Design of a Hybrid Indoor Location System Based on Multi-Sensor Fusion for Robot Navigation

. 2018 Oct 22;18(10):3581. doi: 10.3390/s18103581

Algorithm 1 KNNBP.
Input: FD = $\{F_{1}, F_{2}, F_{3}, \dots, F_{n}\}$
F $_{j}$ = $\{(M A_{j 1}, R_{j 1}), (M A_{j 2}, R_{j 2}), \dots, (M A_{j m}, R_{j m})\}$ (F $_{j} \in$ FD)
F $_{c}$ = $\{(m a_{1}, r_{1}), (m a_{2}, r_{2}), \dots, (m a_{m}, r_{m})\}$
1:	forj = 1 to n do

2:	for i = 1 to m do

3:	if ma $_{i}$ = MA $_{j i}$ then

4:	if \| $r_{i}$ − $R_{j i}$ \| < V then

5:	$P_{i}$ = 1 − $\frac{\| r_{i} - R_{j i} \|}{V}$

6:	end if
7:	end if
8:	end for
9:	$P_{j}$ = $\frac{1}{a} \sum_{i = 1}^{a} p_{i}$
10:	end for
11:	sort $P_{j}$ from largest to smallest

12:	select the first K positions ( $x_{k}$ , $y_{k}$ ) corresponding to $P_{j}$

13:	calculate position (x, y) = $\frac{1}{K} \sum_{j = 1}^{K} (x_{j}, y_{j})$

14:	return (x, y)