Modeling and Scheduling with DAG used for tasks in ME with OVP

Several techniques have been proposed to adapt multimedia and telecommunication applications to resource constraints. These techniques are mostly implemented at the application layer and make simplistic assumptions about the system resources communication and they are often unconvinced to the co-design in embedded system SoC and MPSoC. Moreover, they often assume that the data codec video characteristics and their processing needs are stationary, which is not true in practice especially in hardware for the real target FPGA. In fact, data streams in codec video are highly dynamic and may also experience concept drift, thereby requiring continuous online adaptation of the throughput and quality to each processing task in video and frame. Hence, existing solutions for multimedia and telecommunication applications are often too conservative or too aggressive, especially in video codec. To address these limitations, complexity in video test, quality of frame, precision, send video with a fast time especially in medical applications, airport and military flight problems. we propose offline algorithm scheduling which minimize the execution time (i.e., troughput and output quality of video and frame) of medical and military applications under real time and resources constraints. Our algorithm scheduling uses offline and online reinforcement learning techniques. Moreover, our scheduler is able to detect concept drifts and to smoothly adapt the scheduling strategy (DAG algorithm). Our experiments realized on a chain of tasks modeling real-life streaming application demonstrate that our scheduler is able to learn the scheduling policy and to adapt it such that it minimizes the execution time and optimize the precision, quality of frame as the video codec characteristics are dynamically changing. The strategy of scheduling algorithm is to order the tasks in processor or different processors. This approach remade tow problem in static scheduling algorithm with video and frame processing (the physiological data of a patient and allows in cases of urgent problems to trigger an alarm remotely controlled by an expert to intervene quickly in case of emergency. The key of contribution in this research are the modeling and scheduling tasks in ME blocks in video codec with new approach automatic, implementation and developpement this algorithms virtual Platform “OVP”, we work with tow codec video (H264 and H265) in ME block.