Description
This thesis demonstrates the application of techniques to a embedded system of com- mercial off the shelf (COTS) microcontroller and sensor devices with the intent to create a power aware system of which the primary goal is to maximize data collection on finite power. The target application of this system is to collect environmental data in a remote location, such that the system draws power from finite a energy storage. Power analysis of system features (e.g., Central Processing Unit (CPU) core, Input/Output (I/O) buses, and peripheral sensors) are studied. Software-based power optimization techniques utilizes the power information and pairs up with hardware-assisted power gating to control system features to extend the embed- ded system’s up time in a field, under a deployed finite energy scenario. By doing so the proposed power optimization algorithm can collect more data when compared to a Low Power Energy Aware Processing (LEAP) approach. Simulations resulted in hundreds of additional days of system up time after applying the proposed techniques. Lastly,a 128-bit Advanced En- cryption Standard (AES) algorithm is applied on the collected data using various parameters. A study of this system’s encryption capabilities reveal predictable power trends based on data size and processor settings. Design considerations are discussed to identify conditions and sce- narios that would yield non-optimal results. These finding can be further enhanced by applying techniques to more complex microprocessor systems and Wireless Sensor Network (WSN) of similar composition.
