Author Affiliations
1Key Laboratory of Opto-Electronic Science and Technology for Medicine Ministry of Education, College of Photonic and Electronic Engineering, Fujian Normal University, Fuzhou 350007, Fujian, China2Jinshan College of Fujian Agriculture and Forestry University, Fuzhou 350002, Fujian, China3Fujian Key Laboratory of Photonics Technology, College of Photonic and Electronic Engineering, Fujian Normal University, Fuzhou, Fujian 350007, China4Fujian Engineering Technology Research Center of Photoelectric Sensing Application, College of Photonic and Electronic Engineering, Fujian Normal University, Fuzhou 350007, Fujian, China5Concord University College, Fujian Normal University, Fuzhou 350117, Fujian, Chinashow less
Fig. 1. Experimental setup of the VLP system with a multi-PD receiver
Fig. 2. Time division multiplexing scheme
Fig. 3. Single-hidden layer feedforward network with L hidden neurons
Fig. 4. Conceptual architecture of positioning system. (a) 2-D positioning; (b) 3-D positioning
Fig. 5. Positioning algorithm flow diagram
Fig. 6. CDF of positioning error for different algorithms. (a) 2-D positioning; (b) 3-D positioning
Fig. 7. APE of different algorithms. (a) 2-D positioning; (b) 3-D positioning
Fig. 8. Impact of M on APE. (a) 2-D positioning; (b) 3-D positioning
Fig. 9. Impact of N on APE. (a) 2-D positioning; (b) 3-D positioning
Fig. 10. Impact of Pt on APE. (a) 2-D positioning; (b) 3-D positioning
Parameter | Reference |
---|
Indoor space unit size(L×W×H)/cm | 100×100×150 | Plane range of receiver /cm | (0,0)to(65,70)(resolution:5) | Transmitter power /W | 5,6,7[45,58] | Height of the receiver /cm | 10,20,30 | Position of four LEDs(x, y,z)/cm | LED1(-10,-10,120) LED2(80,-10,120) LED3(80,80,120) LED4(-10,80,120) | Distance between each LED /cm | 90 | The FOV of LED /(°) | 60 | Distance between each PD /cm | 5 | The FOV of PD /(°) | 120 | The effective area of PD /cm2 | 1 |
|
Table 1. Experimental parameter
Algorithm | Parameter |
---|
KNN | Distance metric:Euclidean distance;K=3 | ELM | Number neurons in input,hidden and output:16, adaptive and 1;Activation function:Sigmoid | RF | Tree number:50;Weak classifier:Decision tree | AdaBoost | Learning cycle:100;Weak classifier:Decision tree |
|
Table 2. Parameter of the four machine learning algorithms
Algorithm | S-KNN/KNN | S-ELM/ELM | S-RF/RF | S-AdaBoost/AdaBoost |
---|
2-D positioning APT /s | 0.01/0.01 | 0.01/0.05 | 0.72/1.14 | 0.98/2.43 | 3-D positioning APT /s | 0.02/0.07 | 0.05/0.32 | 5.45/6.82 | 4.45/18.41 |
|
Table 3. APT of different algorithms