Following my undergraduate dissertation on network support for resource-constrained highly mobile embedded devices, my masters project moved on to the exact opposite scenario: supporting physically static devices. There are a broad class of network-connected devices with a physical presence to which location is an intrinsic part of their identity. A networked speaker in, say, the Oval Office is defined by its location: it’s simply the Oval Office Speaker. If the specific device moves location its identity should change with its new location, and if the device is replaced then the replacement should assume the function of its predecessor.
My masters project explored how an augmented reality interface for interacting with these devices could be built and the systems support required for communicating with using the myriad of addresses we use beyond IP. The Domain Name System, the standard for both global and network-local naming, provides a registry for network address that is compatible with the Internet protocol suite. We extended the DNS with algorithms for geospatial queries on this database through DNS resolutions, which we coined the `Spatial Name System`.
We wrote these ideas down in a paper ‘Where on Earth is the Spatial Name System’ in 2023 which was accepted to the 22nd ACM Workshop on Hot Topics in Networks.
Recent work in this area has included Roy Ang’s work on `Building Bigraphs of the Real World`, taking Robin Milner’s Bigraphs and implementing models of OpenStreetMap with Glasgow’s Bigrapher tool written in OCaml.
I’m interested in putting these ideas into practice with Josh Millar’s sensor networks.