MoN16: Sixteenth Mathematics of Networks meeting

James Hodgson (Cambridge University) – Queue-based metamorphosis: Outlining a challenge in reproducing the real-world computer network

To control and manage complex computer networks, we often seek to understand their operation under a variety of scenarios to, for example, check resilience under failure, test new architectures, and evaluate application performance. Yet, increasingly modern computer networks are fun to draw but hard to simulate, difficult to emulate and impossible for those without billion-dollar incomes to replicate. Without reproduction we are doing bad-science. With reproduction - lowering the bar to simulation, emulation, and reproduction - as well as making smoother (intellectual) movement/comparison between these domains can only be a good thing allowing greater community contribution and engagement. I interpret my need as a desire to model many (general) queues with a smaller number of (potentially more complicated) queues; aside from an experimental modelling and validation, I would like to put this work on a stronger theoretical foundation should a suitable foundation exist. If time permits I will also outline a number other challenges in this space of mapping between modelling, simulation, emulation, and reproduction.

Return to previous page

Contact: Keith Briggs (mailto:keith.briggs_at_bt_dot_com) or Richard G. Clegg (richard@richardclegg.org)