Close

HPE Aruba Networking Blogs

Getting the Design Right…The First Time

By Scott Lester, Blog Contributor

You may have heard someone say wireless is easy. This is because wireless is a very forgiving protocol with built-in mechanisms designed to provide a resiliency. Wireless can function adequately in difficult situations. I’ll help you dispel that rumor now by saying that wireless isn’t easy.

Designing a quality wireless network is hard. It takes years of experience and several less-than-perfect designs before the wireless design engineer learns a fraction of the things needed to accomplish what many see as a very easy task.

There are many questions to ask during the design process, some of which you might find yourself asking whether you are the design engineer or customer. If designing networks is so difficult, how does a designer ensure the design being implemented is correct? How do they avoid having coverage gaps or lacking the secondary coverage required in the design? If only there was a wireless survey method that could help avoid these issues.

Predictive surveys have long been a mainstay in the wireless industry. Without them it is a guessing game for where to place the APs. While accomplishing the goal of providing a quality network without predictive software is attainable it is certainly much easier with that software. However, a predictive design is only an educated guess. It is only as accurate as the information put into the design software. Things like correct wall attenuation, accurate physical measurements of the space and ceiling heights, placements of impediments like warehouse shelves, and clearance above those objects can affect the accuracy of a predictive design. While everyone understands predictive surveys and the value they bring to the design process, there is another survey type that is often bypassed or overlooked.

This survey type involves onsite testing in a method known as “AP on a Stick" or APoS. It often provides invaluable feedback to the design process that helps meet the goal of getting the design correct the first time. Let’s take a look at how an APoS survey works.

An APoS survey is completed by placing an AP (preferably the one that will be used in the final network implementation) at several different points within the space that have been identified as AP locations by the software.

Once the AP is placed at each position and configured with the appropriate power output settings, the engineer uses survey software containing the same scaled map used to produce the predictive design to passively survey around the AP.

An APoS survey in healthcare. Source: Author

There will always be things that affect how the wireless signal propagates in every environment that won’t always be known to the design engineer. An APoS survey will help identify and allow corrective action to be taken prior to network installation. You may have already figured out that this process might cost more for the project and you would be correct.

The time involved in completing an APoS does add to the overall project costs. However, those costs can be easily justified by the potential savings of time and money if the implemented design doesn’t meet the network requirements. The APoS method is critical in assisting the engineer in producing a quality wireless network design, that gets it right the first time.

Recently, I found myself in such a situation where an industrial customer wanted to replace an aging wireless system in its warehouse that had been designed using overhead AP placements. While overhead coverage isn’t a bad thing, in this instance, the overhead APs were attempting to provide coverage within the warehouse shelving units that were approximately 30 feet tall (9.1m for my metric fans) and contained inventory made of galvanized steel, PVC and heavy rubber. All of the inventory items were materials that added several dB of attenuation, causing the wireless signal to diminish immensely by the time it reached user devices.

How could it be possible that predictive software didn’t pick up on this? The software simply didn’t have enough information to be able to estimate the signal levels properly. Without an APoS survey being completed, the customer would have purchased a design that contained 3x fewer APs than were required in the final design. Not only would an incorrect number of APs been sold, but it would have been a very poor design as the APoS survey helped determine that an overhead coverage model was not adequate for providing the level of coverage and performance required by the customer.

Ultimately as a design engineer, I always try and follow the mantra of “trust but verify.” By utilizing the APoS survey method, the designer ensures that the predictive design is trusted and verified. Far too often, wireless networks are installed with poor designs as there was a lack of information used during the design process.

As a wireless professional, I can’t stress enough how important it is to have as much information available during the design process. The data gained by performing this extra step of an APoS survey during the design process is extremely valuable. Don’t forget! A predictive design is only as good as the information put into it, and more is always better in this case.

-Scott Lester, CWNE #253

Questions or comments? You can find me on Twitter @theITrebel or visit my blog!