Using Smartphone BLE RSSI Values to Classify Proximity with Deep Neural Networks

Devices used

Device	Name	Model
D1	Xiaomi Redmi Note 10 Pro	M2101K6G
D2	Xiaomi Redmi 9	M2004J19C
D3	Nokia 5.1 Plus	-
D4	Nokia 2.3	-

4 different scenarios

• Data collection with Devices 3 feet apart in an indoor open space.
• Data collection with Devices 9 feet apart in an indoor open space.
• Data collection with Devices 3 feet apart with an interior wall in between.
• Data collection with Devices 9 feet apart with an interior wall in between.

Scenario	Conditions	Target Classification
open3feet	- 3 feet apart - Indoor space	1
open9feet	- 9 feet apart - Indoor space	0
wall9feet	- 3 feet apart - Indoor wall in between	0
wall9feet	- 9 feet apart - Indoor wall in between	0

Results:

Hidden layer size	1 Dense Layers	model	model description
Hidden layer size 64
Hidden layer size 128