In order understand the suitability of high speed network capabilities for videoconferencing applications, we must first establish the general requirements of videoconferencing applications. Draoli et al (1997), admittedly during the early years of videoconferencing when networking was more expensive and less flexible than today (especially when considering “computing power restrictions”), concluded that in a 2-way videoconferencing application the “image size, the frame rate and the number of displayed colors” in addition to a measurement of user acceptance of acceptable delays and losses during the application were the main considerations. User acceptance is a qualitative rather than a quantitative measurement, however it would be safe to assume that minor delays and frame losses would not be considered a major issue.
Asynchronous Transfer Mode (ATM) was developed in the 1970/80s for broadband ISDN connections and provides an effective means of achieving a “fast, real-time and demand-responsive switching for efficient use of network resources” according to Ojesanmi, OA (2009). ATM is perfectly suited to videoconferencing applications as a result as it fulfils all of the necessary technical speed, data rate and user acceptance quality requirements, assuming that the switched point-to-point architecture is standard across the entire network. Gigabit Ethernet has a single use speed that is greater than ATM, however ethernet is not normally considered single use as many applications use the same cabling infrastructure. Therefore the main difference is the way in which the packets are composed and sent in different size packets. Here larger is not necessarily better as Welsh et al (1997) state that ATM is able to manage its smaller packets more effectively to provide a better quality result.
The main differences between ATM and Gigabit Ethernet in terms of videoconferencing applications are flexibility, quality and cost. As ATM requires a specific videoconferencing connection this implies that this technology is limited to those with the budget to have a dedicated videoconferencing infrastructure whereas Gigabit Ethernet is extremely flexible in that any user on a Gigabit infrastructure could effectively use the service from any desktop, similarly to Skype or FaceTime videoconferencing applications (that do not necessarily have to run over Gigabit connections), therefore Gigabit is more flexible.
When Welsh et al (1997) stated that “direct application access to the network interface can provide both low-latency and high-bandwidth communication over commodity networks such as 155Mbps ATM and 100Base-TX Fast Ethernet” (over a slower than Gigabit network), they concluded that ATM’s main advantage was when higher bandwidth and floating-point performance was required. This advantage has largely been negated by the advances in a Gigabit Ethernet environment. Given also the proliferation of Ethernet in use today and the ability and advantage that a single cabling system has multi-uses, this is also an advantage over ATM as infrastructure costs are lower.
Overall I would conclude that except for extremely specialist circumstances where quality requirements or data content security are paramount, Gigabit Ethernet’s advantages vastly outweigh those of ATM in videoconferencing applications. It must be noted however that with advances in videoconferencing technologies, we are seeing that user acceptable quality is currently being achieved over much reduced resources and networking speeds/capabilities.
Draoli, M., Gaibisso, C., Lancia, M. & Mastromartino, EA. (1997) Satisfying High Quality Requirements of Videoconferencing on a Packet-Switched Network [Online]. Available at http://www.isoc.org/inet97/proceedings/F3/F3_2.HTM (Accessed 8 May 2011).
Ojesanmi, OA (2009) ‘Asynchronous Transfer Mode (ATM) Network’, Computer Science & Telecommunications, 2009, 6, pp. 70-82, Computers & Applied Sciences Complete, EBSCOhost, viewed 8 May 2011.
Welsh, M., Basu, A. & Von Eicken, T. (1997) IEEE: ATM and Fast Ethernet Network Interfaces for User-level Communication [Online]. Available at http://www.eecs.harvard.edu/~mdw/papers/hpca97.pdf (Accessed 8 May 2011).