RF Surveys: How do you increase AP density?We often undertake jobs where there's an existing layout of Access Points. It can be at a variety of densities when we start:
Sometimes it's at the lower end (what I call 'hotspot grade') - it only works in specific spots, you can't roam and if you try, devices drop out.
Sometimes it's at a decent Data Grade - it will support laptops anywhere, phones in most spots, but devices probably don't roam perfectly
The challenge is to increase the AP density from either of these up to voice grade (supports WiFi voice and lower powered devices - they can roam reliably), or RTLS (highest grade, supporting a density of APs to triangulate devices) while maintaining the existing AP locations but supporting the grade that was asked.
Methodology for increasing AP Density
We often get asked to come to a site and survey and advise on where to add the extra APs. The problem is, for sites with low density of APs, this is a flawed approach. We can survey, sure, we'll find out how the coverage is, but with the APs all running at maximum power.
Assuming that TPC (Cisco) or ARM (Aruba) are running, APs will be running at maximum power. So, we've surveyed, the APs are running at 200mW, the coverage looks great, but devices are dropping out. We see this all the time.
If the AP density is pretty close to where it needs to be, a post install survey and addition of APs based on this data is a good way to go.
If, for example. the AP density is low, it's too hit and miss. We need to go back to first principles, and design it properly. This means a desktop survey - a predictive model is undertaken and we can use the existing AP locations and add more APs.
This means the customer can retain the value in the existing cabling and we work out how best to add the new APs. The use of predictive modelling may seem counter intuitive when APs are already deployed, but it allows us to quickly test a variety of AP placements, to optimise the design.
No site survey needed then?
This isn't quite true either. We just use it in a different way. Every desktop survey should be counted with an onsite survey. For a greenfield site, we'd bring a rig onsite with a powered Access Point and measure with that. In the case of pre-existing, we can survey and examine the RF drop on either side of the walls - this allows us to confirm the wall density, which feeds back into the predictive model to make it a more accurate representation.
The summary is that for low density deployments, we can't necessarily just do a post-install survey and come up with a magic answer. It needs some more detailed design to come to that - so a blend of post install and predictive survey is needed to really produce the optimised design.
Contact us for more details on sales.