Online task scheduling system
Abstract : In such systems, the efficiency of task scheduling algorithms directly impacts the overall system performance. By using previously proposed 2D scheduling model, existing algorithms could not provide an efficient way to find all suitable allocations. In addition, most of them ignored the single reconfiguration port constraint and inter-task dependencies. Further more, to our best knowledge there is no previous work investigating in the impact on the scheduling result by reusing already placed tasks. In this paper, we focus on online task scheduling and propose task scheduling solution that takes the ignored constraints into account. In addition, a novel “reuse and partial reuse” approach is proposed. The simulation results show that our proposed solution achieves shorter application completion time up to 43.9% and faster single task response time up to 63.8% compared to the previously proposed stuff ing algorithm.
? In this paper, we bring the review of various existing scheduling methodologies for HSCS. The task scheduling models are basically two types called static and dynamic scheduling.
? The SD algorithm is compared with existing duplication TDS, MTDS, and non-duplication scheduling algorithms with respect to Normalized Schedule Length (NSL), Efficiency.
? In this paper, we have demonstrated various High Speed Computing systems which support runtime requirements of applications and also prepared a summary chart for existing scheduling methodologies.
? When number of DRL cells is equal or more than three, the techniques in brings better optimization for shared memory architecture than the local memory architectures.
? The scheduling problem now is similar to the strip packing as presented in. Based on the 2D model, in this paper, we propose our solution to the online task scheduling problem.
? They introduced a convenient 2D model which can convert the online scheduling problem to the strip packing.
? By investigating the related works, we noticed that, firstly, when the online task scheduling problem is handled by using the 2D model, there is no suitable algorithms searching available allocations.
? If the availability of the reconfiguration port on the FPGA is not taken into account, the three tasks are scheduled ,which can be treated as a simple strip packing problem.
• In, the task scheduling is based on cost function whereas the cost function is an attribute of tasks of an application. There are several list scheduling algorithms proposed for microprocessor as follows.
• A Dynamic Critical Path (DCP) Scheduling algorithm proposed for multiprocessors where the DCP intended to find critical path of a task graph and rearranges the schedule on each processor dynamically.
• A dynamic scheduling and placement algorithm has been proposed for RS based on finishing time mobility of the tasks.
• An efficient multi task scheduler for runtime reconfigurable systems proposed a new parameter called Time-Improvement as cost function for compiler assisted scheduling algorithm.
? The efficiency of the online scheduler will directly impact the overall performance of the whole system.
? Although the algorithm focused on a higher task acceptance ratio, it did not show obvious reduction of the completion time of the overall application, which reflects the overall performance of the partially reconfigurable systems.
? The best fit heuristic is to schedule the arrival task into an available sRectangle which results in less fragmentation and better time performance in the 2D model.
? The STRT is an very important character to measure the system performance especially for the real-time systems.
|