Intelligent edge systems constitute a key growth segment within the cloud-backed cognitive IoT marketplace. In this context, connected autonomous and semi-autonomous vehicles constitute one of the most prominent examples, where cars communicate with each other and with the infrastructure through vehicular ad hoc networks (VANETs). The dynamic (ad hoc) nature of VANETs along with the strict performance and security requirements of vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications create unprecedented challenges in terms of visualization, analysis and introspection of wireless vehicular networks. Traditional graph-based network analysis tools with graphical user interfaces fall short of the capabilities needed to capture the behavior of vehicular networks with enough precision and timeliness.
In this work, we propose and evaluate the use of mixed (virtual and augmented) reality to support visualization of wireless software-defined-radio (SDR) networks in automotive applications. Specifically, we build upon XRShark, a mixed reality network introspection platform, to enable real-time visualization of V2V and V2I interactions within a connected car application called ERA. The latter is an open-source workload that drives the IBM-led EPOCHS project under the DARPA-sponsored Domain-Specific System on Chip (DSSoC) program . Through this approach, we unveil more intuitive vehicular network analysis (not possible using state-of-the-art tools) while we open up opportunities for spectrum analysis and threat detection.
|Secondary Topic||SDR Instrumentation and Control|