Adaptive OS noise mitigation for microbenchmarking on mobile platforms
Journal Publication ResearchOnline@JCUAbstract
Microbenchmarking provides developers with flexible tools for fine-grained application performance analysis. This facilitates enhancements to application responsiveness and delivery of mature, robust consumer products. This study investigates the levels of Operating System (OS) noise on Apple iPad mobile devices. OS noise introduces variations in application performance that interfere with microbenchmark results. OS noise manifests in collected data through extreme outliers and variations in skewness. After removal of outliers through an iterative, semi-automated process, our results demonstrate changes in OS noise levels arise from changes in device state. We present an adaptive OS noise mitigation technique that dynamically selects transformation strategies to alter the granularity of measurements, whilst satisfying a set of correctness constraints. We validate this approach using the data collected in our study. Our proposed approach adaptively mitigates the effects of OS noise on microbenchmarking results, whilst maintaining the maximum possible noise-reduced accuracy.
Journal
N/A
Publication Name
N/A
Volume
24
ISBN/ISSN
1041-2808
Edition
N/A
Issue
4
Pages Count
17
Location
N/A
Publisher
Association of Management
Publisher Url
N/A
Publisher Location
N/A
Publish Date
N/A
Url
N/A
Date
N/A
EISSN
N/A
DOI
N/A