Editorial
EURASIP Journal on Wireless Communications and Networking
Biao Chen 0 1 2 3
Wendi B. Heinzelman 0 1 2 3
Mingyan Liu 0 1 2 3
Andrew T. Campbell
0 Department of Electrical Engineering and Computer Science, University of Michigan , Ann Arbor, MI 48109-2122 , USA
1 Department of Electrical and Computer Engineering, University of Rochester , Rochester, NY 14627 , USA
2 Department of Electrical Engineering and Computer Science, Syracuse University , Syracuse, NY 13244 , USA
3 Department of Electrical Engineering and Center for Telecommunications Research, Columbia University , NY 10027 , USA
-
Recent advances in integrated circuits and in digital
wireless communication technologies have enabled the design
of wireless sensor networks (WSN) to facilitate the joint
processing of spatially and temporally distributed
information. Such networks immensely enhance our ability to
understand and evaluate complex systems and environments.
Using wireless connections for sensor networks offers
increased flexibility in deployment and reconfiguration of
the networks and reduces infrastructure cost. These
advantages enable WSN applications in areas ranging from
battlefield surveillance to environment monitoring and control to
telemedicine.
Enormous challenges in the understanding of sensor
networks presently impede deployment of many of the
envisaged applications. In particular, for WSN that employ in situ
unattended sensors, physical constraints, including those of
power, bandwidth, and cost, have presented significant
challenges as well as research opportunities in the field. Of
particular interest to this special issue are topics related to the
communications and networking aspects of WSN. Indeed,
one of the major concerns in sensor networks is maintaining
connectivity and networking functions with geographically
dispersed sensor nodes under stringent resource constraints.
This is further exacerbated by the volume of data generated
by the sensors, which is disproportionately large compared
with the network capacity. The papers in this special issue
are reflections of some of these issues.
Sensor networks are typically built to perform some
system-wide missions, that is, collective inference tasks that
involve all sensor nodes. Examples include detection of an
event and estimation of a parameter or a process. The first
three papers are concerned with designing such WSN. The
first paper, coauthored by Niu and Varshney, considers the
detection of an event in sensor networks with a random
number of sensors. High network density and limited
bandwidth impose a severe constraint on the number of bits each
sensor can transmit, and the authors treat the extreme case
where a single bit is sent from each sensor. Under the
assumption of a Poisson model on the number of sensors, a
simple counting rule is proposed at the fusion center to strike
a balance between performance and requirement on a
priori information. This work demonstrates that for large-scale
heterogeneous sensor networks, heuristics based on intuition
often trump theoretically optimal processing that is typically
too demanding in its requirement. Under the same network
architecture, that is, a number of sensors communicating
with a single fusion center, the second paper, by Krasnopeev
et al. treats an estimation problem where the unknown
signal is corrupted by spatially correlated additive noises. Again,
bandwidth constraints dictate that each sensor sends a finite
number of bits to the fusion center. By exploiting the spatial
correlation of the noise in terms of its covariance matrix, the
minimum energy quantizer design is reformulated as a
convex optimization problem and hence can be solved efficiently
EURASIP Journal on Wireless Communications and Networking
using standard convex programming. Taking the problem
one step further, the third paper, coauthored by Del Ser et al.
deals with the estimation of a random process. Specifically,
two binary sources, whose correlation is modeled by a hidden
Markov process, are transmitted and the receiver is assumed
to reliably recover one of them. This then serves as side
information for the decoding of the other. It was demonstrated
that the hidden Markov model parameters and the
transmitted source can be jointly recovered via iterative decoding.
A perennial problem encountered in large-scale sensor
networks is medium access control (MAC): the lack of a
central node and the stringent bandwidth and other resource
constraints make it an extremely difficult problem. In the
paper by Yang et al., the authors consider information
retrieval and processing problems in the SENMA (sensor
networks with mobile access) network architecture, where data
generated by ground sensors are collected by mobile access
points (e.g., unmanned aerial vehicles). Three MAC
protocols are proposed to produce desired data-retrieval patterns,
so as to minimize the reconstruction distortion. These MAC
protocols integrate random access by the sensor nodes a (...truncated)