Article Preview
TopIntroduction
The last decade has witnessed a growing interest of environment monitoring applications in Wireless Sensor Networks (WSNs) due to their unique potential in the remote detection and prevention of disasters. However, such a critical mission requires a high efficiency to ensure data availability and timely delivery within a reasonable cost. Thus, the deployed WSNs’ connectivity and lifetime along with fault-tolerance and cost effectiveness are key network properties. Due to the dynamic physical environment and possible hardware failures, the raw data collected by the sensor nodes are inherently inaccurate and imprecise. The sensor networks in various fields, for various applications, have attracted the interest of researchers in most fundamental problems. Despite the advanced progress in the field of construction technology (MEMS), Sensor networks suffer from many problems. Among them, we can cite technological problems (limitation of monitoring and communication radii, very limited energy reserve, fragile structure), usage problems (non-deterministic protocols, a polynomial or exponential algorithmic complexity, theoretical protocols that are not feasible and not applicable in reality, random deployments), and many other problems. This leads to unbalanced dispersion, incomplete coverage of the area to be penetrated or controlled, rapid loss of data communication, rapid network depletion, etc. Our recent literature overview in the field of sensor networks has shown that the most fundamental problem in wireless sensor networks stills the energy coverage problem efficiency as stated (Cardei and Wu, 2005; Vijayarani and Ramya, 2016). The real world is uncertain, characterized by: (a) the incompleteness of the knowledge of the real state of the world (if the nodes in sub-zone A are completely, partially or not active?), (b) its inaccuracy (the exact positions of the nodes in surveillance? the position of targets and intruders in the area of interest?), (c) the actions’ results (the targets are moving, the active nodes are crushed, lost or exhausted), etc. By uncertainty, we mean the hostility of the environment where the sensor nodes are deployed. This could be caused by the variations of atmospheric circumstances, the modifications of the deployed sensor network topologies, the unreliability of the communication radio, etc. All these uncertain causes affect the quality of service and decision on real world information. The atmospheric changes, that impact the physical environment, influence on the position accuracy, the communication power and the sensor nodes’ monitoring area in the network. This reality forces us to consider such type of uncertainty. In order to do this, our proposal consists in introducing the fuzziness in the process of scheduling sensor nodes in WSN for several purposes. Among the types of considered uncertainty in WSN, there are:
- •
Uncertainty in radio communication links: the communication power increases if the Euclidean distance increases. In case of a 3D deployment in mobile environment, energy power and connectivity are constraints that prevent the communication of the sensor nodes within the
- •
network.
- •
Uncertainty in the detection links: environmental interference, angle, nonlinear distance, noise, sensor types, and other factors may introduce uncertainty in the detection process in sensor networks.
- •
Detection uncertainty in the data collection: when sensors are deployed in hostile environments, different things can affect the collected or detected data quality, such as node sensibility due to signal interferences caused by environment objects (e.g. foliage) or atmospheric phenomena (e.g. cloud), node physical state due to possible deterioration (wind, soil state, animals, etc.).