Injoo Jeong

SESAME: space-efficient stack allocation mechanism for multi-threaded sensor operating systems

This paper proposes a SESAME, which is a Space-Efficient Stack Allocation MEchanism for multi-threaded sensor operating systems. It adaptively adjusts the stack size by allocating or releasing additional stack frame based on the amount of each functions stack usage information. Our experimental results show that the SESAME significantly minimizes spatial overhead of threads stacks with tolerable time overhead compared with fixed stack allocation mechanism of the multi-threaded sensor operating systems.

Arbiter-Recall : a new function update scheme by arbitration between a remote call and a dynamic update for wireless sensor networks

In this paper, we propose Arbiter-Recall, which is a new kind of function code update management. When a function requires updating, Arbiter-Recall adaptively selects an energy-efficient method, deciding between a remote-call and a local-call after updating a function. The run-time decision is made by analyzing the cost of a remote-call, and a local-call with dynamic function update in terms of the energy consumption. Simulation results show that Arbiter-Recall significantly reduces energy consumption compared to existing schemes.

PreeMe: preemptive real-time task management for event-driven sensor operating systems

This paper proposes PreeMe, which is a preemptive realtime task management scheme for event-driven sensor operating systems. By efficiently allocating independent stack space for each task, PreeMe enables the preemptive task management for the event-driven systems. The size of the minimal stack space for each task is determined at compiletime, and dynamic stack allocation is performed at run-time. Simulation results show that PreeMe significantly reduces scheduling latency with small execution overhead.