Distributed approximation and tracking using selective gossip

D. Üstebay, R.M. Castro, M. Coates, M. Rabbat

Research output: Chapter in Book/Report/Conference proceedingChapterAcademic

Abstract

This chapter presents selective gossip which is an algorithm that applies the idea of iterative information exchange to vectors of data. Instead of communicating the entire vector and wasting network resources, our method adaptively focuses communication on the most significant entries of the vector. We prove that nodes running selective gossip asymptotically reach consensus on these significant entries, and they simultaneously reach an agreement on the indices of entries which are insignificant. The results demonstrate that selective gossip provides significant communication savings in terms of the number of scalars transmitted. In the second part of the chapter we propose a distributed particle filter employing selective gossip. We show that distributed particle filters employing selective gossip provide comparable results to the centralized bootstrap particle filter while decreasing the communication overhead compared to using randomized gossip to distribute the filter computations.
Original languageEnglish
Title of host publicationCompressed sensing & sparse filtering
EditorsL. Mihaylova, S.J. Godsill
Place of PublicationBerlin
PublisherSpringer
Pages325-355
ISBN (Print)978-3-642-38397-7
DOIs
Publication statusPublished - 2014

Publication series

NameSignals and Communication Technology
ISSN (Print)1860-4862

Fingerprint

Communication

Cite this

Üstebay, D., Castro, R. M., Coates, M., & Rabbat, M. (2014). Distributed approximation and tracking using selective gossip. In L. Mihaylova, & S. J. Godsill (Eds.), Compressed sensing & sparse filtering (pp. 325-355). (Signals and Communication Technology). Berlin: Springer. https://doi.org/10.1007/978-3-642-38398-4_10
Üstebay, D. ; Castro, R.M. ; Coates, M. ; Rabbat, M. / Distributed approximation and tracking using selective gossip. Compressed sensing & sparse filtering. editor / L. Mihaylova ; S.J. Godsill. Berlin : Springer, 2014. pp. 325-355 (Signals and Communication Technology).
@inbook{23f9439dbb2847bcb31e9b7bbd609f90,
title = "Distributed approximation and tracking using selective gossip",
abstract = "This chapter presents selective gossip which is an algorithm that applies the idea of iterative information exchange to vectors of data. Instead of communicating the entire vector and wasting network resources, our method adaptively focuses communication on the most significant entries of the vector. We prove that nodes running selective gossip asymptotically reach consensus on these significant entries, and they simultaneously reach an agreement on the indices of entries which are insignificant. The results demonstrate that selective gossip provides significant communication savings in terms of the number of scalars transmitted. In the second part of the chapter we propose a distributed particle filter employing selective gossip. We show that distributed particle filters employing selective gossip provide comparable results to the centralized bootstrap particle filter while decreasing the communication overhead compared to using randomized gossip to distribute the filter computations.",
author = "D. {\"U}stebay and R.M. Castro and M. Coates and M. Rabbat",
year = "2014",
doi = "10.1007/978-3-642-38398-4_10",
language = "English",
isbn = "978-3-642-38397-7",
series = "Signals and Communication Technology",
publisher = "Springer",
pages = "325--355",
editor = "L. Mihaylova and S.J. Godsill",
booktitle = "Compressed sensing & sparse filtering",
address = "Germany",

}

Üstebay, D, Castro, RM, Coates, M & Rabbat, M 2014, Distributed approximation and tracking using selective gossip. in L Mihaylova & SJ Godsill (eds), Compressed sensing & sparse filtering. Signals and Communication Technology, Springer, Berlin, pp. 325-355. https://doi.org/10.1007/978-3-642-38398-4_10

Distributed approximation and tracking using selective gossip. / Üstebay, D.; Castro, R.M.; Coates, M.; Rabbat, M.

Compressed sensing & sparse filtering. ed. / L. Mihaylova; S.J. Godsill. Berlin : Springer, 2014. p. 325-355 (Signals and Communication Technology).

Research output: Chapter in Book/Report/Conference proceedingChapterAcademic

TY - CHAP

T1 - Distributed approximation and tracking using selective gossip

AU - Üstebay, D.

AU - Castro, R.M.

AU - Coates, M.

AU - Rabbat, M.

PY - 2014

Y1 - 2014

N2 - This chapter presents selective gossip which is an algorithm that applies the idea of iterative information exchange to vectors of data. Instead of communicating the entire vector and wasting network resources, our method adaptively focuses communication on the most significant entries of the vector. We prove that nodes running selective gossip asymptotically reach consensus on these significant entries, and they simultaneously reach an agreement on the indices of entries which are insignificant. The results demonstrate that selective gossip provides significant communication savings in terms of the number of scalars transmitted. In the second part of the chapter we propose a distributed particle filter employing selective gossip. We show that distributed particle filters employing selective gossip provide comparable results to the centralized bootstrap particle filter while decreasing the communication overhead compared to using randomized gossip to distribute the filter computations.

AB - This chapter presents selective gossip which is an algorithm that applies the idea of iterative information exchange to vectors of data. Instead of communicating the entire vector and wasting network resources, our method adaptively focuses communication on the most significant entries of the vector. We prove that nodes running selective gossip asymptotically reach consensus on these significant entries, and they simultaneously reach an agreement on the indices of entries which are insignificant. The results demonstrate that selective gossip provides significant communication savings in terms of the number of scalars transmitted. In the second part of the chapter we propose a distributed particle filter employing selective gossip. We show that distributed particle filters employing selective gossip provide comparable results to the centralized bootstrap particle filter while decreasing the communication overhead compared to using randomized gossip to distribute the filter computations.

U2 - 10.1007/978-3-642-38398-4_10

DO - 10.1007/978-3-642-38398-4_10

M3 - Chapter

SN - 978-3-642-38397-7

T3 - Signals and Communication Technology

SP - 325

EP - 355

BT - Compressed sensing & sparse filtering

A2 - Mihaylova, L.

A2 - Godsill, S.J.

PB - Springer

CY - Berlin

ER -

Üstebay D, Castro RM, Coates M, Rabbat M. Distributed approximation and tracking using selective gossip. In Mihaylova L, Godsill SJ, editors, Compressed sensing & sparse filtering. Berlin: Springer. 2014. p. 325-355. (Signals and Communication Technology). https://doi.org/10.1007/978-3-642-38398-4_10