A parameterized compositional multi-dimensional multiple-choice knapsack heuristic for CMP run-time management

Abstract

Modern embedded systems typically contain chip-multiprocessors (CMPs) and support a variety of applications. Applications may run concurrently and can be started and stopped over time. Each application may typically have multiple feasible configurations, trading off quality aspects (energy consumption, audio-visual quality) with resource usage for various types of resources. Overall system quality needs to be guaranteed and optimized at all times. This leads to the need for a run-time management solution that selects an appropriate system configuration from all the application configurations of active applications. This run-time management problem can be phrased as a multi-dimensional multiplechoice knapsack (MMKP) problem. We present a compositional heuristic to solve MMKP, that due to the compositionality is better suited to CMP run-time management than existing heuristics that are all not compositional. Our heuristic outperforms the best-known heuristic to date. The heuristic is parameterized, leading to the additional advantage that it allows to trade off execution time vs. solution quality, and to bound the time needed to compute a solution. The latter makes it particularly well-suited for resource-constrained embedded platforms.

A hierarchical run-time adaptive resource allocation framework for large-scale MPSoC systems

“…To overcome the shortcomings of pure static [4,11,15] and dynamic [21,36] task mapping algorithms, hybrid (semi-static) approaches have become increasingly popular in recent years. Usually, in this kind of approaches, multiple mapping solutions are found at design time and applied at run time based on the current state of the system.…”

Predictable dynamic embedded data processing

“…[15] explores trade-offs between time and memory usage for memory architectures for data storage and transfer. For run-time trade-offs, [16], [17], [18] present heuristics for run-time configuration selection for multiple applications with design-time evaluated resource-usage tradeoffs on shared resources.…”

“…In fact, the problem is very much akin to the well-known Multi-dimensional Multiplechoice Knapsack Problem (MMKP) [33], which is known to be NP Complete. Work has been done to realize fast heuristics for run-time configuration selection [17], [16], [18]. [18] describes a heuristic based on the Pareto Algebraic model.…”

“…Work has been done to realize fast heuristics for run-time configuration selection [17], [16], [18]. [18] describes a heuristic based on the Pareto Algebraic model. The runtime composition of system-level Pareto-optimal configurations from components trade-offs using Pareto-Algebra is a very generic approach.…”

Online Heuristic for the Multi-objective Generalized Traveling Salesman Problem

“…The Compositional Pareto-algebraic Heuristic (CPH) [2] [3] is a heuristic approach that has been specifically designed for multi-objective, multiple-choice optimization problems. CPH has been applied successfully to the Multi-dimensional Multiple-choice Knapsack Problem (MMKP) [3].…”