An Extensible Fault-based Predicate Testing Toolset for Wireless Sensor Network Software Applications

Project: Research

View graph of relations

Researcher(s)

Description

Wireless sensor networks (WSN) are an enabling technology in ubiquitous computing to support Hong Kong as a next-generation services centre. The quality of embedded WSN software is essential for assuring its success. Since the majority of the programs involved are control software, the testing of control predicates is crucial. Nonetheless, this task is difficult because of memory constraint and primitive operating system support. Software engineers spend a disproportionate amount of effort on software testing. We propose to develop a fault-based predicate testing toolset for WSN software. We shall use the programs developed on TinyOS, the de facto WSN operating system, as a testbed.The toolset will benefit the local industry by automatically measuring test completeness, or how well a program is free from a hierarchy of fault classes. The implementation of the hierarchy is extensible to cater for future enhancements. It generates the patterns for desirable test cases and allows software engineers to refine them iteratively. It minimizes the effort for writing test scripts in embedded programming languages. It also significantly shortens testing time and improves on software quality. A major technical challenge is to fit the test instrumentation under stringent memory constraints. No existing toolset provides this capability.

Detail(s)

Project number9449003
Grant typeITF
StatusFinished
Effective start/end date1/09/0630/11/07