RET Summer Internship Program UNT, 2009

waralligatorMobile - Wireless

Nov 21, 2013 (3 years and 6 months ago)

73 views

RET Summer Internship Program UNT, 2009





Funded by NSF Grants:




NSF IIS
-
0844342

DLR 0431818,

CI
-
TEAM 0636421, CRI 0709285







Dawn Chegwidden

Sharon Wood

Introduction


Wireless Sensor Network (WSN) technology has a broad range of present and
future applications for monitoring environmental conditions in the field.



Applications may include water quality, soil moisture, rain events, medical
monitoring, sprinkler systems, agriculture, and volcanic activity [1],[4],


The advantages of these WSN systems are that they can be set up in the field
and remotely monitored from a laptop or office computer.


A working system is cost effective and should require only periodic
maintenance.


Some potential problems that need to be considered when deploying a WSN
system include vegetation humidity, temperature, weather, and topography


Environmental conditions affect the sensor networks thus affecting robustness
of the network system overall


Some of the sensor issues when deploying the WSN in the field are: receiver
signal strength (RSS), packet receiver rate (PRR), battery power and hopping
distance between nodes. [4]






Distance Effects on Received Signal Strength(RSS)
and Packet Receiver Rate (PRR)


Our research will take place at two sites: (1) Discovery Park and (2)
Pecan Creek Wastewater Treatment Facility


Due to the importance of WSN and PRR in the design and
implementation of a WSN in the field, our research will focus on:



How does distance between motes affect RSS and PRR?


How does topology effect the RSS and PRR?


Does temperature and humidity affect the RSS and/or PRR?




Receiver Signal Strength (RSS)


Receiver signal strength is important in WSN deployment


RSS is a measure of the received radio frequency (RF) signal


RF sensitivity is represented by the lowest RSS signal that retrieves complete data
from a neighboring node


Field tests are necessary to determine the “best” range for sensor topology


Environmental habitats vary in topography and vegetation


This can present a problem when deploying nodes in the field


Therefore, an evaluation of the relationship between RSS ,PRR and distance can
help determine the placement of nodes in the field

.


Packet Receiver Rate (PRR)


Packet receiver rate is an indicator of the % data received from a neighboring node.


The % of data being received is related to the RSS and the distance between
sensor nodes.


Data retrieval is critical in maintaining a workable WSN system


If transmission strength is weak then the PRR will decrease and data could be lost
and energy consumption will be higher due to re
-
transmission


The goal is to find a balance between RSS, PRR, and distance so that data retrieval
is close to 100
%

Our Wireless Sensor Technology


Wireless sensors are small and compact.


They contain a data acquisition board (small
computer) and sensor board


The sensor board contains various sensors for
measuring temperature, humidity, RSS, PRR, and
battery voltage


Panasonic Alkaline Plus AA batteries were used
to power the sensors in the field


Solar panels have been used to extend battery life
though they were not used in this experiment


With duty cycling, the sensor alternates between
an awake and sleep period which can extend the
life of the sensor’s batteries.




Wireless Measurement System


This is a wireless communications system
in which every node has router capacity [8]


2.4 GHz


RF power: 3
dBm


Sensitivity:
-
101
dBm


Outdoor Range


300 m


Indoor Range


50 m


Battery


2X AA batteries


Iris XM2110

Oxbow: www.xbow.com

WSN Sensor Mote:

Data Acquisition Board

Features:

MDA 300



A multi
-
functional board with
temperature and humidity sensor


It is designed to be used as the
primary interface between iris board
and external sensors


It can be used for environmental
and habitat monitoring such as
humidity, temperature, and soil
moisture. [8]



Crossbow: www.xbow.com

The Experimental Design: Discovery Park


Discovery Park is located on a large open filed
with little vegetation behind the university
research center.


Our deployment network consisted of a two
row, 150 x 10 meter, network matrix


Each node was approximately 10 meters from
its nearest neighbor nodes, making the furthest
node 150 meters away


Data collection was every 6 minutes


Data collection began at 2 pm on July 25, 2009
and ran for 45 hours
.




10 meters

10 meters

150
meters

Discovery Park WSN Deployment Pictures


WSN nodes were deployed in two


rows of 15


10 meters apart and 10 meters


across
.


Nodes were set out July 24th


Due to a “bug” in the program, nodes were
brought back in on July 25
th
after 12:00
noon


They were re
-
programmed and then
deployed on July 26, at 2 pm


Experimental Design
-

Discovery Park

15

14

13

12

11

10

9

8

7

6

5

4

3

2

1

30*

29*

28

27

26

25

24

23*

22*

21

20

19

18

17

16*

non
-
transparent lid,

transparent lid,

*
Experimental data was retrieved and analyzed from nodes 16,22, 23, 29 and 30







Slide adapted with permission from Laura De
Lemos

Discovery Park: Data Graphs


Node 23


The received signal strength and distance
of node 23 was compared to its
surrounding neighbors


There is a negative correlation between
neighboring sensors which is indicated
by the red line


The packet receiver rate and distance of
node 23 was also compared to its
surrounding neighbors


PRR showed 100% data collection for
most of the nodes up until about 70 m



Discovery Park: Data Graphs


Node 23


This graph demonstrates how
temperature and humidity will affect
RSS


Higher temperature can lead to a slight
decrease in RSS


There was a 2
-
5
dBm

RSS loss as
temperature goes from 25
0
C to 45
0
C.


As humidity increases, the RSS
becomes stronger.


Discovery Park: Data Graphs


Node 23


This graph looks at the RSS from
node 23 to its three neighbors (29,22,
16)


It appears that on 8 PM on Sunday,
July 27, there was 2
-
4
dBm

jump


Node 23 did not receive a signal from
node 16 until 8 PM


It is not clear why signal from node
was not received until after 8:00 PM.


One possibility might be an increase
humidity and decreased temperature
due to rain


Discovery Park: Data Graphs


Node 23


This graph compares the trend between
temperature, humidity, and battery voltage


As temperature decreased humidity
increased


Battery voltage dropped from 3.0 V to 2.6 V


Conclusion: Voltage drops due to usage but
does not appear to be significantly affected
by temperature or humidity



Discovery Park: Data Graphs


Node 30



This graph is similar to Node 23


The RSS drops to
-
90 after 120
meters


The PRR is not stable after 90 meters

Result Discussions:



Evaluating the data, we conclude that a compromise
between distance, RSS and PRR would result in the
following
:


Experimental data seems to show that the RSS like cool, wet weather
and as opposed to warm, dry weather


This being the case, recommended distance between nodes would be
50 meters in areas with low vegetation


RSS sensitivity
-
85
dBm

or less for better PRR


Using the recommended distance would allow for variable environmental
factors associated with Texas summers


















Part 2: Experimental Goals: Pecan Creek
Wastewater Treatment Facility


To measure the signal transmission strength and
packet receiver rate and determine how it is
affected by distance.


To determine if temperature and humidity affect
the signal transmission strength and packet
receiver rate


To compare the results between our initial tests
at Discovery Park with data collected at Pecan
Creek


Recent data collected from Greenbelt Corridor
was compared as well

Experimental Design: Pecan Creek


We deployed 8 nodes and one gateway at
Pecan Creek in an area with moderate
vegetation


The nodes were set up in a clearing
beyond already existing nodes that were
currently being used in another experiment.


Nodes were set up randomly


Sensor # 1 was used as our gateway
sensor.


The nodes were activated at Thursday, 9
AM, July 30, 2009


Nodes were collected on Monday, 9 AM,
August 3, 2009

Pecan Creek Nodes

Distance and Elevation Diagram

Gateway

540 ft

543 ft

539 ft

541 ft

536 ft

540 ft

543 ft

544 ft

544 ft

Transparent lid cover

Non
-

transparent lid
cover

15 m

16 m

20 m

15 m

20 m

17 m

16 m

21 m

16 m

11 m

18 m

20 m

16 m

Elevation


(ft)

# 26

# 25*

#1

# 23

# 27*

# 5*

# 21

# 24

# 22

Node Identification #

*Data evaluated included
notes 5,25, and 27


ID #

1

N33.19587
-
W097.07283

5

N33.19585
-

W097.07265

21

N33.19581
-
W097.07244

24

N33.19595
-
W097.07250

22

N33.19609
-

W097.07257

27

N33.19618
-

W097.07266

26

N33.19615
-

W097.07284

25

N33.19601
-

W097.07284

23

N33.19603
-

W097.07267

1

16

36

32

34

37

31

15

23

5

16

20

17

27

36

37

25

20

21

36

20

16

33

45

53

43

32

24

32

17

16

21

29

38

32

18

22

34

27

33

21

13

26

26

11

27

37

36

45

29

13

17

25

16

26

31

37

53

38

26

17

15

20

25

15

25

43

32

26

25

15

16

23

23

20

32

18

11

16

20

16

Distance between nodes at Pecan Creek using GPS Coordinates

Data evaluated from nodes 5, 21, 25, 26,and 27

Data Graph Comparison: RSS/PRR and Distance


Discovery Par
k
-

little vegetation


Pecan Creek
-

mixed vegetation and clearings


Greenbelt Corridor
-

dense vegetation


D.P.* Node 23

P.C*. Node 27

* D. P (Discovery Park, P.C (Pecan Creek), GBC.
(Green Belt Corridor)

GBC*, Node 4 Dense Vegetation

Data Graph Comparison: RSS, Temperature and
Humidity

P.C. Node 27

D.P. Node 23

Discovery Park
:


2
-

4 temperature
dBm

drop


4


6 % humidity increase


Temperature range: 30
0
C to 45
0
C

Pecan Creek


1
-
2
dBm

temperature drop


1


2 % humidity increase


Temperature range: 20
0
C to 40
0
C

GBC. Node 4

Greenbelt Corridor (Dense
Vegetation)


There was not a significant decrease
in RSS due to temperature


No significant increase
in
RSS in
increased humidity

Data Graph Comparisons: RSS Over Time



P.C. Node 27

D.P. Node 23

GBC. Node 4


Data graphs indicate that the RSS varies over time..


Temperature, humidity, weather, and vegetation can affect the signal strength


The graphs below are an interesting comparison between low, moderate, and dense
vegetation


It indicates the importance of distance and vegetation with respect to both RSS and PRR


Related Work


Outdoor research in the
Sonoran

Desert showed a reduction in RSS during the
hottest times of the day. There was also a noticeable daily variation in the signal
strength. A linear decrease of about a 8
dBm

was noted in RSS for the transition from
25 C to 65 C. While Texas temperature ranges were similar to the
Sonoran

Desert,
there was significantly less humidity and different vegetation.[1]


Research on potato fields also concludes that radio waves propagate better with high
humidity (i.e. night and rain). [4]


UNT research at the Greenbelt Corridor ( area of dense vegetation) indicated that

nodes can transmit 30 m with 95% PRR and 50 m with 80% PRR .Seasonal
variations can also affect the RSS and PRR.[5]


Greenbelt Corridor


Jue

(Jerry) Yang set out 15 nodes in the Greenbelt after
examining our Discovery Park data. It was interesting to note that the graph for
dense vegetation showed that RSS and PRR became unstable at 30 meters or less.
[6]

Conclusions and Recommendations


Pecan Creek and the Greenbelt are the areas being monitored by the TEO site .


Both sites hope to expand their sensors to over 100 for monitoring temperature,
humidity, and soil moisture.


Discovery Park helped establish base line data in an area with very little vegetation
and interference


The Pecan site has some trees closer to the gateway and a wide clearing further
away.


The Greenbelt has many trees and few real clearings.


Experimental data showed a distance between nodes of 70 meters at the Discovery
Park and 45 meters at Pecan Creek.


We suggest that due to the variable temperatures and humidity during Texas
summers that the distance between motes should be 30 meters or less in forested
areas and 50 meters in clearings.


Data indicates that the next step should be to deploy a network using the above
distances and evaluate the relationship between RSS and PRR while collecting
real
-
time environmental data such as soil moisture


References:



[1] Gupta,
Sandeep
, Gianni
Giorgetti

and Kenneth Bannister. "Wireless Sensor Networking for "Hot"
Applications: Effects of Temperature on Signal Strength, Data collection and Localization."
HotEmNets

(2008).


[2] Holland, Matthew M., Ryan G.
Aures

and Wendi B.
Heinzelman
.
Experimental Investigation of Radio
Performance in Wireless Sensor Networks.

Rochester, New York: University of Rochester.


[3]
Srinivasan
,
Kannan

and Philip Levis.
RSSI is Under Appreciated.

Stanford, CA: Stanford University.


[4]
Thelen
, John,
Dann

Goense

and
Koen

Langendoen
.
Radio Wave Propagation in Potato Fields.

July
2009 <http://www.st.ewi.tudelft.nl/~koen/papers/winmee.pdf>.


[5] Yang,
Jue
, et al.
Integration of Wireless Sensor Networks in Environmental Monitoring Cyber
Infrastructure.

Denton, Texas.


[6] Yang,
Jue
. " “Greenbelt Corridor Data” Denton, TX, August 2009.


[7]Fernandez
-
Martinez, Roberto, J,
Ordieres

and A Gonzales
-
Marcos. "Low Power Wireless Sensor
Networks in Industrial Environment."
12th WSEAS International Conference on SYSTEMS

(2008): 643
-
648.


[8] Iris XM 2110/MDA 300 Data Sheets,
http://www. xbow.com


[9]
Hussain
,
Sajid
, and
Md

Shafayat

Rahman
.

Received Signal Strength Indicator to Detect Node
Replacement and Replication Attacks in Wireless Sensor Networks
.
Wolfville
, NS, Canada, Acadia

University.

Acknowledgements:


We started out as two non
-
technical science teachers . However,
through the course of this internship we learned a lot. Not just
about the technical aspects of wireless sensors, but about doing
a research experiment, evaluating data, and presenting our
conclusions, The hands
-
on approach was extremely valuable
not only for us, but for our students. We truly want to thank the
Electrical Engineering Department for the very unique
experience. We plan to carry this back to our individual
campuses and hopefully encourage high school students to
consider a careers in electrical engineering.

Heartfelt thanks for making this
experience possible:


Dr. Miguel Acevedo, Interim Chair, Mechanical and Energy, Engineering


Dr.
Shenli

Fu, Professor, Electrical Engineering


Dr. Oscar Garcia, Professor and Founding Dean of Electrical Engineering


Dr. Rubio Garcia, Associate Dean of Outreach and Public Relations


Dr.
Xinrong

Li, Professor, Electrical Engineering


Dr. Yan Huang, Associate Professor, Computer Science and Engineering


Dr.
Murali

Varanasi, Professor and Department Chair, Electrical Engineering


Mitchell Horton, Graduate Student


Ning

(Martin)
Xu
, Graduate Student


Nitya

Kmdukuri
, Technical Support


Special Thanks to
Jue

(Jerry) Yang


Without your help this learning experience would not have been possible
.










Jerry


You’re the BEST!