University of Bologna. Department of Computer Science and Engineering (DISI)

Welcome to the PADS homepage, you can find some information about this research group, the active projects and the related publications.

A lot of our research work is based on a new Parallel and Distributed Simulation (PADS) middleware (ARTÌS). The GAIA framework provides a migration-based approach to the simulation. The ARTÌS middleware and the GAIA framework are available for educational and research purposes.

Please feel free to contact us for any question or comment.

Parallel And Distributed Simulation (PADS) is a modeling paradigm that enables efficient implementation of large simulation models represented as a collection of interacting entities, called Simulate Entities (SEs). To speed up the simulation, SEs are assigned to independent execution units for parallel execution. Finding the best strategy for allocating SEs to execution units is one of the most important problems in PADS. Ideally, highly interacting SEs should be placed on the same execution unit so that all communications remain local. However, placing too many SEs on the same execution unit would degrade performance, effectively creating a bottleneck for the entire simulation execution. The problem is further complicated by the fact that many simulation models exhibit non-uniform computation and communication patterns among their components, which can change dynamically at unpredictable times during execution. In this paper, we propose a self-clustering heuristic based on the migration of SEs within the PADS execution architecture. The heuristic is designed to identify and exploit communication locality, with the aim of reducing the communication cost during execution. It adapts automatically to changing interaction patterns by migrating SEs, without imposing strict assumptions on the simulation model and while supporting highly heterogeneous execution platforms. A set of tunable parameters allows users to find the best balance between the gain obtained by clustering highly interacting SEs and the cost of running the heuristic and performing the migrations. We conduct an extensive set of computational experiments to assess the effectiveness of the heuristic on a representative simulation scenario, evaluating the role of the main parameters with the aim of guiding users in selecting optimal values, supporting the development of more advanced heuristics, and providing insights into the definition of fully self-tuning mechanisms.

For more information follow this link.

Parallel And Distributed Simulation (PADS) is a well known modeling paradigm that allows efficient implementation of large simulation models as a collection of interacting entities, called Simulated Entities (SEs). SEs can then be assigned to independent execution units for parallel execution, if possible. Allocating SEs to execution units is one of the most challenging problems in PADS: ideally, highly interacting SEs should be placed on the same execution unit so that all communications are local; however, placing too many SEs on the same processor might degrade performance. Furthermore, many simulation models exhibit non-uniform computation and communication patterns among components, that might change during execution at unpredictable times. In this paper we propose a clustering heuristic that exploits communication locality, with the aim to reduce the communication cost experienced by the PADS during the execution. The heuristic adapts automatically to changing interaction patterns by migrating SEs, and can do so without any user-visible modification of the simulation model; there are, however, some parameters that can be used to tune the heuristic. We perform a large set of computational experiments to assess the effectiveness of the heuristic on a real-world scenario, with the aim of guiding the users in selecting optimal values of the heuristic parameters.

For more information follow this link.

In this paper we perform a systematic literature review of multilevel simulation applied to human mobility and behavior, to identify research trends and contribute towards a better understanding of techniques, methodologies and terminology used within this research area. At the end of our analysis we uncover some recurrent modeling and simulation patterns that appear frequently within specific research areas, and we classify these methodological patterns along several dimensions such as modeling type and level of detail.

For more information follow this link.