Résumé:
Wireless Sensor Networks (WSNs) are composed of spatially distributed sensor nodes capable of monitoring physical or environmental conditions such as temperature, humidity, motion, or pressure, and transmitting the collected data to a central unit for analysis. These networks play an essential role in various domains, including environmental monitoring, smart agriculture, industrial automation, and health systems, where real-time data collection is critical for decision making and system control.
Despite their usefulness, WSNs face significant limitations due to the constrained energy resources of sensor nodes, which often operate on limited battery power and are deployed in inaccessible environments. This makes energy efficiency a major concern in WSN design.
In this thesis, we propose a set of analytical models to better understand and optimize the energy consumption of sensor nodes. Our approach is based on queueing theory and incorporates the N-policy strategy, which delays service until a threshold number of packets has accumulated.
We evaluate and compare several models using both analytical techniques based on Continuous-Time Markov Chains and discrete-event simulation. This combined method- ology enables a comprehensive evaluation of energy performance trade-offs under different system conditions. The results contribute to the development of energy aware strategies and offer insights for designing more efficient and sustainable wireless sensor networks.
Keywords: Wireless Sensor Networks, Queues with Vacation, Vacation policies, Continuous-Time Markov Chains (CTMCs), Performance Metrics.
