Light-Weight Run-Time Monitoring System for Examination of Internet-of-Things Devices
Thanks to advances of semiconductor and communication technologies, a number of devices can be connected over a network, which brings an era of Internet-of-Things (IoT). It is expected that the number of Internet-of-Things (IoT) devices reaches 50 billion in 2020. After development and test of individual IoT devices, they should be integrated and eventually deployed. If the system does not work at this moment, the problem should be "examined". The purpose of examination is different from that of debugging. Debugging is to identify the root cause of the problem and to fix it whereas examination is to find operational problems such as configuration errors or users' mistakes and to find a faulty device that should be debugged or replaced.
Existing run-time monitoring tools have been found for debugging, diagnosis, and security audit are not suitable for examination of heterogeneous IoT devices. Therefore, the proposed research will examine the effectiveness of a standardized light-weight examination method which is based on processor-level architectural support. The proposed method has been envisioned to become an industry standard like the Joint Test Action Group (JTAG) debugging interface. The proposed technique monitors function calls and memory accesses by hardware monitoring modules integrated with a processor, which incurs minimal overhead to the device under monitoring. Which functions and which memory locations should be monitored needs to be customized to different devices, but the interface and protocol will be the same for all devices. These novel hardware monitoring modules are capable of run-time connection (hot plug), and provide enough information for examination at minimal performance cost.