Peformance control of tasks in a multiprocessing environment
In time-dependent environments which exist in process control and/or monitoring applications a designer may need to use multiple processors to insure the necessary task performance. The feasibility of using processing power multiplexing techniques to accomplish this is investigated. Processing power is adjusted by time division multiplexing (internal to a processor) and input-data-stream multiplexing (external to a processor). In the m/g/c queuing model studied, each software task is assigned a job queue, and its service rate capability and total system delay time mean and variance are considered. A profit function approach is used to provide a relative basis for distributing the processing power over the tasks, and a numerical optimization strategy developed and tested. 8 references.
- OSTI ID:
- 5139747
- Resource Relation:
- Conference: Sponsored by IEEE, Port Chester, NY, USA, 31 Oct 1983
- Country of Publication:
- United States
- Language:
- English
Similar Records
Multiple-task real-time PDP-15 operating system for data acquisition and analysis
HPC-Colony: Services and Interfaces to Aupport Systems With Very Large Numbers of Processors