On the joint utility accrual model

Summary form only given. We extend Jensen's time/utility functions and utility accrual model with the concept of joint utility functions (or JUFs) that allow an activity's utility to be described as a function of the completion times of other activities and their progress. We also specify the concept of progressive utility that generalizes the previously studied imprecise computational model, by describing an activity's utility as a function of its progress. Given such an extended utility accrual model, we consider the scheduling criterion of maximizing the weighted sum of completion time, progressive, and joint utilities. We present an algorithm called the combined utility accrual algorithm (or CUA) for this criterion. Experimental measurements with an implementation of CUA on a POSIX RTOS illustrate the effectiveness of JUFs in a class of applications of interest to us.

[1]  E. Douglas Jensen.  Asynchronous Decentralized Real-Time Computer Systems , 2000 .

[2]  Sang Hyuk Son,et al.  Feedback Control Real-Time Scheduling: Framework, Modeling, and Algorithms* , 2001, Real-Time Systems.

[3]  C. D. Locke,et al.  Best-effort decision-making for real-time scheduling , 1986 .

[4]  Raymond Keith Clark,et al.  Scheduling dependent real-time activities , 1990 .

[5]  Arkady Kanevsky,et al.  An Adaptive, Distributed Airborne Tracking System ("process the Right Tracks at the Right Time") , 1999, IPPS/SPDP Workshops.

[6]  Binoy Ravindran,et al.  Choir: a real-time middleware architecture supporting benefit-based proactive resource allocation , 2003, Sixth IEEE International Symposium on Object-Oriented Real-Time Distributed Computing, 2003..

[7]  Riccardo Bettati,et al.  Imprecise computations , 1994, Proc. IEEE.

[8]  Ken Chen,et al.  A scheduling algorithm for tasks described by Time Value Function , 1996, Real-Time Systems.

[9]  E. Douglas Jensen.  Asynchronous Decentralized Realtime Computer Systems , 1992, NATO ASI RTC.

[10]  Arkady Kanevsky,et al.  An adaptive, distributed airborne tracking sysem , 1999 .

[11]  Daniel Mossé,et al.  Value-density algorithms to handle transient overloads in scheduling , 1999, Proceedings of 11th Euromicro Conference on Real-Time Systems. Euromicro RTS'99.

[12]  Binoy Ravindran,et al.  Time-utility function-driven switched Ethernet: packet scheduling algorithm, implementation, and feasibility analysis , 2004, IEEE Transactions on Parallel and Distributed Systems.

[13]  Binoy Ravindran,et al.  A utility accrual scheduling algorithm for real-time activities with mutual exclusion resource constraints , 2006, IEEE Transactions on Computers.