Scheduling algorithms for real-time agent systems

The paper addresses the scheduling issues of real time software agent systems. Different from the traditional client/server approach, an agent contains executable codes and is activated on a server computer after the agent is shipped through the network to the server computer. The advantage of the agent approach is to efficiently utilize the resources and the power of the server computers as well as to reduce the load on the network. A real time agent model is proposed to handle distributed real time applications that dispatch intelligent agents to remote server locations. We investigate different scheduling algorithms on the server computer and compare their performances by simulation.