We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

eess.SY

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Electrical Engineering and Systems Science > Systems and Control

Title: Distributed Event-Triggered Algorithms for Finite-Time Privacy-Preserving Quantized Average Consensus

Abstract: In this paper, we consider the problem of privacy preservation in the average consensus problem when communication among nodes is quantized. More specifically, we consider a setting where some nodes in the network are curious but not malicious and they try to identify the initial states of other nodes based on the data they receive during their operation (without interfering in the computation in any other way), while some nodes in the network want to ensure that their initial states cannot be inferred exactly by the curious nodes. We propose two privacy-preserving event-triggered quantized average consensus algorithms that can be followed by any node wishing to maintain its privacy and not reveal the initial state it contributes to the average computation. Every node in the network (including the curious nodes) is allowed to execute a privacy-preserving algorithm or its underlying average consensus algorithm. Under certain topological conditions, both algorithms allow the nodes who adopt privacypreserving protocols to preserve the privacy of their initial quantized states and at the same time to obtain, after a finite number of steps, the exact average of the initial states.
Comments: 12 pages
Subjects: Systems and Control (eess.SY); Optimization and Control (math.OC)
Cite as: arXiv:2102.06778 [eess.SY]
  (or arXiv:2102.06778v1 [eess.SY] for this version)

Submission history

From: Apostolos Rikos [view email]
[v1] Fri, 12 Feb 2021 21:19:32 GMT (2242kb,D)

Link back to: arXiv, form interface, contact.