Field Level Hard Real-time Ethernet Scheduling Algorithm Research And Application,

Functions of real-time computing are very powerful. It is widely applied to aviation , spaceflight , nuclear power energy sources , industry control , military affairs command control , artillery auto control and so on. Especially because of development of real-time computing and computer network technology, Field bus controlling systems continue to be applied at practice. Field buses are the link among digital communication equipment. Controlling equipments are linked together to become open digital and multicast communication controlling local network by the way of Field bus ,and are linked to become controlling system by the way of network node. Because of all kinds of reasons, Field bus controllling equipments of many factories aren't compatible and their bandwidth is also narrow. At the same time, Field bus also can not meet require of real time communication transmitssion controlling. With deeply being applied, the user brings forward eager requires with operator, bandwidth, transmission distance, fault tolerance, reliability and price of modern industry controlling software. It is widely applied because Ethernet have a series of advantages in non real-time network.But it can not be applied to real-time network because Ethernet adopt CSMA/CD technology and exponential increasing mechanism. But hard real-time networks are required to be developed because of social development and requirement of national defense. So the researchers work at research of hard real-time network technology based on bottom Ethernet. In this thesis, the author brings forward the theory of field hard real-time Ethernet based on original knowledge and theory. Especially, the author realizes network scheduling algorithm in field hard real-time theory. .The author roundly analyzes network scheduling algorithm of hard real-time network in this paper and thinks present algorithm have a series of shortcoming. For example, the first is low utilization of CPU, the second is that it can not hold out dynamic changing period task, the third is the low efficence while processing non-period task, the fourth is that it isn't predictibale whether the task computing is successful, the fifth is not confirmation of instantaneous overload. Thus the author...