ANSI SCTE 165-2 pdf free download
ANSI SCTE 165-2 pdf free download.IPCablecom 1.5 Part 2: Audio/Nideo Codecs.
5 BACKGROUND
This section outlines the IPCablecom architecture support elements and the DOCSIS® network infrastructure necessary to deliver quality audio and video service. It is intended to clarify external interfaces and functional requirements necessary to implement the targeted audio and video quality using speech and video codecs.
The key requirement for voice communications using 1P transmission is the ability to attain “toll” or better audio quality. Given the variable nature of shared packet mediums and the stringent human-factor requirements of this quality standard, it is necessary to optimize multiple system parameters to attain this goal. Additionally, IPCablecom has been tasked with offering superior quality, exceeding current PSTN standards where feasible. Key requirements from the IPCablecom product definition requiring architectural optimization for codecs follow.
5.1 lPCablecom Voice Communications Quality Requirements
As defined in the lPCahlecom architecture document 0, requirements for toll-quality voice communications service in IPCablecom include numerous metrics to ensure competitive or superior quality and service to the PSTN. In order to support these requirements. network plant and equipment may have to be groomed. In order to provide guidelines for that grooming, several network implications affecting codec performance are discussed below.
5.2 Network Preparation for Codec Support
The critical areas of network performance, which must be optimized in tandem with codecs, are packet loss, latency, and jitter. Elaboration of network/codec implications for each of these areas follows.
5.2.1 Packet Loss Control
There is a direct correlation between packet integrity and audio quality. Anecdotal codec research suggests initial 3% packet loss rate results, on average, in a reduction in Mean Opinion Score (MOS) scores of 0.5 point, on a scale of 5. Due to less-than-pristine conditions and human-detectable compromises with most codecs, the resulting audio quality for a 3% packet loss rate will be well below PSTN “toll” quality. Above 3%, codec performance falls off rapidly, and resulting voice quality is unacceptable.
Applications and/or codecs may provide error correction or concealment mechanisms, which may increase latency through buffering. Once latency thresholds have been exceeded, the tradeoff between latency and fidelity becomes an untenable situation.
5.2.2.1 Latency Control: Buffering While network jitter and corresponding buffering increase call latency, another source of buffering can be induced by the application as a corrective response to severe packet loss. Although the ultimate solution to additional buffering delay is a pristine network, realistically some packet loss will occur. Accounting for lost packets suggests the need for support concealment or reconstruction of lost data, and in many instances these techniques employ some mechanism of redundant information encoding, temporally shifting and embedding audio frames in the data stream. This not only increases the effective bandwidth requirement, but also creates, in effect, an additional buffer to allow for reassembly, increasing latency. In order to apply certain reconstruction methodologies in an optimal fashion, the application needs accurate data regarding the statistical characteristics of the media stream. Some information is available through real-time control protocol (RTCP) mechanisms, such as a gross measure of packet loss. Additional information, such as burst frequency and predictive time-of-day effects, would improve the potential of the application to make optimal adjustments. Planning for the collection and analysis of this type of network information will allow developers more options in the future, potentially creating applications that will increase network utilization efficiency or quality.
ANSI SCTE 165-2 pdf download.
