Electrical Engineering
      and Computer Sciences

Electrical Engineering and Computer Sciences

COLLEGE OF ENGINEERING

UC Berkeley

Implementation of a Localization System for Sensor Networks

Tufan Coskun Karalar and Jan M. Rabaey

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2006-69
May 18, 2006

http://www.eecs.berkeley.edu/Pubs/TechRpts/2006/EECS-2006-69.pdf

Localization is very important for self-configuring wireless sensor networks. There are two main tasks to performing localization. Assuming availability of reference points, first the relationships to the reference points are established; in this thesis this relationship is the distance to the reference point. Second, using the reference point positions and the relations to these points, an algorithmic computation is carried out to compute the position. In the existing body of research on sensor network localization, the algorithmic aspects of this final position calculation have received the most attention. However there remain significant implementation issues related to both distance measurements and algorithmic computations. In this thesis the implementation issues regarding a sensor network localization system is studied along with some examples. In the first half, the implementation of a distributed, least-squares-based localization algorithm is presented. Low power and energy dissipation are key requirements for sensor networks. An ultra-low-power and dedicated hardware implementation of the localization system is presented. The cost of fixed-point implementation is also investigated. The design is implemented in a 0.13u CMOS process. It dissipates 1.7mW of active power and 0.122nJ/op of active energy with a silicon area of 0.55mm2. The mean calculated location error due to fixed-point implementation is shown to be 6%. In the second part, a radio frequency(RF) signal based Time of Flight (ToF) measuring ranging system for wireless sensor networks is proposed, designed and prototyped. The prototype measurement error is within -0.5m to 2m while operating at 100Msps sampling rate and using a 50MHz signal in the 2.4GHz ISM band. The system accuracy is limited by the sampling rate and can be linearly improved with increasing rates. This RF method is more cost effective than acoustic signal based ranging schemes, as it does not require ultrasonic transducers. The system is multipath resilient and can coexist with 2.4GHz band devices such as 802.11b/g networks. The estimated power consumption for the digital baseband is 2.35mW and its estimated area 0.25mm2 in a 90nm CMOS process.

Advisor: Jan M. Rabaey


BibTeX citation:

@phdthesis{Karalar:EECS-2006-69,
    Author = {Karalar, Tufan Coskun and Rabaey, Jan M.},
    Title = {Implementation of a Localization System for Sensor Networks},
    School = {EECS Department, University of California, Berkeley},
    Year = {2006},
    Month = {May},
    URL = {http://www.eecs.berkeley.edu/Pubs/TechRpts/2006/EECS-2006-69.html},
    Number = {UCB/EECS-2006-69},
    Abstract = {Localization is very important for self-configuring wireless sensor
networks. There are two main tasks to performing localization.
Assuming availability of reference points, first the relationships
to the reference points are established; in this thesis this
relationship is the distance to the reference point. Second, using
the reference point positions and the relations to these points, an
algorithmic computation is carried out to compute the position. In
the existing body of research on sensor network localization, the
algorithmic aspects of this final position calculation have received
the most attention. However there remain significant implementation
issues related to both distance measurements and algorithmic
computations.

In this thesis the implementation issues regarding a sensor network
localization system is studied along with some examples. In the
first half, the implementation of a distributed, least-squares-based
localization algorithm is presented. Low power and energy
dissipation are key requirements for sensor networks. An
ultra-low-power and dedicated hardware implementation of the
localization system is presented. The cost of fixed-point
implementation is also investigated. The design is implemented in a
0.13u CMOS process. It dissipates 1.7mW of active power and
0.122nJ/op of active energy with a silicon area of
0.55mm2. The mean calculated location error due to
fixed-point implementation is shown to be 6%.

In the second part, a radio frequency(RF) signal based Time of
Flight (ToF) measuring ranging system for wireless sensor networks
is proposed, designed and prototyped. The prototype measurement
error is within -0.5m to 2m while operating at 100Msps sampling rate
and using a 50MHz signal in the 2.4GHz ISM band. The system accuracy
is limited by the sampling rate and can be linearly improved with
increasing rates. This RF method is more cost effective than
acoustic signal based ranging schemes, as it does not require
ultrasonic transducers. The system is multipath resilient and can
coexist with 2.4GHz band devices such as 802.11b/g networks. The
estimated power consumption for the digital baseband is 2.35mW and
its estimated area 0.25mm2 in a 90nm CMOS process.}
}

EndNote citation:

%0 Thesis
%A Karalar, Tufan Coskun
%A Rabaey, Jan M.
%T Implementation of a Localization System for Sensor Networks
%I EECS Department, University of California, Berkeley
%D 2006
%8 May 18
%@ UCB/EECS-2006-69
%U http://www.eecs.berkeley.edu/Pubs/TechRpts/2006/EECS-2006-69.html
%F Karalar:EECS-2006-69