Brian Paul Swenson

A New Approach to Zero-Copy Message Passing with Reversible Memory Allocation in Multi-core Architectures

We present a new Zero-Copy approach for message passing in a tightly coupled, multi - process parallel discrete event simulation. Our approach is highly scalable and is suited for large scale distributed simulations. A Zero-Copy approach never copies message content. Rather, messages are created initially in a shared - memory region, and only a smart pointer referring to the shared memory object is passed to a message recipient. The smart pointer can be dereferenced normally, and has reference count semantics allowing memory reuse when no references remain. This approach significantly reduces the amount of data copied between processes and results in considerable improvement in overall application performance as compared to more traditional shared - memory based message passing. We demonstrate the efficiency of our approach using two distributed discrete event simulators using conservative synchronization. However, we also included knowledge of the Global Virtual Time and the Current Simulation Time in our memory management algorithms, allowing rollbacks and reclaiming of memory that was freed prematurely.

Cyber-physical co-simulation of smart grid applications using ns-3

In this paper, we present a simulation tool that supports co-simulation of the cyber and physical aspects of an electric smart grid. This cyber-physical co-simulator combines a state-of-the-art network simulator, ns-3, and a state-of-the-art power system simulator, PowerWorld. This combination has been achieved by extending ns-3 with a power system module, a physical system interface module, and a cyber-system module. We present the details of these new modules as well as a case study of the application of the co-simulation tool using a demand response scenario.

Performance of conservative synchronization methods for complex interconnected campus networks in ns-3

Distributed simulations provide a powerful framework for utilizing a greater amount of computing resources, but require that consideration be taken to ensure that the results of these simulations match results that would have been produced by a single sequential resource. Also, synchronization among the multiple nodes must also occur to ensure that events processed in the simulation maintain a timestamped order across all nodes. This work uses the popular network simulator ns-3. The ns-3 simulator is a discrete event network simulator used for educational and research-oriented purposes which provides two implementation options for distributed simulations based on the null message and granted time window algorithms. We examine the performance of both distributed schedulers available in ns-3 in an effort to gauge specific features of an overall distributed network topology that warrant the use of one synchronization method over the other.

Offload Destination Selection to Enable Distributed Computation on Battlefields

The tactical edge (battlefields, disaster relief operations, etc.) is an ideal domain for utilizing mobile devices for computation. However, these situations require jobs to complete in a timely manner. Efficiently offloading computation can ensure timely job completion, but access to cloud computing infrastructure is typically limited, unreliable, or unavailable. As an alternative, we propose using vehicle-mounted high performance computers, called Tactical HPCs (T-HPCs), as offload targets. A common scenario includes multiple offload targets with different computation capabilities, multiple clients wanting to offload computations, and wireless ad-hoc connectivity. While offloading strategies exist, currently none address the requirements of task scheduling in this setting. This paper presents an algorithm for offloading tasks from multiple sources while taking into account the computation capabilities and current utilization of available offload targets. In addition, distributed scheduling avoids a single point of failure which could be disastrous on the tactical edge. We report the results of simulations showing that our method reduces job completion time compared to alternative offloading schemes.

Implementing explicit congestion notification in ns-3

To detect network congestion, TCP typically relies on detecting packet loss. While this is an effective approach for maintaining high throughput for bulk data transfers, a better approach for interactive, time-sensitive, or loss-sensitive traffic would be to detect congestion prior to packet loss. Explicit Congestion Notification (ECN) is a congestion avoidance strategy that makes use of Active Queue Management to allow TCP endpoints to detect congestion without a corresponding packet drop. This congestion detection strategy is particularly useful for delay-sensitive traffic where packet retransmissions can lead to noticeable delays for the user. In this paper, we present an implementation of ECN as an addition to the TCP protocols in ns-3. We have modified all TCP variants currently in ns-3 to work with this new addition. To validate our work we tested all TCP variants and compared our implementations behavior to previous work.

Designing and Enabling Simulation of Real-World GPU Network Applications with ns-3 and DCE

The ability to execute the original source code for network protocols and applications within a network simulation environment frees the simulation modeler from the time consuming task of having to create, test and debug models representing these applications. This work extends the functionality of the Direct Code Execution (DCE) framework of ns-3 by incorporating the ability to call NVIDIA CUDA kernels from within simulated ns-3 nodes. This new functionality allows researchers to simulate large scale GPU applications in the realm of new and more flexible paradigms such as software-defined networking. Along with presenting this new functionality, this paper examines the different options available within the framework for communicating between the simulated nodes and the GPU. Each implementation is tested with multiple example CUDA kernels to demonstrate how they perform.

PHOLD performance of conservative synchronization methods for distributed simulation in ns-3

The scalability and runtime performance of large-scale discrete event network simulations has been improved previously by spreading processing effort across multiple processors, increasing the provided computational power while decreasing the wallclock execution time of each simulation trial. The popular network simulator ns-3 provides two distributed frameworks that differ in their synchronization implementations. This paper provides those thresholds under which certain selection criteria would deem one synchronization option better than the other in terms of runtime performance. It specifically focuses on the performance of each synchronization method by stripping the model of simulated network topologies and overhead and purely utilizing the synchronization implementations and event scheduler of ns-3. Simulations have been performed across a variety of lookahead values, neighbor selections, and remote traffic percentages, and neighbor connectivity thresholds have been determined that suggest where it is more appropriate to use one option over the other.

Phold performance for distributed network simulation under conservative synchronization methods in ns-3

Jared Ivey School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta, GA 30332-0250 j.ivey@gatech.edu George Riley School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta, GA 30332-0250 riley@ece.gatech.edu Brian Swenson School of Electrical and Computer Engineering Georgia Institute of Technology Atlanta, GA 30332-0250 bswenson3@gatech.edu

Machine to Machine Trust in Smart Cities

In the coming decades, we will live in a world surrounded by tens of billions of devices that will interoperate and collaborate in an effort to deliver personalized and autonomic services. This paradigm of smart objects and smart things interconnected and ubiquitously surrounding us is called the Internet of Things (IoT). Cities may be the first to benefit from the IoT, but reliance on these machines to make decisions has profound implications for trust, and makes mechanisms for expressing and reasoning about trust essential. This paper introduces the project funded by the Georgia Tech Research Institute to look at several dimensions of Machine to Machine Trust in the context of Smart Cities.