Abstract:In previous work we presented the DTRD algorithm, an optimistic synchronization algorithm for parallel discrete event simulation of multi‐agent systems, and showed that it outperforms Time Warp and time windows on a range of test cases. DTRD uses a decision‐theoretic model of rollback to derive an optimal time to delay read event so as to maximize the rate of LVT progression. The algorithm assumes that the inter‐arrival times (both virtual and real) of events are normally distributed. In this paper we present … Show more
“…PDES-MAS uses optimistic synchronisation. Synchronisation algorithms that have developed for PDES-MAS as reported in the wors of (Lees, Logan, and Theodoropoulos 2003), (Lees, Logan, and Theodoropoulos 2004), (Lees et al 2005), (Lees et al 2006a), (Lees et al 2006b), (Lees, Logan, and Theodoropoulos 2008), and (Lees, Logan, and Theodoropoulos 2009). In general, each SSV is associated with a list of Write Periods representing the values taken by the variable at different logical times through the simulation.…”
Multi-Agent Systems (MAS) are increasingly used to solve larger and more complex problems. To provide the computational resources needed to do this, MAS are increasingly distributed over multiple computational platforms. Different approaches for distributing MAS have been proposed over the years. One problem remains central whichever approach is used: how to translate MAS behaviour into a format suitable for the distribution approach used. In this paper we describe the Agent Distributed Shared Memory Interface (ADSMI), an interface between a general MAS description and the PDES-MAS platform as an implementation of a Distributed Shared Memory (DSM) system for distributed MAS simulations. The ADSMI provides a translation of MAS behaviour into event-interactions with PDES-MAS, as well as functionality for handling time progressing in the MAS, and a messaging mechanism between agents.
“…PDES-MAS uses optimistic synchronisation. Synchronisation algorithms that have developed for PDES-MAS as reported in the wors of (Lees, Logan, and Theodoropoulos 2003), (Lees, Logan, and Theodoropoulos 2004), (Lees et al 2005), (Lees et al 2006a), (Lees et al 2006b), (Lees, Logan, and Theodoropoulos 2008), and (Lees, Logan, and Theodoropoulos 2009). In general, each SSV is associated with a list of Write Periods representing the values taken by the variable at different logical times through the simulation.…”
Multi-Agent Systems (MAS) are increasingly used to solve larger and more complex problems. To provide the computational resources needed to do this, MAS are increasingly distributed over multiple computational platforms. Different approaches for distributing MAS have been proposed over the years. One problem remains central whichever approach is used: how to translate MAS behaviour into a format suitable for the distribution approach used. In this paper we describe the Agent Distributed Shared Memory Interface (ADSMI), an interface between a general MAS description and the PDES-MAS platform as an implementation of a Distributed Shared Memory (DSM) system for distributed MAS simulations. The ADSMI provides a translation of MAS behaviour into event-interactions with PDES-MAS, as well as functionality for handling time progressing in the MAS, and a messaging mechanism between agents.
“…To address these issues, we developed the PDES-MAS framework for the simulation of multi-agent systems [7,[14][15][16][17][18][19]. PDES-MAS adopts a standard parallel discrete event approach.…”
Section: Distributed Simulations Of Masmentioning
confidence: 99%
“…exponential distributions) are therefore not necessarily appro-priate models of a situated MAS. See [19] for an evaluation of the performance of DTRD in simulations where the real and virtual inter-arrival times are not normally distributed.…”
Section: Estimating the Probability Of A Straggler Writementioning
We present a detailed analysis of the performance of the Decision Theoretic Read Delay (DTRD) optimistic synchronization algorithm for simulations of multi-agent systems (MAS). We develop an abstract characterization of the access patterns found in MAS simulations based on the simulation's degree of coupling and skew. Using this characterization, we generated stereotypical test cases which we used to compare the performance of the DTRD algorithm with that of Time Warp and time windows. To determine if the test cases reliably predict performance in a real agent simulation, we compared the predictions made by the test cases with performance results from the Boids agent simulation benchmark for a range of simulation parameters. The results indicate that the test cases are useful predictors of performance in real agent simulations, that DTRD adapts to the mixtures of coupling cases found in real agent simulations and is capable of tracking changes in coupling during the simulation.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.