Task scheduling-using neural networks within hardware co-processors

The work described is part of an overall research programme relating to real-time embedded systems. Specifically it is concerned with: the development of scheduling policies based on neural network (NN) principles upgraded with genetic learning; the application of hardware co-processors to task scheduling (J.E. Cooling, 1993); and the integration of hardware and software facilities to provide full real-time executive support to application software (B. Korousic and J.E. Cooling, 1994).<>