Reliability is an important aspect of operating safety-critical applications of networked systems, such as in avionics. In order to ensure sufficient reliability, one approach is to apply redundancy. However, redundancy may be costly if the degree of redundancy is too high. This paper proposes an optimization-based framework for guaranteeing the desired reliability of any graph-based networked system. It focuses on integrated modular avionics architecture, based on the use of minimally redundant components. This framework consists of two steps. The first is to compute the minimum number of components within the architecture, using a geometric program. The second is to determine the topology with the minimal number of connections between these components, using a signomial program. Finally, the method is illustrated on a small example network and a larger network of the A350 avionics architecture.
This paper presents a combined scheduling and path planning scheme for offloading of safety-critical applications in a group of networked vehicles. Due to the dynamic nature of such systems, a scheduler for applications to be offloaded needs to account for the changes of the network parameters. Such a scheduling scheme based on a mixed-integer program (MIP) is detailed with attention to the time-varying duration of message passing operations. Given knowledge of the scheduling requirements of applications to be offloaded, a path planner for a group of networked vehicles can ensure the resulting path will yield a feasible schedule for offloading. An method for generating such a path is described, utilizing the scheduling scheme and a mixedinteger program for path planning with integral communication capacity constraints. The combined scheme is demonstrated in a case study consisting of a pair of UAS equipped with a line-ofsight inter-vehicle a communication link and onboard processors. We show how the scheme can be used to make individual vehicles of the group fault-tolerant to processor failures.Index Terms-communication-aware path-planning, faulttolerance, scheduling, offloadingThe authors would like to thank General Atomics for the gift that supported this work.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.