Research On Causal Consistency Control Technology In Large-Scale Distributed Virtual Environment

With the development of virtual reality and internet techniques, more and more distributed virtual environments (DVE) are employed. As we known, time-space consistency is most important to guarantee the reality of DVEs, and the causal consistency is the first necessary requirement to achieve time-space consistency. Causal consistency is vital to the fidelity and credibility of the DVE, which make it an urgent problem to solve in large-scale distributed virtual environment. This thesis research on the causal relation models and the causal consistency control technologies based on the modeling of causal relations.Firstly, the "happened before"-based causal relation model is analyzed in depth. This model do not capture the actual causal relations and the network overhead caused by the corresponding causal consistency control mechanism is related to the number of nodes in the system, which is fatal to the scalability of the DVE.Then two types of events in DVE are distinguished according to the features of the DVE, there are interaction event and update event, and a causal relation model based on the interaction event is defined. This causality model captures the actual causal relations among events and relaxes the "happened before" relations, remarkable reduce the number of causal related nodes at every logical time advancing step, so make it possible to propose a causal consistency control mechanism independent of the number of nodes in the system.Based on the interaction-based causal relation model, this thesis propose a interaction-based causal consistency control mechanism to maintain the interaction-based causality in DVEs, while improve the responsiveness and scalability of the DVE system. A interaction-based causal consistency control algorithm is implemented based on the HLA/RTI , and a causal consistency control middleware between RTI and applications is designed and developed.To validate the efficiency of the interaction-based causal consistency control algorithm on HLA/RTI platform, we design and implement a validation platform for causal consistency control mechanism based on HLA/RTI. Different causal consistency control mechanism can be compared and validated on this validation platform. The experiment results show that the interaction-based causal consistency control mechanism has a good effect on causal consistency control, eliminate up to 98% causal non-consistency events and reduce about 30% extra message processing time compared to the "happened before"-based causal consistency control mechanism.