Priority inversion at the network adapter when scheduling messages with earliest deadline techniques

In this paper we present a novel approach in the study of the predictability of real-time message transmission and its relationship with the design of network adapters for real-time distributed systems. The aim is to limit the occurrence of large priority inversions among messages, so as to achieve a better degree of predictability. We show that when the proper network adapters are used in conjunction with earliest deadline message scheduling the loss in processor utilization is minimized and predictable.

[1]  Chung Laung Liu,et al.  Scheduling Algorithms for Multiprogramming in a Hard-Real-Time Environment , 1989, JACM.

[2]  Andy J. Wellings,et al.  Analysing real-time communications: controller area network (CAN) , 1994, 1994 Proceedings Real-Time Systems Symposium.

[3]  Domenico Ferrari.  A new admission control method for real-time communication in an internetwork , 1995 .

[4]  Kang G. Shin,et al.  Non-preemptive scheduling of messages on controller area network for real-time control applications , 1995, Proceedings Real-Time Technology and Applications Symposium.

[5]  M. Di Natale,et al.  Earliest deadline message scheduling with limited priority inversion , 1996, Proceedings of the 4th International Workshop on Parallel and Distributed Real-Time Systems.

[6]  Wei Zhao,et al.  The timed-token protocol for real-time communications , 1994, Computer.

[7]  Hermann Kopetz,et al.  TTP - A time-triggered protocol for fault-tolerant real-time systems , 1993, FTCS-23 The Twenty-Third International Symposium on Fault-Tolerant Computing.