Accurate Distance Estimation Using Fuzzy based combined RSSI/LQI Values in an Indoor Scenario: Experimental Verification

Abstract

The received signal strength indicator (RSSI) and the link quality indicator (LQI) are the metrics that are commonly available in commercial off-the-shelf (COTS) sensor hardware. The former has been widely regarded as the main source for distance estimation and node localization. However, experimentally RSSI has been shown to behave in an inconsistent manner, even in ideal scenarios, and serve at best as bounds for distances. The latter is effectively a measure of chip error rate, and can be used to identify higher quality transmissions, and the combination RSSI/LQI can be expected to make more precise estimates with the tradeoff of increased delay and estimation cost. In this paper, we describe our distance estimation system that uses these two metrics and test our hypothesis purely through experimental measurements using sensor nodes. Results indicate that such a combination of metrics can be used to provide a tighter bound on the range of estimated distances. We then quantify the improvement in distance estimation by relying on these two metrics. Through a unique classification using fuzzy logic and TBM, we developed an algorithm that is capable of precise distance estimation within the range of 100cm to 400cm, on at least 80% of the times while reaching accuracy as high as 100%.

Improving the Signal Propagation at 2.4 GHz Using Conductive Membranes

“…We can also find in the related literature RSSI combined with other systems for localization [24] and distance estimation [25] purposes.…”

Experimental study: a LQI-based ranging technique in ZigBee sensor networks

“…The results in (20) indicate that when LQI increases, the probability of correctly estimating distance is increasing. In (21), RSSI/LQI with fuzzy logic and TBM classification methods are used. The authors claimed that they can receive 80% ranging accuracy when devices are less than 4-meter between each other, but errors spread beyond 4 meters.…”

RSSI Prediction in WiFi Considering Realistic Heterogeneous Restrictions

“…As in (1) two terms were distinguished, the general signal trend and the jumps term, the same must be considered in order to evaluate (2). Firstly jumps must be identified for variables ,  J and σ J 2 be calculated.…”

“…Note that a coefficient, according to (2), must be an exponential whos exponent is always negative. Therefore its value has to be limited to a(0,1), although previous a 0 value exceeds this interval, adjusting a to be a minimum  distance over 0 or down 1, according to which limit a 0 has crossed.…”

“…For this reason it is very important to predict efficiently the future RSSI values taking into account the recent historic received values. In general prediction of RSSI can be used in other domains: distance estimator [2] and device localization mechanism [3], adaptive data and power transmission [4], link level quality estimation [5], terminals connectivity indication to Access Point (AP), service disruption anticipation in the event of predicted link degradation or failure to support a handover process [6].…”