According to the terms of reference, this study is concerned with the roll of pricing in
regard to different categories of quality of service over IP networks, with a view to the
commercial viability of real-time IP services. In providing information about this topic the
report provides an analysis of several academic studies (supplied by the ITU) in terms
of their contribution to the topic, and explains the business case for international bypass
offered by the Internet and possibly other cross-boarder IP networks.
When looking at real-time service over IP networks, we are compelled to address
issues of service quality. Real-time service over IP requires certain service quality
characteristics which are problematic for IP networks - especially a public network like
the internet. This report is fairly ambitious in its approach, setting out structural and
technical reasons explaining why service quality is problematic on IP networks, and in
doing so it explains the various categories, grades and classes of service quality which
are possible on IP networks in controlled environments.
In practice quality of service options are not widely available on the Internet at present
although some are making an appearance on private IP networks and campus
networks. However, there are no accounting or payment systems that would enable
users to pay for a higher QoS on the basis of usage or packet throughput. There is thus
much to be done to marry future quality of service developments (the supply-side) with
the introduction of pricing options (the demand management side).
The main quality of service problem that must be addresses in order for real-time
services to be viable, is congestion. The report discusses congestion problems and
explains that when packets that require real-time service quality are randomised with
other packets, as occurs with the Internet, either all packets must receive real-time
service quality, or packets requiring priority treatment must be able to be targeted for
special treatment.