The development of multiprocessor Systems on chips (MPSOC) led to increasing difficulties in system design. The optimization of energy consumption has become a primary concern. The solution is to use various techniques for optimizing energy and power. The purpose of this paper is to integrate a Dynamic Slack Reclamation technique in the application software radio which is an algorithm based on the strategy of slowing down the frequency/voltage couple on the processor. The advantage of this algorithm is that it ensures the detection of task completion and determines the quantity of delay time by comparing the worst execution time with the real time. Experimental results show that the use of a DSR algorithm increases the performance by respecting real-time constraints in terms of power consumption and energy on a real platform, OMAP 3530. In addition, the power and energy consumption is reduced up to 27% and 8% respectively for scenario 1 and scenario 2. The result also show that scenario 1 consumes 80% more energy than scenario 2.