A two-step optimization technique for functions placement, partitioning, and priority assignment in distributed systems

Abstract

Modern development methodologies from the industry and the academia for complex real-time systems define a stage in which application functions are deployed onto an execution platform. The deployment consists of the placement of functions on a distributed network of nodes, the partitioning of functions in tasks and the scheduling of tasks and messages. None of the existing optimization techniques deal with the three stages of the deployment problem at the same time. In this paper, we present a staged approach towards the efficient deployment of real-time functions based on genetic algorithms and mixed integer linear programming techniques. Application to case studies shows the applicability of the method to industry-size systems and the quality of the obtained solutions when compared to the true optimum for small size examples.

Resource Allocation in Software-Defined Vehicles: ILP Model Formulation and Solver Evaluation

“…Mehiaoui-Hamitou et al [16] presented a mixed-integer linear programming (MILP) based approach to solving the automotive deployment and scheduling problem regarding energy efficiency. They considered constraints including timing and utilization values to map tasks on ECUs and considered power consumption for computation and communication as the cost function to enable shutdown of an ECU if no tasks are needed in a specific driving situation.…”

Formulating customized specifications for resource allocation problem of distributed embedded systems

“…OM1: As mentioned in Section 4.1, pseudo logic variable xi = 1 xi = 0 are exceptions of PB propositions. They do not need to be transformed into the normal form as given in (10). Another kind of exceptions are the propositions in form xi = xj or xi = xj.…”

“…A key architectural design task of these systems is to deploy computational and communicational resources on a distributed hardware platform. Mehiaoui et al [10] defined this design task in the AUTOSAR [1] framework as a PPS problem which is short for Placement (functions to execution nodes), Partitioning (functions to tasks and and signals to messages) and Scheduling (assign priorities to both tasks and messages). The complexity of this design task can be attributed to the huge amount of design options.…”