High performance distributed cluster-based individual-oriented fish school simulation

Towards a conceptual framework to support adaptative agent-based systems partitioning

“…Therefore clustering agents into partitions based on the position may not be effective in the cases where agents are acting individually. In [7], the authors tackle the faced challenges when distributing individual-oriented systems through distributed architecture. To obtain an efficient distribution the authors adopt a cluster based method using Voronoi diagrams and covering radius criterion.…”

“…The used cluster based approach shows better performance over a grid based approach in case of large-scale agent simulations. [8] is a continuation for what was done in [7], the authors added a proximity load balancing strategy to maintain an equal load between partitions during the simulation time. In [4], a comparative study between three partitioning algorithms for crowd simulations is presented.…”

“…The main challenge, for MAS designers and developers is how to select the appropriate partitioning mechanism for a given MAS. Partitioning mechanisms can be classified into two families [7][4] [8]: 1) Cluster-based approaches in which the agents are clustered based on given criteria then assigned to different nodes.2) Grid-based approaches, called also region-based approach consists of decomposing the space component (environment) into multiple portions. Each portion together with the set of agents residing in it are assigned to a node.…”

“…For the crowd simulations, many algorithms were evaluated like the k-mean clustering approach in [5], the grid-based partitioning method in [6], as well as the convex Hull algorithm in [4]. Also, in [7][8], the authors presents respectively a cluster-based approach and a proximity load balancing method to partition individual-oriented fish school simulations. In [10], the authors present the Pandora framework to distribute large-scale social simulations on the High Performance Computing (HPC) infrastructure.…”

Impact of ICT-mediated collective awareness on urban mobility

“…To the best of our knowledge, a cellular automata (CA) based city-scale evacuation simulation with hundred of thousands agents has not been attempted before. Many studies for large scale crowd simulation are conducted but they only focus on efficient simulation of agents and ignore the spatial dimension of the scenario (Solar et al 2011;Perumalla et al 2011). We believe that the spatial dimension is of high significance and must be considered in such scenarios.…”