Research Area: Edge Computing
In an edge computing-enhanced Internet of Things (IoT) setup, data can be processed closer to the IoT devices (i.e. at the network edge). However, security and privacy remain two key issues that need to be considered. In this paper, we propose the first multidimensional and multidirectional data aggregation (MMDA) scheme for privacy-preserving edge computing-enhanced IoT communications. In MMDA, the data of each IoT device are described as an n-dimensional vector and m IoT devices’ data are listed as a matrix D of order m × n. MMDA enables an edge device (acting as a gateway) to aggregate the multidimensional data of the m IoT devices in two directions: row aggregation and column aggregation. Such data can then be employed to compute the summation of data in each row and each column of D in a privacy-preserving way. Unlike existing multidimensional data aggregation schemes that have only the column aggregation, MMDA allows an additional row aggregation. This allows the capability to provide more statistical information to an IoT control center for analysis and processing. MMDA also adopts the batch verification technology to reduce authentication costs. Extensive analysis shows that MMDA is practicable in terms of computation cost, security, and fault-tolerance.
Keywords:
Edge computing
Internet of Things
Data aggregation
Multidimensional
Multidirectional
Fault-tolerance
Author(s) Name: Peng Zeng, Bofeng Pan, Kim Kwang Raymond Choo, Hong Liu
Journal name: Journal of Systems Architecture
Conferrence name:
Publisher name: Elsevier
DOI: 10.1016/j.sysarc.2020.101713
Volume Information: Volume 106, June 2020, 101713
