Providing Databases for Different Indoor Positioning Technologies: Pros and Cons of Magnetic Field and Wi-Fi Based Positioning

Table 2

Main characteristics of the data from selected papers. Infrastructure corresponds to the kind of infrastructure required to support the localization (free means no explicit infrastructure), # of WAPs is the number of Wireless Access Points detected/considered/required for positioning, Area is the surface covered for positioning in , # of RefPoints is the number of reference points where the training samples were collected, and  # of Fp/Place is the number of fingerprints taken at each reference points. Multibuilding? and Multifloor? denote if the database includes samples from multiple buildings and multiple floors, respectively.

Paper	Infrastructure	# of WAPs	Area (approx.)	# of RefPoints	# of Fp/Place	Multibuilding?	Multifloor?
[9]	Free	520	110,000 m^2	933	At least 20	YES	YES
[26]	Based	3	22.5 × 43.5	70 × 4	At least 20	NO	NO
[27] Exp. 1	Free	48	27 × 18	216	10,000	NO	NO
[27] Exp. 2	Free	4	27 × 18	48 + 3	3 seconds
[27] Exp. 3	Free	156	140 × 90	1400	10,000	NO	NO
[27] Exp. 4	Free	10/12	140 × 90	101+15	3 seconds	NO	NO
[28]	Free	206	10,000 m^2	96	30	NO	NO
[29]	Free	101	2 buildings	392 × 3	3 × N	YES	YES
[30]	Based; Wi-Fi router	4	450 m^2	41	40	NO	NO
[30]	Based; Nodes	6	800 m^2	56	40	NO	NO
[30]	Based; Wi-Fi router	3 × 4	3 × 450 m^2	3 × 41	40	NO	YES
[30]	Based; Zotac PCs	4	1000 m^2	100	40	NO	NO
Our DB	Free	93	200 m^2	34	20	NO	NO

Approximately, they used grids at 1.5 and 3 m.
They collected 10,000 beacons scanning data during 5 minutes approx.
They collect 30 scans, but then they apply the average function.
Selected with the iLoc algorithm.