This proposed project will further develop the concept of Evidence Based Scheduling as a mechanism to efficiently utilize the communication and network resources of a real-time distributed message-based computer system. Initially the project will enhance and refine the outline evidence based scheduling mechanism developed at Worcester Polytechnic Institute. A theoretical analysis of this scheduling mechanism will be undertaken and its performance potential compared with existing mechanisms such as those proposed by Stankovic (the Spring Project), Blazewicz, and the work at Shin's Real-Time Computing Laboratory. A simulation of an autonomous mobile robot in an industrial environment will then be used to verify the operation of the scheduling and focus- of-attention mechanisms. It is intended that this activity will demonstrate the potenital of evidence based scheduling for real-time applications such as the control of autonomous systems and enable the work to be brought to the attention of the real-time research community by suitable publications.
