Selecting the parameters for your Beacon Study


A beacon sensor can behave differently, depending on its configuration. We measure the difference in terms of  System Performance Metrics , and allow for tweaking using a series of  Experiment  and  Engineering  parameters.
All these parameters will be used at different stages of the creation of your study on the ODIN platform.

System Performance Metrics

As mentioned above, we measure beacon behavior in terms of the following System Performance Metrics:
    Sensitivity of Arrival Detection (iOS)
    Specificity of Arrival Detection (iOS)
    Sensitivity of Departure Detection (iOS)
    Specificity of Departure Detection (iOS)
    Sensitivity of Arrival Detection (Android)
    Specificity of Arrival Detection (Android)
    Sensitivity of Departure Detection (Android)
    Specificity of Departure Detection (Android)
These metrics will be impacted by your scientific choice of Near and Far, and the values you choose for the 4 Engineering Parameters, described below.


Experiment Parameters

    These parameters are determined by contemplating the scientific experiment and its semantics. You must decide on these two numbers before proceeding with the rest of this document.
  • Near (measured in feet): What it means to be "close to" a beacon
  • Far (measured in feet): What it means to be "far from" a beacon


Engineering Parameters

    TxPower (an integral number between 0 and -20): This represents the strength of the signal transmitted by the beacon, 0 being the strongest signal (having the greatest range). This is set in the physical beacon hardware device.
    RSSI Threshold: (received signal strength, an integral number between -40 and -100): This represents the threshold at which the phone will consider the beacon to be "potentially close" to it. This is set in the Arrival or Departure Rule
    Forgetting threshold (received signal strength, an integral number between -40 and -100): This is the maximum difference between the instantaneous and smoothed signal received. If the instantaneous signal is more different than this, the smoothing operation "resets" its historical memory. This is set in the Beacon Proximity Sensor configuration via the ODIN website.
    Alpha: This is a constant that indicates the effect of historical received signal readings on the smoothed value. When alpha = 1, the smoothed = the instantaneous received signal. As alpha ⟶ 0, the smoothed reading is more heavily influenced by historical readings. This is set in the Beacon Proximity Sensor configuration via the ODIN website.




Next Steps:

Once you have your experiment and engineering parameters defined, you should check the list of supported beacons that could be used in your study and how to set it up. The following document shows the supported beacons  Supported Beacons