On-Line Scheduling to Maximize Task Completions

The problem of uniprocessor scheduling under conditions of overload is investigated. The system objective is to maximzze the number of tasks that complete by their deadlines. For this performance metric it zs shown that, in general, any on-line algorithm may perform arbitrarily poorly as compared to a clairvoyant scheduler. Restricted znstances of the general problem for whzch on-line schedulers can provide a guaranteed level of performance are identijied, and on-line algorithms presented for these speczal cases.

• Type: Conference
• Year: 1994
• Where: RTSS
• Authors: Sanjoy K. Baruah, Jayant R. Haritsa, Nitin Sharma
• Keywords: Control Systems, On-line Algorithms, RTSS 1994, Uniprocessor Scheduling, Whzch On-line Schedulers