Data distribution management schemes for HLA-compliant distributed simulation systems

Abstract

Data Distribution Management (DDM), one of the six services provided by High Level Architecture and Run-Time Infrastructure, provides an efficient and scalable mechanism for data routing among hosts in distributed simulations. Traditional, DDM schemes are classified into two main types, region-based methods and grid-based methods. Currently, the time, computation and communication overhead of DDMs are still issues for large-scale simulations. We proposed two new DDM schemes addressing these issues. Our first algorithm, which we refer to as optimized dynamic grid-based DDM scheme, aims at further reducing irrelevant data that might be received by simulation entities in dynamic grid-based approach [11], by enforcing a second level of sender-side data filtering mechanism. Our second algorithm, which we refer to as grid-filtered region-based DDM, uses a threshold value of coverage percentage to determines if exact matching is necessary. In this thesis, we present and discuss the implementation of our proposed DDM algorithms, and report on their performance based on an extensive set of simulation experiments. Last but not least, we present the preliminary work we have done on real-time enabling scheme to RTI for HLA-compliant simulations.