A wireless communication selection approach to minimize energy-per-bit for wearable computing applications

Body Sensor Networks (BSN) provide a way to gather continuous observations of human movements, which has a potential of improving medical care quality, and enabling continuous remote patient monitoring. Despite their potential, BSNs face serious werability constraints. Energy optimization is essential since werability is most affected by the battery size of the device. In this paper, we introduce a burst communication technique that takes advantage of data buffering to achieve lower energy-per-bit cost with a possibly higher packet size or more energy efficient communication scheme suited for higher data rates. Our energy model combines the knowledge of the signal processing required to complete a task with the deadline associated with that task to define the optimal burst transmission schedule. Based on the selected energy model, we formulate an optimization function that minimizes the overall energy cost of communication for a given signal processing task. We demonstrate the effectiveness of our approach in sway monitoring BSN applications with a short, medium, and long deadlines. We further demonstrate the relationship between the task deadline extension and the energy cost of the system. Our results show that the proposed approach can improve the cost of communication 85-95% compared to streaming data to the basestation as it becomes available.