Managing energy & data quality in swarms
Jinseok Yang, Sameer Tilak, Tajana Simunic Rosing

Citation
Jinseok Yang, Sameer Tilak, Tajana Simunic Rosing. "Managing energy & data quality in swarms". Talk or presentation, October, 2014; Poster presented at the 2014 TerraSwarm Annual Meeting.

Abstract
Information network generates information based on the upload raw data which was collected from the deployed sensing platforms. The sensing platforms have the limited battery capacity. Thus, reducing energy consumption is important to support seamless information. There exits two different approaches to decrease energy consumption. The first approach is delaying message transmission. All raw data collections transform to a message. The proposed approach finds out the best transmission instance that minimize energy consumption and probability of message time-out. The approach consumes 26% to 48% less energy than state-of-arts without message timing out. The second approach is data reconstruction. Sensing platform can reduce energy consumption by skipping transmission. The compressive sensing uses sparsity factor for data reconstruction. However, we observed that various sensors and length of sensing interval cause different sparsity factor. We use matrix factorization method to decompose the sparse training set. We extract the interactions among time, location and sensor types. Our approach achieves 1.2 to 4 times better reconstruction error than other approaches

Electronic downloads

Internal. This publication has been marked by the author for TerraSwarm-only distribution, so electronic downloads are not available without logging in.

Citation formats

• HTML

  Jinseok Yang, Sameer Tilak, Tajana Simunic Rosing. <a
  href="http://www.terraswarm.org/pubs/394.html"><i>Managing
  energy & data quality in swarms</i></a>,
  Talk or presentation,  October, 2014; Poster presented at
  the <a
  href="http://www.terraswarm.org/conferences/14/annual"
  >2014 TerraSwarm Annual Meeting</a>.

• Plain text

  Jinseok Yang, Sameer Tilak, Tajana Simunic Rosing.
  "Managing energy & data quality in swarms".
  Talk or presentation,  October, 2014; Poster presented at
  the <a
  href="http://www.terraswarm.org/conferences/14/annual"
  >2014 TerraSwarm Annual Meeting</a>.

• BibTeX

  @presentation{YangTilakRosing14_ManagingEnergyDataQualityInSwarms,
      author = {Jinseok Yang and Sameer Tilak and Tajana Simunic
                Rosing},
      title = {Managing energy \& data quality in swarms},
      month = {October},
      year = {2014},
      note = {Poster presented at the <a
                href="http://www.terraswarm.org/conferences/14/annual"
                >2014 TerraSwarm Annual Meeting</a>.},
      abstract = {Information network generates information based on
                the upload raw data which was collected from the
                deployed sensing platforms. The sensing platforms
                have the limited battery capacity. Thus, reducing
                energy consumption is important to support
                seamless information. There exits two different
                approaches to decrease energy consumption. The
                first approach is delaying message transmission.
                All raw data collections transform to a message.
                The proposed approach finds out the best
                transmission instance that minimize energy
                consumption and probability of message time-out.
                The approach consumes 26% to 48% less energy than
                state-of-arts without message timing out. The
                second approach is data reconstruction. Sensing
                platform can reduce energy consumption by skipping
                transmission. The compressive sensing uses
                sparsity factor for data reconstruction. However,
                we observed that various sensors and length of
                sensing interval cause different sparsity factor.
                We use matrix factorization method to decompose
                the sparse training set. We extract the
                interactions among time, location and sensor
                types. Our approach achieves 1.2 to 4 times better
                reconstruction error than other approaches 	 },
      URL = {http://terraswarm.org/pubs/394.html}
  }
  

Posted by Jinseok Yang on 21 Oct 2014.