Grant: Discovery Grants Program - Individual

Wireless sensor networks (WSNs) are used in many applications, such as security, military, agriculture, automation, and habitat monitoring purposes. Although sensors have been used for decades, due to the latest developments in the electronics industry, it is possible to add some processing, storage, and communication resources with the sensors. However, due to limited battery power supply, it is necessary to design energy efficient communication techniques, which could extend battery life from a couple of days to a few years. Further, as sensors are deployed in harsh and hostile environments, it is necessary to incorporate security, trust management, and fault tolerant techniques; otherwise, the sensor data will not be reliable. I will use a database approach to reduce the burden of the application developer. The user will specify the query in a simple English-like language. However, the proposed middleware will address the following concerns: a) energy efficient communication, b) security and trust management and c) fault tolerance. I will use intelligent techniques such as genetic algorithms, fuzzy logic and simulated annealing to obtain energy efficient communication. Also, I will use endocrine communication system of the human body as a model to design energy efficient heterogeneous WSN. For security, I will develop energy efficient key management techniques to address the following issues: confidentiality, integrity and authentication, as well as several attacks such as man-in-the-middle, collusion and replay. The proposed middleware will be validated and verified for several deployment environments using a WSN testbed (e.g., WINTeR). The proposed middleware will be useful for several applications such as: a) industry process control automation, e.g., CNA Plant (training facility in St. John's, NL) and Hibernia (oil refinery near St. John's, NL) and b) intelligent applications for smart environments (nursing homes and tele-health).