OODIDA: On-Board/Off-Board Distributed Real-Time Data Analytics for Connected Vehicles

Ulm, Gregor; Smith, Simon; Nilsson, Adrian; Gustavsson, Emil; Jirstrand, Mats

doi:10.1007/s41019-021-00152-6

Cited by 12 publications

(1 citation statement)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While in the past companies could potentially afford the central gathering of all the data sensed by an entire fleet of vehicles (see [8], [9] for applications and services in VANETs), a modern vehicle can now generate several gigabytes of data per hour [1], making this approach infeasible in terms of infrastructure and costs -which could be alleviated by enforcing the collection of just enough data from relevant vehicles only. Several recent studies in the literature are focusing on how to avoid general central data gathering by transitioning to models in which the data selection processes, or even the analysis itself, are pushed towards the vehicles [10], [11], [12], for efficient and continuous filtering [13], preprocessing through online compression [14], [7], and conversion of raw data into information in Federated Learning [15]. However, these distributed analysis models often lack mechanisms that attempt to involve only valid vehicles into the analysis, to thus avoid unnecessary computational load and data transfers VOLUME X, 2021 or other hindrances to the analysis.…”

Section: Introductionmentioning

confidence: 99%

Time- and Computation-Efficient Data Localization at Vehicular Networks’ Edge

et al. 2021

View full text Add to dashboard Cite

As Vehicular Networks rely increasingly on sensed data to enhance functionality and safety, efficient and distributed data analysis is needed to effectively leverage new technologies in real-world applications. Considering the tens of GBs per hour sensed by modern connected vehicles, traditional analysis, based on global data accumulation, can rapidly exhaust the capacity of the underlying network, becoming increasingly costly, slow, or even infeasible. Employing the edge processing paradigm, which aims at alleviating this drawback by leveraging vehicles' computational power, we are the first to study how to localize, efficiently and distributively, relevant data in a vehicular fleet for analysis applications. This is achieved by appropriate methods to spread requests across the fleet, while efficiently balancing the time needed to identify relevant vehicles, and the computational overhead induced on the Vehicular Network. We evaluate our techniques using two large sets of real-world data in a realistic environment where vehicles join or leave the fleet during the distributed data localization process. As we show, our algorithms are both efficient and configurable, outperforming the baseline algorithms by up to a 40× speedup while reducing computational overhead by up to 3×, while providing good estimates for the fraction of vehicles with relevant data and fairly spreading the workload over the fleet. All code as well as detailed instructions are available at https://github.com/dcs-chalmers/dataloc_vn.

show abstract