Time-Fluid Field-Based Coordination

Abstract

Emerging application scenarios, such as cyber-physical systems (CPSs), the Internet of Things (IoT), and edge computing, call for coordination approaches addressing openness, self-adaptation, heterogeneity, and deployment agnosticism. Field-based coordination is one such approach, promoting the idea of programming system coordination declaratively from a global perspective, in terms of functional manipulation and evolution in "space and time" of distributed data structures, called fields. More specifically, regarding time, in field-based coordination it is assumed that local activities in each device, called computational rounds, are regulated by a fixed clock, typically, a fair and unsynchronized distributed scheduler. In this work, we challenge this assumption, and propose an alternative approach where the round execution scheduling is naturally programmed along with the usual coordination specification, namely, in terms of a field of causal relations dictating what is the notion of causality (why and when a round has to be locally scheduled) and how it should change across time and space. This abstraction over the traditional view on global time allows us to express what we call "time-fluid" coordination, where causality can be finely tuned to select the event triggers to react to, up to to achieve improved balance between performance (system reactivity) and cost (usage of computational resources). We propose an implementation in the aggregate computing framework, and evaluate via simulation on a case study.

Time-Fluid Field-Based Coordination through Programmable Distributed Schedulers

“…The contribution of this work extends the earlier version presented at COORDINATION 2020 [20], and can be summarised as follows:…”

“…Differently from the work in [20], in which a field computation could actually be assigned a single guard policy, here we extend this approach by allowing many guard policies to be associated to an actual field computation, and since guard policies are particular kinds of field computations as well, we also allow guard policies to be stacked at multiple layers, forming a tree structure. So, essentially, platform triggers schedule low-level guard policies, the leafs of the tree, which are field computations enacting low-level (causality-based) "clocks".…”

“…• first, we propose a model that enriches field-based coordination approaches with the possibility to explicitly program the scheduling of coordination actions, by enabling a functional description of causality, relying on a tree-based composition to express high-level notions of time; • second, we formally define the operational semantics of such a model in the context of the field calculus, as an extension and variation of its network semantics in [3]; • third, we evaluate the proposed model in the context of aggregate programming, showcasing its expressiveness and practicality in achieving improved efficiency and, in some cases, performance. The model enhancement with compositionality, the formalisation, and the detailed evaluation are novel contributions with respect to [20].…”

Time-Fluid Field-Based Coordination through Programmable Distributed Schedulers

“…Differently from the work in [PMVZ20], where a field computation could be assigned a single guard policy, here we extend this approach by allowing many guard policies to be associated to an actual field computation, and since guard policies are particular kinds of field computations as well, we also allow guard policies to be stacked at multiple layers, forming a tree structure. So, essentially, platform triggers schedule low-level guard policies, the leafs of the tree, which are field computations enacting low-level (causality-based) "clocks".…”