Design of energy-efficient wireless sensor networks with censoring, on-off, and censoring and on-off sensors based on mutual information

In wireless sensor networks, owing to the limitation of energy, memory, and computation, it is necessary to construct the networks under the constraints. To save the energy, the censoring sensor that transmits information of its observations only when it is informative was proposed, and the optimal censoring region was derived so that the detection probability was maximized. The on-off sensor that is either awake or asleep (on or off) is also effective to save the observation energy. The collaboration of censoring and sleeping plays a great role in saving more energy. In this paper, we derive the mutual information in a wireless sensor network with i) censoring sensors, ii) on-off sensors, and iii) censoring and on-off sensors. Under the constraint on the average cost of each sensor, we also derive the optimum censoring and sleeping probabilities with respect to the mutual information in the sensor networks. Using the mutual information, we show that the largest mutual information in the sensor network with censoring and on-off sensors can always be achieved. We also show that the mutual information in the sensor networks with on-off sensors and with censoring and on-off sensors are larger than that in the sensor network with censoring sensors when the average cost of each sensor is small and the observation cost is large. Furthermore, we show that the largest mutual information in the sensor networks with censoring and on-off sensors and with censoring sensors can be obtained when the average cost of each sensor is not small and the observation cost is small. Based on the results, we can design very energy-efficient wireless sensor networks by using censoring and on-off sensors.