Parallel Simulation of Hybrid Network Traffic Models

Liu, Jason

doi:10.1109/pads.2007.26

Cited by 11 publications

(8 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The hybrid model can also be parallelized. We have shown both analytically and quantitatively that asymptotically the parallel hybrid model can achieve the same scalability as the pure packet-oriented simulation [29].…”

Section: Hybrid Traffic Modelingmentioning

confidence: 93%

A Primer for Real-Time Simulation of Large-Scale Networks

Liu

2008

41st Annual Simulation Symposium (Anss-41 2008)

Self Cite

View full text Add to dashboard Cite

Real-time network simulation refers to simulating computer networks in real time so that the virtual network can interact with real implementations of network protocols, network services, and distributed applications. In this paper, we present the motivation behind real-time network simulation and compare it against other major networking research tools, including analytical models, physical testbeds, simulation, and emulation. We introduce PRIME, a parallel real-time network simulator, and provide a summary of techniques that allow PRIME to model large-scale networks and interact with many real applications under the real-time constraint. We also discuss ongoing research efforts that will allow real-time network simulation to assume important roles at supporting future networking research.

show abstract

Section: Hybrid Traffic Modelingmentioning

confidence: 93%

A Primer for Real-Time Simulation of Large-Scale Networks

Liu

2008

41st Annual Simulation Symposium (Anss-41 2008)

Self Cite

View full text Add to dashboard Cite

show abstract

“…We also demonstrated that the hybrid model is highly parallelizable [14]. We observed that the time it takes to propagate fluid characteristics (such as the accumulative delay and loss rate) along the flow paths has a lower bound equal to the minimum link delay, according to the governing ordinary di↵erential equations.…”

Section: Integration Of Packet and Fluid Modelsmentioning

confidence: 92%

“…The fluid nodes along the path of a fluid flow are thus broken into multiple segments assigned to di↵erent processors and handled in a parallel fashion. There is a full discussion in [14].…”

Section: Fluid Modelmentioning

confidence: 99%

GPU-assisted hybrid network traffic model

Liu

et al. 2014

Proceedings of the 2nd ACM SIGSIM Conference on Principles of Advanced Discrete Simulation

Self Cite

View full text Add to dashboard Cite

Large-scale network simulation imposes extremely high computing demand. While parallel processing techniques allows network simulation to scale up and benefit from contemporary high-end computing platforms, multi-resolutional modeling techniques, which di↵erentiate network tra c representations in network models, can substantially reduce the computational requirement. In this paper, we present a novel method for o✏oading computationally intensive bulk tra c calculations to the background onto GPU, while leaving CPU to simulate detailed network transactions in the foreground. We present a hybrid tra c model that combines the foreground packet-oriented discrete-event simulation on CPU with the background fluid-based numerical calculations on GPU. In particular, we present several optimizations to e ciently integrate packet and fluid flows in simulation with overlapping computations on CPU and GPU. These optimizations exploit the lookahead inherent to the fluid equations, and take advantage of batch runs with fix-up computation and on-demand prefetching to reduce the frequency of interactions between CPU and GPU. Experiments show that our GPU-assisted hybrid tra c model can achieve substantial performance improvement over the CPU-only approach, while still maintaining good accuracy.

show abstract

“…Another approach, used in [4] [11] and [8], is to treat some of the traffic throughout the network analytically. Then develop methods to explicitly mix both analytic and packet traffic at each multiplexing point throughout the network.…”

Section: Deterministic Modelsmentioning

confidence: 99%

“…There exist methods to improve the fidelity of our example deterministic model, e.g., [4] [11] [8], but these improvements require a priori knowledge of the behavior of the foreground traffic, which in general is not known. This choice is made here for several reasons.…”

Section: A Deterministic/packet Mixing Equationsmentioning

confidence: 99%

Stochastic Process Models for Packet/Analytic-Based Network Simulations

Cole

Riley

Cansever³

et al. 2008

2008 22nd Workshop on Principles of Advanced and Distributed Simulation

View full text Add to dashboard Cite

W E present our preliminary work that develops a new approach to hybrid packet/analytic network simulations for improved network simulation fidelity, scale, and simulation efficiency. Much work in the literature addresses this topic, including [10] [11] [8] [12] [13] and others. Current approaches rely upon models, which we refer to in this paper as Deterministic Fluid Models [9] [12], to address the analytic modeling aspects of these hybrid simulations. Instead we draw upon an extensive literature on stochastic models of queues and (eventually) networks of queues to implement a hybrid stochastic model/packet network simulation. We will refer to our approach as Stochastic Fluid Models throughout this paper. We outline our approach, present test cases, and present simulation results comparing the measured queue metrics from our approach for hybrid simulation to those of a deterministic fluid model hybrid simulation and a full packet-level simulation. We also discuss plans for future areas of research on this approach.

show abstract

Parallel Simulation of Hybrid Network Traffic Models

Cited by 11 publications

References 17 publications

A Primer for Real-Time Simulation of Large-Scale Networks

A Primer for Real-Time Simulation of Large-Scale Networks

GPU-assisted hybrid network traffic model

Stochastic Process Models for Packet/Analytic-Based Network Simulations

Contact Info

Product

Resources

About