Optimizing Data Communications with Contradictory Requirements

As mobility has increased, communication over wireless networks has taken its place in industrial environments. Initially concentrated mainly in warehousing and logistics areas, more and more WLAN installations are being used to provide communications in production and office facilities as well. This expansion is driven not only by the introduction of suitable WLAN security standards, such as WPA and IEEE 802.11i, but also by improved network availability due to mature hardware. According to one user survey, some 50% of the wireless networks currently installed in businesses have more than 10 access points. This complexity makes professional and efficient planning of the WLAN installation indispensable.

A poorly planned network will suffer from gaps in coverage, areas with poor signal quality, slow response times, poor voice fidelity and disconnected calls in telephony over WLAN, and inadequate throughput capacity overall. Because appropriate simulation software for WLAN planning has been quite expensive up to now, many wireless networks have been optimized by setting up individual access points and measuring the results, or by walking through the coverage area only after installation and measuring signal strength. A number of tools are available for such approaches, such as the WLAN planning and testing kit from Psiber Data (Figure 1), which can be used to test access points even in unfinished buildings thanks to its built-in power supply.

However, depending on the actual applications the WLAN is to support, adequate planning needs to take into account special and partly contradictory requirements. These can only be addressed by a simulation of the complete wireless network. Whereas for office applications it is usually sufficient to provide the actual workplaces with moderate data rates, areas such as warehouses where wireless barcode readers are in use require wall-to-wall coverage. In such cases the wireless cells should spatially overlap to ensure sufficient coverage even if one or even several access points should fail. Yet such overlaps can cause interference — both co-channel and near-channel interference, since adjacent channels in the 2.4 GHz band actually overlap in the frequency spectrum.

At the low data rates needed for barcode applications, such interference is not particularly troublesome, since the few data packets transmitted are rarely simultaneous. In telephony-over-WLAN applications on the other hand, because maximum communication bandwidth is needed, interference is important. To provide the necessary bandwidth in this case, the data packets generated by the individual phone calls must be distributed among many access points, and yet the access points’ range, i.e. their transmitter power, must be kept low to prevent interference. Hence such a network consists of many small wireless cells with little spatial overlap.

Most applications lie somewhere between the two extremes just described. In hospitals for example, WLANs are used to locate RFID-tagged doctors and medical equipment: this requires coverage by several access points at every location in the building. Furthermore, in addition optimizing the WLAN for its specific application, planning should also minimize the signal escaping the site in order to prevent unauthorized network intrusion.

Optimum Radio Coverage Through Simulation

Efficient planning can meet these challenges by simulating the wireless network and modeling its applications. Specialized software products are available for this purpose, such as the RF3D WifiPlanner from Psiber Data (Figure 2). This new, low-cost software is the first to provide true three-dimensional computation of the signal strength distribution, taking into account the interference from access points above and below each floor in multi-story buildings, for example. Moreover, the fast 3D algorithms permit on-screen interactive simulation. This means that the changes in the distribution of WLAN signal strength as access points are added or moved can be seen directly on the screen. This allows the user to optimize the number and configuration of access points in the network. Ironically, the most common mistake in WLAN planning is the use of too many access points, which often leads to unnecessary interference between them due to insufficient channel separation. And incidentally, saving even one access point pays for the cost of the planning software.

In planning a wireless network with the RF3D WifiPlanner, a certain sequence of steps should be followed in order to attain the objectives quickly and easily. First, the floor plans of each building level are imported in the form of image files. Because such graphics are often available as manual scans or even digital photos of the building’s emergency escape route plans, they must first be rotated and scaled so that the software can align them vertically. Then the load-bearing walls, partition walls, and ceilings are added to the plans using data from an integrated library. Then the access points, with their specific types of antennas, can be placed in the building using data from another library. The software computes the resulting signal strength in real time and displays it on the screen. (Figure 3)
Figure 3: Visualization of wireless coverage by several access points in one building level

Next comes the most important step: optimizing the WLAN’s coverage and availability. For this purpose, the software can display not only signal strength, but also other key network parameters in an appropriate color scale throughout the building plans: co-channel and near-channel interference, signal-to-noise ratio, data throughput rates, and even worst-case simulations showing the coverage that would remain at each location in the building if the one or two strongest access points for that location should fail. Also during the optimization step, the data load on the simulated network can be set as desired to visualize the effects of interference on throughput rates and coverage. The optimum network configuration for the anticipated conditions can then be determined by adding, removing and relocating access points, and by using and adjusting the appropriate antenna types.

Finally, the resulting displays are printed out in report form as the documentation of the planned wireless network. This documentation can be used both as a specification for outside tenders and as installation instructions for technicians.

Not only mature components and new standards, but also modern planning and diagnostic tools make WLANs suitable for industrial use. This in turn permits more efficient mobile working in administration, production and logistics through continuous network availability with sufficient data rates throughout the company.

About Psiber Data

Psiber Data Systems Inc. was founded in 1994 in San Diego, California. Psiber Data develops, manufactures and markets test instruments for local-area networks. Its products include LanMaster®, Pinger® and Cable Tracker®, used for fault detection, diagnostics and performance testing. Visit Psiber Data on the web at www.psiber.com.