A Dual-Antenna Mobile Robot Orientation Estimation System Based on Rssi Fingerprinting and Machine Learning Techniques

SUMOREK, Mateusz; IDŹKOWSKI, Adam

doi:10.2478/ama-2025-0079

Figures & Tables

Graphical representation of the angle of arrival

The tracker with the IMU orientation marked

The internal structure of the ESP32-WROOM-DA system [16]

Robomaster S1 robot with tracker mounted

The measurement route of the training set with marked measurement locations

Confusion matrix for the Random Forest model

Confusion matrix for the MLP Classifier model

Metrics of individual classes for the k-NN model

Class	Recall	Precision	F₁
‘East’	0.46	0.41	0.43
‘North’	0.34	0.43	0.38
‘South’	0.13	0.33	0.19
‘West’	0.64	0.50	0.56

Sample results of localization based on dual antenna RSSI readouts

Classifier	MAE [m]	RMSE[m]	Stdev [m]
k-NN	1.53	1.90	1.30
MLP Classifier	1.46	1.82	1.06
Random Forest	1.63	1.94	1.35
SVM Classifier	1.54	1.90	1.25

Quaternion mappings [17]

BNO08X physical axis aligned			Mapping quaternion
X	Y	Z	Qw	Qx	Qy	Qz
East	North	Up	1	0	0	0
North	West	Up	(√2)/2	0	0	(√2)/2
West	South	Up	0	0	0	1
South	East	Up	(√2)/2	0	0	-(√2)/2

Class metrics for MLP Classifier

Precision macro	Recall macro	F₁ macro	Micro score	MCC
0.37	0.37	0.36	0.41	0.18

Sample results of localization based on single antenna RSSI readouts

Classifier	MAE [m]	RMSE[m]	Stdev [m]
k-NN	1.61	2.00	1.38
MLP Classifier	1.64	1.99	1.10
Random Forest	1.65	1.97	1.47
SVM Classifier	1.55	1.98	1.43

Metrics of individual classes for the SVC model

Class	Recall	Precision	F₁
‘East’	0.46	0.42	0.44
‘North’	0.31	0.38	0.34
‘South’	0.27	0.22	0.24
‘West’	0.44	0.45	0.44

Class metrics for Random Forest Classifier

Precision macro	Recall macro	F₁ macro	Micro score	MCC
0.31	0.34	0.31	0.41	0.16

Number of classes in the training set and their percentage share in the set

West	East	North	South
256	199	184	97
34,8%	27,0%	25,0%	13,2%

Number of classes in the test set and their percentage share in the set

West	East	North	South
39	28	29	15
35,1%	25,2%	26,1%	13,5%

Metrics for the SVC model

Precision macro	Recall macro	F₁ macro	Micro score	MCC
0.37	0.37	0.37	0.39	0.16

Metrics for the k-NN model

Precision macro	Recall macro	F₁ macro	Micro score	MCC
0.42	0.40	0.39	0.45	0.23

A Dual-Antenna Mobile Robot Orientation Estimation System Based on Rssi Fingerprinting and Machine Learning Techniques

Figures & Tables

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

Fig. 7.

Fig. 8.

Fig. 9.

Fig. 10.

Fig. 11.

Fig. 12.

Fig. 13.

Metrics of individual classes for the k-NN model

Sample results of localization based on dual antenna RSSI readouts

Quaternion mappings [17]

Class metrics for MLP Classifier

Sample results of localization based on single antenna RSSI readouts

Metrics of individual classes for the SVC model

Class metrics for Random Forest Classifier

Number of classes in the training set and their percentage share in the set

Number of classes in the test set and their percentage share in the set

Metrics for the SVC model

Metrics for the k-NN model

Paradigm

My account