the site with the associated antenna, and a physical walkabout is performed.

This type of survey is by far the most accurate because it provides real data transfer in a live environment in a manner similar to that of the actual WLAN user. The survey engineer moves from location to location (within the site) and takes readings to verify RF connectivity.

The downside to a manual survey is the time required to walk the site, move APs from place to place, and verify the coverage. In addition, understanding the results of the survey requires a high level of RF knowledge.

Automated

In an attempt to make installing WLANs easier (and more attractive to those with little or no RF knowledge), the concept of eliminating the necessity for a site survey altogether is being promoted heavily. In an ideal situation, a network engineer could walk into a site, drop APs in the site based strictly on location and density of users, and have the network "self-configure." A great idea, but automated site surveys require some on-site survey work to prevent either putting in too many APs (and increasing the overall system cost) or leaving holes in the RF coverage.

Some automated survey vendors claim that APs that must be located (in the site) and configured manually are bad. Instead, some management system could identify the requirement for RF coverage. And after the APs have been physically installed, this system could then configure the APs to provide the desired coverage.

But, one question continues be raised to those who have looked at this concept in depth: Where do I physically place the APs, and what antennas should I use? The concept for the automated survey is that you just place APs wherever it makes sense to achieve the type of coverage desiredcoverage based on performance, high availability, resilience, and so onand standardize on one simple antenna. In some cases (such as the "coffee cup" survey discussed in Chapter 8, "Discovering Site-Specific Requirements"), this might work just fine. For many types of sites, however, it will not work efficiently or economically.

Consider a warehouse or factory floor where there are certain types of obstacles impeding the RF coverage paths. Directional antennas and critical placement of APs is a requirement. The automated site surveys are not intended to work for these types of installations.

Consider also that sites that need a maximum coverage area but not necessarily maximum bandwidth, such as for data acquisition, RFID, or bar code scanning, present problems for an automated site survey. One common thread is needed for automated site surveys: user density. One vendor describes it as overengineering, or placing more APs than needed in the site. The WLAN management station can just turn down the power of the APs to achieve the required overlapping coverage. If the goal is to have maximum coverage, this method is not economical and can elevate the cost of the WLAN as much as double or triple the actual requirement. The argument here is that the cost of WLAN gear is less than hiring a wireless solution provider to perform a manual survey.

Another downside to most automated survey systems is that there is no feedback from the client. The system uses RF signal information received by APs, from the surrounding APs, to determine what the power levels should be set to. It does not provide a guarantee that there are no dead spots in certain areas, especially along the perimeter of the site, where there may not be another AP there to evaluate the RF levels.

One possible advantage to a system that offers automated surveys is that it provides some ability to reconfigure the RF if the environment changes. If walls are added, new equipment is brought into the factory, or inventory type changes for a certain location, the system might be able to compensate for these changes. However, this can happen only if the APs were installed properly in the first place and the APs are not currently running at full power.

Assisted

Assisted site surveys are a mix between a totally manual survey and a totally automated survey. Assisted site surveys take advantage of the best features from both. However, in doing so, you still have some of the downsides from both methods as well. Some manual surveying is still required, and some minimal RF knowledge of antennas and RF propagation is needed for the initial design. In addition, assisted site surveys are not suited for some sites due to lower RF coverage and the resulting higher cost of systems installed using automated or assisted survey methods.

The concept here is to take a manual survey process and determine optimum coverage in the site for each different type of area. The engineer takes manual measurements in selected areas and then logically extends those results to the remainder of the site. Then a configuration system is used to test and make preliminary adjustments of the APs. Finally a walkabout is done with a client that will work in conjunction with the configuration system, passing data to the configuration system on overall link performance, so that the configuration system can make final changes to the APs configuration.

Although this method comes closer to manual survey results, in most cases it will still result in somewhat more APs than a totally manual survey will, because part of the facility will be overengineered to guarantee coverage. The final walkabout is done to ensure no dead spots.

The assisted survey only works in areas where the user density or application bandwidth requires the AP's coverage area be lower than the maximum capability of the AP and the areas to be covered are somewhat uniform.

Theoretical Surveys

A few companies have developed tools that provide theoretical RF coverage plans for WLANs. For the most part, these tools are intended to remove the necessity for walkabout surveys.

To use a theoretical tool, input such as an aerial photo or scaled or scanned drawing of the site is required. Some permit the inputting of standard graphic files such as BMPs or AutoCAD drawings, whereas others might require the user to actually draw the facility in the supplied software package. The site needs to have a description of the construction and contents, as well as the attenuation factors entered into the system, and specifics about the radios and antennas that are going to be used.

From there the tool enables the design engineer to interactively place APs with a click of the mouse, and the system will determine theoretical RF patterns for the site. The resulting coverage and capacity can then be viewed in a graphical representation.

Most of these tools have already loaded the specifications for the more popular WLAN products, as well as a large assortment of antenna specifications. If the antenna and radio specification are not provided, however, this can be a big issue. Antenna polar plots are not easily obtained, and certain characteristics for radios can be difficult for non-RF engineers to understand.

These types of tools also usually offer features that keep complete and easy-to-use records of all technical and maintenance details for each job. Because most indoor and campus networks involve hidden APs and wiring that are not easy to find and can be lost over time, these graphical model documents can prove very useful for troubleshooting. This is a great aid in the final documentation stage (as discussed in Chapter 13, "Preparing the Proper Documentation").

Although these types of tools actually do work fairly well, they come with a drawback: higher overall cost. First is the initial purchase price of the tool; they tend to be very very expensive. And second, you need to input every piece of the site into the map (including building-wall construction and contents, and attenuation factor assignations for every piece). An experienced survey engineer can usually perform the survey in less time than it takes to import every detail, run the program, and then perform the walkabout verification.

Although these tools make accurate modeling of the site propagation, link budgeting, and modeling of the RF performance a closer reality, successful completion of these tasks is more than just a simple click of the mouse away. They come at a cost, both in dollars and time. And nearly all the theoretical tools recommend some type of final field walkabout to verify the installation.

Types of Site Surveys