Stochastic Allocation and Scheduling for Conditional Task Graphs in MPSoCs

Abstract

Abstract. This paper describes a complete and efficient solution to the stochastic allocation and scheduling for Multi-Processor System-on-Chip (MPSoC). Given a conditional task graph characterizing a target application and a target architecture with alternative memory and computation resources, we compute an allocation and schedule minimizing the expected value of communication cost, being the communication resources one of the major bottlenecks in modern MPSoCs. Our approach is based on the Logic Based Benders decomposition where the stochastic allocation is solved through an Integer Programming solver, while the scheduling problem with conditional activities is faced with Constraint Programming. The two solvers interact through no-goods. The original contributions of the approach appear both in the allocation and in the scheduling part. For the first, we propose an exact analytic formulation of the stochastic objective function based on the task graph analysis, while for the scheduling part we extend the timetable constraint for conditional activities. Experimental results show the effectiveness of the approach.

Energy-aware scheduling on heterogeneous multi-core systems with guaranteed probability

“…A highly complex, non-linear integer programming based method is proposed in [16] for task mapping and scheduling in multiprocessor systems. In [17], the authors attempt to balance the expected energy consumption across processors by partitioning a set of independent tasks.…”

“…In embedded systems, the execution times of some tasks are variable due to conditional instructions (and/or operations) and different inputs [20][21][22][23]. These embedded systems may operate in time-varying environments.…”

Energy efficient multiprocessor task scheduling under input-dependent variation

“…However, the proposed techniques are restricted to CTGs. Expected energy minimization: A highly complex, non-linear integer programming based method is proposed in [1] for task mapping and scheduling in multiprocessor systems. In [2], the authors attempt to balance the expected energy consumption across processors by partitioning a set of independent tasks.…”

Optimizing Tasks Assignment on Heterogeneous Multi-core Real-Time Systems with Minimum Energy

“…The authors of [12] proposed an approach to schedule periodic dependent tasks on multi-core processors and a dynamic programming based pseudo polynomial algorithm is used. In embedded systems, the execution times of some tasks are variable due to conditional instructions and different inputs [13] [14].…”

“…The authors of [6] proposed an approach to schedule periodic dependent tasks on multi-core processors. Task assignment in multiprocessor is NP problem and it cannot get the global optimal solution [7][8][9][10]. Zhang et al [11] proposed a dynamic programming approach to allocate the scalar and array variables on a single processor.…”