Abstract | ||
---|---|---|
Performance of real-time applications on end-to-end packet channels are strongly related to losses and temporal delays. Several studies showed that these network features may be correlated and present a certain degree of memory such as bursty losses and delays. The memory and the statistical dependence between losses and temporal delays suggest that the channel may be well modeled by a hidden Markov model with appropriate hidden variables that capture the current state of the network. In this paper we propose an input/output hidden Markov model that, trained with a modified version of the expectation-maximization algorithm, shows excellent performance in modeling typical channel behaviors in a set of real packet links. The work extends to case of variable inter-departure time the previous proposed hidden Markov model that well characterizes losses and delays of packets from a periodic source. |
Year | DOI | Venue |
---|---|---|
2004 | 10.1109/MMSP.2004.1436597 | MMSP |
Keywords | Field | DocType |
input output,hidden markov models,hidden markov model,expectation maximization algorithm,statistical analysis,hidden variables,internet,real time systems | Forward algorithm,Computer science,Real-time computing,Input/output,Artificial intelligence,Hidden variable theory,Hidden semi-Markov model,Pattern recognition,Markov model,Network packet,Algorithm,Variable-order Markov model,Hidden Markov model | Conference |
ISBN | Citations | PageRank |
0-7803-8578-0 | 2 | 0.46 |
References | Authors | |
4 | 5 |
Name | Order | Citations | PageRank |
---|---|---|---|
Pierluigi Salvo Rossi | 1 | 328 | 27.27 |
Athina P. Petropulu | 2 | 1995 | 135.28 |
Jie Yu | 3 | 2 | 0.46 |
Francesco Palmieri | 4 | 1713 | 182.92 |
Giulio Iannello | 5 | 414 | 46.75 |