Spectra: Robust Estimation of Distribution Functions in Networks

  • Miguel Borges
  • Paulo Jesus
  • Carlos Baquero
  • Paulo Sérgio Almeida
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7272)

Abstract

The distributed aggregation of simple aggregates such as minima/maxima, counts, sums and averages have been studied in the past and are important tools for distributed algorithms and network coordination. Nonetheless, this kind of aggregates may not be comprehensive enough to characterize biased data distributions or when in presence of outliers, making the case for richer estimates.

This work presents Spectra, a distributed algorithm for the estimation of distribution functions over large scale networks. The estimate is available at all nodes and the technique depicts important properties: robustness when exposed to high levels of message loss, fast convergence speed and fine precision in the estimate. It can also dynamically cope with changes of the sampled local property and with churn, without requiring restarts. The proposed approach is experimentally evaluated and contrasted to a competing state of the art distribution aggregation technique.

References

  1. 1.
    Almeida, P.S., Baquero, C., Farach-Colton, M., Jesus, P., Mosteiro, M.A.: Fault-Tolerant Aggregation: Flow-Updating Meets Mass-Distribution. In: Fernàndez Anta, A., Lipari, G., Roy, M. (eds.) OPODIS 2011. LNCS, vol. 7109, pp. 513–527. Springer, Heidelberg (2011)CrossRefGoogle Scholar
  2. 2.
    Borges, M., Jesus, P., Baquero, C., Almeida, P.S.: Spectra: Robust estimation of distribution functions in networks. Tech. Rep. CoRR abs/1204.1373v1, HASLab / INESC TEC & Universidade do Minho (2012), http://arxiv.org/abs/1204.1373v1
  3. 3.
    Cheng, S., Li, J., Ren, Q., Yu, L.: Bernoulli Sampling Based (ε, δ)-Approximate Aggregation in Large-Scale Sensor Networks. In: 29th IEEE Conference on Information Communications (INFOCOM), pp. 1–9 (2010)Google Scholar
  4. 4.
    Ganesh, A.J., Kermarrec, A.M., Le Merrer, E., Massoulié, L.: Peer counting and sampling in overlay networks based on random walks. Distributed Computing 20(4), 267–278 (2007)CrossRefGoogle Scholar
  5. 5.
    Haridasan, M., van Renesse, R.: Gossip-based Distribution Estimation in Peer-to-Peer Networks. In: 7th International Conference on Peer-To-Peer Systems (IPTPS). USENIX Association (2008)Google Scholar
  6. 6.
    Jelasity, M., Montresor, A.: Epidemic-style proactive aggregation in large overlay networks. In: Proceedings of the 24th International Conference on Distributed Computing Systems (ICDCS 2004), pp. 102–109. IEEE Computer Society (2004)Google Scholar
  7. 7.
    Jelasity, M., Montresor, A., Babaoglu, O.: Gossip-based aggregation in large dynamic networks. ACM Transactions on Computer Systems 23(3), 219–252 (2005)CrossRefGoogle Scholar
  8. 8.
    Jesus, P., Baquero, C., Almeida, P.S.: Fault-Tolerant Aggregation for Dynamic Networks. In: 29th IEEE Symposium on Reliable Distributed Systems, pp. 37–43 (2010)Google Scholar
  9. 9.
    Jesus, P., Baquero, C., Almeida, P.S.: A Survey of Distributed Data Aggregation Algorithms. Tech. Rep. CoRR abs/1110.0725, HASLab / INESC TEC & Universidade do Minho (2011), http://arxiv.org/abs/1110.0725
  10. 10.
    Lynch, N.A.: Distributed algorithms, pp. 17–23. Kaufmann Publishers Inc. (1996)Google Scholar
  11. 11.
    Sacha, J., Napper, J., Stratan, C., Pierre, G.: Adam2: Reliable distribution estimation in decentralised environments. In: 2010 IEEE 30th International Conference on Distributed Computing Systems (ICDCS), pp. 697–707 (June 2010)Google Scholar

Copyright information

© IFIP International Federation for Information Processing 2012

Authors and Affiliations

  • Miguel Borges
    • 1
  • Paulo Jesus
    • 1
  • Carlos Baquero
    • 1
  • Paulo Sérgio Almeida
    • 1
  1. 1.HASLab / INESC TEC & Universidade do MinhoBragaPortugal

Personalised recommendations