Fast simulation of rare events in queueing and reliability models
ACM Transactions on Modeling and Computer Simulation (TOMACS)
Queueing networks and Markov chains: modeling and performance evaluation with computer science applications
Proceedings of the 34th conference on Winter simulation: exploring new frontiers
Proceedings of the 34th conference on Winter simulation: exploring new frontiers
Limit theorems for the multilevel splitting algorithm in the simulation of rare events
WSC '05 Proceedings of the 37th conference on Winter simulation
Performance analysis of mobile communication networks with clustering and neural modelling
International Journal of Mobile Network Design and Innovation
Rare Event Simulation using Monte Carlo Methods
Rare Event Simulation using Monte Carlo Methods
Introduction to Rare Event Simulation
Introduction to Rare Event Simulation
| Hi-index | 0.00 |
The paper recommends an effective approach to estimate probability of buffer overflow in wireless communication networks. The buffer overflow probability in queuing systems is defined as a rare event and can be estimated using rare event simulation with Markov chains. Two-node queuing networks are considered in this paper; and an event of buffer overflow at the second node is studied. Two efficient rare event simulation algorithms, based on the importance sampling and cross-entropy methods, are developed and applied to accelerate the overflow probability simulation with Markov chain modelling. Numerical examples and simulation results are provided.