Improved Adaptive Smoothing Filter for Indoor Localization Using RSSI

Abstract

Abstract:In the indoor location estimation system, which has recently been actively studied, the received signal strength indicator contains a high level of noise when measuring the signal strength in the range between two nodes consisting of a receiver and a transceiver. To minimize the noise level, this paper proposes an improved adaptive smoothing filter that provides different exponential weights to the current value and previous averaged one of the data that were obtained from the nodes, because the characteristic signal attenuation of the received signal strength indicator generally has a log distribution. The proposed method can effectively decrease the noise level by using a feedback filter that can provide different weights according to the noise level of the obtained data and thus increase the accuracy in the distance and location without an additional filter such as the link quality indicator, which can verify the communication quality state to decrease the range errors in the indoor location recognition using ZigBee based on IEEE 802.15.4. For verifying the performance of the proposed improved adaptive smoothing filter, actual experiments are conducted in three indoor locations of different spatial sections. From the experimental results, it is verified that the proposed technique is superior to other techniques in range measurement.

Do WiFi Probe-Requests Reveal Your Trajectory?

“…Solutions using RSSI for estimating users' trajectories contain localization errors arising from inaccurately measured distances between the user and the regarded sniffer. Past works have tried to mitigate such errors to some extent using smoothing techniques for RSSI [9]. However, smoothing itself causes bias in the estimated distance.…”