Omaya™

Factor Affecting Accuracy

51 views August 11, 2017 April 24, 2018 admin 0

User Devices

Is user device connected to
Wi-Fi?
Much better! More RSSI samples for Omaya™ to eliminate errors. But sometimes, the client can change transmit power frequently,
throwing accuracy off at those times. Omaya™ cannot plan or account for such variations. Clients not connected to Wi-Fi, only probe once in a while (average once per minute, varies by manufacturer and model). That implies less data sample for location engine to work with. Refer to Device state at terminology.

AP Deployment Design

Designing Wi-Fi with location accuracy in mind? In many situations, location requirement is an afterthought. It is an add on to an age-old AP deployment that was deployed with the objective of providing good Wi-Fi coverage. Customer does not plan to add APs or change AP placements as those costs are budgeted for. Months go by, from location demo to actual deployment and then, customer is unhappy that location accuracy SLA is not being met. Everyone forgets that location accuracy has a lot to do with the design in place, and understanding the variations and unpredictability that exists.
Planning is important for an successful deployment.
3-channel setup in 2.4 and 5GHz radios 3 or more APs must see the same client for location calculation. The
more channels are enabled in the venue, the less the likelihood that a
Wi-Fi device can be seen by 3 APs.
Distance between APs? 3 or more APs must see the same client for location calculation. The
ideal distance varies for venue types. Open spaces, vs, narrow pathways
with thick walls all will require intelligent AP placements for better
accuracy results.
AP install height Ideal in normal situations to be installed 12-15ft from ground. Slightly
more distance is OK for APs with narrow-beam antenna. AP should not be placed too low on the walls – that would cause the signal from client
to AP to get impacted with increase of people traffic in-between. These
would cause inaccurate RSSI reading to location engine and run-time
location results would degrade.
AP Deployment APs should preferably be positioned along and within the perimeter on an enclosed
area. Depending on the size of the area, more APs may be needed inside (not on the
perimeter).For example, assuming an area of 10000ft2 (930m2), the following figure shows the
ideal deployment (“good deployment”) and a deployment that should be avoided
(“bad deployment”).In a large area, distribute the APs evenly, so that APs are placed at relatively equal distances from each other.
Geometric Dilution of Precision The geometric shapes formed by the distribution of APs affect accuracy.
APs arranged in equilateral triangles yield better accuracy than APs that form an
obtuse triangle.Similarly, APs arranged in squares yield better results than APs that form rectangles.

Interference

Object and material effect: Large water bodies, big metal structures or concrete walls all create RF challenges (absorption, reflection…). Some of these can be captured during fingerprinting of the venue, but redesign and structural changes to the venue can introduce inaccuracies to the location solution.

People effect: An empty hall has a different RF propagation than a full hall. Human bodies (which is 60% water) absorb RF and depending on how many such bodies are between the device and the APs, the dynamic variation of the signal strength can throw the location accuracy further

Summary

Set your expectations. Consider factors above.

  • 1-3meter accuracy is not practical. Too many dependencies and almost impossible for all of those factors to be right at all times.
    Rarely any deployment starts with location expectation in mind. Even then, 5-10m accuracy throughout the entire venue, and at all times, is unrealistic.
  • Percentage in accuracy at 10m accuracy its 90% thus its consider user is detected accurately within the 90% of the 10 meter radius.
  • Determine why you want that type of accuracy? Do you have alternative option that is cost effective and meets your targets? Wi-Fi based locationing is great for analytics, and all analytics have certain degree of error built into them. Wi-Fi analytics can get you visibility into 50% or more of your total visitors. Engagement with a customer is best delivered with a personal touch, on their mobile devices, with the installation of a mobile app. Mobile app penetration is < 10% currently. Mobile app with BLE, Wi-Fi location API, and utilization of the phone’s internal sensors can deliver a highly accurate personal experience for those customers that download and use your application.
  • For Omaya™ presence the minimum required Wi-Fi AP is 1 while for Zonal edition the minimum will be 4 unit, As the density of APs grows, the accuracy of the location reports increases.

Increase Accuracy via Fingerprinting.

Omaya™ solutions to improve accuracy with the above factor is via using our mobile calibration apps that perform RF fingerprinting calibration with Omaya™ platform. This greatly improve the accuracy of device within the venue.

Was this helpful?