Abstract

Multi-core processors, despite their technical and economic advantages, are yet hesitantly adopted in safety-critical embedded application domains such as automotive and avionics. A key issue is the potential interference on shared resources, such as interconnect and memory, between applications of different criticality which are running on a Multi-Processor System-on-Chip (MPSoC) with tens of individual cores. In this paper we propose the introduction of established protection switching, known from synchronous data networks, to a hybrid Network-on-Chip (NoC) in order to provide fault-tolerance for critical connections. Our hybrid NoC combines configurable Time-Division-Multiplexing (TDM) for critical task traffic with conventional packet switching for Best-Effort (BE) traffic. We analyze three different protection switching schemes for their worst case latencies in case of faulty NoC links and their resource overheads. Simulations with random traffic and 10% reserved resources for TDM connections reveal that the degradation of BE traffic performance due to the proposed TDM protection switching for critical traffic remains limited to about a 5% lower injection rate even in case of 1+1 protection, which can hence be considered affordable. We conclude that the proposed hybrid NoC is a suitable way to provide both hard real-time guarantees and fault-tolerance for critical connections in advanced MPSoCs.

Full Text
Paper version not known

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call

Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.