D2R-TED: Data-Domain Reduction Model for Threshold-Based Event Detection in Sensor Networks

Abstract

The reduction of sensor network traffic has become a scientific challenge. Different compression techniques are applied for this purpose, offering general solutions which try to minimize the loss of information. Here, a new proposal for traffic reduction by redefining the domains of the sensor data is presented. A configurable data reduction model is proposed focused on periodic duty⁻cycled sensor networks with events triggered by threshold. The loss of information produced by the model is analyzed in this paper in the context of event detection, an unusual approach leading to a set of specific metrics that enable the evaluation of the model in terms of traffic savings, precision, and recall. Different model configurations are tested with two experimental cases, whose input data are extracted from an extensive set of real data. In particular, two new versions of Send⁻on⁻Delta (SoD) and Predictive Sampling (PS) have been designed and implemented in the proposed data⁻domain reduction for threshold⁻based event detection (D2R-TED) model. The obtained results illustrate the potential usefulness of analyzing different model configurations to obtain a cost⁻benefit curve, in terms of traffic savings and quality of the response. Experiments show an average reduction of 76 % of network packages with an error of less than 1%. In addition, experiments show that the methods designed under the proposed D2R⁻TED model outperform the original event⁻triggered SoD and PS methods by 10 % and 16 % of the traffic savings, respectively. This model is useful to avoid network bottlenecks by applying the optimal configuration in each situation.

Keywords: WSN; data compression; event detection.

Figure 1

Tree structure.

Figure 2

$\Phi$ domain reduced to a set of ${\Phi }_{\mathrm{r}\delta }$ domains (j = k = 1).

Figure 3

Example of RGB color domain reduction. Components of color coded with 5, 3 and 1 bit.

Figure 4

Upper left: Original signal $F\left(x\right)$, in black; and a constant threshold (Threshold), in red. Bootom left: Boolean response of the condition ($F\left(x\right)>\mathrm{Threshold}$) applied to the original signal and threshold. Upper right: Reduced signal $F^{\prime }\left(x\right)$, in black; and a constant threshold (Threshold), in red. Bootom left: Boolean response of the condition ($F^{\prime }\left(x\right)>\mathrm{Threshold}$) applied to the reduced signal and threshold. In red, Boolean variations with the obtained in the bottom left image.

Figure 5

Illustration of the metrics analysis process. (A) Transition recall analysis with $\tau =0$. In red, unmatched transitions; (B) transition recall analysis with $\tau =1$. In green, matched transitions; (C) extended transition recall analysis with $\tau =1$. In green, matched transitions; in red, unmatched transitions; in purple, unmatched samples.

Figure 6

Case study II. LP, TP, and TR metrics for training and experimental data.

Figure 7

Representation of increasing linear threshold ‘K’, defined for the temperature signal extracted from the Oropa station.

Figure 8

Case study III. With event-based sampling techniques SoD, PS, A-SoD, and A-PS previously detailed, the condition $T_i>K_i$ has been evaluated for each temperature signal of each station of the Arpa Piemonte repository [53]. The resulting Boolean signals are analyzed through metrics described in Section 3.2 for comparison purposes. The numeric identifier of each station in the repository is as follows: Oropa 0, Cameri 1, Alessandria 2, Vercelli 3, Pallanza 4, Montaldo Scarampi 5, and Bobes 6.