Modelling the Effect of Human Body around User on Signal Strength and Accuracy of Indoor Positioning

Abstract

WLAN indoor positioning system (IPS) has high accurate of position estimation and minimal cost. However, environmental conditions such as the people presence effect (PPE) greatly influence WLAN signal and it will decrease the accuracy. This research modelled the effect of people around user on signal strength and the accuracy. We have modelled the human body around user effects by proposed a general equation of decrease in signal strength as function of position, distance, and number of people. Signal strength decreased from 5 dBm to 1 dBm when people in line of sight (LOS) position, and start from 0.5 dBm to 0.3 dBm when people in non-line of sight (NLOS) position. The system accuracy decreases due to the presence of people. When the system is in NLOS case, the presence of people causes a decrease in accuracy from 33% to 57%. Then the accuracy decrease from 273% to 334% in LOS case.

A Light Solution for Device Diversity Problem in a Wireless Local Area Network Fingerprint Indoor Positioning System

“…The closer to 0 dBm, the stronger the signal is. The RSSI value is very fluctuating (not constant) depending on multipath fading, environmental conditions, and the distance between the sender and receiver [25]. On IPS fingerprint, each coordinate point (x, y) is represented by a set of RSSI values from several access points (radiomap database).…”

“…The kNN algorithm is a method for classifying objects based on learning data closest to the object [24] and to solve user orientation problem [25]. The value of K is the smallest amount of Euclidean distance between each AP and the point of location that was not known beforehand.…”