New LBASense Algorithm For Improved Crowd Analytics

new algorithm


July 12th, 2018  – Singapore


Specialised in location analytics for Smart Cities, event organisers and retail owners, DFRC operates LBASense mobile phone detectors in order to build detailed, yet privacy friendly, Crowd Analytics. In order to keep updated with the constant evolution of mobile phone technologies, our engineers have developed a new algorithm, allowing to detect phones that are randomising their MAC address.

 

MAC addresses anonymised to aggregated data

But why do LBASense sensors need to detect MAC addresses in the first place? WiFi MAC address is what LBASense sensors capture. Soon after the acquisition, data is transferred to a local server, over a secure link, and converted with a strong encryption key into an internal code. Such codes are aggregated based on time and region of the collection, whereas the raw data is deleted immediately after the conversion. Aggregated figures are elaborated with a statistical approach to determine the so-called Crowd Analytics.

LBASense Data Treatment
LBASense Data Treatment

.

More Accurate Analytics Thanks To A New Algorithm

LBASense system now better supports MAC randomisation, being able to detect phones like Samsung’s Galaxy S9 and Motorola, which implement a new method for the MAC address randomisation on Android Operating Systems. Indeed, the new software is capable of recognising if the detected MAC addresses are not real because randomised; as such, the system marks those detections and refines the statistical elaboration taking such portion into account. Thanks to this improved algorithm, the final figures of Crowd Analytics are more accurate.

.

What is MAC Derandomisation?

MAC Derandomisation is a proprietary technology developed by DFRC in order to overcome the problem of mobile devices frequently changing their MAC address while searching for WiFi networks. DFRC’s technology is capable of determining how many phones are behind the randomisation, as well as the dwell time that those phones are staying in, within a specific geographic area. As a result, more accurate stats can be elaborated and delivered to customers.

 

Interested in learning more about this unique technology? Contact us for a tailored scenario.

Photo credits: Featured image by Matt Quinn; LBASense image: all rights reserved to DFRC.