Content uploaded by Hossein Anisi
Author content
All content in this area was uploaded by Hossein Anisi on Sep 27, 2015
Content may be subject to copyright.
A survey of wireless sensor network approaches
and their energy consumption for monitoring
farm fields in precision agriculture
Mohammad Hossein Anisi •Gaddafi Abdul-Salaam •Abdul Hanan Abdullah
Published online: 11 September 2014
ÓSpringer Science+Business Media New York 2014
Abstract Precision agriculture (PA) is the use of information and communication
technology together with best agricultural practices for farm management. PA requires the
acquisition, transmission and processing of large amounts of data from farm fields. A
wireless sensor network (WSN) is a system for monitoring agriculture fields. Several
researchers have used WSNs to collect the required data from the regions of interest for
their intended usages in various applications. In a WSN, the energy consumption of the
sensor nodes is the main issue, due to its direct impact on the lifetime of the network. Many
approaches have been proposed to address this issue using different power sources and
types of nodes. Specifically, in PA, because of the extended time period that is required to
monitor fields, using an appropriate WSN approach is important. There is a need for a
comprehensive review of WSN approaches for PA. The aim of this paper is to classify and
describe the state-of-the-art of WSNs and analyze their energy consumption based on their
power sources. WSN approaches in PA are categorized and discussed according to their
features.
Keywords Wireless sensor networks Energy consumption Topologies Power source
Abbreviations
AG Above ground
BS Base station
CP Center pivot
ET Evaporation–transpiration
FW Full-wave
GW Gateway
M. H. Anisi (&)
Department of Computer System and Technology, Faculty of Computer Science and Information
Technology, University of Malaya, 50603 Kuala Lumpur, Malaysia
e-mail: anisii@gmail.com; anisi@um.edu.my
G. Abdul-Salaam A. H. Abdullah
Faculty of Computing, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia
123
Precision Agric (2015) 16:216–238
DOI 10.1007/s11119-014-9371-8
GPRS General packet radio service
GPS Global positioning system
IMS Irrigation management system
iPAGAT Intelligent precision agriculture gateway
IS Irrigation station
ISSPA Integrated wireless sensor networks solution for PA
KIP-AF Knowledge information processor for agriculture sensor data and fire-sensor
data
LNA Low noise amplifier
LOFAR Low frequency array
LOS Line of sight
LQI Line quality indicator
MAC Medium access control
PA Precision agriculture
PC Personal computer
PHP Hypertext preprocessor
RF Radio frequency
RFID Radio-frequency identification
RMS Remote monitoring station
RSSI Received signal strength indicator
RTAS Real-time alert system
RTK-DGPS Real-time kinematic differential global positioning system
SEA Single ended elliptical antenna
SIM Subscriber identity model
T-MAC Time-out medium access control
TDMA Time division multiple access
TinyOS Tiny operation system
UG Underground
VRI Variable rate irrigation
WC Wireless co-ordinator
WED Wireless end device
WSN Wireless sensor networks
WR Wireless router
WUSN Wireless underground sensor networks
XML Extensible mark-up language
ZC ZigBee co-ordinator
ZED ZigBee end device
ZSIM ZigBee enabled SIM
Introduction
Precision agriculture (PA) is a concept that combines information technology with
agricultural principles to manage spatial and temporal variability in the agricultural
production process. This arrangement may boost agricultural yields whilst reducing
harmful impact to the environment. One of the efficient technologies that is used to
monitor and collect data in PA is a wireless sensor network (WSN). WSNs collect data of
Precision Agric (2015) 16:216–238 217
123
essential spatial and temporal variables that are necessary for decision making in agri-
cultural farm management (Coates et al. 2013;Dı
´az et al. 2011; Dong et al. 2013;Yu
et al. 2013; Zhang et al. 2011).
WSN is a recent novel technology that has prospects for improving many applications
as well as creating novel systems in PA and several other applications, such as global-scale
environmental monitoring, pest and disease control and animal tracing (Stankovic 2008;
Vellidis et al. 2008; Vicaire et al. 2009; Wood et al. 2006). In PA, WSNs have been
deployed to acquire micro-climatological data in farm fields, including temperature,
humidity, soil moisture content and wind speed. For example, in the Luster project (Selavo
et al. 2007), a WSN was used in an environmental science system for measuring the effect
of sunlight on plant growth. Furthermore, WSNs have been used to measure crop growth,
irrigation and several other applications (Baggio 2005; Garcia-Sanchez et al. 2011; Shaikh
et al. 2010; Vellidis et al. 2008).
In a WSN, the lifetime of the network is dependent on the power source requirement
of the nodes. Thus, selecting a suitable power source for a PA application is essential.
The main types of power sources are battery or solar power; however, each has its own
issues. To the best of our knowledge, there are not yet surveys that categorize WSNs for
PA based on the power source. Ruiz-Garcia et al. (2009) discussed the general appli-
cation area of WSNs in agriculture. They explained different systems that are available
based on radio frequency identification (RFID) and WSNs. Similarly, Wang et al. (2006)
explained the available WSN technologies that are applicable to agriculture and the food
industry.
In this paper, first, WSNs and the available network topologies are described. Second,
the sensor node power sources are categorized and discussed in terms of homogeneous and
heterogeneous nodes, and their advantages and disadvantages are noted. Current approa-
ches and their energy consumption are reviewed and described based on these classifica-
tions. Third, the approaches are summarized and discussed according to their features,
applications and technical contributions.
Wireless sensor networks
A WSN is a distributed network of relatively small, lightweight wireless nodes that are
equipped with sensors (Can and Demirbas 2013; Owojaiye and Sun 2012; Rathnayaka
and Potdar 2013; Van Hoesel et al. 2013). Sensor nodes are the basic building blocks of
a WSN. A typical node is composed of a transceiver, sensor(s), external memory and
power source. Figure 1shows the basic architecture of the nodes. The node has the
ability to sense, process and communicate spatial and temporal variables of interest.
Such nodes are deployed in areas of interest to work together for environmental or
system monitoring (Peng and Liu 2012; Stankovic 2008; Vicaire et al. 2009; Wood
et al. 2006).
WSN topologies
Notable topologies mentioned in the literature include the single-star topology (Srbinovska
et al. 2014), mesh topology (Watras et al. 2014), and cluster tree-based topology (Senturk
and Akkaya 2014).
218 Precision Agric (2015) 16:216–238
123
Star single-hop topology
In star single-hop topologies, each node (or ‘client’) must communicate to other nodes
(‘clients’) through a central hub (or a ‘server’) or cluster head. The clients do not relay
packets through other nodes. Instead, each node directly transmits its packet to the cluster
head.
Tongtong et al. (2011) proposed a method of sensing and monitoring the temperature
and humidity in a greenhouse environment for various distances. It consists of one more
sensor nodes, some repeat nodes, one main node, PC Terminal, MYSQL (an open source
relational database management system), web service and one real time alarm system.
Based on this system, the temperature and humidity parameters can be sensed and mon-
itored. They then used the database and web service to record the parameters at specific
times. Finally, a curve showing the measurement differences from the sensing module is
drawn by the web server. If the curve depicts abnormal changes in the temperature and
humidity measurements, an alarm system is triggered, which informs the monitor to take
required action. In this system, although the wireless transceiver distance was limited by
the battery power to just 1.26 m, the distance could be increased in steps of 1.26 m for
each node. Thus, the total distance could be calculated as 1.26 m 9[1 ?N(repeat
nodes)].
Shaikh et al. (2010) designed WSN nodes with a protocol that ensured low power
consumption in their development of a cost-efficient crop irrigation control system. The
nodes are categorized into sensor, actuator and sink nodes, to create a functional archi-
tecture. The system collects sensor data packets wirelessly via a ZigBee wireless tech-
nology (Baronti et al. 2007) configuration and transfers them to a connected PC through a
serial interface. It then sends the request and actuator packet that was generated by the
acquisition and control/decision support system through the WSN network. The network
protocol described makes the protocol in this experiment a star single-hop topology. With
the use of these components and topology, a broadcast based protocol called ‘‘ProtoSense’’
was then developed. In this protocol, data request is made one level at a time. If the target
node is not at one level, it would be forwarded to the next level. Hence, this protocol can
efficiently optimize power by reducing extra communication. It also introduces a request
acknowledgement pattern to ensure reliability.
Fig. 1 Basic architecture of a sensor node
Precision Agric (2015) 16:216–238 219
123
Cluster tree-based multiple-hop topology
In this type of topology, two or more star networks communicate through a central hub
called a ‘root node’. Each sensor node transmits data within a limited radius to reach a
nearby cluster head. The lower cluster heads then refine the data packets and act as relays
to further transmit the refined data to a super cluster head or a root node. There can be
communication among the lower cluster heads via peer-to-peer topologies. This type of
network structure is implemented in a data aggregation method to reduce the energy
consumption in the WSN deployment. For example, Nesa Sudha et al. (2011) organized
sensor nodes into clusters, with one node acting as a cluster head and a base station (BS)
acting as a super cluster head. This experiment combined six sensors per cluster, and there
were three clusters in all. The data transmission between the BS and cluster heads and also
communication between the clusters was by time division multiple access (TDMA). The
data transmission was based on a single-hop from one node to a nearby node. Overall, the
network structure of this deployment is a cluster tree-based multiple hop topology. Each
node is assigned a unique identification address and is known to the BS. At any point in
time, TDMA scheduled any one of the nodes as the cluster head and communicated this
information to all of the nodes in the cluster. All of the nodes transmit their power source
information to the BS, and the node with the highest energy level in the cluster is assigned
to be head of that cluster. Peng and Liu (2012) described the implementation of an auto
system scheme in which the sensor network nodes are combined with fuzzy controller
nodes to schedule irrigation. In the design structure of the system, the WSN is composed of
a sensor node cluster, a co-ordinator node and a controller node. Irrigation pipelines are
laid across the irrigated areas, and electric control valves are installed on the pipelines. The
sensor nodes collect humidity and temperature data and form sensor clusters based on the
plant-water status. In the network structure, each sensor node gathers the real-time data of
its surrounding areas and sends it to the cluster heads. The information is further sent to the
co-ordinator nodes within a set time period. In the co-ordinator node is an embedded fuzzy
controller. This controller controls the pump valves and decides when to activate the
irrigation and for how long. The WSN deployment from the data collection to the decision-
making point makes the network structure a cluster tree-based multiple hop topology.
Mesh topology
In mesh topology, each sensor node communicates to every other sensor node and acts as a
relay to transmit data from the other nodes in the network. Gao et al. (2013) proposed an
intelligent irrigation system based on a WSN and fuzzy control. The objective was to solve
issues of soil fertility loss and water wastage in irrigation. The system was composed of
sensor nodes and the controller node; additionally, soil moisture sensors, irrigation pipes,
spray irrigation and irrigation control valves were deployed. The ZigBee network was
adopted in the mesh network topology. While the sensor nodes acted as terminal nodes in
the field, some sensors nodes were selected as routers. These router nodes collected data
from the terminal nodes and transmitted them to a head node, called sink node, which
gathers and controls data collected by other nodes. A control panel was introduced as the
control node. Data were acquired at periodic intervals; the control valve opened and closed
the irrigation process depending on the command received. Vellidis et al. (2013) reported
on use of a mesh network-based soil moisture sensing WSN in combination with variable
rate irrigation (VRI) at the 2013 European conference on precision agriculture. Majone
et al. (2013) monitored soil moisture dynamics in the top soil to analyze the interplay
220 Precision Agric (2015) 16:216–238
123
between soil moisture dynamics and plant physiology. Their experiment used a multi-hop
WSN that was connected across several soil and temperature nodes that were deployed
across 27 locations. The sensor nodes were connected to an input/output interface through
a node called TinyNode 584 (Dubois-Ferrie
`re et al. 2006) WSN platform, which was
developed by SHOCKFISH company, Lausanne, Switzerland.
Power sources
WSNs have been proposed to detect and monitor spatial and temporal parameters (Garcia-
Sanchez et al. 2011; Peres et al. 2011; Zhang et al. 2012). The results are then used to
control the levels of parameters to suit specific plant soil requirements and to provide event
services. An example is the use of a WSN for intelligent diagnosis services and to schedule
irrigation management.
Power is required for WSN devices throughout their period of operation (Anastasi et al.
2009; Anisi et al. 2013b). There are two main types of power source: battery and solar.
Battery sources
Although batteries are relatively low cost and small size, continuous exposure to high
temperature levels in fields have other undesirable effects, such as a shortened cycle life
and an increased rate of self-discharge (Park et al. 2005). Short time battery life is a
limitation on long-term monitoring. Rechargeable batteries can be used as a source of
power to make up for the energy depletion in the sensor nodes.
Homogeneous battery-based power approaches
In homogenous WSNs, all of the sensor nodes possess the same network attributes in terms
of their computational capability, battery power and transmission radius (Hou et al. 2005).
All of the sensor nodes are identical, and all of the nodes might have the same lifetime. The
advantage of using homogeneous nodes is that they are less complicated and uniform in
their maintenance requirements. All of the sensor nodes could also be low-cost. In terms of
cluster formation, any one of them can be selected as a cluster head. However, due to the
continuous processing of data in terms of aggregation and protocol co-ordination and the
long-range communication that are maintained with the remote BS, the co-ordinator nodes
could become overloaded. As they have the same power resource as the other sensor nodes,
their energy depletes faster resulting in premature failure (Mhatre and Rosenberg 2004).
Vellidis et al. (2008) proposed a system in which multiple homogenous sensor nodes
were used for monitoring soil moisture and temperature to schedule irrigation in a cotton
farm. There was single-hop data transmission of data via RFID to a fixed central receiving
unit. A laptop connected to the receiver was placed at one end of the field. The sensor
nodes were composed of sensors, a sensor circuit board, and an RFID tag. At periodic
intervals, the smart sensor board acquired sensor values and transmitted those values
wirelessly to the receiver. The presence of plant biomass, terrain and objects imposed
transmission difficulties during field testing of the prototype. This difficulty was solved
when tags were removed from the electronic boards and mounted on hollow flexible
fiberglass rods at approximately 1.2 m above ground level. A large number of sensor nodes
were deployed to cover a large area. However, such a large number of nodes can cause
Precision Agric (2015) 16:216–238 221
123
interference problems which may result in more packet loss. A sensor circuit board which
was called ‘‘smart sensor board’’ was supported by battery power. The micro-controller
switched to sleep mode between sensor readings and data transmission. Sub-circuits also
went ON and OFF as required. The microcontroller transmitted an alarm code when the
voltage dropped below an acceptable threshold. This was meant to send an alarm and
thereby avoid regular inspections. This method of power management increased the battery
life through the growing season.
Although some energy consumption reduction methods were implemented to conserve
power, the approach was still not energy efficient. This is because the approach gathered
data by using single-hop transmission of data from the multiple nodes in the field to the
centrally fixed receiving unit. This method does not balance the energy expenditure to the
multiple sensor nodes in the field. The energy levels of the nodes located farther away from
the sink would deplete faster due to more transmission power required for long trans-
mission ranges. Most of those sensor nodes would fail before their counterparts closer to
the sink. Therefore, the energy of the network is unbalanced and thus, the network lifetime
may be short.
Baggio (2005) instrumented a potato field to monitor micro-climate: humidity, tem-
perature and weather. This action was intended to reveal when the crop was at risk of
developing a phytophtora disease. The experiment was also to collect statistics on the
behavior of the WSN in a real-world experiment. Furthermore, the robustness of the
energy-efficient T-MAC (Van Dam and Langendoen 2003) protocol was also tested. The
sensor nodes used in the experiment were called TNOdes. These nodes were deployed
across the field with some of them acting as relays. These relays were without sensors.
These homogeneous sensor nodes measured the climatological conditions. A weather
station was also installed on the field to register luminosity, precipitation, wind strength
and direction. Soil moisture, which is a critical factor for micro-climate and development
of diseases, was measured by a number of sensors. A sensor to measure the height of the
groundwater table was installed. The nodes collected and transmitted infield data to a so-
called field gateway, which also forwarded the data to a low frequency array (LOFAR)
gateway by Wi-Fi. LOFAR is the low-frequency array for radio astronomy built by
ASTRON organization in the Netherlands. The LOFAR gateway was connected to the
internet through cable, and data were uploaded to a LOFAR server and further distributed
to a couple of other servers using the extensible mark-up language (XML) format. The data
on the behavior of the WSN, which was gathered during the experiment, served as a
foundation for WSN algorithm simulations to check failed connections and dead nodes. A
number of methods were adopted to ensure optimal utilization of the power source and to
extend the lifetime of the sensor nodes. For example, the nodes reported data only every
10 min. The data sent over the wireless link was minimized by the delta encoding tech-
nique (Mogul et al. 1997) which is a data compression technique for efficient data storage.
Data compression over radio communication lines was also considered. An energy-effi-
ciency protocol called T-MAC was introduced, and it was implemented on the radio with a
duty cycle of 7 %. Furthermore, some methods also ensured reliable data delivery to the
LOFAR gateway, in such a way as to avert data loss. A multi-hop routing protocol
(MintRoute) was proposed and used to maintain reliability of the WSN. This method
reveals when the crop is at risk of developing disease and informs the farmer to appro-
priately treat the field or parts of it with fungicide.
Their approach, with nodes reporting data every 10 min, was designed to save power.
Furthermore, to balance energy, the multi-hop routing protocol of TinyOS (Levis et al.
2005), an application-specific operation system for WSN, was used to send data. Their
222 Precision Agric (2015) 16:216–238
123
approach was energy-efficient as it combined both energy balancing and energy con-
sumption methods.
Nesa Sudha et al. (2011) proposed an energy-efficient TDMA-based algorithm for
wireless sensor communications in an automatic irrigation system. Their experiment
involved a header node that collected data directly from all of the nodes in the field and
passed them to the BS. Such a network arrangement made it a star single-hop topology.
Direct communication is maintained between the sensor nodes and the BS in the data
accumulation process, whereby the header node does not work on the collected data except
to directly transmit it to the BS. The process starts by the BS assigning unique identifi-
cation addresses to each of the nodes. The nodes went into a sleep state (or into a receiving
mode) after acquiring the address. To obtain data at a specific location, the BS sent a
broadcast request to all of the nodes, but only the addressed node responded with its
present temperature and moisture values. The BS received values from all of the areas by
repeating this process for all of the nodes in its address table.
In this approach, the medium access control (MAC) (I.C.S.L.M.S. Committee 1997)
protocol was used to control energy consumption. Sensor nodes turned ON/OFF their radio
according to a schedule to save energy. The approach verifies that data fusion (aggregation)
in which nodes were grouped into clusters was more energy efficient than direct com-
munication in which each node transmitted data directly to the sink. In energy conserva-
tion, a distributed TDMA-based scheme which leads to collision free transmission over the
data channel was used. This reduced the energy consumption by each node thereby
reducing the overall energy consumption. This approach could therefore extend the net-
work lifetime.
Heterogeneous battery-based power approaches
In a heterogeneous WSN, nodes in the network have different network attributes (Hou
et al. 2005). This type of network is usually associated with a hierarchical sensor network
in which relatively low resource sensor nodes are deployed at the lower level. Additionally,
higher resource sensor nodes are made into co-ordinator and sink nodes. In fact, in het-
erogeneous networks, only the co-ordinators and cluster heads might require higher
resources. This arrangement is to enable the cluster head to receive data from sensor nodes
and process it for the BS, to support data aggregation and protocol co-ordination. Providing
more power source for cluster heads extends their lifetime under high loads. In addition,
this arrangement could increase the cost and complexity.
Zhang et al. (2012) proposed a WSN-based system to monitor the soil moisture and
analyzed the temporal and spatial variability of soil moisture for variable irrigation.
According to Fig. 2, the system architecture included heterogeneous sensor nodes and sink
nodes and a remote management platform. The sensor nodes in the field collected and sent
soil moisture data periodically to the sink nodes. Sink nodes that acted as gateways to the
general packet radio service (GPRS) sent the data to the remote management platform via
the TCP-IP standard protocol. The platform that was based on the BS mode could collect,
store and analyze data and output the report forms. The sensor node positions were
measured with a global positioning system (GPS). Additionally, the distance between the
nodes was established according to the real field situation, the signal transmission distance
and the spatial variability in the soil moisture. The core module of the sensor node was a
wireless microprocessor from Crossbow Company, California, USA. The frequency
domain measurement method to determine the dielectric constant of soil was used to
calculate soil moisture. Both the sensor nodes and the soil moisture sensors were
Precision Agric (2015) 16:216–238 223
123
waterproofed to withstand harsh conditions. The power source to sustain the network
operation was from a battery. Additionally, the soil moisture sensors were always in sleep
mode and were activated only when data were to be collected. Multi-hop and self-orga-
nizing nodes were also used to guarantee network connectivity. The authors designed and
developed a multi-channel interface, peripheral bus and solar power system.
The data collected was processed and analyzed by classical statistical methods. With a
combination of soil moisture monitoring sensors, WSN and spatial analysis software, their
method gathers and transmits soil moisture data to remote management systems auto-
matically. The method used in the research to develop the soil distribution system was
simple, and the system offers valuable information as the basis to make irrigation
decisions.
Soil moisture sensor nodes sent data to the sink node every 30 min. This would reduce
the energy consumption of the sensor nodes. Soil moisture sensors were activated to collect
data only when the data was needed which would also reduce the energy consumption. The
multi-hop method of transferring data enables the nodes to transmit data to neighbor nodes
for onward forwarding to the sink, instead of the case in single-hop data transmission in
which the power source of remotely located nodes would be exhausted quickly.
Yu et al. (2013) proposed a hybrid architecture involving WSN and wireless under-
ground sensor networks (WUSN). An implementation framework was then developed for
real-time soil property monitoring. The impact of soil parameters, node depth, signal
frequency and attenuation on transmission were studied. Reflection, scattering and dif-
fraction may occur in the soil and at the soil–air interface during data transmission. The
experiment revealed that the frequency of the electromagnetic signals and the water
content of soil affect the path loss. As depicted in the Peplinski principle (Peplinski et al.,
1995), c=a?jb, with
a¼xffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
le0
2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1þe00
e0
2
s1
2
43
5
v
u
u
u
tð1Þ
b¼xffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
le0
2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1þe00
e0
2
sþ1
2
43
5
v
u
u
u
tð2Þ
where x=2pf is the angular frequency, lis the magnetic permeability, and e0and e00 are
the real and imaginary parts of the dielectric constant, respectively.
The system architecture used CC2430, a wireless transceiver chip based on ZigBee by
Texas Instruments, USA, which was developed to gather soil terrestrial information. The
WUSN used NRF905 wireless chip manufactured by Nordic Semi-Conductor Company,
Oslo, Norway, to collect and transmit information underground (UG). The hybrid system
topology is a combination of terrestrial and WUSN structures and is illustrated in Fig. 3.
The WSN was installed above 0.4 m depth while the WUSN was below 0.4 m. The sink
node of the WUSN is placed on the ground. All WUSN nodes transmit data to the
terrestrial sink node with their location based on the applications. The sink node is also said
to be static or mobile provided it is within the communication range. WUSN is set up by
placing underground sensors in the field, while deploying data sinks above the ground
located around the field. Mobile data sinks are carried by people or machines in the field.
The results of the experiment reveal that at low frequency signals and low volumetric water
224 Precision Agric (2015) 16:216–238
123
content, signal attenuation and bit error rate are minimal. The path loss is also influenced
by depth of the WUSN.
In the authors’ test model, sensors are buried at depths of 0.8, 1, 1.6 and 2 m, which
implies a large power source requirement. The UG nodes would have to expend a lot of
energy in a single-hop to transmit data to the sink node located aboveground. Thus the
nodes power sources could deplete quickly thereby shortening the network lifetime.
Kim et al. (2011) designed and developed a sensor network architecture using auton-
omous robots with beacons to monitor and detect fires and air pollution in fields. The
authors built a conceptual linear topology and implemented it in a tree-based network
topology. Sensor modules were developed using a RF transceiver with model number
CC2420, a single-chip 2.4 GHz IEEE 802.15.4 designed for low power and low voltage
wireless applications by Texas Instruments, USA. Also, a micro controller with model
number ATMega128L by Atmel corporation, California, USA was used. The application
was to monitor the temperature, gas, smoke, humidity and illumination. The data collected
was forwarded to a co-ordinator node every few seconds. Then, the co-ordinator sends the
data to a specific processor called a knowledge information processor for agriculture-
sensor data and fire-sensor data (KIP-AF) which shows the measured values in real time.
For any data that arrives at the KIP-AF, the server transmits an alert message to the end
user while using the ZigBee enabled SIM (ZSIM). Three types of nodes used in the scheme
were the wireless co-ordinator (WC), the wireless router (WR), and the wireless end device
(WED). The WC was the sink node and gateway, and it transmitted periodic beacon frames
as well as collected data from the WRs and WEDs. The study used heterogeneous sensor
nodes made up of one co-ordinator node, eight routers or parent nodes and 50 sensor nodes
(WED) and a tree-based network topology with a maximum of three hops. A conceptual
network topology that models the implementation architecture of the system is shown in
Fig. 4. All of the sensor nodes transmitted packets in every transmission interval to process
emergency data. At every interval, all of the sensor nodes transmit their sensing data to the
Fig. 2 Architecture of the soil moisture real-time monitoring system (Zhang et al. 2012)
Precision Agric (2015) 16:216–238 225
123
WC, which then forwards it to the KIP-AF. This processor maintained the agricultural
database for intelligent decisions and further analysis. When KIP-AF detects emergency
conditions such as fire and dangerous air pollution, it immediately informs the remote user
via a ZSIM terminal. To gather wide-area data from the farm, the study introduced mobile
robots to cover areas that were inaccessible to the fixed nodes. Therefore, the sensor nodes
in the field were heterogeneous, comprising mobile and static nodes. However, mobility
leads to link failures due to network handover or loss of line of sight (LOS). The solution to
this problem was for the mobile robots to scan for an alternative path to move to the WC.
This action was performed by using the best link quality indicator (LQI) among the
scanned WR candidates. Figure 5shows the handover scenario of the mobile nodes. The
second approach was to increase the RF power of the robot by adopting power amplifier
and low noise amplifier (LNA) functions in such a way as to double the transmission range.
Batteries were used as the power source.
The packet loss rate from the WED to the WC was measured to be less than 4 %, and
the numerical expression to calculate it is given as:
NFail
NFail þNRec 100 ð3Þ
where NFail and NRec are the number of failed packets to transmit to the co-ordinator and
the number of successfully received packets in the co-ordinator, respectively. Additionally,
at any time that the mobile node executes a handover procedure, the average handover
latency THO is calculated by using the expression:
THO ¼Tbeacon loss þTasc req þTack þTdata req þTack þTasc res þTack ð4Þ
where Tbeacon loss is the interval of the beacon loss due to the handover, Tasc req is the
transmission time of the association request frame, Tasc resis the transmission time of the
association response frame, Tdata req is the transmission time of the data request frame and
Tackis the transmission time of the acknowledgement frame.
Fig. 3 Topology structure of WUSN
226 Precision Agric (2015) 16:216–238
123
The most important concern is for reliable data transmission of real-time conditions
without energy losses at the sensor nodes which could affect the network lifetime. Duty
cycle is fixed at 100 %; this suggests that all sensor nodes transmit data to the WC at every
transmission window in order to process data for onward forwarding to the sink. This may
increase network traffic thereby causing packet collisions; retransmissions could lead to
faster power source depletion.
Solar power
Batteries can limit the functionality of WSNs due to their low capacities but a solar energy
harvesting module could fully meet the needs of practical applications in PA. Sensor
network survival no longer depends on battery life. However, a solar harvesting system is
more expensive to run than battery-based approaches and energy from the sun is not
User
terminal
KIP server
Co-ordinator
Relay Router
(WR, parent)
End node (WED, child)
Ethernet
Fig. 4 System implementation architecture
Fig. 5 Mobile handover scenario for router recovery (Kim et al. 2011)
Precision Agric (2015) 16:216–238 227
123
always available. Moreover, it has the disadvantage of being obtrusive. The solar panel
must be mounted above the canopy and generally is an impediment to field operations.
Homogeneous solar-based power approaches
Hedley et al. (2012) developed a WSN soil moisture mapping and monitoring method to
provide information for irrigation scheduling. Data were presented in digital format for
incorporating into the VRI controlling software. The research was conducted in a 75 ha
wheat field. The irrigation was by a center pivot (CP) irrigator with VRI modification.
Each sprinkler was controlled individually by digital maps uploaded to a central con-
troller. Sensors were combined with a real-time kinematic differential global positioning
system (RTK-DGPS) to measure soil and land parameters. Data loggers were mounted
on a vehicle to acquire high resolution vertical mode datasets in two separate surveys.
The field was segmented into three classes, and homogeneous sensor nodes were
deployed in each class. The nodes were organized in a mesh-structured topology and had
a range of up to 1 km in LOS. For energy efficiency, the nodes switched into a sleep
mode when they were in an idle state. Each node was attached with: (1) two Delta-T
SM300 moisture sensors installed at 0.2 and 0.5 m depth to monitor the volumetric soil
moisture content (v/v), (2) a Spectrum Technologies Watermark soil matric potential
sensor that was installed at 0.2 m depth to monitor the soil moisture tension, and (3) a
tensiometer equipped with an absolute pressure transducer, to assess the depth of the
water table, was installed at 1 m depth. A rain gauge was also attached to one node, to
monitor the irrigation and rainfall events. Data were relayed to a BS every 15 min; it
was processed in real-time and converted to the necessary format and immediately made
available through a 3G cellular modem via the internet to a webpage. End users could
access the web site. This WSN method provided a direct measure of the soil moisture
status and the need for irrigation. It has the advantages of using site-specific real-time
soil moisture data and site-specific climatic conditions to schedule the irrigation.
Therefore, (1) there is no need to know the plant type, (2) there is no need to know the
plant growth stage, (3) uncertainties (errors) when estimating the evaporation–transpi-
ration (ET) are avoided, (4) there is no need to know any of the variables needed to
estimate the ET, such as the air temperature, relative humidity (RH), radiation, soil
temperature and wind conditions. In order to conserve power, the sensor nodes relayed
data to the BS every 15 min.
Mafuta et al. (2013) designed a WSN for PA in Malawi and called it the WiPAM. This
system was intended to automate the irrigation process and implement an irrigation
management system (IMS). The authors investigated variability in soil moisture. Infor-
mation on the variability was then used by the irrigation controller to initiate irrigation
events. For precise irrigation, an irrigation event was triggered occasionally. The sensor
nodes were homogeneous, and each was composed of a ZigBee end device (ZED) and a
ZigBee co-ordinator (ZC). In the star topology, three sensors nodes were configured as
ZED (in field-nodes) and one as a gateway node. One other node was also configured as ZC
with the function to aggregate data and actuate the irrigation valves. The network topology
is shown in Fig. 6.
To operate the network, ZED collected soil moisture and temperature data at predefined
time intervals. Sampled data were sent and stored in a co-ordinator node. The co-ordinator
node then sent the data to a gateway node, which also transmitted the data to a remote
monitoring station (RMS) through a cellular network. The power source to sustain the
operation of the network was solar power. This approach was complemented by an energy-
228 Precision Agric (2015) 16:216–238
123
efficient management approach. In other words, sensor nodes went to sleep when idle and
woke up to repeat the previous steps when required. The irrigation valves opened or closed
depending on the values that were stored in the co-ordinator node. The WiPAM was
composed of two sections: the irrigation station (IS) and the RMS. The sections were
linked through a cellular network, as shown in Fig. 7. The RMS captured the performance
parameters of the IS at the remote site. The parameters were the soil moisture level, the soil
temperature, the battery voltage levels of the sensor nodes, the quality of the wireless links,
and the valve status. The main contribution of this paper is the design, implementation and
performance enhancement of a low-cost and efficient IMS that combines sensors and
actuators in a wireless sensor/actuator network.
To attain low power consumption, the protocol, operating at low data rate (250 kbps at
2.4 GHz), was used to gather data. Other technologies such as Bluetooth (802.15.1) and
Wi-Fi (802.11) consume more battery power although they offer higher data rates.
Moreover, in PA, some sensor data do not require wide bandwidth since it is not necessary
to continuously monitor soil moisture and temperature. Using ZC to aggregate data before
actuating an irrigation valve could reduce energy consumption. The approach further
conserved power by putting the ZEDs to sleep and only waking them up to make mea-
surements. The sensor data were collected at intervals of 30 min when the system was idle
and 2 min when irrigation was taking place. This arrangement could further reduce energy
consumption. In general, this approach could be said to be energy sufficient as it incor-
porated a number of energy reduction methods.
Majone et al. (2013) described the deployment of a WSN in an apple orchard. The goal
was to monitor soil moisture dynamics in the top soil to analyze interaction with plant
physiology. In the experiment, several homogeneous sensor nodes were deployed in a
multi-hop wireless mesh topology. The sensor nodes were connected across several soil
and temperature nodes and were deployed across 27 locations. The sensor nodes were
connected through an input/output interface to a TinyNode 584 WSN platform. A solar
power source supplied energy to the network. Some energy-efficient management
approaches were adopted to minimize energy consumption. Real-time monitoring of the
system began with measurement at each node of the battery and solar panel voltages and
Fig. 6 The architecture of the irrigation system
Precision Agric (2015) 16:216–238 229
123
the internal voltage as well as the temperature of the micro-controller. On-board software
routines were based on the TinyOS-2.x operating system. The process to acquire data
started every 10 min, and between every two consecutive measurements, the sensor nodes
went into sleep mode. Unnecessary modules [such as radio frequency (RF), module and
sensor] were switched off. This action decreased the power consumption to a minimum of
approximately 80 lA at 6 V. The energy consumption fluctuated from the start to end of
data collection. Sensor nodes sent data to the sink node. The sink nodes then measured the
received signal strength indicator (RSSI) value of every packet and forwarded the data
packets to a PC gateway (GW) over the serial port. The packets were assigned a date and
time, and the raw data were forwarded to a MySQL database for real time and historic
management. These data were eventually converted and calibrated by specific equations
and were displayed by dynamic PHP web pages. The experiment showed that sensors at
shallow depths responded more quickly than those at larger depths. The nodes which were
placed at lower depth measured high values of volumetric water content compared to the
nodes at higher depths. Moreover, the nodes at higher depth had lower response time due to
higher attenuation.
It appears that this approach employed two contrasting methods in terms of energy
efficiency. On the one hand, the approach may have conserved energy by putting the sensor
nodes into low power mode (also called sleep mode) and switching off every unnecessary
module (such as RF, module and sensor). However, a mesh network topology to enable
multi-hop data transmissions was used and all the data were sent over a radio channel to a
sink node. This means that all the nodes were awake and actively sending or providing
pathways for one another to the sink node, thereby wasting power source in the process.
The inference made is that the data were sent in raw form without any form of data
compression or aggregation and this could have generated packet losses and retransmis-
sions. Again since all nodes sent data to only one sink, the sink could over work thus
depleting power source faster. This approach may not be energy efficient and, therefore,
could have short network lifetime.
Fig. 7 Irrigation management system architecture
230 Precision Agric (2015) 16:216–238
123
Heterogeneous solar-based power approaches
Peres et al. (2011) designed an intelligent autonomous gateway, which was called intel-
ligent precision agriculture gateway (iPAGAT). The goal of iPAGAT was to function as an
interface device that processes in-field data by performing data aggregation and integra-
tion. The processed information would then be transmitted to high-level decision support
systems for farm managers to make decisions. iPAGAT used Bluetooth, IEEE 802.11 and
GPRS to communicate with local and remote users. The power source for the system was
solar power with a rechargeable battery, to increase the network lifetime. In the experi-
ment, the communication was based on five layers. The layers were labeled as L1–5. L1
involves the function of using WSN to gather environmental data from the vineyard
management zones. L2 devices function to co-ordinate the WSN and process the L1 data
by means of data aggregation techniques. The L2 layer also acts as an access point to a site-
specific management application and serves as a gateway to L3, which is the layer that has
the function to manage the entire farm. Layers L4 and L5 are farmer associations and
country policies, which are outside the scope of this paper. There are three methods of
communication in the system set up: (1) the link between the WSN and gateway (i.e., the
L1–L2 connections), (2) the link between the gateway and local user (with a smartphone),
through Bluetooth and/or IEEE 802.11, and (3) the link between the gateway and farm
management office through GSM/GPRS and/or IEEE 802.11. In the experiment, small
stationary and heterogeneous data acquisition nodes were deployed across the field in a
mesh network topology according to the ZigBee standard. The co-ordinator of this network
is situated at the gateway of the infrastructure; it manages the WSN and also acts as a sink
node for the collected data. It fed data into the gateway database which aggregated the
collected data and made the compressed data available to local and remote users through
web services. The solar panel uses Bluetooth and IEEE 802.11 technologies to commu-
nicate with the system devices. The iPAGAT served multiple functions: an environmental
repository for the collected data, a sink node for the WSN, an aggregation engine and a
Real-time alert system (RTAS). The management levels of the system are shown in Fig. 8.
The RTAS embedded in the IPAGAT for disease prediction was a large advantage of
the system. However, iPAGAT was rather complex and not easy to understand. Data were
collected through multi-hop mesh network with clusters. In this way, the load was dis-
tributed among the cluster heads without overloading some specific nodes. Moreover, the
collected data was summarized by a data aggregation method before transmission to the
next layer. In this way, only compressed data was transmitted and the system avoided
frame collisions and packet losses. In the light of the above, this approach could be
described as being energy efficient since it combined both energy balancing and energy
reduction methods in the data collection process.
Garcia-Sanchez et al. (2011) proposed a system to monitor crop fields. The system could
detect and identify procedures of a WSN video surveillance system. It also monitored, detected,
identified and transmitted data over long distances in distributed crops. The authors designed an
integrated wireless sensor networks solution for PA (ISSPA) and added it to the basic PA
mechanisms. The system integrated video transmission to networks which hitherto only
monitored crops. The system was, therefore, able to reduce the cost, increase the cycle lifetime
of the devices and networks, and avoid duplicate infrastructure. The ISSPA synchronized the
duty and sleep phases for the heterogeneous sensor nodes in the system. This goal was
accomplished by the use of message forwarding techniques and multi-interface routing devices.
All of the communications issues conformed to IEEE802.15.4 technology and allowed co-
operation between several communication nodes. Another contribution of the system was the
Precision Agric (2015) 16:216–238 231
123
design and implementation of IEEE 802.15.4 real-devices that avoided frame collisions in the
message transmissions and performance of their specific functions (i.e., crop monitoring, video
surveillance) within the agronomic cycle for the crop of interest. This system provided the
necessary hardware and software for monitoring farm field parameters such as temperature,
humidity and soil nutrient level. It also detected and identified the procedures of a WSN video-
surveillance system. Devices used in the study were categorized into five groups based on their
functionality. The categories were the monitoring group, detection group, identification group,
crop-gateway and farmer Co-op-gateway. All of the components of these nodes and gateways
were encased for protection against dust and rain. For each type of device, appropriate com-
mercialized hardware was selected, and appropriate software was developed to achieve func-
tionality. These assorted nodes incorporated appropriate sensors together with wireless
communication modules. The resulting devices were able to acquire data from different sen-
sors, such as salinity, PH, temperature and soil moisture. Furthermore, motion detectors and
identification sensors existed for video surveillance and crop security and control. Sensors
collected appropriate data and transmitted it to the Gateway through a communication module,
which then delivered the information to the Farmers’ Co-operative. The Co-operative then
provided each farmer with information on his crops in real time. The farmer received the
information at home via the internet on a PDA or Mobile handset. The system architecture is
shown in Fig. 9. Solar power supplied the required power for the devices.
The approach used cluster-tree network topology to collect the required data from the
sensor nodes. Then the ISSPA co-ordinated the information transmission from the cluster
nodes and so reduced power consumption and frame collision probability. To further
conserve power, IEEE 802.15.4 real devices were implemented. This suggests there were
no retransmissions, thereby conserving power. Video-surveillance, which was used over
Fig. 8 Management level scenario (Peres et al. 2011)
232 Precision Agric (2015) 16:216–238
123
the crop areas to capture, process and send pictures to the BS, may imply high energy
consumption. The video messages, along with the monitoring data transmission coming
from many sensors placed in separated crops could also provoke frame collisions and,
consequently, losses. The issue could be more critical when several cameras are consid-
ered. The retransmission of monitoring messages may also cause more energy consump-
tion. To mitigate the shortcomings, an ISSPA mechanism was again used. The mechanism
provides an appropriate synchronization of duty and sleep phases for all nodes in the
system using message forwarding techniques and multi-interface routing devices. Even
though the approach used power-hungry devices such as video cameras in its implemen-
tation, it co-ordinated the data transmission from the nodes of the cluster network.
Additionally, it reduced frame collision probability which prevented packet retransmission;
thereby, mitigating the power consumption.
Dong et al. (2013) provide a proof-of-concept towards an autonomous PA system by
integrating a CP irrigation system with a WUSN. The system provides autonomous irri-
gation management capabilities by monitoring the soil conditions in real time using
WUSN. The results are used to evaluate empirical channel models for soil–air commu-
nications. Their experiment proved that the concept of WUSN-CP was feasible. They
designed an UG antenna to improve the communication range. In the experiment, above-
ground nodes (AG nodes) were installed on the arm of a CP irrigation system, while UG
nodes were buried at depth of 0.35 m in the field in a circle. The CP rotated in both
clockwise and anti-clockwise directions. At suitable range, the AG nodes established
communication with the UG nodes. Two antenna schemes were used: firstly, a full-wave
(FW) dipole antenna for the AG node and secondly, a single ended elliptical antenna
(SEA) for the UG. While the former has a gain of 3 dB, the latter used a directional Yagi
antenna (Pozar 2000) for the AG node and a circular planar antenna for the UG nodes. The
Yagi antenna has a higher directivity with a maximum gain of 10 dB. Both the SEA
Fig. 9 System architecture (Garcia-Sanchez et al. 2011)
Precision Agric (2015) 16:216–238 233
123
Table 1 Features of WSN approaches in PA
No. Paper Application Technical
contribution
Topology Homogeneous/
heterogeneous
Energy source Data
aggregation
1 Vellidis et al. (2008) Irrigation scheduling Reliability Star single-hop star Homogeneous Battery No
2 Peres et al. (2011) Middleware infrastructure-
iPAGAT
Middleware Cluster tree-based
multi hop
Heterogeneous Solar/battery Yes
3 Nesa Sudha et al. (2011) TDMA scheduling irrigation Energy Single hop Star Homogeneous Battery Yes
4 Hedley et al. (2012) Irrigation scheduling Reliability Single hop Star Homogeneous Battery/solar No
5 Zhang et al. (2012) Irrigation scheduling Reliability Cluster tree-based
multi hop
Heterogeneous Battery No
6 Baggio (2005) Monitor climatological
conditions
Energy/reliability Tree-based multi
hop
Homogeneous Battery No
7 Shaikh et al. (2010) Irrigation control Energy Star single hop Heterogeneous Battery Yes
8 Dubey et al. (2011) WSN ?DTMF irrigation
control
Scheduling/energy Star single hop Homogeneous Battery No
9 Gao et al. (2013) WSN and fuzzy irrigation
control
Reliability Mesh Heterogeneous Solar No
10 Xu et al. (2012) Irrigation control Reliability Star single hop Heterogeneous Battery No
11 Kim et al. (2011) Monitoring fire and air pollution Reliability Cluster-tree Heterogeneous Battery Yes
12 Majone et al. (2013) Irrigation Schedule Reliability Mesh multi hop Homogeneous Solar No
13 Garcia-Sanchez et al.
(2011)
Monitoring and video
transmission
Energy and
transmission delay
Cluster-tree Heterogeneous Solar/battery No
14 Fernandes et al. (2013) Data integration platform Transducer-to-
network
interoperability
Mesh Heterogeneous Solar/wind moving
water in irrigation
pipes
Yes
15 Peng and Liu (2012) Irrigation scheduling Scheduling Cluster tree Heterogeneous Battery No
16 Mafuta et al. (2013) Irrigation scheduling Scheduling Star single hop Homogeneous Solar Yes
234 Precision Agric (2015) 16:216–238
123
antenna and the circular planar antenna are customized according to the operation range.
A TinyOS application was developed to analyze the experiment without the need to
reprogram the sensors.
An AG node located on the CP maintains unicast communication with a UG node
buried UG within their communication window. This suggests that the power source was
only spent in transmitting data directly from one node to the other. Since data was not
broadcast, it implies that other sensor nodes would not participate in data transmissions
thereby conserving power. Moreover, no intermediate nodes were required in the data
transfer; the source and sink nodes were therefore peer-to-peer. From the system structure,
there could be delays in the spinning of the CP to reach a particular node. However, this
delay may be acceptable in the application area of PA, since such delay is insufficient to
cause significant difference in the measured parameters. This approach could therefore, be
said to be energy efficient in the light of the foregoing observations.
Discussion
Table 1compares WSN approaches based on their applications, features and contributions
as used in PA. The table shows that the proposed WSN applications solutions are in
irrigation scheduling, although other areas such as pollution and fire detection are also
exploited. The major contributions of the discussed approaches could be categorized into
data transmission reliability, energy efficiency and efficient scheduling of the nodes spe-
cifically for irrigation control. Several approaches (Fernandes et al. 2013; Peres et al. 2011)
have adopted data aggregation techniques to collect the required data which is a technique
known for the reduction of communication overhead and energy expenditure of nodes in
WSN (Anisi et al. 2013a; Nesa Sudha et al. 2011; Theodoridis et al. 2012). Moreover,
WSN approaches must consider the type of topology to use and whether the sensor nodes
are homogeneous or heterogeneous. Finally, a critical consideration identified was the use
of appropriate power source to extend the network lifetime.
In general, energy-efficient routing for extending the network lifetime is provided by
two methods: (1) energy balancing, in which overloading and multi-functioning of certain
specific nodes is prevented, and the residual energy of the sensor nodes is monitored; and
Table 2 Energy efficient methods
No. Paper Energy balancing Energy consumption reduction
1 Vellidis et al. (2008)H
2 Baggio (2005)HH
3 Nesa Sudha et al. (2011)H
4 Zhang et al. (2012)HH
5 Yu et al. (2013)
6 Kim et al. (2011)
7 Hedley et al. (2012)H
8 Mafuta et al. (2013)H
9 Majone et al. (2013)H
10 Peres et al. (2011)HH
11 Garcia-Sanchez et al. (2011)H
12 Dong et al. (2013)H
Precision Agric (2015) 16:216–238 235
123
(2) energy consumption reduction to mitigate the average energy consumption of the
network using methods such as clustering, reducing the number of messages, reducing
overhead and establishing co-operative communications. In Table 2, the approaches have
been classified according to the energy efficient methods used.
Conclusions
A WSN is an efficient monitoring technique for use in PA. In this paper, the state-of-the-art
of WSNs has been discussed and reviewed. To achieve this goal, first, the various topol-
ogies of WSN approaches, including star single-hop, cluster tree-based multi-hop and mesh
networks, were described. Due to the vital role of energy in a WSN, the sources of power
in WSN approaches, including battery and solar power, were described. Second, the
current WSN approaches that are used in PA based on the power source and the type of the
node were categorized and reviewed. Third, the approaches based on the applications,
features and technical contributions of the WSN were outlined.
References
Anastasi, G., Conti, M., Di Francesco, M., & Passarella, A. (2009). Energy conservation in wireless sensor
networks: A survey. Ad Hoc Networks, 7, 537–568.
Anisi, M. H., Abdullah, A. H., Coulibaly, Y., & Razak, S. A. (2013a). EDR: Efficient data routing in
wireless sensor networks. International Journal of Ad Hoc and Ubiquitous Computing, 12, 46–55.
Anisi, M. H., Abdullah, A. H., & Razak, S. A. (2013b). Energy-efficient and reliable data delivery in
wireless sensor networks. Wireless Networks, 19, 495–505.
Baggio, A., 2005. Wireless sensor networks in precision agriculture, In: A. Dunkels, B. Ahlgren, P. Gun-
ningberg, S. Janson, C. Rohner and T. Voigt (Eds.), ACM Workshop on Real-World Wireless Sensor
Networks (REALWSN 2005). ACM, New York, NY, USA, pp. 1-2.
Baronti, P., Pillai, P., Chook, V. W., Chessa, S., Gotta, A., & Hu, Y. F. (2007). Wireless sensor networks: A
survey on the state of the art and the 802.15.4 and ZigBee standards. Computer Communications, 30,
1655–1695.
Can, Z., & Demirbas, M. (2013). A survey on in-network querying and tracking services for wireless sensor
networks. Ad Hoc Networks, 11, 596–610.
Coates, R. W., Delwiche, M. J., Broad, A., & Holler, M. (2013). Wireless sensor network with irrigation
valve control. Computers and Electronics in Agriculture, 96, 13–22.
I.C.S.L.M.S. Committee (1997). Wireless LAN medium access control (MAC) and physical layer (PHY)
specifications. IEEE Std 802.11.
Dı
´az, S. E., Pe
´rez, J. C., Mateos, A. C., Marinescu, M.-C., & Guerra, B. B. (2011). A novel methodology for
the monitoring of the agricultural production process based on wireless sensor networks. Computers
and Electronics in Agriculture, 76, 252–265.
Dong, X., Vuran, M. C., & Irmak, S. (2013). Autonomous precision agriculture through integration of
wireless underground sensor networks with center pivot irrigation systems. Ad Hoc Networks, 11,
1975–1987.
Dubey, V., Dubey, N., & Chouhan, S. S. (2011). Wireless sensor network based remote irrigation control
system and automation using DTMF code. In G. Tomar, A. Abraham, D. Bhatnagar, & A. Pandit
(Eds.), IEEE international conference on communication systems and network technologies (CSNT)
(pp. 34–37). Piscataway, NJ.
Dubois-Ferrie
`re, H., Fabre, L., Meier, R., & Metrailler, P., (2006). TinyNode: A comprehensive platform for
wireless sensor network applications. In J. Stankovic, P. Gibbons, S. Wicker & J. Paradiso (Eds.),
Proceedings of the 5th international conference on Information processing in sensor networks (pp.
358-365). New York, NY: ACM.
Fernandes, M. A., Matos, S. G., Peres, E., Cunha, C. R., Lo
´pez, J. A., Ferreira, P., et al. (2013). A framework
for wireless sensor networks management for precision viticulture and agriculture based on IEEE 1451
standard. Computers and Electronics in Agriculture, 95, 19–30.
236 Precision Agric (2015) 16:216–238
123
Gao, L., Zhang, M., & Chen, G. (2013). An Intelligent irrigation system based on wireless sensor network
and fuzzy control. Journal of Networks, 8, 1080–1087.
Garcia-Sanchez, A.-J., Garcia-Sanchez, F., & Garcia-Haro, J. (2011). Wireless sensor network deployment
for integrating video-surveillance and data-monitoring in precision agriculture over distributed crops.
Computers and Electronics in Agriculture, 75, 288–303.
Hedley, C., Ekanayake, J., Roudier, P. (2012). Wireless soil moisture sensor networks for precision irri-
gation scheduling. In L. D. Currie & C. L. Christensen (Eds.), Workshop abstracts, advanced nutrient
management: Gains from the past-goals for the future. (p. 85), New Zealand: Massey University.
Hou, J., Li, N., Stojmenovic, I., (2005). Topology construction and maintenance in wireless sensor net-
works. Handbook of sensor networks: Algorithms and architectures (pp. 311-341). New York, NY:
Wiley.
Kim, Y.-D., Yang, Y.-M., Kang, W.-S., & Kim, D.-K. (2011). On the design of beacon based wireless sensor
network for agricultural emergency monitoring systems. Computer Standards & Interfaces, 36, 288–299.
Levis, P., Madden, S., Polastre, J., Szewczyk, R., Whitehouse, K., Woo, A., et al. (2005). TinyOS: An
operating system for sensor networks. In W. Weber, J. Rabaey, & E. Aarts (Eds.), Ambient intelligence
(pp. 115–148). Berlin: Springer.
Mafuta, M., Zennaro, M., Bagula, A., Ault, G., Gombachika, H., & Chadza, T. (2013). Successful
deployment of a wireless sensor network for precision agriculture in Malawi. International Journal of
Distributed Sensor Networks, 2013, 1–13.
Majone, B., Viani, F., Filippi, E., Bellin, A., Massa, A., Toller, G., et al. (2013). Wireless sensor network
deployment for monitoring soil moisture dynamics at the field scale. Procedia Environmental Sciences,
19, 426–435.
Mhatre, V., Rosenberg, C. (2004). Homogeneous vs heterogeneous clustered sensor networks: a compar-
ative study, In: H. Sari (Ed.), IEEE international conference on communications (IEEE ICC), (pp.
3646–3651). Piscataway, NJ.
Mogul, J. C., Douglis, F., Feldmann, A., & Krishnamurthy, B. (1997). Potential benefits of delta encoding
and data compression for HTTP. In M. Steenstrup (Ed.), Proceeding of the ACM SIGCOMM’97
conference on applications, technologies, architectures, and protocols for computer communication
(pp. 181–194). New York, NY: ACM.
Nesa Sudha, M., Valarmathi, M., & Babu, A. S. (2011). Energy efficient data transmission in automatic
irrigation system using wireless sensor networks. Computers and Electronics in Agriculture, 78, 215–221.
Owojaiye, G., & Sun, Y. (2012). Focal design issues affecting the deployment of wireless sensor networks
for pipeline monitoring. Ad Hoc Networks, 11, 1237–1253.
Park, C., Lahiri, K., & Raghunathan, A. (2005). Battery discharge characteristics of wireless sensor nodes:
An experimental analysis. In E. Belding-Royer, M. Krunz & T. Nandagopal (Eds.), IEEE international
conference on sensing, communication and networking (IEEE SECON) (pp. 20–21). Piscataway, NJ.
Peng, X. & Liu, G. (2012). Intelligent water-saving irrigation system based on fuzzy control and wireless
sensor NETWORK. In X. Luo, Z. Luo, L. M. Ni & R. Wang (Eds.), Fourth IEEE international
conference on digital home (ICDH) (pp. 252–256). Piscataway, NJ.
Peplinski, N. R., Ulaby, F. T., & Dobson, M. C. (1995). Dielectric properties of soils in the 0.3–1.3 GHz
range. IEEE Transactions on Geoscience and Remote Sensing, 33, 803–807.
Peres, E., Fernandes, M. A., Morais, R., Cunha, C. R., Lo
´pez, J. A., Matos, S. R., et al. (2011). An
autonomous intelligent gateway infrastructure for in-field processing in precision viticulture. Com-
puters and Electronics in Agriculture, 78, 176–187.
Pozar, D. M. (2000). Microwave and RF design of wireless systems. New York, NY: Wiley.
Rathnayaka, A., & Potdar, V. M. (2013). Wireless sensor network transport protocol: A critical review.
Journal of Network and Computer Applications, 36, 134–146.
Ruiz-Garcia, L., Lunadei, L., Barreiro, P., & Robla, I. (2009). A review of wireless sensor technologies and
applications in agriculture and food industry: State of the art and current trends. Sensors, 9, 4728–4750.
Selavo, L., Wood, A., Cao, Q., Sookoor, T., Liu, H., Srinivasan, A., Wu, Y., Kang, W., Stankovic, J., &
Young, D. (2007). Luster: Wireless sensor network for environmental research. In S. Jha (Ed.), Pro-
ceedings of the 5th international conference on embedded networked sensor systems (SenSys) (pp.
103–116). New York, NY: ACM.
Senturk, I. F., & Akkaya, K. (2014). Mobile data collector assignment and scheduling for minimizing data
delay in partitioned wireless sensor networks. In P. Bellavista, A. Mellouk & M. H. sherif (Eds.),
Proceeding of the 5th international conference on ad hoc networks (ADHOCNETS) (pp. 15–31).
Heidelberg: Springer.
Shaikh, Z., Yousuf, H., Nawaz, F., Kirmani, M., & Kiran, S. (2010). Crop irrigation control using wireless
sensor and actuator network (WSAN). In IEEE International conference on information and emerging
technologies (ICIET) (pp. 1–5). Red Hook, NY.
Precision Agric (2015) 16:216–238 237
123
Srbinovska, M., Gavrovski, C., Dimcev, V., Krkoleva, A., & Borozan, V. (2014). Environmental parameters
monitoring in precision agriculture using wireless sensor networks. Journal of Cleaner Production.
doi:10.1016/j.jclepro.2014.04.036.
Stankovic, J. A. (2008). When sensor and actuator networks cover the world. ETRI Journal, 30, 627–633.
Theodoridis, E., Chatzigiannakis, I., & Dulman, S. (2012). Post-processing in wireless sensor networks:
Benchmarking sensor trace files for in-network data aggregation. Journal of Network and Computer
Applications, 35, 548–561.
Tongtong, Y., Wenjie, F., & Zheying, L. (2011). Temperature and humidity wireless sensing and monitoring
systems applied in greenhouse. In J. Wang (Ed.), IEEE International conference on computer science
and network technology (ICCSNT) (pp. 857–861). Red Hook, NY.
Van Dam, T., & Langendoen, K. (2003). An adaptive energy-efficient MAC protocol for wireless sensor
networks. In I. F. Akyildiz, D. Estrin, D. E. Culler & M. B. Srivastava (Eds.), Proceedings of the 1st
international conference on embedded networked sensor systems (SenSys) (pp. 171–180). New York,
NY: ACM.
Van Hoesel, L., Tu
¨Ysu
¨Z Erman, A., Dilo, A., & Havinga, P. J. (2013). Geo-casting of queries combined
with coverage area reporting for wireless sensor networks. Ad Hoc Networks, 11, 104–123.
Vellidis, G., Tucker, M., Perry, C., Kvien, C., & Bednarz, C. (2008). A real-time wireless smart sensor array
for scheduling irrigation. Computers and Electronics in Agriculture, 61, 44–50.
Vellidis, G., Tucker, M., Perry, C., Reckford, D., Butts, C., Henry, H., Liakos, V., Hill, R. W., & Edwards,
W. (2013). A soil moisture sensor-based variable rate irrigation scheduling system. In J.V. Stafford
(Ed.), Precision Agriculture 2013—Proceedings of the 9th European conference on precision agri-
culture (9ECPA) (pp. 713–720). The Netherland: Wageningen Academic Publishers.
Vicaire, P., He, T., Cao, Q., Yan, T., Zhou, G., Gu, L., et al. (2009). Achieving long-term surveillance in
vigilnet. ACM Transactions on Sensor Networks (TOSN), 5,9.
Wang, N., Zhang, N., & Wang, M. (2006). Wireless sensors in agriculture and food industry—recent
development and future perspective. Computers and Electronics in Agriculture, 50, 1–14.
Watras, C., Morrow, M., Morrison, K., Scannell, S., Yaziciaglu, S., Read, J., et al. (2014). Evaluation of
wireless sensor networks (WSNs) for remote wetland monitoring: design and initial results. Envi-
ronmental Monitoring and Assessment, 186, 919–934.
Wood, A., Virone, G., Doan, T., Cao, Q., Selavo, L., Wu, Y., Fang, L., He, Z., Lin, S., & Stankovic, J.
(2006). ALARM-NET: Wireless sensor networks for assisted-living and residential monitoring. Uni-
versity of Virginia Computer Science Department Technical Report 2, VA.
Xu, Z., Lou, B., & Shao, G. (2012). An intelligent irrigation system for greenhouse jonquil based on ZigBee
wireless sensor networks. Internet of things. (pp. 200–207). Berlin Heidelberg: Springer.
Yu, X., Wu, P., Han, W., & Zhang, Z. (2013). A survey on wireless sensor network infrastructure for
agriculture. Computer Standards & Interfaces, 35, 59–64.
Zhang, R.-B., Guo, J.-J., Zhang, L., Zhang, Y.-C., Wang, L.-H., & Wang, Q. (2011). A calibration method of
detecting soil water content based on the information-sharing in wireless sensor network. Computers
and Electronics in Agriculture, 76, 161–168.
Zhang, M., Li, M., Wang, W., Liu, C., & Gao, H. (2012). Temporal and spatial variability of soil moisture
based on WSN. Mathematical and Computer Modelling, 58, 826–833.
238 Precision Agric (2015) 16:216–238
123
