Database Reference
In-Depth Information
and arithmetic coding. Arithmetic coding achieves compression rates
close to the best possible, for a particular statistical model, which is
given by the information entropy. On the other hand, Huffman com-
pression is simpler and faster but produces poor results.
Lossless compression techniques, however, are not adequate for a num-
ber of reasons: (a) as experimentally found in [39], gzip lossless compres-
sion achieves poor compression (50%) compared to lossy techniques, (b)
lossless compression and decompression are usually more computation-
ally intensive than lossy techniques, and (c) indexing cannot be employed
for archived data with lossless compression.
6. Summary
In this chapter, we presented a comprehensive overview of the various
aspects of model-based sensor data acquisition and management. Pri-
marily, the objectives of the model-based techniques are ecient data
acquisition, handling missing data, outlier detection, data compression,
data aggregation and summarization. We started with acquisition tech-
niques like TinyDB [45], Ken [12], PRESTO [41]. In particular, we
focused on how acqusitional queries are disseminated in the sensor net-
work using routing trees [44]. Then we surveyed various approaches for
sensor data cleaning, including polynomial-based [73], probabilistic [21,
63, 52, 65] and declarative [31, 46].
For processing spatial, temporal and threshold queries, we detailed
query processing approaches like MauveDB [18], FunctionDB [64], par-
ticle filtering [33], MIST [5], etc. Here, our primary objective was to
demonstrate how model-based techniques are used for improving various
aspects of query processing over sensor data. Lastly, we discussed data
compression techniques, like, linear approximation [34, 39, 48], multi-
model approximations [39, 40, 50] and orthogonal transformations [1,
22, 53, 7].
All the methods that we presented in this chapter were model-based.
They utilized models - statistical or otherwise - for describing, simpli-
fying or abstracting various components of sensor data acquisition and
management.
Acknowledgments
This work was supported by the OpenSense project (Nano-Tera ref-
erence number 839 401), NCCR-MICS (
http://www.mics.org
), and by
the OpenIoT project (EU FP7-ICT 287305).
