| Sensor networks are being deployed in a multitude of application domains. The purpose of the sensors is to monitor an environment, be it an animal habitat, a battlefield or a human body. The network formed by the sensors also transports the monitored data to a central location where it can be analyzed. Accordingly, the sensors have to be deployed in such a manner that they can monitor the entire domain. In technical terms, the deployed sensors must provide complete coverage for the domain. For the purpose of transportation of the monitored data to the central location, the network formed by the sensors must be connected. Consequently, the connected coverage problem in sensor networks has emerged as an important topic in the research community. Different domains of sensor networks impose their own characteristics and requirements. The challenges posed by the formation of a sensor network in a temperature sensitive environment have been studied. In order to transport monitored data to the central location, the radio transmitters co-located with sensors, have to transmit. Radio transmission dissipates energy, thereby increasing the temperature in the vicinity of the transmitter. This increase in temperature must not exceed a threshold value, if the transmitter is embedded in a human body. This requirement to a large degree can be ensured by maintaining a minimum separation distance between every pair of sensors/transmitters. Several versions of the connected coverage problem in sensor networks under the minimum separation distance constraint have been studied. Almost all versions of the problem are computationally hard (NP-complete) and for some versions a polynomial time approximation algorithm cannot be developed, unless P = NP. Heuristics for some versions of the problem have been developed and efficacy of the heuristics evaluated through extensive simulations. For the deployment scenario where a complete coverage is impossible, the notion of scheduling, where only a subset of transmitters are turned on at any given time slot have been introduced. A mobile sensor environment has also been considered. Analytical and experimental results related to two different environments have been presented, where the sensors are mobile. |
