Abstract: Distributed interactive simulation has broad applications in various fields, such as military, engineering, commerce etc . because of its advantages of efficiency, economy, safety and security. Large scale distributed interactive simulation based on Internet is the future direction. But it faces the problem of system scalability because of its requirements for high network bandwidth and computation capacity, which is caused by redundant data when the system’s scale increases. This situation can be alleviated to great extent by data filtering based on multicast. In this paper the key factors are analyzed, which limit the system scalability from the point of view of system resources and determine a criterion which can be used to evaluate the performance of a filtering scheme. Based on the above work, a new scheme called adaptive region updating algorithm is proposed to balance the filtering spending and filtering efficiency.