A Context-Aware Reflective Middleware Framework for Adaptive Real-time Vehicle Applications

Software has become crucial to develop vehicle systems. Future unmanned intelligent vehicle safety systems will increasingly rely on situational contexts collected at runtime through temporally built ad-hoc and dynamic networks for vehicle-to-vehicle and vehicle-to-roadside communications and dynamic adaptation to the contexts to improve vehicle safety and reduce traffic congestion. Context-aware reflective middleware, which can measure real-time contexts and accordingly reconfigure the behavior of supported applications, is an important technique to enhance the affordability, flexibility, and adaptability of the future vehicle safety systems. However, the long reconfiguration time of existing context-aware reflective middleware cannot satisfy the stringent real-time requirement of the vehicle systems and thus limits its adoption. In this paper, we present MARCHES, a context-aware reflective middleware framework, which improves the reconfiguration efficiency for engineering adaptive real-time vehicle applications in dynamic environments. Different from traditional single component-chain based middleware, MARCHES supports an original structure of multiple component chains to reduce local behavior change time. Further, according to the new structure, a novel synchronization protocol using active messages is proposed to reduce distributed behavior synchronization time. Experimental results show that the reconfiguration time of MARCHES is reduced from seconds (s) to hundreds of microseconds (μs). Evaluations demonstrate that MARCHES is also robust and scalable and generates small memory footprint, which makes it suitable for supporting real-time vehicle applications.