Energy-Efficient Control with Harvesting Predictions for Solar-Powered Wireless Sensor Networks

Abstract

Wireless sensor networks equipped with rechargeable batteries are useful for outdoor environmental monitoring. However, the severe energy constraints of the sensor nodes present major challenges for long-term applications. To achieve sustainability, solar cells can be used to acquire energy from the environment. Unfortunately, the energy supplied by the harvesting system is generally intermittent and considerably influenced by the weather. To improve the energy efficiency and extend the lifetime of the networks, we propose algorithms for harvested energy prediction using environmental shadow detection. Thus, the sensor nodes can adjust their scheduling plans accordingly to best suit their energy production and residual battery levels. Furthermore, we introduce clustering and routing selection methods to optimize the data transmission, and a Bayesian network is used for warning notifications of bottlenecks along the path. The entire system is implemented on a real-time Texas Instruments CC2530 embedded platform, and the experimental results indicate that these mechanisms sustain the networks' activities in an uninterrupted and efficient manner.

Keywords: energy prediction; shadow detection; solar cells; wireless sensor network.

Figure 1

(a) Illustration of a shadow region at some time; (b) Illustration of the moving and shape-changing shadow region half an hour later.

Figure 2

Illustration of battery discharge.

Figure 3

Illustration of the harvested energy of two solar cells at different locations.

Figure 4

(a) An example of the piecewise linear curve fitting estimation; (b) An example of the piecewise quadratic curve fitting estimation; (c) An example of the fourth-order polynomial curve fitting estimation; (d) An example of the fifth-order polynomial curve fitting estimation; (e) An example of the B-spline curve fitting estimation; (f) An example of the Bezier curve fitting estimation.

Figure 5

(a) Time cost of the fitting methods; (b) Mean absolute difference for the fitting methods.

Figure 6

Estimation procedure of the extended Kalman filter (EKF).

Figure 7

Prediction procedure of the solar harvested energy.

Figure 8

Example of the prediction procedure

Figure 9

(a) q is density-reachable from p; (b) p and q are density-connected.

Figure 10

(a) An example of the DBSCAN algorithm addressing 200 random nodes; (b) Another example of the DBSCAN algorithm addressing 200 random nodes.

Figure 11

(a) Sensor node coded from west to east; (b) Sensor node coded from north to south; (c) Example of shadow moving from west to east; (d) Example of shadow moving by a certain angle.

Figure 12

(a) Energy consumption of the 10% duty cycle with the LED on; (b) Energy consumption of the 10% duty cycle with light sampled once; (c) Energy consumption of the 50% duty cycle with idle listening; (d) Energy consumption of the 100% duty cycle with the sending of data.

Figure 13

Illustration of a clustering mechanism.

Figure 14

Illustration of the route selection algorithm.

Figure 15

(a) Illustration of the routing situation. (b) Illustration of double clusterheads.

Figure 16

Hardware illustration of the sensor node.

Figure 17

Illustration of the deployment of TSEP sensor nodes.

Figure 18

Average battery energy levels of the three experimental groups.

Figure 19

Standard deviation of the battery levels of the experimental groups.

Figure 20

Illustration of deployment for the shadow experimental groups.

Figure 21

Average battery energy levels of the shadow experimental groups.

Figure 22

(a) Illustration of transmission topology of SCM group; (b) Illustration of transmission topology of DCM group.

Figure 23

(a) Loss rate of the clusterheads during the experiment; (b) Average battery capacity of the clusterheads during the experiment.