Multiprocessor trade off in real time industrial systems

The authors project a possible tradeoff between the multiple-processor requirement and permissible delay in execution of tasks, without missing the deadlines. An analytical model using semi-Markov processes is developed and used for analysis. The system parameters involved in the modeling are the average execution times, average interrequest times, scheduling time for the tasks, and number of critical/noncritical tasks present in the system. The results obtained show the possible balance between the number of processors and the average delay in execution of critical and noncritical tasks in a typical system. This balance is useful in increasing the processor utilization, reducing the system complexity, and making it cost-effective and fail-safe against the missing of deadlines.<<ETX>>

[1]  Hideyuki Tokuda,et al.  A Time-Driven Scheduling Model for Real-Time Operating Systems , 1985, RTSS.

[2]  Karsten Schwan,et al.  High-performance operating system primitives for robotics and real-time control systems , 1987, TOCS.

[3]  Trevor N. Mudge,et al.  A semi-Markov model for the performance of multiple-bus systems , 1985, IEEE Transactions on Computers.

[4]  Victor K. L. Huang,et al.  A Scheduler for Real-Time Task Control in Microcomputers , 1978, IEEE Transactions on Industrial Electronics and Control Instrumentation.

[5]  J. Karjalainen.  A classification scheme for embedded control systems , 1988, Proceedings.14 Annual Conference of Industrial Electronics Society.

[6]  Harold S. Stone.  Introduction to Computer Architecture , 1980 .

[7]  R. Z. Zurawski,et al.  Microcomputer Based Real-Time Control , 1987, Other Conferences.

[8]  Krithi Ramamritham,et al.  Dynamic Task Scheduling in Hard Real-Time Distributed systems , 1984, IEEE Software.

[9]  Pranay Chaudhuri,et al.  A Real-Time Process Scheduler for a Ring-Type Microcomputer Network , 1985, IEEE Transactions on Industrial Electronics.