Spring Simulation Multiconference 

Abstract: This paper describes the use of the Levels of Conceptual Interoperability Model (LCIM) as a framework for conceptual modeling and its descriptive and prescriptive uses. LCIM is applied to show its potential and shortcomings in the current simulation interoperability approaches, in particular the High Level Architecture (HLA) and Base Object Models (BOM). It emphasizes the need to apply rigorous engineering methods and principles and replace ad-hoc approaches.

Abstract: Discrete Event Specification (DEVS) formalism has been used to study dynamics of discrete event systems. DEVS environments are typically open architectures that have been extended to execute on various middleware such as CORBA, Grid computing, P2P networks, RMI and others. The present work aims to provide another development environment using the Service Oriented Architecture (SOA) framework. The proposed DEVS Modeling Language (DEVSML) is built on XML and provides model interoperability among DEVS models located at remote locations. The DEVSML environment is built on client-sever paradigm and the simulation is executed at the server's end. The proposed DEVS atomic and coupled DTDs are open to standardization from the community for successful model sharing and collaboration. The DEVSML framework provides the needed feature of run-time composability of coupled systems using the SOA framework. DEVSML also provides the capability to translate model to and from XML and JAVA programming language leading to model composability and validation. This paper will demonstrate the client application as well as the server architecture underlying the DEVSML framework.

Abstract: Programmable graphics processing units (GPUs) have emerged as excellent computational platforms for certain general-purpose applications. The data parallel execution capabilities of GPUs specifically point to the potential for effective use in simulations of agent-based models (ABM). In this paper, the computational efficiency of ABM simulation on GPUs is evaluated on representative ABM benchmarks. The runtime speed of GPU-based models is compared to that of traditional CPU-based implementation, and also to that of equivalent models in traditional ABM toolkits (Repast and NetLogo). As expected, it is observed that, GPU-based ABM execution affords excellent speedup on simple models, with better speedup on models exhibiting good locality and fair amount of computation per memory element. Execution is two to three orders of magnitude faster with a GPU than with leading ABM toolkits, but at the cost of decrease in modularity, ease of programmability and reusability. At a more fundamental level, however, the data parallel paradigm is found to be somewhat at odds with traditional model-specification approaches for ABM. Effective use of data parallel execution, in general, seems to require resolution of modeling and execution challenges. Some of the challenges are identified and related solution approaches are described.

Abstract: An important goal of software security is to ensure sensitive/secret data owned by a program shall be exclusively accessible by the program. An obstacle to such security goal is that modern commodity operating systems (OS) for the sake of speed and flexibility have a unified linear address space--any OS kernel program can access all the linear addresses. As a result, rootkits or malicious system software are able to control the OS virtual address space, harvest the sensitive data used by software programs on the compromised computer, and report the data to remote entities controlled by hackers.In this paper, we present a holistic approach against sophisticated malware. Instead of focusing on the security of various abstraction layers of OS, we utilize the hardware techniques to directly provide the trust services to software programs. Without modifying OS, we leverage the virtual machine monitor technologies to create a lightweight hypervisor for fine-grain software runtime memory protection. As a result, a program's memory could be hidden from other high privilege system software in a single commodity OS. In addition, we propose the data locker component in the hypervisor, which prevents the sensitive data of software program in persistent storage from leaking to rootkits or other malware. For the performance evaluation, the implementation based on hardware-assisted x86 virtualization technology is presented and experimental results are reported.