CSIT

https://doi.org/10.1007/s40012-020-00289-1

ORIGINAL RESEARCH

Wireless sensor networks: a review of motes, wireless

technologies, routing algorithms and static deployment strategies
for agriculture applications
Pawan Kumar1 • S. R. N. Reddy2

Received: 31 January 2019 / Accepted: 21 May 2020

Ó CSI Publications 2020

Abstract From the last few decades, a lot of research work Keywords Deployment
Wireless sensor network
Sensor
has been conducted to develop wireless sensor networks nodes
Wireless technologies
Mote’s operating system

(WSN) for wide variety of applications such as agriculture Routing protocols
Deployment techniques
and environment monitoring, healthcare, military applica-
tions etc. Nowadays, the academicians and researchers are
working on resolving various research issues in this tech- 1 Introduction
nology domain. A few of the research issues faced by
researchers’ are designing low cost and low power motes, Recent advances in wireless networking, data communi-
network topologies, deployment strategies, and energy cation and micro electromechanical systems (MEMS) has
efficient data routing protocols. A lot of research has led to the development of low cost, low power, and small
already been done in each of the mentioned areas indi- size motes or sensor nodes [1]. Hundreds or thousands
vidually but only a few research articles or literature dis- numbers of these nodes when networked and deployed
cusses research work about two or more areas collectively. together in a random or predetermined manner, can sense,
The existing literature explores the simulation work only capture and process the data related to any physical phe-
and a few has given the description of complete design, nomena and can also wirelessly communicates the same to
development and deployment of a WSN. This paper pre- distant locations for analysis and evaluation. The sensor
sents the detailed discussion in form of literature review of nodes have self-configuration capability [2]. There are
each aspect of WSN design starting from mote design and three types of nodes in wireless sensor network- sensor
ending to associated application. Since a lot of applications node, router node and sink node (or base station) as shown
are using WSNs for monitoring purpose, this paper focuses in Fig. 1.
on one of the most crucial application for livelihood i.e. The sensor node senses the sensory information and
crop production or agriculture domain. passes it to neighboring nodes or router nodes in multihop
fashion through state of art wireless technologies such as
ZigBee, BLE, Wi-Fi etc. The Router nodes transfer or
routes the data towards the other nodes closer to the sink
node or Base station. This Base station collects all the pre-
& Pawan Kumar processed information from the deployed nodes in that
pawan1234@gmail.com specific area and forwards the same to remote station where
S. R. N. Reddy a database server stores the incoming information for fur-
rammallik@yahoo.com ther analysis and evaluation. After the data analysis done
1 by agronomist and researchers, they provide decision
University School of Information Communication
Technology, Guru Gobind Singh Indraprastha University, support on the registered mobile, tab and laptop devices
Dwarka, Delhi, India etc.
2
Department of Computer Science and Engineering, Indira All the nodes in WSN are having their own protocol
Gandhi Delhi Technical University for Women, Delhi, India stack which handles data communication, data integration

123
 CSIT

Fig. 1 Wireless sensor

network’s architecture

with networking protocols and routing protocols. The developed and deployed WSNs in precision agriculture
protocol stack of sensor nodes consists of physical layer, application. This area is chosen as it affects real life sce-
data link layer, routing layer, transport layer and applica- narios to meet the rising demand of food grains production
tion layer [3]. The physical layer gives modulation, trans- and resource management. The WSN improve the quality
mission, demodulation process. The data link layer and quantity of crop production and also use the resources
contributes for MAC protocols and error correcting codes. with proper planning.
The network layer is used for routing the data packets. The The academicians and researchers are working on var-
application layer is used to analyze the data senses by the ious research issues of WSNs such as sensor node design,
nodes and stored in the database server. The transport layer wireless technology, and operating system for sensor
is used to control the data packet flow and it is basically nodes, routing protocols and deployment techniques. There
used to take data using TCP/IP protocols form gateway in exists several research papers that are written on all these
case of using Internet of things (IoT) technology. research area individually but most of them are concen-
The WSN has different routing protocols and algorithms trated on simulation work and literature lacks the detailed
as that of Adhoc networks because of high density of review of WSN in terms of constituting components of
sensor nodes, limited power, limited data processing, WSNs like sensors, motes, data frame formats, wireless
smaller data rate, limited bandwidth and prone to failure standards and technologies, routing protocols, deployment
[4]. Also, due to the faults or application requirements, the strategies and applications specific to precision agriculture
sensor node changes its topology very fast. Therefore, the domain. The detailed review is also presented in terms of
routing protocols should be scalable. The sensor nodes do building blocks of motes such as processors, radios, power
not have global IDs because large number of packet units and sensing unit. A comparative study of existing
overheads will be increased in doing so. Therefore, the motes has also been presented in one of the section of
WSN routing protocols are different from conventional paper. The following sub section shall discuss wireless
routing protocols. The WSN architecture and its node standards and technologies, routing protocols, data frame
architecture are application dependent. The deployment of formats, data aggregation technique, deployment strategies
sensor nodes in a WSN is done by the single authority or and applications specific to precision agriculture domain.
user whereas Adhoc networks are commonly deployed and The paper is divided in various sections. Section 2 gives
shared by multiple users and organizations. details about the various components of sensor node.
There are multiple applications of WSNs such as health Section 3 gives the comparison between various motes
monitoring [5–7], military [8, 9], forest monitoring available in market and the designing aspects of sensor
[10, 11], environmental monitoring [12, 13], precision nodes are given in Section 4. The details of various wire-
agriculture [14, 15], horticulture applications [16], bal- less technologies used for communication is given in
anced fertilization [17], pest management [18], smart irri- Section 5. The various types of routing protocols used in
gation system [19–21], green house monitoring [22, 23], WSN are discussed in Section 6. The data packet format
aquaculture [24] and precision viticulture [25] etc. This and data aggregation technique is given in Section 7. The
paper focuses on presenting the review of designed, static node deployment techniques are discussed in

123
CSIT

Section 8 whereas some related work of design, develop- used to convert electrical signals (such as voltage, current
ment and deployment of WSN for agriculture applications etc.) into digital data for data logging and further pro-
is given in Section 9. cessing. The data calibration, signal condition are also the
part of sensor unit. The data calibration is the process of
changing the raw sensor readings into the actual readings
2 Review of components of sensor node when sensors come into the actual environment. Designing
of hardware and software calibration techniques is also a
The sensor node is one of the key components of WSN. research issue that needs attention. Table 1 shows com-
The sensor node is also named as mote. A sensor node monly used sensors for measuring multiple parameters in
generally consists of sensors, processing unit, and tran- agricultural domain.
sreciever and power unit as shown in Fig. 2. The power There are two types of sensor available in market, that
generator, location finding system and mobilizer are the is, multifunctional sensor and single sensor. The multi-
optional component of sensor node which can be added to function sensor senses multiple properties as shown in
the sensor node depending upon application requirement Table 1. These sensors have some limitations also. Such
[3]. The location finding system is used to find out the sensors cannot be directly connected to any sensor node,
actual and exact position or location of sensor nodes and since their output usually is very specific linear analogue
mobilizer is used to move the sensor node from one loca- voltage or current loop signals. They require digital output
tion to another. The power generator is used in energy interfaces, for example SPI, I2C, RS-232, RS-485 and SDI-
harvesting techniques such as solar, wind etc. The next 12 bus. Also, the time taken to read the data by protocol
subsection shall discuss both hardware and software of a such as SDI12 is high nearly one second. The main dis-
mote. advantage of digital output interfaces such as SDI-12 is
high power consumption [28]. Therefore the selection of
2.1 Sensors and sensing unit sensors for precision agriculture application depends on
data conversion rate, latency,power consumption accuracy
The sensors work as transducers and are used to convert and reliability over a deployment period.
physical entities (such as temperature, humidity etc.) into
electrical signals (such as voltage, current etc.). The sen- 2.2 Processors
sors give data to the processor unit of sensor nodes. Some
sensors are inbuilt in mote and some has to connect The processor unit takes the sensor data in digital form and
externally. The accuracy of the WSN is dependent on the converts this data into packet frames for feeding it to
cost, precision, accuracy and compatibility of sensors to transceivers for data communication. All the data pro-
environmental conditions. Therefore sensors should be cessing, data storage, forward error correcting codes are
cheap, accurate, sustainable to environment and highly done by the processor unit. The processor used for motes
reliable over a longer period of deployment. WSN nodes are generally from 16- bit to 32 bit processors. As the
have sensing unit where both sensors and analog to digital frequency increases, its power consumption as well as cost
converters (ADC) are situated on board. The ADCs are also increases. The algorithm for data processing is stored

Fig. 2 Architecture of a mote

123
 CSIT

Table 1 Sensors and measureable entities for agriculture domain [26, 27]
Sensor Temperature Moisture Dielectric Rain/water Water Conductivity Salinity
permittivity flow level

Hydra probe II soil sensor Y Y Y Y Y Y Y

Pogo portable soil sensor Y Y Y Y – Y –
MP406 soil moisture sensor Y Y Y – – – –
ECH2O soil moisture sensor Y Y Y – Y Y –
EC sensor (EC250) Y Y – Y – Y Y
ECRN-50 low-REC rain gauge – – – Y – – –
ECRN-100 high-REC rain gauge – – – Y – – –
Tipping buket rain gage – – – Y – – –
107-L temperature sensor (BetaTherm Y – – – – – –
100K6A1B Thermistor)

into the FLASH and the sensing data is processed into the Therefore, it can be kept in OFF state when it is not
RAM. In some cases, the sensory data with date and time required. The RF short range radio is not dependent on line
of sensing is also stored in EEPROM for local storage of sight. They work with free unlicensed band Industrial
application. The flash memory is cost effective memory Scientific Medical (ISM) of 2.4 GHz. The distance
and hence mostly used in processors. Some commonly between two nodes for data transmission should be as
used processors in mote are shown in Table 2. minimum as possible because large distances for data
As the clock frequency as well as memory data capacity transmission require more power to be consumed [30]. The
increases, the cost of processor is also increased. Therefore, radios of sensor nodes use multihop technique. The RSSI
ATmega128L is mostly used in low cost agriculture signal strength of radios should not be effected by reflec-
applications of WSN. tion, obstacles etc. The designing of radio’s having low
duty cycle, low power consumption and high RSSI signal
2.3 Transrceivers strength over the distance is another important area of
research. Table 3 Shows the selection of a radio for agri-
The transreciever transmits the data frame from one node culture application depends upon multiple parameters.
to another. The transceiver unit can be infrared, optical, RF The radio of communication system is controlled by
short range radios, ultrasonic and inductive fields. Among MAC protocols. The MAC protocol has the control on
all these technologies, RF short range radios are cost radio to change its duty cycle and reduce the power con-
effective and they have low power consumption also. The sumption due to the unnecessary awaking of the radio. The
RF transceiver does the modulation, band pass filtering, MAC protocol is used to avoid packet collision, idle lis-
multiplexing, error control mechanism and demodulation tening, over emitting, over hearing and increasing packets
processes during the data transfer. The sensor nodes overheads. The packets are lost due to the data packets
deployed in fields, are close to the ground, therefore, the transmitted by different nodes on the same time or on the
propagation path losses are from fourth order exponent. same channel. The idle listening is the process of a
The data rate is required of nearly few Hz for many unnecessary awaking of radio when there is no packet
applications and after that there is no requirement of data transmission or reception is required by that radio. The
transfer for few minutes. The maximum power of sensor over emitting is the process when receiver radio is not
node is consumed by transceiver in data transmission. ready to take the data packets. The over hearing is the

Table 2 Various types of processor and microcontrollers used in designing sensor nodes [29]
Processor/microcontroller Architecture (no of bits) Clock frequency (Mhz) ROM (KB) RAM (KB) EEPROM (KB)

ATmega128L 8 7.328 128 4 512

TI MSP 430 16 8 48 10 1024
ARM7TDMI 32 12 64 – 512
XSCALEPX271 32 13–416 256 32 32 9 1024

123
CSIT

Table 3 Various radios used in designing sensor nodes [31]

Parameters Microchip MRF24j40-I/ Free scale semiconductor MC ATMEL AT86RF232- Texas instruments
ML 13213 ZX CC2430F32RTCR

Data rate (Kbps) 625 250 250 250

Frequency max (Mhz) 2.483 2.48 2.405 2.4835
Frequency min (Mhz) 2.4 2.405 2.48 2.4
Min supply voltage 2.4 2 18 2
(V)
Price per unit $2.86 $ 8.80 $ 1.22 $ 3.56
Receiving current 19 42 118 27
(mA)
Sensitivity - 95 - 92 - 100 - 92
supply voltage 3.6 3.4 3.6 3.6
max(V)
Transmitting 23 35 138 27
current(mA)

process of receiving data packets by radios of a sensor node harvesting techniques such as solar energy harvesting,
which are dedicated for the different nodes. The packet wind energy harvesting etc. [32, 33] (Table 4).
over heads are increased by either transmitting and The selection of battery depends upon power con-
receiving multiple copies of same data packets or using the sumption of mote, deployment period of motes in fields and
communication scheme where unnecessary packets are also on cost. The Lead Acid battery is widely used in
advertising and acknowledgement. The MAC protocol designing nodes where the deployment period of nodes is
should be scalable so that the changes in node density and more than 4 months and the deployed WSN solution is
topology can be adjusted by the network. The MAC pro- needed to be cost effective.
tocol should be designed in such a way the data trans-
mission latency should be minimum and data throughput 2.5 Operating system (OS)
should be maximum. Therefore, more real data monitoring
can be possible by wireless sensor network. The long data The OS is the main framework which bridges the hardware
packets should be broken into the small packets so that the and software gap. The OS is responsible for resource
sensor node having more data packets to transmit will have management, memory management, power management,
more access to the channel. B-Mac, ZMAC, Time-MAC data routing and task management for sensor nodes. The
and WiseMac are the some examples of MAC protocols for operating system should be hardware independent and it
wireless sensor networks. should have various functional modules such as task
management, interrupt handling, networking management
2.4 Power resources and concurrency control. Designing a hardware indepen-
dent OS and which gives real time data response for all
The power unit is one of the key elements of sensor node. above functional modules is also one of the important
The power can be taken by battery or by the use of energy research issue for WSN. A lot of operating system for

Table 4 Various types of batteries used for mote [32, 34]

Battery Power density (W/Kg) Efficiency (%) Self-discharge (%/month) Recharge cycles

Lead acid 150–400 75–85 20–30 250–1000

Li-polymer 3000 99.8 \ 10 500–1000
Lithium-ion 1800 99.9 \ 10 1200
Nickel-metal hydride 250–1000 66 20 1000
Nickel–cadmium 150 70–90 10 1500
Sealed lead acid 180 70–92 20 500–800

123
 CSIT

motes are also designed and developed such as TinyOs, should be affordable by the user (or farmers in case of
Contiki, SOS, Mantis OS, Nano RK, RETOS and LiteOS agriculture application). Table 6 provides the technical
etc. There are various challenges such as real time comparison of motes used for agriculture applications. The
response, node configurability, ease of programming etc. in designing of sensor node having hardware of low power
designing small size OS. A comparison of existing OS that consumption, high RSSI over distance and low power
are being utilized in designing motes is given in the modulation data transfer communication scheme is one of
Table 5. TinyOS is the most common OS used in various the research issues.
mote such as MicaZ, IRIS etc. The inbuilt sensors on mote are arranged in sequence
number for Table 6 and these are Light (1), temperature
(2), humidity (3), barometric(4), pressure (5), accelerom-
3 Comparative study of sensor nodes eter (6), GPS(7), RH (8), microphone (8), sounder (9) and
magnetometer (10). The MICA and TelosB sensor nodes
There are a lot of sensor nodes are already available in are mostly used in designing and developing WSN because
market. These sensor nodes are categorized on the basis of of their low cost and support for multiple OS.
physical characteristics, cost, sensing, data processing, data
communicating and power consumption parameters. The
physical characteristics are the mote dimensions such as 4 Designing and deployment aspects of motes
area, weight, operating temperature range etc. The sensing
parameters are dependent on the inbuilt sensors on the The designing of a sensor node for specific applications
sensor node board or the provision of connecting the depends upon a lot of factors such as reliability, scalability,
external sensors. The output interface and data reading node cost, operating environment, sensor network topol-
time from sensor is also an important parameter which ogy, hardware constraints, transmission media etc. [3, 4].
effects the power consumption of mote. If more time is These parameters also affect the design and development
required to read the sensor data, more power will be con- of routing protocols and deployment techniques.
sumed. The mote should have data pre-processing capa-
1. The reliability of a sensor network is a property which
bility such that only meaning and aggregated data can be
is affected by the failure of a sensor node. A sensor
transmitted. The radio of mote should be in a manner such
node can be dead due to the loss of power or due to
that less power should be consumed during data trans-
some physical damage. The probability of fault
mission. The cost of sensor nodes is also an important
parameter because the total cost of wireless sensor network

Table 5 Various types of operating system used for mote [35–38]

Features TinyOS Contiki SOS Mantis NanoRK RETOS LiteOS

Publication ASPLOS EmNets MobiSys MONET RTSS IPSN IPSN

Year of 2000 2004 (2005) (2005) (2005) (2007) (2008)
publication
Design Monolithic Modular Modular Layered Monolithic Modular Modular
File system Single level Coffee No No No N0 Hierarchical
Language Ne C C C C C C LiteC ??
support
Memory Static Dynamic Dynamic Dynamic Static Dynamic Dynamic
managment
Networking Active message uIP, uIPv6, Message ‘‘comm’’ Socket Three-layer File-assisted
Rime architecture
Program Event and Event and Event Thread and event Thread Thread Thread and event
model thread threads
Resource Virtual and Serialized – Through Mutexes and – Through
sharing complete access Semaphores semaphores synchronized
events primitives
Scheduling FIFO Interrupt – 5 priority classes Rate monotonic – Round Robin
performance and preferences harmonized scheduling
scheduling

123
CSIT

Table 6 Various types of motes used in WSNs [39–42]

Feature MICA MICA2DOT MICA Z Imote2 TelosB IRIS CRICKET

Microcontroller ATmega128L ATmega128L ATmega128L MARVELL/ TIMSP430 ATmega128L ATmega128L

used XSCALEPX271
Battery (AA) 2 3Vcoin cell 2 3 2 2 2
Power 75 75 75 100 8 24 75
consumption
(lw) in sleep
mode
Inbuilt sensors 1–10 1–2, 6 1–10 1–3, 6 1–3 1–10 1–10
available
Transceivers RFM TR1000 Chipcon series TI CC2420 TI CC2420 Chipcon AT86RF230 CC1000
868/916 MHz series
Operating system TinyOS,MantisOS TinyOS, TinyOS, TinyOS TinyOS, TinyOS TinyOS
MantisOS MantisOS Contiki,
MantisOS

tolerance of a sensor node is given by Poisson with other slaves. Bluetooth can form network with only
distribution: few limited number of nodes in network, therefore, this
Ri ðtÞ ¼ expð ki tÞ wireless technology is not used in large wireless sensor
network applications. On the other hand, ZigBee wireless
where Ri(t) is the reliability of an ith sensor node technology has low power consumption, low data rate, very
between time 0 to t and k is the failure rate of sensor small wake up time and high reliability. The IEEE 802.15.4
node. Therefore, the node should have fault tolerance standard defines the lower layers (PHY and MAC) of
capacity and robust design for field deployment [3] ZigBee device and allows to create a sensor network of
2. The deployment of sensor nodes must be scalable to almost 65,535 sensor nodes. The IEEE 802.15.4 standard
few nodes to thousands of nodes. All the nodes do not assumes the data to be transfer at small distances and with
participate all the time in the sensor network depend- low duty cycles by the nodes. Therefore this suits more to
ing upon application. Sometimes, some more nodes are precision agriculture application because a low data
also deployed in the network after some time of transfer rate between nodes, is required after a sampling
deployment to reduce power consumption and frequency of 5 to 30 min in precision agriculture. The
improved data sensing accuracy [4]. Zigbee standards are defined for high layers of ZigBee
3. The cost of sensor nodes should be kept as low as device. The ZigBee defines the network layer to form three
possible because if the total setup cost of wireless types of topologies such as mesh, star and tree. In mesh
sensor network is large as compared to the traditional topology, all the nodes are directly connected and it
method used by user for the particular application, then increases the complexity of network architecture. Zigbee
the user will not preferred the WSN [3, 4]. also defines the types of devices participated in sensor
network. The Full functional device (FFD) acts as router
and reduce function device(RFD) acts as end device.
5 Wireless standards and technologies

There are various communication technologies available 6 Routing protocols in WSN

which support different frequency, data rate, range, power
consumption etc. The wireless technologies are compared The sensor nodes have limited power and limited band-
with respect to these important factors are shown in width constraints. Therefore, the sensor nodes collects data
Table 7. from target deployed region and forwarded these data
The IEEE 802.11 standard has high data rate transfer packets to the router nodes in the direction of base station.
capability but it demands a lot of power for communica- Hence, the data packets are routed by different sensor
tion. The IEEE 802.15.1 (Bluetooth) standard has the nodes in mutihops using routing protocols. There are sev-
capability to form Pico nets and scatter nets with high data eral challenges in designing routing protocols such as
rates. The one master device does the synchronization work sensor nodes location finding, low power, low memory and

123
 CSIT

Table 7 Various wireless

Technologies Frequency (ISM) Max data rate Range Power consumption
technologies used in WSN
[43–53] IEEE 802.15.4 (ZigBee) 868/915 MHz, 2.4 GHz 250 kbps 100 m Low
UWB IEEE 802.15.4a 3.1–10.6 GHz 110 Mbps 10 m Low
Blue-tooth 2.4 GHz 3 Mbps 10–100 m Low
BLE 2.4 GHz 1 Mbps 200 m Ultra-low
Z-wave Sub-1 GHz 40 kbps 30 m Low
ANT 2.4 GHz 1 Mbps 10 m Ultra-low
Wavenis 868, 915, 433 MHz 100 kbps 1–4 km Ultra-low
Dash7 433 MHz 200 kbps 2 km Low
EnOcean 868, 315 MHz 125 kbps 300 m Ultra-low

low processing capability of sensor nodes, random APTEEN and TTDD are the few hierarchical routing
deployment of nodes, redundant data creation by nodes, protocols [57–59].
scalability, quality of service, fault tolerance, latency etc.
[4, 54]. Therefore, the routing protocols for WSN are dif- 6.3 Location based routing protocols
ferent from conventional routing also. The routing proto-
cols are categorized into three types for static node The sensor nodes are able to compute their location in this
deployment in agriculture application. location based routing protocols and then directly send data
to the BS. In topology based routing protocols, the prob-
6.1 Data centric routing protocols lems such as high end to end delay, more packet overhead
and less packet delivery ratio occurred in routing. There-
In Data centric protocol, all the sensor nodes have equal fore, to avoid these problem locations based routing pro-
power distribution, processing and communication capa- tocols are used. LAR, GLAR and ALERT are the few
bility. All the sensor nodes can take data from any target location based routing protocols [60, 61].
event or node. The data centric routing protocols are fur-
ther divided in three types (1) Traditional flooding mode
(2) Event-driven mode (3) Query-drive mode [55, 56]. In 7 Data frame formats and aggregation techniques
Traditional flooding, the sensor node broadcast the data and
sent to BS via router nodes whereas in event-driven mode, The data packet format consist of three parts (1) Packet
when a specific event happens, then only sensor node Header (2) Data/Payload (3) Packet Trailer. The packet
senses data and broadcast it. In query drive mode, the BS Header and trailer consist of Destination Node address and
broadcast a query and sends to specific node through router source Node address; network address, size of payload,
nodes by flooding. The sensor node replied to Base station data communicated error detection and correcting values.
through a specific path of router nodes. Flooding, Gossip- The data packet optimization is also one of important key
ing, SPIN, EAR, GBR are the examples of data centric research issue in WSNs. The following important features
protocols. are required to be considering during the design of data
frame format and establishment of communication
6.2 Hierarchical routing protocols scheme in between sensor nodes.

In Hierarchical routing protocol, the WSN is divided into 7.1 Packet payload length and transmission
the small cluster. Then, the selection of cluster head takes overhead
place and the cluster head take the sensed data from all
member sensor nodes in a cluster and then directly transmit Increasing the data packet size leads to more chances of bit
to the BS. The sensor node having high battery power is communication error, whereas small data packet size
selected as cluster head is to balance the power dissipation means a large ratio of overhead of data packet header will
of all nodes in a cluster and hence, the overall network occur and it increases more power consumption
lifetime increases. The data aggregation is take place at [62, 63].Therefore, the packet size should be optimum for
cluster head in case of Hierarchical routing protocols. The energy efficient WSN.
data packet format and data aggregation technique is
explained in next section. LEACH, PEGASIS, TEEN,

123
CSIT

7.2 Packet reception rate

Continuously transmission of data packets increases

channel congestion and also require more bandwidth for
communication. The packet transmission or reception rate
should be such that it covers all the data set of target
physical phenomena monitoring and also save the data
transmission power.

7.3 Dynamic packet length adaptation

In Dynamic Packet Length Adaptation (DPLA), the data

packet length changes according to channels traffic. If the
data transmission channel is empty, then large packets are
sent but if the channel is congested then smaller packets are
Fig. 4 Data aggregation at coordinator node level
sent. This process gives overall throughput and efficiency.
Therefore, DLPA is more suitable for developing data
The size of data packets and data transmission frequency
communication scheme and data packet formats.
also changes with the utilization of data aggregation
The data aggression techniques are also important factor
technique. The analysis of monitored parameters is done in
for designing and deploying WSNs. The eighty percent of
more accurate way. There are various types of data
battery power of sensor node is consumed during data
aggregation approach for routing protocols such as Spatio-
communication. Therefore, the BS has capability to pre-
temporal correlation, Scale Free Aggregation in Sensor
process the data packets before transmitting to base station
Networks and HEED etc.
[64]. The purpose of preprocessing is to remove the
redundant data, and it is also called as data aggregation.
In Fig. 3, the Nodes 1, 2, 3 and 4 are transmitting data
8 Node deployment strategies and network
packets to the coordinator node or cluster node. This
topologies
coordinator node or cluster head sends these data packets
as it is to the base station (BS). Therefore, the data packet
To cover the maximum deployment area with minimum
traffic increases at BS and also the end to end delay for data
number of sensors and reliable connectivity between the
packets increases. The alternate way of this, the collection
sensor nodes are the two major deployment challenges of
of data from nodes 1,2,3 and 4 and applied the aggregation
wireless sensor network. When the sensor nodes are ran-
functions such as average (AVG), maximum (MAX),
domly deployed in an area, then the sensor node does not
minimum (MIN) value function as required by application
cover the whole deployment region properly and some
and then send to the base station as shown in Fig. 4.
uncovered region will exist that is called Coverage Hole.
Similarly, when the data sensed from sensor nodes is
routed via router nodes towards BS, few router node
transfer many nodes data to base station and few router
nodes transfer less data to base station. Therefore, the
router nodes communicating more data, consumes more
power and its battery depleted fast. After depletion of
complete power, the sensor node gets into the off state and
called Dead node. The Dead node does not contribute to
the deployed WSN and the covered region of Dead node is
called Energy hole. The process of unbalanced power
consumption among sensor nodes is also a major challenge
in WSN. The problem is solved by using proper deploy-
ment techniques. The static node deployment technologies
are more preferred over mobile node deployment tech-
niques for agriculture application of wireless sensor net-
work because the static deployment of nodes does not harm
the crop and also the connectivity and communication
Fig. 3 Normal data flow b/w nodes and base station

123
 CSIT

tradeoffs between nodes are more effectively handled by 8.1.3 Random-Gaussian deployment strategy [69]
these techniques. Although maximum coverage is achieved
by the mobile node deployment techniques but the static In case of random uniform sensor nodes deployment, dif-
node deployment technologies are more preferred over ferent kind of coverage holes in a network can be created.
mobile node deployment techniques for agriculture appli- The existence of the coverage-hole in the monitored area
cation of wireless sensor network. The static deployment of means that every point in the target area is not being
nodes does not harm the crop growing process. The con- covered. This situation is similar to dead sensor node and
nectivity and communication tradeoffs between nodes are the energy-holes occur due to the many-to-one traffic pat-
also more effectively handled by static node deployment tern in WSNs. The sensor nodes which are close to sink
techniques. These static node deployment strategies are node carry more traffic loads irrespective of the data
classified mainly in two groups (1) Non-corona based transmission technique (multi- hop, direct, or cluster). This
sensor node deployment strategies, (2) Corona based sen- pattern results in more energy depletion within the area
sor node deployment strategies. closer to the sink and the formation of what is called
‘‘energy-holes’’. If once it is created, this process is irre-
8.1 Non-corona based sensor node deployment versible and the network operation of the remaining system
trategies is worthless because data can no longer be delivered to the
sink on the optimized paths. Therefore, a modification of
These deployments of sensor nodes for agriculture appli- the random uniform deployment strategies is needed for
cations are done in other than coronas shape and not con- mitigating the energy-hole problem. This is done by
trolled deployment techniques. increasing the number of nodes neighboring to the sink
which can decrease the energy-hole occurrence near the
8.1.1 Random-uniform deployment strategy [65–67] sink locality thereby achieving the non-corona-based
Gaussian deployment.
Manual deployments are not feasible in some application
of wireless sensor networks, for example, deployment in 8.2 Corona-based sensor node deployment
natural floods to check the intensity of floods. In such strategies
cases, nodes are deployed by helicopter, boats etc. The
deployment is done in random phase but try to make a In corona sensor node deployment strategy, multiple
uniform distance between nodes to cover the entire region. coronas or circles of different radius and having same
The nodes which are more closed to sink node, acts as center is created. The sensor nodes are deployed at the
router nodes and transmit more data packets of other nodes various points on the circumferences of coronas keeping
to sink node as compared to the other nodes which are far sink node at the center of corona. This corona will be
away from sink nodes. Therefore, the energy of nodes will encircled by the corona having also sensor nodes at its
be depleted and after the depletion of energy, these sensor circumference and the same sink node as center. Similarly
nodes will act as dead node. These dead nodes creates the entire sensor nodes are deployed at the circumferences
unbalanced energy utilization and reduces the overall net- of circles. The sensor nodes in inner most circle or corona
work lifetime. communicated more sensed data as router nodes because
all the sensed data from sensor nodes at outer corona are
8.1.2 Engineered-uniform deployment strategy [67, 68] communicated in multiple hop via inner corona’s sensor
node to base station placed at common center of all coro-
This type of deployment strategy is used for Controlled na’s. These Engineered corona-based sensor node deploy-
deployment. The sensor nodes are uniformly deployed (for ment strategy (ECSDS) are used in the application where
example in grid or triangular manner) in the application node placement is in control.
area. This technique provides less power dissipation among
sensor nodes. It also minimizes the data latency and 8.2.1 Engineered-Gaussian deployment strategy [70]
enhances the total throughput. This technique is better than
random uniform deployment but this is very difficult to In this deployment strategy, the sensor nodes in the
implement because geographical areas are always not in innermost corona are in large number as compared to the
uniform nature. outermost corona’s sensor node. The sensed data from
outer sensor nodes are sent to sink node in multiple hops
via in between various coronas’ sensor node. If one sensor
nodes dies in the inner corona, then the data is relayed via
neighbor node to died sensor node in the inner most corona

123
CSIT

and send to the base station. In this way, the energy whole consecutive terms (Ci, Ci?1) is same d as between the next
problem is solved but engineered Gaussian deployment two consecutive terms (Ci, Ci?1). In this strategy, the
strategy is hard to implement. geometric proportion is taken as 2 or 3.

8.2.2 ECSDS using minimum sensor node for full coverage 8.2.4 ECSDS using optimum number of nodes in each
[70, 71] corona [69, 71]

This deployment strategy is used with the objective of The equal number of sensor nodes are deployed in corona
covering maximum deployment region with the minimum in uniform coronas-based sensor node deployment strategy.
number of nodes. In this strategy, the required minimum The nodes in inner coronas transmitted or routes more
number of sensor network is calculated by length of cir- packets as compared to the nodes deployed in outer coro-
cumference of corona and communication range of sensor nas. Therefore, the nodes in energy coronas depleted more
node. energy and again many to one route condition occur.
Therefore, the author [69] proposed deployment strategy in
Ai ¼ pr2i
which the sensor node in every coronas is calculated with
Here, Ai and ri represents the circumference length of the following equation
corona and the radius of the ith corona respectively. X
R
Therefore, the sensor node density with full coverage in the Dj ¼ 2pri =Rc For 1  j\R
ith corona is calculated by the ratio of circumference length i¼j
number of sensor nodes in the ith corona is calculated
Here, Dj represents the number of nodes in the jth corona, ri
Di ¼ 2pri =Rc is radius of the ith corona whereas Rc is the communication
range of the node. This equation shows the ratio the ratio
The problem with this strategy is that it doesn’t solve
between two consecutive terms (Ci, Ci?1) is same as
energy hole problem.
between the next two consecutive terms (Ci, Ci?1).
Therefore, balance energy depletion is possible by this
8.2.3 ECSDS using arithmetic and geometric proportion
strategy. The disadvantage of this strategy is that a very
[69, 71]
dense deployment occurs in the inner coronas.
This strategy uses the arithmetic and geometric proportion.
8.2.5 ECSDS using routernodes [69, 72]
The difference between two adjacent numbers in arithmetic
proportion is constant. For example If C1 is the first
The router nodes are used to route the data packets in the
number of arithmetic proportion series and the difference
network and deployed in such a way as to avoid redundant
between two adjacent numbers is d, then the ith number of
data transmission towards the sink and distributed with the
the arithmetic progression is given by equation
sensor nodes in such a pattern as to maximize the balanced
Ci ¼ C1 þ ði  1Þd energy consumption. The authors proposed two types of
In this strategy, the sensor nodes equals to first term of nodes. The first type is used to sense the environment,
arithmetic progression is placed at the outer most corona whereas the second type; the router node is used to router
and the nodes required to place on the second last outer data packets in a network. In this strategy, full coverage
corona is calculated by adding the difference to the nodes can be achieves using a minimum number of sensor nodes
at first corona. Similarly, the nodes at third last outer cor- (as described in last section) and router node is used for
ona is equal to the nodes at second last outer corona plus data trans-mission to avoid redundant data transmission
the arithmetic progression and so on. For example if 7 towards the sink. The router nodes are distributed with the
nodes are in outer most corona and the arithmetic pro- sensor nodes in such a pattern as to achieve the balanced
gression is 6, then 13, 19, 25 nodes will be placed at second energy depletion. According to the radius of the corona, the
last outer, third last outer, fourth last outer corona respec- density of the sensor nodes increases from inner to outer.
tively. The density of the sensor nodes increases from the X
R

outermost corona to inner most corona. This strategy is also Dj ¼ 2pri =Rc For 1  j\R
i¼jþ1
solve energy hole problem and it is alternative to the ran-
dom-Gaussian deployment strategy. This strategy assumes The sensor nodes are distributed in the coronas and each
that all the nodes are having same initial energy and pro- corona’s sensor node can communicate with its neighbor-
cessing capability. This strategy also uses geometric ing corona’s nodes directly through one hop. As the inner
propagation. According to which the ratio between two part of the network has a heavier traffic load than the outer

123
 CSIT

Table 8 Various deployed WSN for agriculture applications

Author (year) Challenges addressed Node Operating Wireless Deployment WSN
system technology topology architecture

Seong-eun Yoo Multiple sink data collection A-node EOS IEEE Mesh Two tier
et al. [73] 802.15.4 topology
Durga et al. [74] Energy efficiency and cost effectiveness. MicaZ TinyOS RF Mesh Two tier
topology
Rodrigo et al. Low cost WSN Raspberry Pi Node JS Zigbee Mesh Two tier
[75] topology
Schin et al. [76] Energy efficiency Raspberry Pi, Debian WiFi Star Two tier
ardiuno
Yousef et al. [77] Automatic and manual control actuators Arduino Arduino Bluetooth star Single tier
IDE
George [78] Localization using flooding technique PIC16F877A Microchip RF Grid Two tier
based IDE
Taun et al. [79] 77.5 and 67.6 accuracy in monitoring Texas Texas IDE nfr24L01 RF star Two tier
temperature and humidity MSP430G2553
based
Zhen et al. [80] Valid data rate 95 percentage IRIS mote TinyOS RF star Two tier

part, so the density of the router nodes increases from the nodes in one square km agriculture field that are
outer to the inner coronas as given by equation. equipped with EC5 and GS3 for monitoring moisture,
temperature and pH and place in the fields according to
the variability in soil properties. The author claimed
9 WSN deployed for agriculture application that the mentioned WSN is cost effective and power
efficient solution for multiple crop applications.
A number of wireless sensor networks are already deployed 3. Rodrigo et al. [75] deployed a WSN for sugarcane crop
for agriculture applications. The users who deployed the in Sao Paulo, Brazil. The author used multiple sensors
wireless sensor network, faced various challenges such as to measure air humidity, air temperature, luminous
sensor calibration, signal condition, power consumption by intensity with soil temperature, humidity and electric
nodes, signal obstruction and low rate data packets conductivity etc. The author claimed that designed
throughput etc. The author studied few deployment of nodes gives soil properties and atmospheric data very
wireless sensor network for agriculture application and close to the data collected from center for weather
their contribution, challenges faced during deployment. station and climate studies organization.
4. Sachin et al. [76] designed a WSAN prototype named
1. Case Study-1(A2S): Seong-eun Yoo et al. [73]
FARMNET in Nirma University, India for one month.
deployed an agriculture automatic system (A2S) for
The nodes consist of five sensors moisture, sensor, rain
1 month in Korea to give services on laptop or PDA
sensor, GL5528 light sensor, DHT temperature and
devices of farmers. The A2S based WSN consist of 25
humidity sensor and PIR sensor. This research work
nodes, 1 actuator node and 3 sink nodes. The sink node
discusses the variation of data packets variation based
transmitted data over TCP/IP protocol to management
on temperature and humidity conditions. The author
subsystem having database server, application server
also claimed that using sleep awake technique the life
and web server. The system run uses Li-ion batteries
time of sensor node can be increased to 13 months.
and uses sleep-awake technique to make energy
5. Yousef et al. [77] designed a greenhouses smart system
efficient system. The node senses luminance, temper-
at gaza strip in Palestine. The author used NTC
ature and humidity for 20 s and again goes into sleep
temperature sensor and HR202 humidity sensor to
for 5 min. The author explained the system architec-
monitor temperature and humidity to automatically
ture but lacks explanation on low power routing and
control fan and water pump. The author also designed
communication schemes.
android application to control the fan and water pump
2. Durga et al. [74] deployed a WSN for multiple crop
manually.
such as maize, groundnut and green gram in India. The
WSN consists of 64 numbers of solar powered sensor

123
CSIT

6. George [78] designed a WSN to monitor soil moisture 6. Skubic M (2005) Assessing mobility and cognitive problems in
and soil temperature. The author found that the elders. In: AAAI 2005 fall symposium, workshop on caring
machines: AI in Eldercare, Arlington, Virginia
location of sensor nodes are found through flooding 7. Brunetti F, Moreno JC, Ruiz AF, Rocon E, Pons JL (2006) A new
method and after that the data packets send to sink platform based on IEEE802.15.4 wireless inertial sensors for
node through Modified source routing protocol. motion caption and assessment, EMBS’06. In: Proceedings of the
7. Taun et al. [79] designed and deployed a WSN for 28th annual international conference of the IEEE engineering in
medicine and biology society, New York, USA, pp 6497–6500
monitoring temperature humidity, light and soil mois- 8. Madhu A, Sreekumar A Wireless sensor network security in
ture in Vietnam for 6 months. The author showed the military application using unmanned vehicle. IOSR J Electron
data of temperature and humidity is collected by an Commun Eng (IOSR-JECE). e-ISSN: 2278-2834, p-ISSN:
accuracy of 77.5 and 67.6 percentages respectively. 2278-8735, pp 51–58
9. Ahmad I, Shah K, Ullah S (2016) Military applications using
8. Zhen et al. [80] designed a WSN for monitoring soil wireless sensor networks: a survey. Int J Eng Sci Comput 6:7039
volumetric water content, temperature, and bulk elec- 10. Tao H, Zhang H (2009) Forest monitoring application systems
trical conductivity at different depths such as 5 cm, based on wireless sensor networks. In: Third international sym-
15 cm, 30 cm and 60 cm in wheat fields. The valid posium on intelligent information technology application
workshops
data packet delivery rate is found to be near to 95% 11. Awang A, Suhaimi MH (2007) RIMBAMON: a forest monitor-
(Table 8). ing system using wireless sensor networks. In: International
conference on intelligent and advanced systems
12. Othmana MF, Shazali K (2012) Wireless sensor network appli-
cations: a study in environment monitoring system. Int Symp
10 Conclusion Robot Intell Sens Procedia Eng 41:1204–1210
13. Valverde J, Rosello V, Mujica G, Portilla J, Uriarte A, Riesgo T
(2012) Wireless sensor network for environmental monitoring:
There is much more research is required to make low cost, application in a coffee factory. Int J Distrib Sens Netw 8 Article
low power WSN solution to agriculture applications such ID 638067, 18. https://doi.org/10.1155/2012/638067
that it is affordable by farming community and they can 14. Bendigeri KY, Mallapur JD (2015) Advanced remote monitoring
of a crop in agriculture using wireless sensor network topologies.
increase their crop production and crop quality. Therefore,
Int J Electron Commun Eng Technol 6(9):30–38
to design, develop and deploy of an affordable wireless 15. Yoo SE; Kim JE; Kim T; Ahn S; Sung J; Kim D (2007) A2S:
sensor network for agriculture applications requires the automated agriculture system based on WSN, Consumer Elec-
complete knowledge of survey of sensors, sensor node tronics, 2007. Isce 2007. In: IEEE international symposium on,
June 2007, pp 1–5, 20–23
design, wireless technology, and operating system for
16. LópezRiquelme JA, Soto F, Suardı́az J, Sánchez P, Iborra A,
sensor nodes, routing protocols, data packet format, data Vera JA (2009) Wireless sensor networks for precision horti-
aggregation technique and deployment techniques. This culture in Southern Spain. Comput Electron Agric 68:25–35
paper covers all these important areas. This paper also 17. Warpe ST, Pippal RS (2016) A study of fertilizer distribution
system for agriculture using wireless sensor network. Int J
covers research areas and various problems faced by the
Comput Appl 147:2. https://doi.org/10.5120/ijca2016911005
users or researchers during the development and deploy- 18. Azfar S, Nadeem A, Basit A (2015) Pest detection and control
ment of WSN for agricultural application. techniques using wireless sensor network: a review. J Entomol
Zool Stud 3(2):92–99
19. Goumopoulos C, Oflynn B, Kemas A (2014) Automated zone
specific irrigation with wireless sensor/actuator network and
adaptable decision support. Comput Electron Agric 105:20–33
References 20. Khriji S, Houssaini D, Jmal MW, Viehweger C, Abid M, Kanoun
O (2014) Precision irrigation based on wireless sensor network
1. Sohraby K, Minoli D, Znati T (2007) Wireless sensor networks: IET Sci Meas Technol 8(3):98–106
technology, protocols, and applications. Wiley, New York 21. Nesasutha M, Valarmathi MI, Babu AS (2011) Energy efficient
2. Krishnamachari B (2005) An introduction to wireless sensor data transmission in automatic irrigation system using wireless
networks’’ tutorial. In: Presented at the second international sensor networks. Comput Electron Agric 78:215–221
conference on intelligent sensing and information processing 22. Martinović G, Simon J (2014) Greenhouse microclimatic envi-
(ICISIP), Chennai, India ronment controlled by a mobile measuring station. NJAS
3. Akyildiz IF, Su W, Sankarasubramaniam Y, Cayirci E (2002) Wageningen J Life Sci 70–71:61–70
Wireless sensor networks: a survey. Comput Netw 38:393–422 23. Mancuso M, Bustaffa F (2006) A wireless sensors network for
4. Singh SK, Singh MP, Singh DK (2010) Routing protocols in monitoring environmental variables in a tomato greenhouse
wireless sensor networks—a survey. Int J Comput Sci Eng Surv 24. Simbeye DS, Zhao J, Yang S (2014) Design and deployment of
(IJCSES) 1(2):63–83 wireless sensor networks for aquaculture monitoring and control
5. Kim H, Jarochowski B, Ryu D (2006) A proposal for a home- based on virtual instruments. Comput Electron Agric 102:31–42
based health monitoring system for them elderly or disabled, vol 25. Matese A, Di Gennaro S, Zaldei A, Genesio L, Vaccari F (2009)
4061/2006. Lecture notes in computer science. Springer, Berlin, A wireless sensor network for precision viticulture: the NAV
pp 473–479 system. Comput Electron Agric 69(1):51–58

123
 CSIT

26. Ojha T, Misra S, Raghuwanshi NS (2015) Wireless sensor net- 48. Subrahmanyam C, Reddy KC, Sattar SA (2011) Enhancement of
works for agriculture: the state-of-the-art in practice and future throughput for single hop WPAN’s using UWB-OFDM physical
challenges. Comput Electron Agric 118:66–84 layer. Int J Comput Sci Inf Secur 9(5):114
27. Rehman A, Abbasi AZ, Islam N, Shaikh ZA (2014) A review of 49. Gomez C, Oller J, Paradells J (2012) Overview and evaluation of
wireless sensors and networks’ applications in agriculture. bluetooth low energy: an emerging low-power wireless technol-
Comput Stand Interfaces 36:263–270 ogy, sensors, ISSN 1424-8220. Sensors 12:11734–11753. https://
28. Lopez JA, Javier A, Sanchez G, Soto F, Iborra A (2010) Felipe doi.org/10.3390/s1209117342012
Garcia-Sanchez, Joan Garcia-Haro. Design and validation of a 50. Hansen MS, Støa S (2006) Practical evaluation of IEEE 802.15.4/
wireless sensor network architecture for precision horticulture ZigBee medical sensor networks. Thesis, Master of Science in
applications, Springer, LLC Communication Technology, Deptt. Of ETCE, Norwegian
29. Beutel J, Romer K, Ringwald M, Woehrle M Deployment tech- University of Science and Technology June 2006
niques for sensor networks 51. Nixon M (2012) A comparison of WirelessHARTTM and
30. Gowrishankar S, Basavaraju TG, Manjaiah DH, Sarkar SK ISA100.11a, white paper: HCF_SPEC-xxx, Revision 1.0, Pre-
(2008) Issues in wireless sensor networks. In: Proceedings of the liminary A, Release Date: July 1, 2012
world congress on engineering 2008, 2008, London, UK 52. Wang G (2011) Comparison and evaluation of industrial wireless
31. http://www.newark.com/rf-transceivers-zigbee sensor network standards ISA100.11a and wireless HART.
32. Penella MT, Albesa J, Gasulla M (2009) Powering wireless Master of Science Thesis, Communication Engineering, Report
sensor nodes: primary batteries versus energy harvesting. In: No. EX036/2011, Department of Signals and Systems, CHAL-
I2MTC 2009—International instrumentation and measurement MERS UNIVERSITY OF TECHNOLOGY, Sweden
technology conference, Singapore, 5–7 May 2009 53. Baronti P, Pillai P, Chook V, Chessa S, Gotta A, Hu Y (2007)
33. Sarkar S, Shanavas HI, Bhavani V (2014) Energy scavenging Wireless sensor networks: a survey on the state of the art and the
technology and components in WSN. Int J Comput Netw Inf 802.15.4 and ZigBee standards. Comput Commun
Secur 3:44–51 30(7):1655–1695
34. Li Y, Shi R (2015) An intelligent solar energy-harvesting system 54. Wan J, Yuan D, Xu X A review of routing protocols in wireless
for wireless sensor networks. EURASIP J Wirel Commun Netw sensor networks
2015:179 55. Ghaffari Z, Jafari T, Shahraki HE (2013) 2013 International
35. Farooq M, Kunz T (2011) Operating systems for wireless sensor conference on communication systems and network technologies
networks: a survey. Sensors 11(12):5900–5930 56. Zia Q (2015) A survey of data-centric protocols for wireless
36. Hwang J, Shin C, Yoe H (2010) Study on an agricultural envi- sensor networks. J Comput Sci Syst Biol 8:127–131
ronment monitoring server system using wireless sensor net- 57. Xu DW, Gao J (2011) Comparison study to hierarchical routing
works. Sensors 10(12):11189–11211 protocols in wireless sensor networks. In: 2011 3rd International
37. Distributed and Object Systems Lab Department of Computer conference on environmental science and information application
Science and Engineering Indian Institute of Technology Madras. technology (ESIAT 2011), procedia environmental sciences, vol
Operating Systems for Wireless Sensor Networks: A Survey 10, pp 595–600
Technical Report 2007 58. PatelBhoomika D, PatelAshish D (2015) Hierarchical routing
38. Ojha T, Misra S, Raghuwanshi N (2017) Wireless sensor net- protocols in wireless sensor network. Comput Technol Appl
works for agriculture: the state-of-the-art in practice and future 6(5):847–851
challenges 59. Manap Z, Ali BM, Ng CK, Noordin NK, Sali A (2013) A review
39. Gautam KK, Gautam NR, Agrawal PC (2014) Memory required on hierarchical routing protocols for wireless sensor networks.
for wireless sensor nodes on the basis of characteristics and Wirels Pers Commun 72:1077–1104. https://doi.org/10.1007/
behaviour when using TinyOS. Int J Emerg Technol Adv Eng s11277-013-1056-5
4(1) 60. Iche AH, Dhage MR (2015) Location based routing protocols: a
40. Lynch JP, Loh KJ (2006) A summary review of wireless sensors survey. Int J Comput Appl 109(11):28–31
and sensor networks for structural health monitoring. Shock Vib 61. Kumar A, Shwe HY, Wong KJ, Chong PHJ (2017) Location-
Digest 38(2):91–128 based routing protocols for wireless sensor networks: a survey.
41. Nittel S (2009) A survey of geosensor networks: advances in Wirel Sens Netw 9:25–72
dynamic environmental monitoring. Sensors 9:5664–5678 62. Lade AM, Dorle SS (2015) Optimization of dynamic packet
42. Rodrigues JJPC, Neves PACS (2010) A survey on IP-based length in wireless sensor network. In: Proceedings of 25th IRF
wireless sensor network solutions. Int J Commun Syst international conference, 26th April 2015, Pune, India. ISBN:
23:963–981 978-93-85465-00-0
43. Rawat P, Singh KD, Chaouchi H, Bonnin JM (2013) Wireless 63. Leghari M, Abbasi S, Dhomeja LD (2013) Survey on packet size
networks: a survey on recent developments and potential syner- optimization techniques in wireless sensor networks
gies. Springer, New York 64. Maraiya K, Kant K, Gupta N (2011) Wireless sensor network: a
44. Buratti C, Conti A, Dardari D, Verdone R (2009) An overview on review on data aggregation. Int J Sci Eng Res 2(4). ISSN
wireless sensor networks technology and evolution. Sensors ISSN 2229-5518
1424–8220(9):6869–6896. https://doi.org/10.3390/s90906869 65. Khan NM, Khalid Z, Ahmed G (2009) GRAdient cost estab-
45. Decuir J (2010) Changing the way the world connects—bluetooth lishment (GRACE) for an energy-aware routing in wireless sen-
4.0: low energy. http://chapters.comsoc.org/vancouver/BTLER3. sor networks. EURASIP J Wirel Commun Netw 2009:1–2
pdf 66. Petrushin VA et al (2006) Multiple-sensor indoor surveillance
46. Villegas J Bluetooth low energy version 4.0, Helping create the system. In: The 3rd Canadian conference on computer and robot
‘‘internet of things’’ vision
47. Lee JS, Su YW, Shen CC (2007) A comparative study of wireless 67. Zaidi SAR et al (2009) On minimum cost coverage in wireless
protocols: bluetooth, UWB, ZigBee, and Wi-Fi. In: The 33rd sensor networks. In: Proceedings of the 43rd annual conference
annual conference of the IEEE industrial electronics society on information sciences and systems CISS
(IECON), Taipei, Taiwan 68. FangZ W (2010) Hybrid approximation for minimum-cost target
coverage in wireless sensor networks. Optim Lett 4(3):371–381

123
CSIT

69. Rahman AU et al (2016) Corona based deployment strategies in 75. Filev R, Tran ALH Precision agriculture using remote monitoring
wireless sensor network: a survey. J Netw Comput Appl systems in Brazil
64:176–193 76. Gajjar S, Kothari D, Upadhyay M, Dhingra V (2017) FARM-
70. Brooks A et al (2006) Implementation of an indoor active sensor NET: agriculture support system using wireless sensor and
network experimental robotics IX, Springer, Berlin, pp 397–406 actuator network, IEEE
71. Rahman AU, KhanHH OU (2011) Energy holes mitigation 77. Hamouda YEM, Elhabil BHY Precision agriculture for green-
techniques in sink’s proximity using sensor deployment in houses using a wireless sensor network. https://doi.org/10.1109/
wireless sensor networks. Res J Inf Technol 3(3):167–180 picict.2017.20
72. Rahman AU, Hasbullah H, Sama NS (2012) Impact of Gaussian 78. John GE (2016) A low cost wireless sensor network for precision
deployment strategies on the performance of wireless sensor agriculture, IEEE
network. In: Proceedings of the international conference on 79. Le TD, Tan DH et al Design and deploy a wireless sensor net-
computer and information science (ICCIS) work for precision agriculture. In: 2nd National foundation for
73. Yoo SE, Kim JE, Kim T, Ahn S, Sung J, Kim D (2007) A2S: science and technology conference on information and computer
automated agriculture system based on WSN, Consumer Elec- science
tronics, Isce 2007. In: IEEE international symposium on June 80. Li Z, Wang N, Franzen A, Taher P, Godsey C, Zhang H, Li X
2007, pp 1–5, 20–23 (2014) Practical deployment of an in-field soil property wireless
74. Durga P, Narayanan G, Gayathri B, Ramesh MV, Divya P (2017) sensor network. Comput Stand Interfaces 36:278–287
Modelling a smart agriculture system for multiple cropping using
wireless sensor networks, IEEE

123