Complex software systems are commonly developed by integrating multiple, occasionally Off-The-Shelf (OTS), components. This process results into a more modular design and reduces development costs; however, it raises new dependability challenges in case of safety critical systems. Testing activities conducted during the development of the individual components might be not enough to ensure a proper safety level after the integration. The failures of the components and their impact on the overall system safety have to be assessed in critical scenarios. This paper proposes a method to support component integration in complex software systems. The method uses (i) the knowledge of the architectural dependencies among the system components, and (ii) the results of failure-modes emulation experiments, to assess both error propagation phenomena within the system and the criticality of the components in the system architecture. This information is valuable to design effective error-mitigation means and, when needed, to select the most suitable OTS item if multiple equivalent options are available. The method is applied to a real world Air Traffic Control system, developed in the context of an academic-industrial collaboration.
Criticality-Driven Component Integration in Complex Software Systems
PECCHIA, ANTONIO;
2011-01-01
Abstract
Complex software systems are commonly developed by integrating multiple, occasionally Off-The-Shelf (OTS), components. This process results into a more modular design and reduces development costs; however, it raises new dependability challenges in case of safety critical systems. Testing activities conducted during the development of the individual components might be not enough to ensure a proper safety level after the integration. The failures of the components and their impact on the overall system safety have to be assessed in critical scenarios. This paper proposes a method to support component integration in complex software systems. The method uses (i) the knowledge of the architectural dependencies among the system components, and (ii) the results of failure-modes emulation experiments, to assess both error propagation phenomena within the system and the criticality of the components in the system architecture. This information is valuable to design effective error-mitigation means and, when needed, to select the most suitable OTS item if multiple equivalent options are available. The method is applied to a real world Air Traffic Control system, developed in the context of an academic-industrial collaboration.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
