Abstract-Network traffic generators allow video and audio streams to be modelled. 'Bursty' traffic patterns imply packetby-packet rather than time-averaged analysis to determine the impact on network routers. Due to computer operating system time event scheduling dependencies, the ability to generate accurate packet delivery is compromised. A real-time operating system on an x86 commodity machine is shown to result in greater accuracy and range, but with some loss in precision outside its operating range, compared to a general-purpose Linux kernel.
Abstract-Traffic at varying bit rates, packet lengths and packet rates are passed across a bottleneck link. The router response is determined, especially the onset of instability due to excessive CPU load. The paper concludes that it is primarily packet rate that governs the router response. The results are relevant to multimedia streaming applications.
The H.263 codec is an efficient way to stream variable bit-rate video sequences. This letter proposes that, for equivalent bandwidth and frame rate, a two-slice packetization scheme results in superior peak signal-to-noise ratio, rather than the conventional oneslice scheme, and that constant inter-packet gap rather than 'bursty' delivery is preferable. Introduction: Studies of packetization schemes for the H.263+ video codec, for example [1], tend to assume the one slice per packet recommendation contained in RFC 2429 section 3.2 [2], whereas in this letter a two-slice scheme is proposed. In [3], the burst length is identified (for H.264) as a source of peak signal-to-noise ratio (PSNR) degradation, as much as the average packet loss rate. In this letter, the method of delivery, either per frame bursts, or uniform (constant) inter-packet gap (IPG) is shown to
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.