120 Int. J. Simulation and Process Modelling, Vol. 5, No. 2, 2009 Distributed simulation of large-scale and detailed models G. D’Angelo* and M. Bracuto Department of Computer Science, University of Bologna, Mura Anteo Zamboni 7, 40127 Bologna, Italy E-mail: gdangelo@cs.unibo.it E-mail: bracuto@cs.unibo.it *Corresponding author Abstract: We present a new approach for the distributed simulation of large-scale and detailed models. Our approach increases the simulator speed jointly addressing two main problems of distributed simulation: the reduction of the communication overhead and the load-balancing in the execution cluster. The proposed method dynamically reconfigures the simulation, considering the performance of each part of the execution architecture. In this way, commercial-off-the-shelf hardware can be used for fast and cost-effective simulations. The performance evaluation, based on the 802.11 DCF protocol, demonstrates that this approach is feasible for the detailed simulation of very large-scale models such as wireless networks. Keywords: simulation; distributed simulation; load balancing; wireless networks. Reference to this paper should be made as follows: D’Angelo, G. and Bracuto, M. (2009) ‘Distributed simulation of large-scale and detailed models’, Int. J. Simulation and Process Modelling, Vol. 5, No. 2, pp.120–131. Biographical notes: Gabriele D’Angelo received the Laurea Degree (summa cum laude) in Computer Science in 2001, and a PhD Degree in Computer Science in 2005, both from the University of Bologna, Italy. He is an Assistant Professor at the Department of Computer Science of the University of Bologna. His research interests include parallel and distributed simulation, distributed systems, internet games and computer security. He is the author of several publications on these topics. Michele Bracuto received the Laurea Degree in Computer Science in 2002, from the University of Bologna, Italy. He is a research associate at the Department of Computer Science of the University of Bologna, Italy. His research activity includes parallel and distributed simulation, distributed systems and quality-of-service of multimedia on internet. 1 Introduction A simulation is a system that represents or emulates the behaviour of another system over time (Fujimoto, 2000). Nowadays, the simulation technique is of primary importance in the design, implementation and performance evaluation of many real world systems. The systems of interest are often very large, as number of entities, and characterised by very fine-grained models. In detail, many of them are composed of many entities or parts, and characterised by a very dynamic nature and progress. In designing simulations, one key factor is the level of detail of the simulated model. That is the complexity of the representation of real world entities and interactions, within the simulation. The importance of the level of detail is twofold: first of all, the correctness of the simulation results is deeply influenced by the amount of details involved in the representation of the simulated system (Perrone et al., 2003). In a performance evaluation, an inadequate amount of details in the model representation can lead to misleading or wrong results (Cavin et al., 2002). On the other side, the level of detail affects the time required to run the simulation (Hedidemann et al., 2001). An increased amount of details translates to many factors: more computation is required to evolve the simulation, more memory is used to represent the modelled system and an increased amount of communication among the simulated entities is necessary. The common approach to simulators implementation is based on monolithic design, that is a single execution unit manages the evolution of the whole simulation. In the scenario described above this approach often gives unsatisfactory results. Firstly, a single execution unit is unable to provide the scalability required for the simulation of large-scale models. A lot of memory is Copyright © 2009 Inderscience Enterprises Ltd.