Are your customers getting the most from their wireless LANs? Sadly, many IT managers and system administrators are discovering that their wireless networks are inefficient and unreliable. Like most technical challenges, these inefficiencies occur because network designers used incomplete data during their initial design processes. In fact, many wireless network design methods may actually create more problems than they solve.

While the typical WLAN design is sufficient for basic e-mail and Web access, this approach fails for high-bandwidth applications such as voice. Introducing these applications over a poorly designed wireless network can degrade quality of service, cause disruptions or even produce outages. You may think adding more access points solves the problem, but how much equipment should you buy and exactly where should you place it? What are the important issues you must consider when designing a network for demanding wireless applications?

DESIGNING FOR COVERAGE

Wireless coverage is the area around an AP within which a client can connect to the network. The importance of coverage really depends on the type of applications you are running. For example, if a user moves from his cubicle to the conference room, his e-mail connection may drop due to a coverage hole a minor annoyance. Dropping a voice call, however, would be unacceptable, especially if the call was to a key customer.

When planning a wireless network, you must first consider which areas you want the network to service. Do you want coverage in the parking lot, or just the office cubicles? Is it important to cover the stockroom as well as the produce aisle? Do you want coverage in the hospital lobby as well as the nurses station? The answers to these questions are usually the starting point for most WLAN designs.

Because APs use radio waves, there is a limited coverage area surrounding the AP. You may discover most AP manuals show circular coverage areas of a specific radius. Unfortunately, this is misleading because AP coverage is not typically circular or uniform. Why? Walls, tinted glass windows, metal shelving, ceilings and other physical obstructions can block or impede the wireless signals a critical factor often overlooked during the WLAN design process.

Many wireless networks are deployed using trial-and-error AP placement. As coverage holes are detected, network designers add new APs a costly and unpredictable process. Some designers place APs in a regular pattern or grid to eliminate coverage holes. Unfortunately, this approach does not take into account physical obstructions, and may provide WLAN services to areas where you may not want them, like outside the building.

Another commonly used WLAN design method uses site survey devices to measure wireless signal strength. Network designers install test APs, then walk around measuring the coverage area. In reality, this technique is exhausting and does not predict precise AP requirements.

Keep in mind site surveys are only as good as the collected data. If you require coverage in a given area, you will have to collect readings in that area. Otherwise, the obstructions in the physical environment may block wireless signals. Since the physical environment plays such a major role in determining coverage, the site-survey approach falls short.

Even though you may have 100 percent wireless coverage in your target area, your WLAN may not have enough capacity to keep users connected and productive.

DESIGNING FOR CAPACITY

There is another challenge when planning a wireless network: capacity. Wireless network capacity is the maximum amount of data or number of users that can simultaneously be on the network. The more users on the network, the more data you can expect the network to carry. As with cellular telephone networks, if the amount of data the network must carry exceeds its capacity, problems such as packet loss, timeouts, or loss of connectivity will occur.

Because each AP has a maximum bandwidth, the total bandwidth divides across the number of users currently using that AP. Each new wireless device connecting to the AP causes the total bandwidth available for each user to shrink. This is especially critical when users add high-bandwidth applications such as voice or streaming media.

Designing for capacity is more complex than designing for coverage. Of course, all of the same challenges posed by the physical environment still apply. Then you need to consider three things: the number of people the network will service, the user locations, and what applications they will use.

For example, lets say that you expect 20 engineers in the cubicle area will need e-mail and Web access, while five users in the sales office will need voice service. Since all applications require a certain bandwidth, the number of application users in a given area defines the total bandwidth requirement. To satisfy high-bandwidth requirements, you can place APs with overlapping coverage, thus increasing the available bandwidth within the overlap area. However, this also introduces the chance for intranetwork interference as APs whose coverage overlaps can interfere with one another if you do not take the proper care in assigning channels to the APs.

In summary, iterative trial-and-error AP placement and exhaustive site surveys make the traditional WLAN design process unrealistic for networks targeted for high bandwidth or real-time applications. In fact, you may find that the labor costs associated with these iterative approaches may be higher than the actual WLAN equipment costs.

Fortunately, there is a wireless network design solution that allows you to predict coverage and capacity before deploying hardware resources.

TAKING CONTROL

A better way to solve the WLAN design challenge is by using a predictive design approach that leverages knowledge of the intended use of the network and the physical environment. Decades of research have led to the development of predictive design software for simulating the coverage area and capacity of wireless devices when physical environment data is available. These techniques are the most effective way of planning AP placement that guarantees WLAN coverage, signal quality and capacity.

With this approach, the WLAN design process begins by importing data about your facility into the software. Import formats include CAD files, scanned images, and digital pictures. Next, you take into account physical obstructions by tagging each wall in the facility map with material types such as sheetrock or brick from a library of common building materials. Then, you place APs within the site map and simulate the coverage areas. The software predicts the effects of physical obstructions and visually displays a coveragestrength map. The network designer can adjust the design to address any identified coverage holes.

Most importantly, this software makes design for capacity efficient and predictable. To do this, you simply indicate expected client locations and application requirements within your environment. With this information, the simulation engine predicts whether the network capacity is sufficient given the current equipment placement and configuration. This allows you to quickly determine whether your existing network design can sustain high data rate applications and meet the needs of your subscribers.

Figure A shows the designation of the unique user requirements. The red area will serve 200 Internet users in a conference hall. The yellow area will need to handle four users with steaming video, and six employees using e-mail and the Internet will be in the blue area.

In Figure B , the software has evaluated the impact of both the building environment and the user requirements to recommend the best placement of APs in the facility. The color contours represent different signal power levels; green indicates the signal will be sufficient to service the applications and users in the various regions.

With predictive design software, you can solve your WLAN coverage and capacity challenges by generating design guidance based upon location, number of users, bandwidth requirements and physical obstructions. Now you can design and deploy your wireless network with confidence and ease.

Roger Skidmore is vice president and chief product officer for Wireless Valley, which provides software for the design, documentation, monitoring and management of wireless networks.