On Execution Platforms for Large-scale Aggregate Computing

   page       BibTeX_logo.png   
Proceedings of the 2016 ACM International Joint Conference on Pervasive and Ubiquitous Computing: Adjunct, pages 13211326
UbiComp '16
ACM, New York, NY, USA

Aggregate computing is proposed as a computational model and associated toolchain to engineer adaptive large-scale situated systems, including IoT and wearable computing systems. Though originated in the context of WSN-like (peer-to-peer and fully distributed) systems, we argue it is a model that can transparently fit a variety of execution platforms (decentralised, server-mediated, cloud/fog-oriented), due to its ability of declaratively designing systems by global-level abstractions: it opens the possibility of intrinsically supporting forms of load balancing, elasticity and toleration of medium- and long-term changes of computational infrastructures. To ground the discussion, we present ongoing work in the context of scafi, a language and platform support for computational fields based on the Scala programming language and Akka actor framework.

keywordsaggregate computing, cloud computing, execution platforms, internet of things, large-scale systems