Optimizing Component-Based Embedded Software

Abstract

As functionalities required by embedded systems increase, higher levels of abstraction become necessary to fulfill development exigences. At the same time, traditional physical and economical constraints such as size, memory usage and energy consumption subsist, making embedded systems development even more complex as it must face two contradictory requirements. In this paper, we expose how a component-based development framework is able to support these two conflicting demands. We follow a component-based approach to benefit from reusability and separation of concerns at design-time, producing flexible systems. We propose a set of flexibility-oriented properties allowing architects to generate, for a same architecture, a set of systems with different flexibility capabilities; these optimization techniques allow developers to pay for flexibility only where it is actually desired. Experiments conducted on several use-cases confirm the effectiveness of our approach.

Minimal yet effective reconfiguration infrastructures in component-based embedded systems

“…Scenario 1, our standard highly flexible system, uses 61% more memory space than the static, not reconfigurable scenario 5. This confirms previous results in [20,21] and ratify the need of producing a lighter infrastructure for system evolution. Scenario 2 data show that architectural optimizations only reduced memory footprint size by 1,57%.…”

“…We use GCC compiler and, as we are interested in memory footprint size, we force GCC to optimize accordingly (-Os flag on). As it has been previously shown in [20], it exists a direct relationship between memory footprint overheads introduced by the Think compiler components reification and other performance metrics like interfaces methods execution times; consequently, in this paper we focus on memory footprint size measurement.…”

“…The Think compiler makes it possible to generate different binary images having different performance versus flexibility trade-offs, using architectural optimization properties [20], including:…”

A Hardware/Software CBSE Framework for RTOS Services: The Timing Service Case Study

“…At the SW level, a layered architecture was designed to separate different levels of service and technological concerns. However, during code-generation, adequate optimizations (constant propagation, functions in-lining, dead code elimination) could be performed [9] fro a better solution.…”

Towards Balancing Determinism, Memory Consumption and Throughput for RTSJ-Based Real-Time Applications

“…RELATED WORKS There are currently several research works focusing on quantitative analysis of real-time embedded applications [13,14,15,16]. Wandeler et al [13] have proposed an approach to calculate upper and lower bounds on performance measures of real-time applications, such as throughput and resource utilization, in an analytical way.…”