Presentation 11/29: Ecolocation: A Sequence Based Technique for RF Localization in Wireless Sensor Networks
8 views
Skip to first unread message
Su Kim
Nov 27, 2007, 2:16:41 PM11/27/07
to ASU:CSE535 FALL 07 Mobile Computing
Authors: Kiran Yedavalli, Bhaskar Krishnamachari, Sharmila Ravula,
Bhaskar Srinivasan,
Title: "Ecolocation: A Sequence Based Technique for RF Localization in
Wireless Sensor Networks,"
Published: IPSN 2005
Bhaskar Srinivasan,
Title: "Ecolocation: A Sequence Based Technique for RF Localization in
Wireless Sensor Networks,"
Published: IPSN 2005
weiji...@asu.edu
Dec 1, 2007, 1:26:28 AM12/1/07
to ASU:CSE535 FALL 07 Mobile Computing
Bellow, please find my critique
Critique for "Ecolocation: A Sequence Based Technique for RF
Summary:
In the paper a new sequence-based RF localization algorithm,
ecolocation, was proposed and examined using simulation, indoor
experiments and outdoor experiments. Besides examining the ordered
sequence of RSSI, a constraint-based approach was also adopted to
provide certain amount of robustness. In the end, the author claims
that the proposed approach generates better results than other state
of the art approaches in terms of localization accuracy and precision,
which lacks enough datum and results in a weak point.
Pros:
The technique is novel. Though it will not generate good results some
times plus it is more suitable for simulation rather than real world
experiments, new techniques or new attempts are always encouraged.
The technique is adaptive to both indoor and outdoor localization.
Techniques for localization are always specially optimized for indoor
or outdoor environment. However, the mobile host might move from
indoor to outdoor, or vice versa. And at this point, the techniques
that are suitable for both indoor/outdoor localization are highly
appreciated.
The author really spends a lot of efforts to do the simulation and
those indoor or outdoor experiments which increase the strength of the
argument.
Through a constraint-based approach, the technique does achieve some
degree of robustness to signal propagation and shadowing ect.
Cons:
The argument that the technique outweighs the state of the art
approaches is not strongly supported by the actually experiment
results.
Though through adopting a constraint-based approach, the technique
achieve some degree of robustness, still the technique uses RSSI to
indicate distance parameter, which means the it will suffer from
unstable signal strength and signal propagation.
It is not specially optimized for indoor or outdoor localization, it
could be an advantage. However,
on the other hand, this also indicates that many factors that will
impact the results were not taken
into account so that the technique is not fully optimized.
Critique for "Ecolocation: A Sequence Based Technique for RF
Localization in Wireless Sensor Networks"
Weijia Che
Summary:
In the paper a new sequence-based RF localization algorithm,
ecolocation, was proposed and examined using simulation, indoor
experiments and outdoor experiments. Besides examining the ordered
sequence of RSSI, a constraint-based approach was also adopted to
provide certain amount of robustness. In the end, the author claims
that the proposed approach generates better results than other state
of the art approaches in terms of localization accuracy and precision,
which lacks enough datum and results in a weak point.
Pros:
The technique is novel. Though it will not generate good results some
times plus it is more suitable for simulation rather than real world
experiments, new techniques or new attempts are always encouraged.
The technique is adaptive to both indoor and outdoor localization.
Techniques for localization are always specially optimized for indoor
or outdoor environment. However, the mobile host might move from
indoor to outdoor, or vice versa. And at this point, the techniques
that are suitable for both indoor/outdoor localization are highly
appreciated.
The author really spends a lot of efforts to do the simulation and
those indoor or outdoor experiments which increase the strength of the
argument.
Through a constraint-based approach, the technique does achieve some
degree of robustness to signal propagation and shadowing ect.
Cons:
The argument that the technique outweighs the state of the art
approaches is not strongly supported by the actually experiment
results.
Though through adopting a constraint-based approach, the technique
achieve some degree of robustness, still the technique uses RSSI to
indicate distance parameter, which means the it will suffer from
unstable signal strength and signal propagation.
It is not specially optimized for indoor or outdoor localization, it
could be an advantage. However,
on the other hand, this also indicates that many factors that will
impact the results were not taken
into account so that the technique is not fully optimized.
shayok.ch...@asu.edu
Dec 4, 2007, 12:05:21 PM12/4/07
to ASU:CSE535 FALL 07 Mobile Computing
Hi,
Please find my critique for this paper below:
Ecolocation: A Sequence Based Technique for RF Localization in
Wireless Sensor Networks
that is based on the ordered sequence of received signal strengths at
different reference nodes placed in the deployment area. The algorithm
is insensitive to absolute signal strength amplitudes which renders it
robust to signal fluctuations and fading. A wide range of experiments
have been carried out under different conditions which prove the
superiority of Ecolocation over other localization techniques.
Strong Points:
1. The proposed approach is robust to signal strength fluctuations and
multipath fading.
2. The authors have provided a distinct analysis of the time
complexity of their algorithm.
3. The proposed approach has been tested under a wide variety of
channel conditions, node deployment parameters and indoor and outdoor
environments.
4. The algorithm has also been compared to other localization
algorithms like proximity, centroid, MLE which prove the superiority
of Ecolocation.
Weak Points:
1. No special measure was taken to optimize the algorithm in case of
indoor or outdoor localizations.
2. The proposed method performs better than other techniques but does
not totally take care of multipath fading and other signal
unstabilizing factors. Also, there are some cases where other schemes
like proximity perform better than Ecolocation. This is unexplained.
3. The calculation of the distance Da and the estimation of the
average location error as a percentage of this distance is not very
convincing.
4. The implementation described in the paper is not optimized for
space or time complexity.
Regards,
Shayok.
On Dec 4, 2:52 am, "Shiraz Saleem" <shirazsaleem...@gmail.com> wrote:
> Pls find my crtical analysis attached
>
> Shiraz
>
> 16KDownload
0 new messages