Figure 3 - uploaded by Adelinde M. Uhrmacher
Content may be subject to copyright.
Source publication
Compartments play an important role in molecular and cell biology modeling, which motivated the development of BETA-BINDERS, a formalism which is an extension of the pi-CALCULUS. To execute BETA-BINDERS models, sophisticated simulators are required to ensure a sound and efficient execution. Parallel and distributed simulation represents one means t...
Similar publications
The problem of scheduling multiproduct and multipurpose batch processes has been studied for more than 30 years using math programming and heuristics. In most formulations, the manufacturing recipes are represented by simplified models using state task network (STN) or resource task network (RTN), transfers of materials are assumed to be instantane...
In operating system the CPU scheduler is designed in such a way that all the resources are fully utilized. With static priority scheduling the scheduler ensures that CPU time will be assigned according to the highest priority but ignores other factors; hence it affects the performance. To improve the performance, we propose a new 2-stage vague logi...
Citations
... Processors use EBs to record the number of events sent and received by LPs, as well as the LVT of each LP and the timestamps of events in transit. GVT computation via the EB approach can avoid message acknowledgement, which is a common solution for the problem of transient event in distributed GVT algorithms [16,17,20,21]. In our algorithm, the size of Event-Bulk is negatively correlated with the accuracy of GVT computation but positively correlated with the performance of GVT computation. ...
... In GVT algorithms for distributed memory platforms, LPs communicate their LVTs and the timestamps of events in transit with each other by exchanging messages. GVT algorithms have to tackle two problems: the LVT simultaneous reporting problem and the transient message problem [20,21]. The LVT simultaneous reporting problem is that it is difficult to measure LVTs of all LPs running concurrently on different processors (or threads) at the same wall-clock time. ...
Global Virtual Time computation of Parallel Discrete Event Simulation is crucial for conducting fossil collection and detecting the termination of simulation. The triggering condition of GVT computation in typical approaches is generally based on the wall-clock time or logical time intervals. However, the GVT value depends on the timestamps of events rather than the wall-clock time or logical time intervals. Therefore, it is difficult for the existing approaches to select appropriate time intervals to compute the GVT value. In this study, we propose a scalable GVT estimation algorithm based on Lower Bound of Event-Bulk-Time, which triggers the computation of the GVT value according to the number of processed events. In order to calculate the number of transient messages, our algorithm employs Event-Bulk to record the messages sent and received by Logical Processes. To eliminate the performance bottleneck, we adopt an overlapping computation approach to distribute the workload of GVT computation to all worker-threads. We compare our algorithm with the fast asynchronous GVT algorithm using PHOLD benchmark on the shared memory machine. Experimental results indicate that our algorithm has a light overhead and shows higher speedup and accuracy of GVT computation than the fast asynchronous GVT algorithm.
... It is optimized to the situation that the LA is short, but the possibility, that some external events should be exchanged right after the time interval defined by the LA, is very small. In other words, the mechanism is optimized to the situation that the estimated lookahead could be stochastically too conservative (multiagent system, stochastic Petri-net, cell biology [3] etc.). ...
Distributed discrete event simulation is a very important method today to analyze the behavior of large models. We investigate the practical implementation of distributed discrete event simulation with conservative synchronization and its acceleration through dynamic estimation of process-to-process look-ahead. Since the dynamic look-ahead changes with time, we have to face the situation, that the look-ahead between some logical processes is decreased temporarily. The shortened lookahead has a very negative influence to the performance of the simulation and it is hard to avoid. However, this effect can be reduced by introducing some extra mechanisms in the simulation. In this paper, we present a mechanism to optimize the simulation for the situation that the look-ahead between some processes is very short. This mechanism is based on exchange of conditional look-ahead and broadcast of invalidation announcement. Our evaluation shows reduction of the execution time of a majority of distributed simulations, especially when the estimated look-ahead is stochastically too conservative.
... Conservative synchronization is performed using the NMA and its variant techniques. Some recent advancements and interesting results in the optimistic algorithms can be found in [7][8][9][10]. Conservative protocols fundamentally maintain causality in event execution by strictly disallowing the processing of events out of time-stamp order [4]. ...
This paper presents a new logical process (LP) simulation model for distributed simulation systems where Null Message Algorithm (NMA) is used as an underlying time management algorithm (TMA) to provide synchronization among LPs. To extend the proposed simulation model for n number of LPs, this paper provides a detailed overview of the internal architecture of each LP and its coordination with the other LPs through sub-system components and models such as communication interface and simulation executive. The proposed architecture of LP simulation model describes the proper sequence of coordination that need to be done among LPs though different subsystem components and models to achieve synchronization. To execute the proposed LP simulation model for different set of parameters, a queuing network model is used. Experiments will be performed to verify the accuracy of the proposed simulation model using the pre-derived mathematical equations. Our numerical and simulation results can be used to observe the exchange of null messages and overhead indices.
... Both rounds define an interval for GVT computation. In addition to define a lower bound on the unprocessed events, a GVT algorithm must also address the following two problems: transient message and simultaneous reporting problems [22]. Transient messages are those that have been sent but have not been yet received [1]. ...
Time Wrap algorithm is a well-known mechanism of optimistic synchronization in a parallel discrete-event simulation (PDES) system. It offers a run time recovery mechanism that deals with the causality errors. For an efficient use of rollback, the global virtual time (GVT) computation is performed to reclaim the memory, commit the output, detect the termination, and handle the errors. This paper presents a new unacknowledged message list (UML) scheme for an efficient and accurate GVT computation. The proposed UML scheme is based on the assumption that certain variables are accessible by all processors. In addition to GVT computation, the proposed UML scheme provides an effective solution for both simultaneous reporting and transient message problems in the context of synchronous algorithm. To support the proposed UML approach, two algorithms are presented in details, with a proof of its correctness. Empirical evidence from an experimental study of the proposed UML scheme on PHOLD benchmark fully confirms the theoretical outcomes of this paper.
... Algorithms commonly known under the term stochastic simulation algorithms (SSA) are based on sampling the Chemical Master Equation (CME): a partial differential equation describing the time evolution of the system's state probability distribution [14]. To address the problem of efficiency different strategies have been pursued: by introducing improved scheduling algorithms or data structures, [13], by trading accuracy for efficiency [6] , by combining numerical integration and stochastic discrete event approaches [50], or by parallel and distributed simulation [5, 32]. One way of considering space within simulation is partitioning space into sub-volumes and extending the master equation with a diffusion term, which leads to the reaction-diffusion master equation (RDME). ...
Regenerative systems are able to overcome significant perturba-tions, and maintain autonomously their functionality in dynamic and uncertain environments. More and more this ability of biologi-cal systems plays a role in designing technical systems, e.g., in sen-sor networks, as well. Important properties of regenerative systems are their dynamic structures and their operation on different spatial and temporal scales. Those propel the development of new model-ing, simulation, and visualization methods. Among them, variants of the π-calculus formalism, a portfolio of Gillespie related spatial simulation algorithms, means for automatically configuring simu-lators, and the integrated visualization methods, that make use of innovative layouts and linked and coordinated views target chal-lenges in analyzing biological regenerative systems. They provide a basis for analyzing regenerative systems in general by means of simulation.
... In addition to define a lower bound on the unprocessed events, a GVT algorithm must also address the following two problems: transient message and simultaneous reporting problems [22]. Transient messages are those that have been sent but have not been yet received [1]. ...
Distributed discrete event simulation is an important approach for enabling the modeling and analysis of the behavior of large systems. This approach presents a major problem, namely, the possible low performance due to the excessive overhead in synchronizing the distributed logical processes. To counter this, our approach to distributed discrete event simulation involves conservative synchronization and its acceleration using dynamic estimation of process‐to‐process look‐ahead with a feedback mechanism. This mechanism allows for the estimation of a larger look‐ahead, which may be invalidated and recalculated during the course of the simulation, if one of the processes obtains more detailed knowledge.
In this work, we extend the dynamically estimated look‐ahead, on the basis of the local state of the logical processes, by exchanging conditional look‐aheads, in conjunction with the broadcast of invalidation announcements. A notable reduction in runtime in various cases is thus achieved, especially when the estimated look‐ahead is stochastically too conservative. Copyright © 2016 John Wiley & Sons, Ltd.
We consider the problem of simulating spatially distributed entities which can move, see each other, and react accordingly. We provide centralized reference algorithms for both time-stepped and discrete-event simulation. Under reasonable assumptions, we then proceed to distribute the simulation among several nodes by assigning each node a subregion of the simulation space. A main characteristic of our approach is that the subregions do not form a partitioning, but a covering. That is, they partially overlap, hence causing some duplicated computation, which is apparently redundant. The amount of overlapping is a tunable parameter of our algorithms, which affects the overall performance in a non-trivial way. Through an analytical model as well as experimental results we discover a trade-off. Choosing a small overlapping requires to perform frequent synchronizations, which negatively affect performance. However, a large overlapping leads to more duplicated work, which also decreases performance. Balancing the amount of overlapping is then required to optimize performance.
The π-Calculus is a modeling formalism for concurrent processes. Realized as part of the plug-in based modeling and simulation framework JAMES II, we propose an architecture for π-Calculus-based modeling and simulation, which supports both flexibility and efficiency. Facilitating the design of new π-Calculus-based formalisms and simulators is of particular relevance in the field of computational systems biology, for which many different π-Calculus dialects and simulators have been and still are being developed. Therefore, a flexible representation of π-Calculus models is used, which is illustrated by a mapping from the biochemical variant of the π-Calculus to the representation. Simulation engines are exchangeable and even automatically configurable according to the task at hand. Moreover, we present three different simulator implementations, working on the model representation. Efficiency denotes that our architecture supports the implementation of high-performance simulators. In order to assess efficiency, we perform experiments with these simulators and compare the results to the current cutting edge implementation in the field, the Stochastic Pi Machine.
We consider a class of models describing generic agents (e.g. macromolecules, small organisms) which are able to travel in space, can sense the surrounding environment, and can react accordingly. In these models, we focus on individual-based simulation. We start with defining a simple centralized simulation algorithm, which we then improve so to develop a distributed algorithm producing the same output. An analytical model is given to estimate the expected speedup of our distributed algorithm depending on several parameters. A main aspect of our approach is that it trades computation time for synchronization time. That is, we allow each node to perform apparently redundant computation whenever this reduces the amount of needed synchronization in such a way that the overall performance improves.