Title :
A predictable real-time kernel for distributed multisensor systems
Author :
Lee, Insup ; King, Robert B. ; Paul, Richard P.
Author_Institution :
Dept. of Comput. & Inf. Sci., Pennsylvania Univ., Philadelphia, PA, USA
fDate :
6/1/1989 12:00:00 AM
Abstract :
The authors present a real-time kernel developed to support a distributed multisensor system encountered in robotics applications. To ensure predictability, the kernel provides services with bounded worst-case execution times. In addition, the kernel allows the programmer to specify timing constraints for process execution and interprocess communication. The kernel uses these timing constraints both for scheduling processes and for scheduling communications. To illustrate the kernel, the authors describe a multisensor system being developed on their distributed real-time system. They present the measured performance of kernel primitives along with conclusions and remarks regarding distributed real-time systems.<>
Keywords :
distributed processing; network operating systems; real-time systems; scheduling; Ethernet; distributed multisensor systems; distributed real-time system; interprocess communication; kernel primitives; microVAX II; predictability; predictable real-time kernel; process execution; robotics applications; scheduling; sensor fusion; timing constraints; Arm; Intelligent robots; Job shop scheduling; Kernel; Multisensor systems; Operating systems; Processor scheduling; Real time systems; Robot sensing systems; Timing;
