Addendum to Voltage-Clock-Scaling Adaptive Scheduling Techniques for Low Power in Hard Real-Time Systems

ae IN Section 4 (page 1591, column 1) of this paper [2], some sentences describing the updating of the online unfinished work were inadvertently omitted from the materials submitted for publication. These sentences (which should be inserted after " If both these conditions are satisfied, the processor runs at high voltage ") are as follows: The management of the online unfinished work must be done according to slack times and task deadlines. In [1], we showed a way in which, when a task finishes, the online unfinished work can be updated and the released time reclaimed. In this paper, we adopt the following strategy: When a task finishes without using its worst-case execution time, the slack time can be released to other tasks which have deadlines greater than that of the finishing task. Assume that task T A is scheduled to run at time t. the difference between offline_task(t) and online_task(t) can be simply represented as the sum of two terms. The first term is the slack time that can be released to the task T A at time t. The second term is the difference between the unfinished work of T A in the offline schedule (with inflated execution time) and the unfinished work of the real task T A .