Abstract | ||
---|---|---|
Packet switching is an effective technology for integrating voice and data in a single network. An example for such a kind of a network is Internet. New service is emerging, and it is a Voice over Internet Protocol (VoIP). An important aspect of packet-switched voice is the reconstruction of a continuous stream of speech from the set of packets that arrive at the destination DTE. Because of the statistical nature of packet switching, each packet can encounter a different amount of delay in traversing a packet network. The problem described in this paper is how to design an appropriate voice reconstruction strategy that reproduces acceptable quality speech from packets that arrive with varying delay. The basic idea proposed in the paper is to measure a delay jitter between several points on the Internet. Our voice traffic generator, implemented on a workstation, sends packets that are 20ms long. Receiver measures interarrival time, keeps track of mixed sequences and lost packets. The measurement results are given in a form of average delay jitter. Receiver also keeps a record of all relevant packet information in a text file that is used as an input to a delay jitter predictor. Traffic prediction element predicts talkspurt characteristics and directly influences playout delay. For a prediction element, we propose Adaptive Neuro-Fuzzy Inference System (ANFIS). |
Year | Venue | Keywords |
---|---|---|
2000 | ESM | optimal delay buffer,voip application |
Field | DocType | ISBN |
Computer science,Computer network,Voice over IP | Conference | 1-56555-204-0 |
Citations | PageRank | References |
0 | 0.34 | 1 |
Authors | ||
3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Viktor Matic | 1 | 0 | 0.34 |
Alen Bazant | 2 | 4 | 1.45 |
Mladen Kos | 3 | 21 | 3.42 |