Wireless Sensor
Network for Water Quality
Monitoring
Ákos Milánkovich and Krisztina Klincsek
BME-Infokom Innovátor Nonprofit Kft., Magyar tudósok krt. 2. H-1117 Budapest, Hungary
{milankovich.akos, klincsek.krisztina}@bme-infokom.hu,
http://www.bme-infokom.hu
Abstract. This book chapter introduces a European Union-backed water quality
monitoring and early warning system implemented in the cross border region of
Hungary and Slovakia achieved by Hungarian and Slovak partners. The meas-
urement system consists of buoys which carry sensors and send measurement
data wirelessly to a gateway on the riverside. The gateway transmits the data to
a server, where it is saved and processed. Based on the collected data, alerts can
be sent in case of anomalies. Utilizing this solution, an early warning system for
water pollution can be deployed, which decreases water quality monitoring
costs.
1 Introduction
Wireless Sensor Network for Water Quality monitoring (WSN-Aqua) is a joint re-
search and development project financed by the European Union for creating a water
quality monitoring sensor network and early warning system achieved by Hungarian
and Slovak partners. The project started in July 2013 and is due to end in June 2015.
During the two years of the project, the partners will cooperatively develop and im-
plement all the necessary components to create the water monitoring system. The
main output of the project is a test system of 8 buoys placed in the cross border region
of Hungary and Slovakia on the River Ipoly.
This book chapter is organized as follows: Section 2 defines the goals of the pro-
ject; Section 3 introduces the project partners along their responsibilities. Section 4
describes the considered environmental aspects and the selected location of the first
demo system. Section 5 presents the details of the implementation and Section 6 dis-
cusses research results. Finally, Section 7 concludes this book chapter.
2 Aims and Outcomes of the Project
The overall goal of the joint research project is to foster intensive cooperation be-
tween Hungarian and Slovak R&D institutions to increase the economical
28
Milà ˛ankovich à ˛A. and Klincsek K.
Wireless Sensor Network for Water Quality Monitoring.
DOI: 10.5220/0006164600280047
In European Project Space on Information and Communication Systems (EPS Angers 2015), pages 28-47
ISBN: 978-989-758-155-7
Copyright
c
2015 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
competitiveness and capacity of cross-border regions.
The project-specific goal is to develop a sensor network which could be utilized
for water quality measurements and contribute to the environmental protection of
rivers and lakes and be able to function as an early warning system for the authorities
in a cost efficient way. Such a system would be beneficial for both countries.
The project’s subject and goals are directly harmonizing with the priorities and
goals of the HUSK program, especially with “Cooperation in science, research and
innovation” priority (HUSK 1101/1.2.1) and the aim to increase the economical com-
petitiveness of the cross-border region.
The main result of the joint research project is a new product capable of measuring
and monitoring water quality. To implement such a system, a radio communication
protocol will be developed, a buoy will be designed and a measurement test system
will be deployed as a proof of concept.
The project reinforces the social and economical integration of the cross-border
region of Hungary and Slovakia, more specifically it builds partnership in the scien-
tific research community.
The outcomes of the project include a new radio communication protocol, hard-
ware board designs, buoy design, measurement test system on Ipoly and documenta-
tion.
3 Project Partners and Responsibilities
This chapter introduces the partners and their responsibilities (work packages, WPs)
in the WSN-Aqua project.
3.1 BME-Infokom Innovator Nonprofit Ltd.
1
BME-Infokom Innovátor Nonprofit Kft. is a nonprofit organization of technology
Research and Development meeting market demands.
We take on various technolog-
ical challenges and carry innovative ideas into execution by utilizing our long-term
R&D and project experiences, as well as professional staff. Our company offers out-
standing cooperation opportunities and reliable partners, such as the Budapest Univer-
sity of Technology and Economics.
BME-Infokom is responsible for the following work packages:
WP2 Development of an air interface for the environmental monitoring system
Wireless transmission is a critical part of the monitoring system. There will be two
different radio technologies applied in the system. The nodes communicate with the
central server with GPRS/UMTS data transmission. The nodes use another ISM band
1
www.bme-infokom.hu
29
Wireless Sensor Network for Water Quality Monitoring
29
radio solution to provide data transmission. This work package defines the wireless
interfaces and their parameters.
WP5 Development of a wireless communication protocol for the environment pro-
tecting monitoring system
This working package defines the communication protocol that makes possible to
transmit the sensory data via the air interface, defined in WP2. During the design of
the protocol energy efficiency was a major factor. It is important that the protocol
works as efficiently as possible. This phase specifies the data packet structure, where
the different types of sensor data are considered. The various message types during
communication are also to be defined.
WP7 Analysis of network topologies and alternative energy sources
The work package has two goals: first, the definition of network topologies, on which
the system will operate, second, to develop a module utilizing renewable energy
sources, which provides long operating time for the system. The topology largely
determines the protocol being developed, so it is connected to the processes defined
by WP5. The various situations require very different topological solutions, in order
that we have to prepare for more scenarios.
WP8 Establishment of the system monitoring system
The system is placed open terrain, therefore it is exposed to weather effects, ship
traffic and possible damage of vandals. The intermittent failure of a device is una-
voidable. The aim of this WP is to create a centralized remote monitoring system that
collects the status information of the devices. The information should be collected
with energy efficiency taken to account. The protocol has to be implemented in a way
that it uses as few messages as possible, but also provides satisfactory amount of
status information.
3.2 Technical University of Kosice
2
Technická Univerzita V Košiciach (TUKE) was founded in 1952 and is currently
active in nine scientific fields as a research and educational regional center. Currently
there are more than 10000 students at the university. Besides educational activities,
TUKE has successfully completed many R&D projects.
TUKE is responsible for the following work packages:
WP1 Analysis and selection of sensors for water quality monitoring in distributed
sensor network nodes
During this activity, we analyze and select a set of suitable sensors to be used for cost
effective monitoring of water quality in distributed highly embedded network nodes.
There is a wide variety of sensors and transducers at the market, which can measure
temperature, pH, oxygen ratio, water flow volume, etc. Several key parameters of
selected sensors will be optimized during the selection process.
2
www.tuke.sk
30
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
30
WP3 Sensor node hardware components, architecture and embedded software de-
velopment
This activity deals with the selection of suitable hardware components, composition
of the sensor node hardware in terms of used embedded devices, such as particular
microprocessors, particular radio frequency transceiver chips (RF chips), various
sources of energy, and also with efficient energy management.
WP4 Implementation of data confidentiality and integrity algorithms and protocols
During this activity, we analyze and implement a suitable set of cryptographic algo-
rithms and protocols that ensure confidentiality and integrity of transmitted sensors
data via the air interface, defined in the second activity and to be compatible with
available communication protocol data structures defined in the fifth activity. Confi-
dentiality and integrity of protected data will be based on a suitable combination of
symmetric and asymmetric algorithms implemented on an embedded hardware plat-
form defined in the third activity.
WP6 Development of sensor data processing methods for the comprehensive
evaluation of the water quality state
The application of the sensor network to be developed within the project is to provide
corresponding specialists and authorities with the relevant information on the water
quality state within the monitored water flow. The activity is intent on the design of
the final content and form of reporting the water quality state based on processing of
data provided by the particular nodes of the sensor network.
Both partners are responsible for the following work packages:
WP9 Development and evaluation of the environmental monitoring test system
This WP aims at integrating all the components created in the previous 8 WPs and
also tests the whole system in operation in real-word conditions.
WP10 Dissemination of the project results
The project results can be summarized as a contribution to environmental protection
(social contribution), the sensor network applied for the water quality monitoring
(technical solution), the novel scientific and engineering methods applied for sensor
network design and development (research results) and by the excellent example of
Hungarian-Slovak cross-border cooperation. Each of these different kinds of the pro-
ject results will be disseminated by the methods, such as popularization presentations,
seminars, papers in journals and conference proceedings, the developed sensor net-
work demonstration and education process.
4 Environmental Aspects
Each unit of the environmental monitoring system implemented within the framework
of the project will include sensors capable of measuring 4 parameters, enabling real-
time processing and visual display of these data. Thanks to the communication meth-
od of the new sensor system, users (authorities, water experts) will be able to locate
contaminations, physical and chemical changes or other events, and to monitor or
model these in real-time, or to provide estimates and forecasts.
In this project, we examined the temperature, pH, conductivity and the level of dis-
solved oxygen (mg/l) in a given river sector. The sensors will provide us with five
31
Wireless Sensor Network for Water Quality Monitoring
31
further parameters: redox potential, dissolved oxygen (%), pressure, salinity, total
dissolved salts (kcl). Later on, replacing and extending the sensors can enable us to
monitor the locally significant parameters of a chosen area.
The disadvantage of the current monitoring methods (measurement stations or ana-
lyzing water samples in a laboratory) is that experts receive the results with a delay,
rendering immediate intervention in case of a sudden damage impossible. Another
hindrance is that the presence of water experts is constantly required, but – due to the
lack of resources – this is impossible to provide.
Using radio communication technology, the data of the wireless sensor network
used in the project is sent to other units for processing and evaluation. Thanks to this,
if the level of contamination reaches a limit, experts or authorities can be informed
immediately. Although the results are of course not as accurate as in a laboratory,
after a real-time alert, further analysis can be carried out in order to decide whether
immediate intervention is required, which can play an important role in effectively
mitigating damages to the environment.
The standalone mode of the sensor network and technology also enables hard-to-
reach areas to be monitored. This possibility is available thanks to the energy efficient
and renewable energy based operation of the devices.
4.1 Criteria for Choosing the Monitoring Area
Before choosing the monitoring area, we had examined a number of river sectors. As
the project is a Hungarian-Slovak joint task, we tried to choose cross-border rivers
(Hernád, Sajó) or rivers that are natural borders and their drainage basin and tributar-
ies are present in both countries (Danube, Ipoly)
We wanted to choose a river which lacks shipping routes as ship traffic can dam-
age the buoys, the obligatory illumination of buoys results in higher energy consump-
tion – further increasing the size of the required solar panel –, and the authorization
process takes longer.
After installation, the buoy system can be damaged by natural phenomena as well
as human factors. We examined the number of police calls and the population density
in the areas of the aforementioned rivers, because we wanted to choose the safest,
most sparsely populated area.
In terms of natural phenomena, rivers can be most severely affected by floods.
The hydrometeorological data of previous years enables us to predict the temporal
distribution of floods.
It is also very important to choose a location that is easily accessible.
The aim of the monitoring system is to detect damages and contaminations that
negatively affect the condition of the river. This is why we tried to choose an area
which has potential water pollution sources nearby.
4.2 Limit Values of Water Quality
The results of river quality analysis have to be compared with the current limit values
of water quality. The water quality classes are as follows: I Excellent, II Good, III
Acceptable, IV polluted, V severely polluted. In cooperation with the authorities, we
32
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
32
set the alert threshold for the system. Decree 10-2010 lays down the following limit
values (listed in Table 1):
Annex II of Decree 10/2010. (VIII. 18.) of the Ministry of Rural Development
3
1. LIMIT VALUES OF WATERS
1.1. WATER QUALITY LIMIT VALUES OF RIVERS
Table 1. Water quality limit values of rivers.
Type of water body based on the rele-
vant le
g
islation
1 Physical and chemical characteristics Small plain rivers (type 11, 12, 15, 18)
2 pH 6.5-9
3 Conductivity (µS/cm) <1000
4 Chloride (mg/l) <60
5 Oxygen saturation (%) 60-130
6 Dissolved oxygen (mg/l) >6
7 BOI5 (mg/l) <4
8 KOIcr (mg/l) <30
9 NH4-N (mg/l) <0.4
10 NO2-N (mg/l) <0.06
11 NO3-N (mg/l) <2
12 Total N (mg/l) <3
13 PO4-P (mg/m3) <200
14 Total P (mg/m3) <400
We consulted with our Slovak partner about the water quality limit values set by
Slovak authorities. As there is a slight difference between the limits of the two coun-
tries, we agreed on creating two different settings in terms of limit values. The first
buoy cluster is set according to the Slovak limits, while the second cluster is in har-
mony with the Hungarian limits.
Based on this, the monitoring system alerts in the cases summarized in
Table 2:
Table 2. Alert limits for the monitoring system.
SK – 1
s
t
buoy cluster
(average/moderate status)
HU – 2
n
d
buoy
cluster
pH <6 and 9< <6.5 and 9<
conductivity [µS/cm] >700 >1000
dissolved oxygen [mg/l] <6.5 <6
T [°C] >24 (>24)
3
http://net.jogtar.hu/jr/gen/hjegy_doc.cgi?docid=A1000010.VM
33
Wireless Sensor Network for Water Quality Monitoring
33
4.3 Parameters Assessed during the Project
Each buoy of the WSN-AQUA monitoring network will be equipped with 3 Ponsel
sensors, which are capable of measuring pH/redox, conductivity, dissolved oxygen
(mg/l) and temperature.
pH (pondus Hydrogenii): Chemical characteristic of aqueous solutions which
shows the acidity or basicity of the solution. A pH of 7 is neutral. A pH less than 7 is
acidic and a pH greater than 7 is basic. The pH is defined as the decimal logarithm of
the reciprocal of the hydrogen ion activity. Hydronium is produced by protonation of
water, in which giving up a proton results in hydroxide (OH) and hydronium
(H3O+) ions.
pH = -log
10
[H
3
O
+
] = -lg [H
3
O
+
]
Redox potential: ORP – Oxidation Reduction Potential, a value measured in mV. It
is the tendency of a chemical species to acquire electrons and thereby be reduced. The
more positive the potential, the greater the species' affinity for electrons and tendency
to be reduced.
Dissolved oxygen [mg/l] (OPTOD – Optical Dissolved Oxygen technology): It is
an important parameter in assessing water quality because of its influence on the or-
ganisms living within a body of water.
Conductivity and temperature: Conductivity is a measure of the ability of water to
pass an electrical current. For this, we use a Ponsel sensor to assess electrical current
(µS/cm). As conductivity increases by 2% in case of a 1 °C temperature rise, the
measurement has to be standardized to 25 °C.
Although all three sensors measure temperature, this factor is very important when
measuring conductivity, therefore the conductivity sensor’s temperature data will be
considered as reference value.
4.4 The Prioritized Area
After preliminary consideration, from the Ipoly, Sajó and Hernád river sections we
decided to further examine the Ipoly area. We considered the security of the area, but
the distance between possible locations was also a different factor as it greatly influ-
ences setup and maintenance costs, and shorter distances mean that the area can be
reached faster in case of an alert. The description of the river basin management plan
(RBMP) of Ipoly and a debate report for its review from 2014 provide us with useful
information about the area.
The Ipoly meets the criteria of our project because it acts as a natural border for
almost 100 km and the quality of the water flowing into it from its river basin and
tributaries can indicate pollution along the river basin.
Due to river control, the Ipoly is now only 212.5 km long, its river basin is 5151
km2, out of which 1502 km2 is located in Hungary. The Slovak Ore Mountain, the
Selmec Mountains, the Karancs, the Cserhát and the Börzsöny are all located in its
river basin.
The valley of the Ipoly is characterized by basins separated by natural formations
that hinder erosion. In several areas, it has a gentle gradient with tributaries of steep
gradient. The river bed consists of impermeable rocks, but it is not an abundant river
34
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
34
due to little rainfall and small river basin. Its fluctuation is however considerable, any
significant amount of rainfall triggers flood waves that last no longer than few hours
or 1-2 days maximum. The affected areas are mainly protected by embankments,
dams and pumping stations.
4.5 Protected Areas
4
According to the EU Water Framework Directive, every area receives protected status
where any legislation identifies surface waters and groundwater or water dependent
species to be protected.
In the Ipoly subunit of the river basin management plan, we can come across the
following categories of protected areas:
Protection of drinking water, nutrient or nitrate sensitive areas, natural bathing
sites, areas protected due to their natural values.
4.6 Flood Risk
The fluctuation of rivers is another important factor when installing a buoy system.
The fluctuation of the Ipoly is fairly wild and there are two flooding periods during
the year, one at spring snowmelt and another in summer due to heavy rainfall. The
water level is low in August and September. Floods are also possible in winter due to
fast snowmelt. According to observation data collected at Nógrádszakál, the lowest
water level was 33 cm, while the highest was 362 cm.
The Middle-Danube Valley Water Directorate prepares monthly hydrometeorolog-
ical reports of its territories. The reports provide an analysis of the meteorological
situation of the given month and the temperature changes as well as the connection
between these two factors and their effect on river fluctuation. These reports contain
water level and discharge parameters of the Ipoly, the Zagyva and the Danube. The
reports also provide information regarding groundwater level and the water manage-
ment situation. The latter includes data about the condition and water level of reser-
voirs operated by the Directorate. There is also information regarding the level of
flood preparedness in certain areas as well as water quality data (dissolved oxygen,
ammonia, nitrite, nitrate, minerals and water quality classification).
5
We examined the river fluctuation data of hydrometeorological reports available on
the website (dating back to July 2011). Our aim was to provide an overview of river
fluctuation characteristics of the Ipoly, which is a required measure before installing
the buoy system.
Installing the monitoring system is due to late April and early May. As earlier ob-
servations suggest, we have to be prepared for sudden flood waves, so we intend to
install the buoys considering meteorological data and hydrometeorological forecasts
of the Directorate. Generally, flood waves are not frequent in the onset of summer.
4
http://vizeink.hu/files3/1_8_Ipoly.pdf – available in Hungarian
5
http://www.kdvvizig.hu/ Middle-Danube Valley Water Directorate
35
Wireless Sensor Network for Water Quality Monitoring
35
Our findings suggested that the Ipoly is the most suitable of the possible areas for
our purposes. In order to set the exact location of the monitoring system, further re-
search is required, and circumstances of carrying out the installation and maintenance
have to be considered as well.
4.7 Choosing the Monitoring Site, Field Inspection
In order to choose the most suitable site for the monitoring system, we carried out
preliminary water quality surveys during our field inspections of the areas that were
deemed suitable based on our background research. In the future, the results of these
surveys can be used as reference values to compare with results measured by the
monitoring system.
Firstly, we inspected three fields on the lower Ipoly section on the river sections
along Letkés, Ipolydamásd and Szob. A representative of Aqua-Terra Lab Ltd., the
company responsible for designing and manufacturing the buoys, and a colleague of
the Institute of Environmental Engineering of Pannon University also participated in
the field inspection.
During the field inspection, we assessed the accessibility of the sites as well as the
suitability of the river section regarding buoy installation and maintenance. We also
tested the Ponsel sensors to be used during the project, and we collected water sam-
ples for laboratory measures at Pannon University. We wanted to choose a site where
contamination or external impacts that influence water quality are highly possible.
This is why we tried to choose confluences of tributaries, areas close to settlements or
river sections near arable lands.
After inspecting the lower Ipoly section, we considered the characteristics of all
three sites and arrived at the conclusion that the area between Letkés and Salka is the
most suitable site to install a monitoring system. The site is easily accessible, there is
hourly gauge data available both at Ipolytölgyes and Salka, which can be followed at
http://www.vizugy.hu/
6
and the Slovak website, http://www.shmu.sk/
7
.
The next step was to consult the experts of the Middle-Danube Valley Directorate.
They expressed their support of the monitoring system. Since the project field is lo-
cated on a river that acts as natural border between two countries, we had to contact
the competent Slovak water management authority. The experts of the Directorate
also helped us in this case, as they established close cooperation and strong relation-
ship with the Slovak colleagues who are competent in this area.
After prior consultation, we inspected the Letkés section of the river and we dis-
cussed our plans regarding the buoy system. The meeting was attended by two experts
of the Middle-Danube Valley Directorate, two representatives of the competent Slo-
vak water management directorate, the leader of the river warden service of the sec-
tion, two river wardens, an expert of Aqua-Terra Lab and of the Institute of Environ-
mental Engineering of Pannon University as well as the engineers of BME-Infokom.
6
//www.vizugy.hu/
7
http://www.shmu.sk/
36
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
36
During the meeting, we presented the aims of the project to the Slovak experts, and
they also expressed their support. The Slovak and Hungarian directorates issued writ-
ten permit to install the monitoring system.
During the field inspection, we also tested whether sufficient radio communication
is possible in the area. The data collected from the two gateways and the measuring
results show that decent data transmission is possible in the area. This means that the
transfer – to gateways – of data collected by the buoy sensors will be perfectly feasi-
ble.
The monitoring points are to be installed in two separate chains because, according
to the information of the Directorate, the island under the bridge at Letkés is going to
be demolished next spring. Since we would like to avoid the river section affected by
the earthworks, we intend to install two set of buoys (in an upper and a lower section).
4.8 Devices used for On-the-Spot Measurement
The following sensors were used in the surveyed areas to assess temperature, pH,
redox potential, conductivity, dissolved oxygen, oxygen saturation and turbidity:
NOETEK-PONSEL pH/redox measuring instrument: pH, redox potential
NOETEK-PONSEL OPTOD optical dissolved oxygen measuring instrument: dis-
solved oxygen, oxygen saturation
NOETEK-PONSEL C4E: temperature, conductivity
NOETEK-PONSEL digital nephelometric turbidity measuring instrument: turbidi-
ty (NTU)
4.9 Standards and Regulations Followed
MSZ EN ISO 5667-1:2007 Water quality. Sampling. Part 1: Guidance on the design
of sampling programmes and sampling techniques
MSZ EN ISO 5667-3:2004 Water quality. Sampling. Part 3: Guidance on the
preservation and handling of water samples
MSZ EN ISO 5667-6:1995 Water quality. Sampling. Part 6: Guidance on sampling
of rivers and streams (superseded standard)
MSZ 448-2:1967 Drinking water analysis. Determination of temperature, colour,
turbidity and limpidity (superseded standard)
MSZ EN ISO 10523:2012 Water quality. Determination of pH (ISO 10523:2008)
MSZ 448-32:1977 Drinking water analysis. Determination of specific electrical
conductivity (superseded standard)
MSZ EN 25814:1998 Water quality. Determination of dissolved oxygen. Electro-
chemical probe method
MSZ EN ISO 7027:2000 Water quality. Determination of turbidity
4.10 Code Table of Colour, Odour and Weather of Surface Waters
Simplified summary (see Table 3) of categories and codes used when sampling sur-
face waters (exception: Lake Balaton and its surrounding rivers).
37
Wireless Sensor Network for Water Quality Monitoring
37
Table 3. Code tables for categories.
Weather
First digit Precipitation Second digit Level of cloudiness
no precipitation 0 sunny 1
humid 1 cirrostratus cloud 2
foggy 2 cloudy 3
rain 3 very cloudy 4
shower, thunderstorm 4 Third digit Wind
ice pellets 5 no wind 0
snow 6 wind 1
snow shower 7 strong wind 2
hail 8 storm-force wind 3
Colour
First digit Colour strength Third digit secondary colour
colourless 0 colourless 0
colour intensity 1-7 yellowish 1
greenish 2
Second digit Dominant colour blueish 3
colourless 0 greyish 4
yellow 1 brownish 5
green 2 blackish 6
blue 3 reddish 7
grey 4 murky 8
brown 5
black 6
red 7
Odour Ice
odourless 00 no ice 1
chloric odour 01 ice along the coastline 2
dirty odour 02 drift-ice 3
tarry odour 03 ice cover 4
oily odour 04
fishy odour 05
decaying, swampy odour 06
odour resembling medicines 07
4.11 Conclusion, Choosing the Area
During the preliminary surveys and field inspections, we examined a number of pos-
sible sites that could be suitable for a monitoring system. Based on the criteria de-
scribed above, we can conclude that the 2 km long river section between Letkés and
Salka is the most suitable area for our project, and the local competent water man-
agement authorities – both Hungarian and Slovak – support our initiative.
38
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
38
The buoy system to be installed will be able to monitor and detect any possible
contamination that gets into the River Ipoly from the Letkés stream as the upper buoy
group provides reference values for the lower buoy group – which detects the contam-
ination. Since buoys of the lower buoy group are 200 m away from each other (see
Fig. 1), the system can monitor the rate at which the contamination dilutes.
Fig. 1. Location of the buoys and gateway units.
5 Implementation
This chapter presents the implementation of the water quality measurement system.
First the network topology and the components necessary to achieve the project goals
are introduced. Next, the internals of the network elements (buoys) are demonstrated.
39
Wireless Sensor Network for Water Quality Monitoring
39
5.1 Network Topology
The sensor network created for the WSN-Aqua project is based on data-gathering
sensor-equipped buoys placed on the surface of the water. The buoys send their col-
lected data to a Gateway module located on the shore. In case a buoy is not in the
radio range of the Gateway, they can send data in multi-hop fashion as they automati-
cally create a mesh-network.
The radio communication between the buoys and the gateway is conducted on 433
MHz ISM band, therefore the nodes have about 500 m line of sight in case of 10 mW
output power. This frequency is free to use in the EU for such radio systems, but the
communication protocol has to comply with output power and duty cycle restrictions.
The gateway module sends the data collected from the buoys via a cellular con-
nection (EDGE or 3G depending on coverage) to a server module in an aggregated
way. However, in case of an emergency (i.e. thresholds are exceeded), the gateway
sends the data to the server immediately upon arrival, so the server can send alert
messages (SMS and email) to the corresponding authorities.
The management server communicates with the Business Intelligence (BI) server
(via internet), which is responsible for gathering statistics and create multiple views
and reports of the measured data.
The BI server can send configuration information and new firmware all the way
down to the buoys via the private management interface.
The flow of the data in the system is presented in Fig. 2.
Fig. 2. Data flow diagram of the WSN-AQUA project.
5.2 Buoy and Gateway
The buoy (Fig. 3, Fig. 5, Fig. 6) consists of the following components:
Buoy body (80 x 90 x 50 cm) with lockable opening made from plastic,
3 sensors according to Table 4 to gather data,
Atmel ATxmega128A3 [1] microcontroller to conduct measurements and con-
trol the communication,
TI CC1101 [2] radio module for data transmission,
433 MHz antenna for the radio module placed on top of the buoy body for better
40
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
40
range,
5 W solar panel to be able to operate without battery replacement,
Battery capable of operating for a month without solar power,
Charge controller board to switch between solar panel and battery power, de-
veloped for this project.
Table 4. Sensors used in the WSN-Aqua project.
Model
8
Ponsel PHEHT Ponsel OPTOD Ponsel C4E
Measured Property
pH/redox dissolved oxygen conductivity
Protocol
Modbus RS485 via adapter Modbus RS485
Scope
wastewater, surface water, etc.
Measuring range
0,00...14,00 pH, 0... 50 °C,
-1000...+1000 mV
0,00 – 20,00 mg/l, 0 –
200% saturation
0...200 mS/cm, 0...2000
mS/cm, 0...20 mS/cm,
0...200 mS/cm
Measurement
accuracy
± 0,1 pH, ± 0,1
o
C, ± 2 mV ± 0,1% ± 0,5%
Compensation
Temperature
Pressure, temperature,
salinity
Temperature
Calibration
5 point, it could be manually calibrated
Fig. 3. Sensor node hardware, embedded
ternals.
Fig. 4. Gateway node hardware in- into the
buoys.
The Gateway module (Fig. 3) consists of the following components:
Embedded PC for controlling the module,
Cellular communication module integrated for sending data to the server,
TI CC1101 [2] radio module for data transmission with the buoys,
433 MHz antenna for the radio module placed on top of the buoy body for
better range,
100 W solar panel to be able to operate without battery replacement,
8
www.ponsel-web.com
41
Wireless Sensor Network for Water Quality Monitoring
41
Battery capable of operating for a week without solar power,
Charge controller board to switch between solar panel and battery power,
developed for this project.
Fig. 5. The buoy was designed and manufactured in association with Aqua-Terra Lab Ltd.,
Hungary
9
.
Fig. 6. Deployed buoy during testing.
9
www.aquaterra.hu
42
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
42
5.3 Business Intelligence
The data gathered from the sensors are sent to a management server via the gateway
module. This management server can send alerts in case of an emergency situation
(e.g. the thresholds of the measured parameters are exceeded). The data is stored in a
database and made available for a data analyzer component. This Business Intelli-
gence server can present the data via a web interface to end users and expert (mainte-
nance) users as well, and is able to collect statistics and generate various reports.
As part of the Business Intelligence, a management system is developed, which is
capable of monitoring the status of every node in the network (e.g. battery status, alert
configuration, diagnostics, etc.) and initialize firmware upgrade for future require-
ments. Administrators of the system can monitor and support the system via the man-
agement system running on the management server.
6 Research Results
This chapter is an excerpt from [3] and was presented at the Sensornets 2015 confer-
ence in Angers, France.
Creating wireless sensor networks requires a different approach than traditional
communication networks because energy efficiency plays a key role in sensor net-
works which consist of devices without external power. The amount of energy used
determines the lifetime of these devices. In most cases, data packets are less sensitive
to delay, thus can be aggregated, making it possible to gather more useful information
reducing the energy required to transmit information.
This chapter discusses the energy efficiency of different Forward Error Correction
(FEC) [4] algorithms and presents a method to calculate the optimal amount of aggre-
gation of the data packets in terms of power consumption, while taking into account
the Bit Error Rate (BER) characteristics of the wireless channel. In addition, we pro-
vide a formula to determine the optimal wakeup signal length, which minimizes the
energy consumption.
Fig. 7. shows the gain (the ratio of not using aggregation and using it) that can
achieved by using aggregation without FEC. The graph line representing BER = 4
10

is jagged because the number of required retransmissions is growing as the
Fig. 7. for different BER values without FEC.
43
Wireless Sensor Network for Water Quality Monitoring
43
aggregation number n is increasing. The number of retransmissions is the same in the
neighboring points that follow each other without a jump in their values.
In the figures of this section, the =1 values are marked with a red line to indi-
cate the level above which the use of aggregation is more efficient.
Remark. This phenomena can be observed in case of other BER values, e.g. for
 = 4 10

the first jump is at 200, which is above the data aggregation
value we considered worthy to examine.
Fig. 8. has the same setup as Fig. 7., with the only difference that Hamming [5],
[6] codes were applied. The graphs show that in case of medium quality channel
( = 4 10

) and good quality ( = 4 10

) channel, there is no difference;
the calculated points are perfectly aligned.
Fig. 9 and Fig. 10 show similarity of the values of for different BER levels with
respect to aggregation number n for Reed-Solomon [7], [8] and BCH [9], [10] FEC
codes.
Fig. 8. for different BER values with Hamming code.
Analyzing Fig. 9 and Fig. 10 it can be noticed, which in case of a poor quality
channel ( = 4 10

) for every aggregation number n we got better gain val-
ues than in case of better channel. This is because more powerful FEC codes provide
more benefits compared to the same aggregation numbers in case of poor quality
channels. The better BER channels result in the same gain .
Also, the graphs look like stages because the block length of Reed-Solomon codes
Fig. 9. for different BER values with Reed-Solomon code.
44
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
44
is fixed. Therefore, if the payload is not long enough, padding is used to fill the rest of
the block, which is inefficient.
In Fig. 10 in case of aggregation number <40, the poor quality channel gains
more using aggregation and BCH code than the better quality channels. Also, at better
quality channel, there is significant gain compared to baseline (no aggregation, no
FEC) just like in the case of Reed-Solomon codes.
Fig. 10. for different BER values with BCH code.
The next three figures (Fig. 11, Fig. 12 and Fig. 13) compare the cases of different
FEC codes grouped by channel quality ( = 4 10

,410

4 10

) in
respect to n. For every diagram, a table is included which shows the optimal aggrega-
tion number (the highest point of the graphs and the corresponding number of re-
quired retransmissions.
Remark. The optimal aggregation number can be much higher is case of BCH and
RS codes, but the authors considered n<100 aggregation numbers are worth dealing
with because higher aggregation numbers would cause much higher delays. For ex-
ample, if the aggregation number is n=100 and the packets are generated on an hour-
ly base, then the aggregation delay can be as high as 100 hours. For most real-world
scenarios, the delay should be within a day.
No FEC Hamming BCH RS
Opt. aggr. no. 2 9 37 18
No. of reps. 6 8 4 3
Fig. 11. Comparison of FEC codes at BER=4E-3.
45
Wireless Sensor Network for Water Quality Monitoring
45
Fig. 11 compares FEC codes on the worst quality channel. This scenario shows
that the energy cost of different FEC codes is the best. The graph emphasizes that not
using any FEC is the worst, and BCH and Reed-Solomon codes perform as the best. It
can be seen that in case of lower aggregation numbers (<10), Reed-Solomon is the
best solution, and from 20 < < 40 RS and BCH are at the same level. When fur-
ther increasing the aggregation number, RS code is the most efficient again.
Fig. 12 compares FEC codes on a channel with  = 4 10

. It can be seen that
in case of n>20 aggregation numbers, FEC codes provide more energy efficient op-
eration. The FEC codes perform similarly.
No FEC Hamming BCH RS
Opt. aggr. no. 18 98 94 100
No. of reps. 2 2 2 2
Fig. 12. Comparison of FEC codes at BER=4E-4.
According to Fig. 13, in good quality channels there is no benefit of using FEC
codes because for every aggregation number, the case without FEC performs the best.
The FEC codes just converge to the graph of no FEC case.
No FEC Hamming BCH RS
Opt. aggr. no. 100 98 94 100
No. of reps. 2 2 2 2
Fig. 13. Comparison of FEC codes at BER=4E-5.
46
EPS Angers 2015 2015 - European Project Space on Information and Communication Systems
46
7 Conclusions, Future Plans
This book chapter presented the involved partners, implementation details and re-
search results of the WSN-Aqua project. The verification of the proof of concept test
system, which is the main output of the project, will be carried out during May to
June 2015, therefore the final results cannot be published in this article.
The Middle-Danube Valley Water Directorate is interested in the utilization of the
project results with more measured parameters.
Future plans – after the project has been successfully finished – also include further
cooperation between the parties involved in new HUSK-CBC tenders.
Acknowledgements. This project has been supported by BME-Infokom Innovator
Nonprofit Ltd., http://www.bme-infokom.hu.
This project has been sponsored by The European Union’s Hungary-Slovakia Cross-
border Co-operation Programme. Building Partnership.
www.husk-cbc.eu, www.hungary-slovakia-cbc.eu
The content of this paper does not necessarily represent the official position of the
European Union.
References
1. “AVR XMEGA A3 Device Datasheet,” Dec. 2010.
2. Texas Instruments, Incorporated, “CC1101 Low-Power Sub-1 GHz RF Transceiver (Rev.
I),” Nov. 2014.
3. Á. Milánkovich, G. Ill, K. Lendvai, S. Imre, and S. Szabó, “Research on Payload Aggrega-
tion of Packets in WSNs,” Sensornets '15, Angers, Feb. 2015.
4. P. Faraj, J. Leibrich, and W. Rosenkranz, Coding gain of basic FEC block-codes in the
presence of ASE noise, vol. 2. IEEE, 2012, pp. 80–83 vol.2.
5. N. A. Ali, H. M. Elsayed, M. El-Soudani, and H. H. Amer, Effect of hamming coding on
WSN lifetime and throughput. IEEE, 2011, pp. 749–754.
6. Z. Jianwu, S. Yibing, and L. Yanjun, Software Implementation of a Novel Approach to
Improving Burst Errors Correction Capability of Hamming Code. IEEE, 2007, pp. 2–499–
2–503.
7. S. B. Wicker, V. K. Bhargava, I. C. Society, and I. I. T. Society, Reed-Solomon codes and
their applications. Inst of Electrical &, 1994.
8. L. Biard and D. Noguet, “Reed-Solomon Codes for Low Power Communications,” JCM,
vol. 3, no. 2, pp. 13–21–9, Jan. 2008.
9. S. S. Sarnin, N. F. Naim, and W. N. S. W. Muhamad, Performance evaluation of phase shift
keying modulation technique using BCH code, Cyclic code and Hamming code through
AWGN channel model in communication system. IEEE, 2010, pp. 60–65.
10. R. C. Bose and D. K. Ray-Chaudhuri, “Further results on error correcting binary group
codes,” Inf Contr, vol. 3, no. 3, pp. 279–290, Dec. 1959.
47
Wireless Sensor Network for Water Quality Monitoring
47