EDISON: An Edge-Native Method and Architecture for Distributed Interpolation

Abstract

Spatio-temporal interpolation provides estimates of observations in unobserved locations and time slots. In smart cities, interpolation helps to provide a fine-grained contextual and situational understanding of the urban environment, in terms of both short-term (e.g., weather, air quality, traffic) or long term (e.g., crime, demographics) spatio-temporal phenomena. Various initiatives improve spatio-temporal interpolation results by including additional data sources such as vehicle-fitted sensors, mobile phones, or micro weather stations of, for example, smart homes. However, the underlying computing paradigm in such initiatives is predominantly centralized, with all data collected and analyzed in the cloud. This solution is not scalable, as when the spatial and temporal density of sensor data grows, the required transmission bandwidth and computational capacity become unfeasible. To address the scaling problem, we propose EDISON: algorithms for distributed learning and inference, and an edge-native architecture for distributing spatio-temporal interpolation models, their computations, and the observed data vertically and horizontally between device, edge and cloud layers. We demonstrate EDISON functionality in a controlled, simulated spatio-temporal setup with 1 M artificial data points. While the main motivation of EDISON is the distribution of the heavy computations, the results show that EDISON also provides an improvement over alternative approaches, reaching at best a 10% smaller RMSE than a global interpolation and 6% smaller RMSE than a baseline distributed approach.

The Many Faces of Edge Intelligence

“…Latency as communication speed [34] Performance as model-building time [35] KPIs for full operational loop; low-latency network reconfiguration; latency of milliseconds Localized intelligence First steps toward edge-capable AI, federated learning; no generalization over multiple problems [36] Cloud-centric AI/ML capabilities; lack in real-time demands [1], [37] Distributed model building, sharing, and cooperation between different application verticals Where is the edge Local computational resources; tasks distribution over the network [38] Cloud computing paradigm [20] Dynamic collaboration and resource sharing between enduser devices, specific application-domain devices, and cloud services Ubiquitous resources Sensing for context-awareness and localized actions [38], [39] Application composition solutions, ML and reasoning to support the end-users [40], [41] Enabling self-management properties (e.g. migration, service continuity, application self-healing)…”

“…On the positive side, such dependency structures can offer a way to distribute the model. For example, Lovén et al [36] propose a distributed interpolation method that takes advantage of spatio-temporal dependencies and partitions data for local model learning along boundaries projected on the spatial dimension.…”

“…Environmental sensing The EDISON project, studied in Oulu, Finland, proposes an edge-native method and architecture for distributed interpolation [36]. EDISON assumes a large fleet of mobile sensors, collecting environmental data.…”

Edge Computing: The Computing Infrastructure for the Smart Megacities of the Future

“…We use the EDISON method for finding the optimal edge server placement as it has been shown to provide improved results over other state-of-the-art methods. 19 We first divide the area into a grid of 100 × 100 cells and then count the number of observations in each cell.…”

A dark and stormy night: Reallocation storms in edge computing

“…However, edge computing architectures come with several already envisioned challenges, including computational optimization and physical placement of the edge servers in dynamic scenarios with mobile users [5,6]. Particularly load balancing has seen as a mission-critical challenge for any computing service from cloud to local networking capabilities [7].…”