This paper proposes a low power real-time dynamic voltage scaling algorithm with low power consumption for the energy-efficient design in the applications of wireless sensor networks which require collaborative processing executed in parallel on sensor nodes. In this algorithm, a single-task processing model of nodes is established and used to reflect the influences of associated communication events on the processing of tasks, and an optimal intratask voltage scheduling strategy is introduced to obtain the ideal operation frequency （or voltage ） of each basic block in the model. Afterwards, a frequency split method is proposed to effectively eliminate the slack time between the task completion time and the deadline. The implementation method of the proposed algorithm is also presented and the effectiveness of the algorithm in reducing the energy consumption of nodes is finally verified through a simulation.