Xing Kexing¹, Zuo Decheng¹, Zhou Haiying¹ and HOU Kun-Mean², ¹Harbin Institute of Technology, China and ²University of Blaise Pascal, France

Wireless Sensor Networks (WSNs) are used in many application fields, such as military, healthcare, environment surveillance, etc. The WSN OS based on event-driven model doesn’t support real-time and multi-task application types and the OSs based on thread-driven model consume much energy because of frequent context switch. Due to the high-dense and large-scale deployment of sensor nodes, it is very difficult to collect sensor nodes to update their software. Furthermore, the sensor nodes are vulnerable to security attacks because of the characteristics of broadcast communication and unattended application. This paper presents a task and resource self-adaptive embedded real-time microkernel, which proposes hybrid programming model and offers a two-level scheduling strategy to support real-time multi-task correspondingly. A communication scheme, which takes the “tuple” space and “IN/OUT” primitives from “LINDA”, is proposed to support some collaborative and distributed tasks. In addition, this kernel implements a run-time over-the-air updating mechanism and provides a security policy to avoid the attacks and ensure the reliable operation of nodes. The performance evaluation is proposed and the experiential results show this kernel is task-oriented and resource-aware and can be used for the applications of event-driven and real-time multi-task.

WSN, event-driven, thread-driven, scheduling strategy, over-the-air, security