Discussant(s): Richard Hanley Introducer(s): Frank Witlox - Ghent University Session Description: Over the past few years, the empirical study of the geography of the Internet traffic has been largely based on capacity data such as backbone bandwidth size. Researchers thereby argue that, because it is virtually impossible to obtain measures of the actual volume of data flows between/in geographic locations, this backbone bandwidth size indicator (or more precisely: the partial bandwidth capacity) has been the best proxy around. However, for the second year in a row now, the total international Internet capacity grew faster than the actual Internet traffic. Furthermore, the degree of utilisation varies significantly geographically: while utilisation on international links to Europe and Asia fell in 2008, they rose in the US and Canada and Latin America (even to the degree that it outpaced the deployment of new Internet bandwidth). Overall, this implies that is increasingly difficult to sustain the notion that digital inter-city connections can be measured properly based on this bandwidth indicator. Taken together, this suggests that we need a number of new approaches for measuring the Internet traffic between/in geographic locations. The formative purpose of this session is to think about how other approaches (e.g. VoIP traffic, geo-coding of IP-addresses, exchange points traffic, generated traffic at/between places in cyberspace) may provide us with alternatives for bandwidth volume as a measure for revealing the geography of Internet traffic, including alternative conceptual approaches for studying urban digital accessibility.
