Battery life determines the duration and the extent of mobility for the battery-powered portable embedded systems and is a key performance metric for scheduling tasks on these multiprocessor systems. The main challenge is to space out the tasks to be executed on the multiprocessor system with the objective of shaping the battery discharge profile so as to enhance battery life, while at the same time, meeting the real-time timing constraints of the tasks. The current monograph addresses battery-aware static scheduling of precedence constrained task graphs with non-negligible intertask communication onto fully connected multiple processors in battery-powered mobile embedded systems. We propose a scheduling scheme comprising of genetic algorithm and simulated annealing techniques, both of which can efficiently explore the search space of all possible legal schedules to find a battery-efficient one. We demonstrate the effectiveness of our scheduling scheme over existing approaches on example task graphs and study the impact on battery life using a battery model. Our experimental results shows considerable improvement in battery life over existing approaches.