Modeling of H.263 encoded low bitrate video traffic for tactical video conferencing applications
Robert E. Parker, M. Tummala
Abstract
Robert E. Parker, M. Tummala
Abstract
Growing demand for networked video applications has led to increased interest in VBR traffic, since VBR traffic allows statistical multiplexing to conserve bandwidth. However, transporting VBR traffic poses challenges in guaranteeing QoS and efficiently utilizing resources relative to CBR traffic. Modeling VBR traffic aids in estimating QoS metrics, evaluating scheduling algorithms, and determining statistical multiplexing gains. While most VBR traffic models have focused on high-bandwidth traffic, we propose a traffic model for H.263 encoded video sequences, appropriate for low-bit-rate video. The traffic model uses a multi-state Markov chain representing a pseudo-histogram of observed bit-rates. To reduce correlation between multiplexed sequences and enhance multiplexer performance, each stream is deterministically smoothed over one frame length.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Growing demand for networked video applications has led to increased interest in VBR traffic, since VBR traffic allows statistical multiplexing to conserve bandwidth. However, transporting VBR traffic poses challenges in guaranteeing QoS and efficiently utilizing resources relative to CBR traffic. Modeling VBR traffic aids in estimating QoS metrics, evaluating scheduling algorithms, and determining statistical multiplexing gains. While most VBR traffic models have focused on high-bandwidth traffic, we propose a traffic model for H.263 encoded video sequences, appropriate for low-bit-rate video. The traffic model uses a multi-state Markov chain representing a pseudo-histogram of observed bit-rates. To reduce correlation between multiplexed sequences and enhance multiplexer performance, each stream is deterministically smoothed over one frame length.
Key concepts: Variable bitrate, Computer science, Statistical time division multiplexing, Multiplexing, Computer network, Real-time computing, Quality of service, Constant bitrate