Hindawi
Journal of Sensors
Volume 2022, Article ID 1628537, 21 pages
https://doi.org/10.1155/2022/1628537
Review Article
WSN Protocols and Security Challenges for Environmental
Monitoring Applications: A Survey
Kofi Sarpong Adu-Manu ,1 Felicia Engmann,2 Godwin Sarfo-Kantanka,1
Godwill Enchill Baiden,1 and Bernice Akusika Dulemordzi1
1Department of Computer Science, University of Ghana, Legon, Accra, Ghana
2School of Technology, Ghana Institute of Management and Public Administration, Accra, Ghana
Correspondence should be addressed to Kofi Sarpong Adu-Manu; ksadu-manu@ug.edu.gh
Received 25 March 2022; Revised 26 June 2022; Accepted 20 July 2022; Published 21 August 2022
Academic Editor: Zhenxing Zhang
Copyright © 2022 Kofi Sarpong Adu-Manu et al. This is an open access article distributed under the Creative Commons
Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work
is properly cited.
In recent years, communication technology has improved exponentially, partly owing to the locations and nature of the
deployment of sensor nodes. Wireless sensor networks (WSNs) comprise these sensor nodes and can provide real-time
physical and environmental measurements. The sensor nodes have limited power, which reduces their lifespan, especially when
placed in human-inaccessible locations. This paper reviews energy-efficient protocols for environmental monitoring
applications and energy harvesting-wireless sensor networks. The dynamic deployment and communication challenges
associated with environmental monitoring applications (EMAs) make this paper take into account the WSN protocol stack,
focusing on the physical layer, network layer (routing), and medium access control (MAC). The paper will delve deeper into
the security challenges of deploying sensor nodes for environmental monitoring applications (EMAs). The paper further
describes scientific approaches that churn out innovative and engineering applications that must be followed to improve
environmental monitoring applications.
1. Introduction forming a wireless sensor network (WSN) is used to sense
the changes in the environment and wirelessly communicate
Environmental monitoring is aimed at identifying the status the sensed data to a base station for processing.
of a changing environment using data collection tools. Data WSNs are distributed systems comprised of several
collection tools employed for determining the status of the nodes and base stations (BS) to monitor physical environ-
changing environment rely on data acquisition systems mental conditions [5]. Each sensor node contains a wireless
(DAS). DAS uses measurement devices that are designed radio transceiver to communicate with other nodes and the
to allow for gathering representative samples. The samples BS. Sensor nodes communicate via insecure radio channels
depend on the device’s intrusiveness, sampling accuracy, prone to interference and fading [6].
and sample storage [1–3]. These measurement devices have WSNs are used in several applications and deployed on
varying degrees of impact depending on the application land (terrestrial), underground, and underwater. Recently,
and the data gathering method. For example, the sensitivity they have been used in multimedia applications and mobile
of the measured physical value to external influences may applications such as animal tracking, air quality monitoring,
vary depending on the type of application. The traditional forest fire warning, and flood detection, among the terres-
way of collecting the status of the changing environment trial applications of WSNs. WSNs monitor soil conditions
has proven ineffective, having data reliability, delays, and underground, particularly for agricultural and mining pur-
security challenges [4]. Hence, in recent times, the introduc- poses. Underwater applications include freshwater quality,
tion of technological devices such as sensor nodes capable of climate change, ocean monitoring for aquatic life, and
2 Journal of Sensors
assessing coral reef alterations underwater—WSNs track essential feature of the WSN is the maintenance of a secured
events such as video, audio, and imaging in multimedia network [18]. The different applications in EMAs require
applications. They are utilised in real-time monitoring of that their security solutions are provided with the objectives
hazardous compounds, target tracking, and rescue and and application needs in mind.
search applications in mobile applications [7, 8]. Despite The paper explores how WSNs for EMAs are affected by
the capabilities presented by WSNs, they are also associated new security vulnerabilities at the physical, network, and
with challenges in various application domains. Table 1 data link layers. To appreciate the security challenges in
summarises some of these challenges and their associated EMAs, the paper discusses the WSN protocol stack,
recommended solutions when WSNs are utilised in the dif- energy-efficient protocols, and energy harvesting protocols
ferent application domains. suitable for environmental monitoring applications. There
Researchers have studied the design of protocols that can are further discussions on the design requirements, simula-
address these challenges in a variety of application domains, tion environments for EMA protocol designs, quality-of-
taking into account the challenges of the application [9–13]. service requirements, and network topology requirements.
The protocols are intended to operate at the sensor network Finally, the paper presents the security issues in WSN for
protocol stack (physical, link, network, and transport layers). EMAs, detailing the threats at the nodal and network levels
Protocol design is aimed at helping with data collection, and their prevention and countermeasures.
aggregation, processing, and communication to maximize
network lifetime and uptime [14]. Protocols govern the 2. WSN Protocol Stack
operation of sensor nodes in a sensor network, specify the
requirements and guidelines for operation, and ensure that WSNs are distinguished by their adaptable network topol-
the sensor network fulfils its intended use [8]. A wide range ogy, which various networking protocols enable at multiple
of protocols is designed in communication networks to over- layers. Designing efficient and reliable communication pro-
come the challenges discussed in Table 1 and improve net- tocols for WSNs for EMAs is difficult due to different con-
work performance. These protocols extend the operationality straints on the sensor platform and the different
of the network to perform some intended function. environments’ lack of certainty and dynamics [19]. An anal-
In WSNs, data packets are transmitted to the BS in two ysis of the design requirements of protocols for WSNs for
ways: single-hop or multihop. The node sends the generated EMAs is provided in this section. Physical, data link, net-
packet directly to the base station in a single hop. In con- work, transport, and application are the five core layers of
trast, in multihop, source nodes send packets to the BS via the WSN protocol stack, of which three (physical, data link,
a multipath, with each node in the path forwarding the and network) will be explored in this section.
received (or, in the case of the source node, generated) The physical, data link, network, transport, and applica-
packet to another node until the packet reaches the BS [6]. tion layers of the wireless sensor network protocol stack are
WSNs face some challenges, which include energy consump- similar to the classic open system interconnection (OSI) par-
tion, sensor node deployment, routing algorithms, energy adigm. In each of these layers, several activities are under-
efficiency, cluster-head (CH) selection, resilience, etc. taken. The physical layer handles frequency selection,
Researchers have developed several routing and medium carrier frequency production, signal detection, modulation,
access control (MAC) protocols to address these issues. and data encryption. The data link layer handles the multi-
Optimisation algorithms have also been designed to deter- plexing of data streams, data frame detection, medium
mine the best path between the transmitter and receiver access, and error correction. In a communication network,
nodes to save energy and extend the network lifetime. it enables reliable point-to-point and point-to-multipoint
Designing efficient communications and network proto- connections. The data given by the transport layer is routed
cols for WSNs for EMAs manages sensor node operation in by the network layer [8]. In WSNs, the network layer design
their deployable environment and achieves successful sensor must consider energy consumption, communication, aggre-
node objectives [15]. The variability of EMAs and their gation, and other factors. The transport layer aids in data-
peculiar characteristics should be considered when design- flow maintenance and may be necessary if WSNs are
ing efficient protocols suitable for gathering accurate and accessed over the Internet or other external networks.
timely data from sensors in the field. A unique channel is Depending on the sensing duties, different forms of applica-
frequently used to communicate between wireless sensor tion software can be set up and employed at the application
nodes. The channel has the property that only one node layer. The following section discusses the routing protocols
can send a message at any time. As a result, shared channel for managing the increasing energy requirement of sensor
access necessitates the implementation of a MAC protocol nodes to monitor environmental applications such as photo-
among the sensor nodes [16]. The MAC protocol is aimed synthesis, soil carbon flux, and soil salinity.
at managing access to the shared wireless medium to meet
the underlying application’s performance requirements. 2.1. Routing Protocols. Routing protocols in EMAs determine
On the other hand, routing protocols are essential during optimum dynamic routes for exchanging information
data transmission to create optimum paths from sensed data between sensor nodes depending on the application-specific
to be transmitted from source to destination [17]. Maintain- requirements. These application-specific requirements of the
ing optimum paths in WSNs for EMAs is critical to maxi- routing protocols include throughput, capacity, coverage, net-
mizing the nodes’ lifespan and data throughput. Another work performance, end-to-end delay, real-time delay, and
Journal of Sensors 3
Table 1: Challenges in WSN application domains.
Application
Challenges Recommended solutions
domain
Energy minimisation techniques
Design of efficient routing protocols
Limited power supply
Use of energy harvesting
Terrestrial Data redundancy
Implement an effective node deployment
Data latency
strategy (e.g., multihop)
Short transmission range
Difficulty in deployment
High signal losses
Design of efficient data communication
Underground High levels of attenuation
protocols
Higher energy cost
Difficulty in battery replacement
Limited bandwidth of the acoustic channels
Low link quality of acoustic channels
Significant propagation delays due to the speed of sound
Design of efficient underwater data
Underwater Energy limitation with sensor nodes
communication precools
Creation of the Doppler effect due to the relative motion of the
transmitter and the receiver
Nodes are susceptible to corrosion
High bandwidth demand
High energy consumption
In-network processing
Provisioning of quality of service due to variable delays
Filtering
Multimedia and channel capacity
Design of efficient compression techniques
High demand for data processing
Design of cross-layer protocols
Challenges with cross-layer design
Data compression
Unreliable data transfer Design of efficient mobility-aware protocols
Uncontrolled mobility results in poor network performance for message exchanges
Mobile
Localization and coverage Design of efficient mobility-aware power
Unstable contact detection due to shorter contact durations management protocols
collision. They are helpful, especially in mobile applications (1) Proactive Protocols. Proactive protocols are table-driven
such as battlefields, disaster zones, animal tracking, water protocols. Proactive routing techniques store the routes
monitoring (freshwater/ocean), and air quality. Some EMAs without any route matching. These protocols keep track of
are time-sensitive, while others have bandwidth constraints. every sensor node’s connectivity to other nodes in the sensor
The sensor nodes are either powered by fixed-energy batteries network at any given time. Proactive routing protocols
or rechargeable batteries. Battery-powered WSNs for EMAs enable every sensor node to send periodic updates and have
are expected to operate for at least two (2) years without failure a clear and consistent network topology view. Proactive
[20]. However, most applications of EMAs are deployed in routing keeps a fresh list of destinations and associated paths
areas where changing or recharging batteries is a difficult task. by frequently disseminating routing tables over the sensor
Generally, routing protocols are classified into route discovery network. Examples of proactive routing protocols include
protocols, network organization protocols, and protocol oper- destination sequence vector (DSDV), optimised link state
ations, as shown in Figure 1. routing (OLSR), and wireless routing protocol (WRP). Other
There are three route discovery protocols: reactive, pro- proactive protocols proposed in the literature include source
active, and hybrid. There are four network organization pro- tree adaptive routing (STAR); global state routing (GSR);
tocols: flat-based, hierarchical-based, location-based, and cluster head gateway, switch routing (CGSR); and fisheye
data-centric. Based on their operation and how the protocols state routing (FSR) [23–25]. The popular examples discussed
work in a deployable environment, routing protocols are in this paper are DSDV and OLSR.
classified as negotiation-based, multipath-based, query-
based, quality-of-service-based, and coherent-based proto- DSDV is a proactive table-driven protocol that uses the
cols [21, 22]. The following sections cover each category in Bellman-Ford routing technique. Through sequence num-
detail, providing examples for each. bers, DSDV ensures loop-free operation. Every mobile node
in the network keeps a routing table that lists all of the net-
2.1.1. Discovery Protocols. Discovery routing protocols are work’s possible destinations and the number of hops
discussed in this section. Developments of discovery routing required to reach each one. A sequence number is assigned
protocols are presented in Figure 2. to each entry by the destination node. The sequence
4 Journal of Sensors
Routing protocols in where the source/destination pair with time. It is also suit-
WSN able for applications that do not allow for long delays and
operate in dense networks. However, it is disadvantageous
in mobile networks where the frequent change in topology
increases the exchange of control overhead messages,
Discovery protocols Network organizationprotocols Protocol operation increasing the cost of energy and memory use [27].
Reactive Flat-based Negotiation-based (2) Reactive Protocols. Reactive protocols are designed, so
sensor nodes within the network initiate a route discovery
Proactive Hierarchical- Multipath-based process only when a route to a destination node is required.based The routes are subjected to computation because the sensor
Hybrid Location-based Query-based node computes its route based on demand [20]. The routes
that have been established are created and maintained in
Quality of service- two stages: route discovery and route maintenance. TheData-centric based route discovery occurs on-demand by flooding the network
with route request (RRQ) packets.When a route is discovered,
Coherent-based the destination responds with a route reply (RREP) containing
the route information traversed by the RREQ. Ad hoc on-
Figure 1: Classification of routing protocols. demand distance vector (AODV), ad hoc on-demand multi-
path distance vector (AOMDV), associativity-based routing
(ABR), sequence number-based secure routing (SNSR), signal
Discovery protocols stability routing (SSR), and dynamic source routing (DSR) are
popular examples of reactive protocols [25, 26, 28]. More pop-
ular reactive protocols are discussed here. They are best suit-
Reactive Proactive Hybrid able for applications where the sensors are mobile and have
their routes changing frequently.
AODV DSDV ZRP
AODV reduces control traffic by generating path
AOMDV OLSR TORA requests on-demand and constructs its routes without prior
ABR WRP TZRP knowledge using an RREQ and RREP request loop between
the source and destination nodes. RREQ packets are broad-
SNSR FSR HWMP cast as part of the route-building process. When a node
CGSR receives an RREQ packet, it checks its record to see if it
has previously received an RREQ packet. If a packet is not
STAR logged when received, the node resends it [28]. Ad hoc on-
demand multipath distance vector (AOMDV) is a modified
GSR version of the AODV routing protocol. AOMDV is a proto-
col with multiple routes, disjoint paths, and no loops from
Figure 2: Discovery routing protocols. source to destination. AOMDV uses hop count during the
advertisement. AOMDV is designed to keep multiple dis-
numbers let the mobile nodes distinguish between old and joint loop-free paths in route discovery. The AOMDV pro-
new routes, preventing routing loops [26]. tocol reduces energy consumption and packet loss in
OLSR is a proactive protocol where frequent topology WSNs. Compared to AODV, it performs better in applica-
changes cause flooding of the network with erratic updates. tions with high traffic loads. The essence of the AOMDV
The OLSR reduces the maximum time interval for updating protocol is guaranteeing that multiple pathways discovered
control overhead messages to optimise the link-state infor- are loop-free and discontinuous and identifying such paths
mation. The critical concept of OLSR is the multipoint relays fast utilising a flood-based route discovery method.
(MPRs), which are selected nodes that forward broadcast AOMDV route update rules are critical for maintaining
packets during the flooding of the network. The MPR loop-freeness and applying disjointness properties locally at
reduces the control overhead of flooding since only MPRs each node [29].
generate link-state information. Therefore, the flooding of When a source node has packets to send to a destination
the network with control overhead messages reduces when node, dynamic source routing (DSR) checks its cache first to
OLSR is employed [27]. OLSR only transmits partial link- see if it has a route to that destination. A new packet header
state information between MPRs and their MPR selectors is constructed to include the destination’s path if a route is
for route calculations. This approach reduces the computa- available. Suppose no route is found in the cache. In that
tional cost of route calculations and hence minimal delay case, the node initiates the discovery process by sending an
in establishing routes. This protocol is best for networks RREQ broadcast packet with the source and destination
where larger subsets of the network communicate with other node identifiers of the route to be discovered and a unique
larger subsets of the same network. It is suitable for networks identifier for the RREQ. Associativity-based routing (ABR)
Journal of Sensors 5
is an on-demand routing protocol initiated by the source Network organization
node. ABR employs both point-to-point and broadcast rout- protocols
ing techniques. In ABR, the destination node decides on a
route based on the property of “associativity.” The chosen
route is used, and all other routes are discarded because
the decision is based on the property of “associativity,”
resulting in long-lived routes. ABR is divided into three Hierarchical Flat/Data-centric Location-based
stages (route discovery, route reconstruction, and route
deletion). LEACH DD GAF
GEAR
(3) Hybrid Protocols. Hybrid routing methods combine the HEED SPIN
benefits of proactive and reactive routing strategies and their TEEN FG MECN
drawbacks. In some scenarios, hybrid routing is preferable.
These hybrid protocols can achieve consistency across pro- PEGASIS COUGAR SPAN
active and reactive protocols. Because nodes must keep
high-level topological information, these protocols have a APTEEN RR SPEED
drawback in that they require more memory and power.
Examples of hybrid routing protocols include zone routing BCDCP CADR LAR
protocol (ZRP), two-zone routing protocol (TZRP), tempo-
CCS ACQUIRE DECA
rarily ordered routing algorithm (TORA), and hybrid wire-
less mesh protocol (HWMP). A description of the most EECS EAR EELIR
popular hybrid routing protocol is provided [25, 26]. ZRP
was created to work in a zoned network. The node in ZRP SOP ALERT
maintains routes to all the routing zone’s destinations. Intra-
zone routing protocol (IARP), interzone routing protocol Figure 3: Network organization protocols.
(IERP), and bordercast resolution protocol are the three sub-
protocols of ZRP (BRP). When the route is within the zone,
intrazone routing protocol (IZRP) is used, and when it is efficient gathering in sensor information systems (PEGASIS),
outside the zone, interzone routing protocol (IERP) is and hybrid energy-efficient distributed (HEED) clustering.
used [26]. Also, we have protocols such as the threshold-sensitive
energy-efficient sensor network (TEEN), base station-
2.1.2. Network Organization Protocols. Sensor nodes collabo- controlled dynamic clustering protocol (BCDCP), concentric
rate to complete the tasks of the app for which they were clustering scheme (CCS), and adaptive threshold sensitive
deployed. Sensor nodes with extreme energy constraints energy-efficient protocol (APTEEN). Finally, hierarchical
have limited computation, storage, and communication routing protocols also designed for large-scale networks
capabilities. Some network-level protocols are designed to include energy-efficient clustering scheme (EECS), self-
address the limited computing, storage, and communication organizing protocol (SOP), energy-balanced chain-cluster
capabilities of WSNs. These include flat-based routing pro- routing protocol (EBCRP), chain-based hierarchical routing
tocols, hierarchical protocols, data-centric protocols, and protocol (CHIRON), energy-aware data aggregation tree
location-based protocols. In flat-based protocols, all nodes (EADAT), balanced aggregation tree routing (BATR),
are treated as peers. Hierarchical protocols are a type of pro- power-efficient data gathering and aggregation protocol
tocol that is based on clusters. Data-centric protocols are (PEDAP), and enhanced tree routing (ETR).
intended to disseminate information throughout the net-
work, whereas location-based protocols are aimed at Hierarchical routing protocols are divided into four
addressing sensor network concerns by utilising node posi- types: cluster-based (LEACH, TEEN, HEED, APTEEN),
tion information. In each of these categories, several routing chain-based (PEGASIS, CHIRON, EBCRP, CCS), tree-
algorithms have been proposed. The following sections pro- based (EADAT, BATR, ETR, PEDAP), and grid-based
vide brief descriptions of the various network organization (SOP, BCDCP). Figures 4–7 illustrate the four hierarchical
protocols, as illustrated in Figure 3. routing protocols employed for environmental monitoring
applications depending on the requirements.
(1) Hierarchical Protocols. Hierarchical routing protocols We describe two popular hierarchical protocols (LEACH
enable large-scale network deployment through self- and PEGASIS) suitable for environmental monitoring appli-
organization capabilities. The primary goal of the hierarchical cations due to their flexibility in handling the routing infor-
routing protocol is to keep sensor nodes’ energy consumption mation. LEACH (low-energy adaptive clustering hierarchy)
as low as possible by performing data aggregation and fusion is a routing algorithm that collects and delivers data to a sink
to reduce the amount of data transmitted to the base station node or base station. LEACH is the most widely used
[30]. In Figure 3, a list of hierarchical routing protocols is energy-efficient cluster-based hierarchical routing protocol
designed for use in WSNs. Details of these protocols are low for EMAs in WSNs. It employs localized coordination to
energy adaptive clustering hierarchy (LEACH), power- enable scalability and robustness in dynamic networks.
6 Journal of Sensors
In PEGASIS, nodes are connected in a chain, with one
node connected to the opposite and the last node connected
H to the base station. The advantage of PEGASIS is that it
reduces the amount of data transferred between contiguous
S nodes by moving information obtained by detector nodes
from one node to the next until the last node transmits the
data to the final terminal and restricts the number of trans-
H missions and receptions between nodes. Two main impedi-
ments challenge the use of PEGASIS in environmental
monitoring applications. First, the base station acts as a sin-
gle leader, which can cause bottlenecks. Secondly, PEGASIS
Figure 4: Chain-based. generates excessive disruption for remote nodes in the chain.
These two challenges can be solved by enabling concurrent
transmission between neighbouring nodes, which reduces
the latency incurred during connection with the base station.
Other ideas include using signal coding or only permitting
data to be transmitted simultaneously between far apart
nodes [31].
S
(2) Flat/Data-Centric Protocols. It is not practical to issue
global identifiers to each node in many sensor network
applications. The lack of global identifiers and the random
deployment of sensor nodes make selecting a set of sensor
nodes to interrogate problematic. As a result, data is often
Figure 5: Tree-based. transported to a deployment location with a high level of
redundancy. The solution to these challenges led to develop-
ing a data-centric routing protocol [21]. Examples of flat/
data-centric routing protocols include gradient-based rout-
ing (GBR), COUGAR, and constrained anisotropic diffusion
S H routing (CADR). The others are directed diffusion (DD),
flooding and gossiping (FG), energy-aware routing (EAR),
H active query forwarding in sensor networks (ACQUIRE),
and rumor routing (RR). Two popular examples (SPIN
and directed diffusion) of data-centric protocols discussed
H are provided in this paper.
SPIN stands for “sensor protocols for information via
negotiation,” a data-centric, negotiation-based family of
Figure 6: Grid-based. WSN information dissemination protocols. The fundamen-
tal goal of these protocols is to efficiently distribute data col-
lected by source nodes to the rest of the network’s sensor
nodes. Simple protocols such as flooding and gossiping are
frequently recommended to achieve information distribu-
H tion in WSNs. Each node in the network must send a copy
of the data packet to its neighbours until the information
S reaches all nodes in the network [6, 9, 22]. Directed diffusion
H is a data-centric routing approach. There must be a list of
attribute-value pairs (interval, name of objects, duration,
H and area). If a node is interested in specific data, the sink
broadcasts the inquiry to its neighbours. The interest is
cached by the nodes that get it to use later. The data is com-
pared to the values in the cached interests. Within the inter-
Figure 7: Cluster-based. est, there are additional gradient fields. The gradient is a
reply link to the neighbour who sent the interest [22, 32].
LEACH incorporates data fusion into the routing protocol This data-centric strategy is used to acquire and deliver
to reduce the information transmitted to the base station information because it is energy efficient, saving energy,
and organizes the network into clusters using a hierarchical and extending the network’s lifespan. The directed diffusion
approach. A cluster head is responsible for each cluster. routing protocol does not require addressing because all
The cluster head carries out multifaceted tasks [10, 30, 31]. communication is node-to-node [22].
Journal of Sensors 7
(3) Location-Based Protocols. A source node sends a packet in directed diffusion. As a result, GEAR saves more energy
to a destination node in a location-based routing (LBR) pro- than directed diffusion.
tocol, and the destination node appends to each packet by Geographical adaptive fidelity (GAF) is a location-based
the source node. Packets received by intermediate nodes routing technology considering energy consumption. Sensor
along the path to the destination node use the location infor- networks will find it informative and applicable. Algorithms
mation in the packet and deliver it to the next one-hop control the network’s nodes, turning them on and off to save
neighbours. They are geographically closest to the destina- energy while maintaining high fidelity. For the covered
tion. The operation is repeated until the data packets arrive space, GAF develops a virtual grid. Each node receives its
at the destination node. Because of the locality, location- current location through GPS, associated with a virtual grid
based routing necessitates the minor state in each node. point. Regarding packet routing costs, nodes belonging to
Because advertisements of routing tables, as in traditional the same grid are regarded as the same.
routing protocols, are not required, it has a low communica-
tion overhead. As a result, route creation and maintenance 2.1.3. Protocol Operation Routing. Protocol operation is
are no longer needed with location-based routing. another way to categorise routing protocols in WSNs. This
Location-based routing is used in more extensive networks category can be divided into five major sections based on
where node positions change and the destination node’s protocol processes. Some subcategories include multipath,
location is known to the source [33]. LBR makes use of node query, negotiation, quality-of-service, and coherent-based
location information to improve efficiency and scalability. In routing protocols. Examples of these protocols are presented
LBR, each node in the network must be aware of its location in Figure 8.
information, obtained via GPS or other techniques. Also, In WSNs, data processing is critical, and several strate-
each node must be informed of the position of its one-hop gies are applied to lower processing costs to save energy. In
neighbour node, and the source must know where the desti- this case, the obtained data can be handled logically. In
nation node is located. coherent data processing-based protocols, the sensor nodes
perform the bare minimum of processing locally [34]. Data
LBR generally requires accurate location information, is sent to sink nodes after minimal data processing in coher-
which can be acquired through some sort of localization ent routing. Tasks like time-stamping and duplicate elimina-
technique. Because location information is critical for many tion are often included in the minimum processing.
EMA WSNs, such as animal tracking and forest fire moni- Coherent processing is frequently used to produce energy-
toring, it is expected that each wireless sensor node in the efficient routing [35]. Examples of protocols include the
network will be equipped with some form of localization multiple winner algorithm (MWA) and single winner algo-
device. Location-based routing is categorised into GPS- rithm (SWA).
based and non-GPS-based protocols. The sensor nodes The protocol may use various methods to transmit data
may be deployed as mobile or static in each category. LBR from source to destination in multipath routing. Multiple
employs greedy algorithms to forward packets from the routes improve network fault tolerance while significantly
source node to the destination node. It is critical in preserv- increasing energy consumption and protocol overhead. An
ing the energy of the sensor nodes. Examples of location- extension of the method evaluates only the path with the
based routing protocols are illustrated in Figure 3. Details highest energy nodes. When a better path is found, the pro-
of these protocols are geographic adaptive fidelity (GAF), tocol changes to it. The network’s reliability can be strength-
geographical and energy-aware routing (GAER), minimum ened using the multipath routing protocol in severely
energy communication network (MECN), sensor protocols unstable conditions. A large packet can be broken down into
for information via negotiation (SPAN), SPEED (a real- smaller chunks and sent via numerous channels. A message
time routing protocol for sensor networks), location-based can still be produced even if one of the subpackets is lost
energy-efficient intersection routing (EELIR), anonymous owing to connection issues [16]. Some multipath routing
location-based efficient routing protocol (ALERT), energy- protocols include energy-constrained multipath routing
efficient geographic forwarding algorithm for wireless ad (ECMP) and multiconstrained quality-of-service multipath
hoc and sensor network (DECA), improved hybrid routing (MCMP).
location-based ad hoc routing protocol (IHLAR), location- A node initiates a query and propagates it across the net-
based routing protocol (LBRP), selective bordercast in ZRP work in query-based routing. The query is sent to each node;
(SBZRP), and location-based selective bordercast in ZRP only the node with matching data receives responses. Rather
(LBZRP). In this paper, GAF and GEAR are described. than disseminating the queries throughout the network, the
The geographic and energy-aware routing (GEAR) algo- node could send them down a random path and wait for a
rithm employs geographic information to route queries to response. If none of the other nodes responds, the node
the most relevant places. In many location-aware systems, can broadcast it to the entire network [36]. Quality-of-
notably sensor networks, disseminating information to a service (QoS-based) routing ensures that a wireless network
geographic region is valuable. GEAR uses an energy-aware will deliver the expected results. Latency (delay), throughput,
and geographically informed neighbour selection algorithm error rate, and energy consumption are a few quality-of-
to route a packet to the target region instead of flooding service parameters in WSNs, which differentiates traffic
the query or packet across the whole network. On the con- flows by treating packets differently based on their nature.
trary, interest is inundated throughout the entire network The quality-of-service routing protocol is also in charge of
8 Journal of Sensors
Protocol operation
routing
Coherent Multipath Query-based Negotiation-based QoS-based
SWE MMSPEED DD SPAN SPEED
MWE SPIN SPIN SAR SAR
MCMP COUGAR DD MMSPEED
ECMP ACQUIR SPIN DACR
DD RR ROL
Figure 8: Protocol operation routing protocols.
prioritising data flows to maintain a predefined performance 3.1. Energy-Efficient Physical Layer Protocols. The Zigbee
level. When delivering data in applications where parame- protocol suite is the preferred physical layer protocol for
ters such as delay, resources, and bandwidth are critical, environmental monitoring. It is widely considered due to
the routing protocol must maintain the quality and specifi- its low duty cycle for energy-efficient implementations. The
cations of the required parameter. The quality-of-service communication range of Zigbee is within 100m, with a
routing protocol balances energy consumption and other reduced communication range of 30m for indoor applica-
metrics [37]. Sequential assignment routing (SAR) is a typi- tions [38]. Using precision agriculture and farming applica-
cal QoS-based routing protocol. Flooding and gossiping in tions, Zigbee enabled supervised irrigation, water quality
WSNs can cause the network to implode; thus, a single node management, pesticide and fertiliser control, and total field
may get several copies of data. Negotiation-based protocols surveillance [39]. Other implementations of Zigbee include
are intended to prevent duplicate packets from propagating. investigating signal propagation and strength distribution
In wireless sensor networks (WSNs), sensor nodes exchange characteristics of wireless sensor networks in date palm
negotiation messages to send redundant data to the next orchards with application-efficient specific parameters such
node. It reduces network congestion and conserves energy. as signal strength on the spacing between nodes, the leaf
Routing protocol categories (route discovery protocols, density, and antenna height of the base station. Cattle
network organization protocols, and protocol operation) grazing-field implementation [40, 41], greenhouse [42, 43],
are beneficial in specific application areas. These protocols livestock monitoring [44], and smart beehives [45] are some
provide myriad challenges and energy requirements when current implementations using Zigbee.
used for specialised purposes. Table 2 highlights most of
the application areas, challenges, and energy requirements 3.1.1. LoRa (Long Range Radio). LoRa is a digital wireless
for the various forms of routing protocols. data communication technology spread-spectrum radio
modulation (EP2763321 from 2013 and US7791415 from
2008) derived from chirp spread spectrum (CSS) technology.
3. Energy-Efficient Protocols It uses the unlicensed free radio frequency bands 169MHz,
433MHz, and 868MHz in Europe and 915MHz in the US.
Data communication is a critical challenge in many WSN LoRa is presented in two ways: the physical layer shown
applications. As a result, it is advantageous to construct as LoRa on devices and the upper layers presented as the
and develop easily accessed resources to give data in long range wide area network (LoRaWAN) [46]. LoRa pro-
WSNs. The number and position of nodes in WSNs make vides a low-cost, secure bidirectional mobile communication
recharging or replacing batteries impossible. As a result, for IoTs and other machine-to-machine (M2M) for high
energy consumption is a universal design concern for throughput systems. These systems typically run on 3G or
WSNs. Researchers have concentrated on reducing energy Ethernet networks, WiFi, or cellular technologies. The LoRa
dissipation at all stages of system design, from hardware to has been implemented in smart bee monitoring farms to
protocols to algorithms. As a result, it is critical for the ensure communication of the colony activities [47]. Green-
network to carefully define the parameters of the protocols houses measure air temperature, humidity, light intensity,
in the network stack to achieve the required energy effi- and soil moisture [48]. LoRa’s advantage is its scalability to
ciency and meet the quality-of-service requirements. The several nodes and adaptability to interference and noise.
following is a discussion of some energy-efficient protocols Another implementation involves real-time monitoring
used in the WSN architecture at the physical, link, and using a multisensor combination module (MSCM) and
network layers. LoRa. In the implementation, wetland parameters include
Journal of Sensors 9
Table 2: Application areas, challenges, and energy requirements of routing protocols.
Categories of routing
Application areas Challenges Energy requirements
protocols
Continuous energy source, size of
Energy harvesting sources,
Reactive Animal tracking, habitat monitoring sensor nodes, frequent change in node
rechargeable batteries
position, latency, security
Fixed energy batteries,
Significant overhead for storing
Route Infrastructure systems, hospice care, capacitors, rechargeable
Proactive routing tables in sensor nodes, not
discovery weather forecasting, military surveillance batteries. Mobile energy
suitable for mobile applications
transmitters
Mobile energy
Require more memory and power,
Hybrid Sewage and gully pot monitoring transmitters, energy
mobility
harvesting nodes
Animal tracking, military surveillance, crop
Network Size of sensor nodes, energy source, Fixed energy batteries,
farming, forest fires, disasters. Habitat
organization number of nodes, modulation schemes rechargeable batteries
monitoring, disaster monitoring
Smart farming, structural monitoring, Energy consuming, topology control, Supercapacitors,
Protocol earthquake monitoring, water quality maintaining energy-neutral operation rechargeable batteries,
operations monitoring, air quality monitoring, climate (ENO), mobility of nodes, antenna energy harvesting sources
change monitoring sensitivity (primarily solar)
water temperature, pH, conductivity, turbidity, dissolved various mobile network operators. Sigfox provides a
oxygen, and water level [11]. A recent work by authors con- software-based communication platform that allows the com-
siders the inability of the elderly and disabled farmers to plexities of network computing to be carried out on the cloud
monitor and oversee agricultural tasks on farms. In their instead of the devices. Their unique setup with mobile cellular
implementation, LoRa radios were used to run several soft- networks provides an extensive network of global devices
ware applications from speech processing software, voice transmitting data without setting up or maintaining a connec-
recognition, and other assistive artificial intelligence (AI) tion [52]. Therefore, the Sigfox setup removes network bottle-
features. However, LoRA may only transmit data not necks such as signalling overheads, providing a robust and
exceeding a few kbps (not suitable for video capturing) over optimised network protocol. It uses the radio frequency chan-
longer distances (not exceeding 3 km). The work by the nel on the 100Hz band, with a data rate of 100bps. It is based
authors included smartphones and tablets intercepting voice on the ultra narrow band (UNB). It uses a differential binary
commands generated to be transmitted to a remote agent’s phase shift keying (DBPSK) modulation scheme, with scat-
device, usually at the farm’s end. Applications on the agent’s tered nodes accessing the network using random frequency
device are designed using raspberry pi and other program- time division multiple access (RFTDMA) [53].
ming facilities like the Arduino Uno unit with a WiFi radio Narrowband Internet of Things (NB-IoT) is developed
module. Countries with existing implementations include by the 3rd Generation Partnership Project (3GPP) to scale
smart gardens, vineyard crops in Spain, irrigation on kiwi up WSN applications and make them more dependable.
and corn farms in Italy, and banana farming in Columbia. NB-IoT uses an unlicensed frequency band in long-term
evolution (LTE) and consumes more power than LoRaWAN
3.1.2. Bluetooth. Bluetooth is another short-range communi- due to its constant need for synchronisation. It implements
cation protocol in environmental monitoring applications. It orthogonal frequency-division multiplexing (OFDM) and
is usually implemented between movable portable devices frequency division multiple access (FDMA), increasing its
such as laptops over 10m distances. Due to its availability power consumption. However, its applications require low
on most handheld devices, it can be used for multilevel agri- latency and high data rates.
cultural applications. Agricultural implementations include Other less implemented physical layer protocols in envi-
weather information, soil moisture, temperature, and irriga- ronmental monitoring include Bluetooth LE and mobile cel-
tion. Bluetooth is beneficial due to its low energy consump- lular technologies such as GPRS, 3G, and 4G. Due to their
tion, wide availability of devices, and ease of use. It has also high energy consumption, mobile technologies are limited
been used in other monitoring environments, such as disas- in IoTs and environmental monitoring applications. In con-
ter prediction and monitoring [49], food storage systems trast, the high data rate they provide may not be helpful in
[50], and environmental monitoring in small spaces [51]. many applications [54–61]. Table 3 summarises the charac-
teristics of energy-efficient physical layer protocols for envi-
3.1.3. Sigfox. Sigfox is the name of the company and the low ronmental monitoring applications.
power wireless area network (LPWAN) operator. It is one
protocol gaining popularity in IoT and environmental mon- 3.2. Energy-Efficient Routing Protocols. Energy-efficient rout-
itoring applications in several countries, cooperating with ing is aimed at increasing the network lifetime by
10 Journal of Sensors
Table 3: Characteristics of energy-efficient physical layer protocols.
Power Security
Protocol Standard Applications Limitations Range Topology
consumption capability
IEEE 802.15.1 Fire and disaster
(no more). The Low monitoring, food Line of sight between Point-to-
Bluetooth 128-bit AES 1-10m
current standard (100mW) storage, greenhouse communicating devices point
is Bluetooth 5.3 monitoring
Temperature and
fire monitoring in
underground mines,
Low Line of sight between 1-75m and
Zigbee IEEE 802.15.4 128-bit AES agriculture, cattle Mesh
(36.9mW) communicating devices more
grazing, bee hives,
and greenhouse
monitoring
Greenhouse monitoring,
2-5 km
Low voice detection Scalability, the maximum Star of
LoRa IEEE 802.15.4g AES CCm 128B (urban) and
(100mW) techniques in farming, data rate of 250 kbps stars
15 km (rural)
irrigation monitoring
Air pollution, earthquake
detection, temperature High power consumption,
IEEE High
WiFi 128-bit AES and humidity sensing, security, long access time 100m Star
802.11a,b,g,n (835MW)
humiture and optical (13.74 s)
sensing
Mobility of nodes can
Key generation, Water quality prediction, only be deployed in a
message air quality monitoring, few countries, and
Sigfox Sigfox Low 100 km Star
encryption and optimum farming communication is limited
sequence parameters from the base station
to the nodes
Water quality monitoring,
High power, high
NB-IoT 3GPP Medium NSA/AES 256 air pollution, industrial Star
data rate
environment
considering the energy cost of the communication path. The require adaptive routing to ensure efficient real-time data
routing protocols generally are categorised based on cluster- delivery. The routing protocols for such emergency services
ing, the mode of the protocol’s functionality, the node’s par- include the real-time routing protocol with proposed load
ticipation, and the network structure [20]. The general distribution [65] and improvement [66].
challenges for environmental monitoring include security,
scalability, node deployment strategies, connectivity, and 3.3. Energy-Efficient MAC Protocols. Using energy-efficient
coverage. In mitigating these challenges, energy consump- MAC protocols in WSN is aimed at meeting the challenges
tion is integral in implementing these solutions. of general WSNs such as latency, throughput and fairness,
The challenges of WSN applications are diversified based channel utilisation, and scalability. Latency refers to the time
on the application areas, which influences the routing proto- it takes from a source node to reach the destination node. Its
cols and quality-of-service parameters. For example, under- requirements in WSN are application dependent. The
water communications in underwater wireless sensor throughput is also a measure of successful data received by
networks (UWSN) use acoustic signals for propagation, the destination node. Fairness here refers to the ability of
unlike radio frequency (RF) signals that are used in terres- the destination node to receive a fair amount of data from
trial wireless sensor networks (TWSN). These acoustic sig- each sensor node in the network. Therefore, the MAC proto-
nals are at lower magnitudes of 1500m/s, five times lower cols ensure optimal results, with energy efficiency integral to
than in TWSNs. Quality parameters such as delay are critical its operations. To ensure energy efficiency, MAC protocols
in acoustic mediums, which may be negligible for terrestrial must overcome the challenges of multiple transmissions in
environments. A centralised routing protocol, proposed by the networks, such as energy losses due to control overheads,
authors in [62], is based on a full-duplex communication idle listening, collisions, and overhearing.
that implements network management (gateway managers) Environmental monitoring applications require MAC
and routing agents. These agents periodically probe the net- protocols that adapt to mission-critical applications. These
work for the statuses of the nodes to allow the gateways to applications require a quick response time. The applica-
determine a priori the optimum path between neighbouring tions may be deployed in inaccessible human environ-
nodes to avoid congestion for high traffic applications. Other ments. Hence, efficient energy management systems are
protocols implemented in [63, 64] were based on water needed to prolong the network lifetime. Examples include
depth and temperature. Fire hazard monitoring applications volcanic eruption-prone areas, surveillance applications,
Journal of Sensors 11
environmental monitoring and control systems, and health The network setup consists of MICAz WSN nodes deployed
care systems. The quality-of-service (QoS) parameters identi- at fixed locations in a mapped-out area on a smart farm.
fied for MAC protocol implementation in any set of mission- Sensor nodes are connected to a gateway/base station pow-
critical system is similar and may be applied to other systems. ered by the main supply. Sensor nodes transmit their data
ADMC-MAC is a mission-critical MAC protocol using to relays through multihop transmits for forwarding to the
regression techniques to decide the duty cycle of sensor gateway. The network supports bidirectional communica-
nodes [67]. tion from the gateway directly with the deployed sensor
nodes.
4. Energy Harvesting-Based Protocols For a comprehensive review of up-to-date energy har-
vesting WSNs (EH-WSN), the reader may refer to [14] for
The slow development in battery technologies makes energy references. An energy harvesting–based clustering protocol
harvesting (EH) a viable solution to the energy challenge in is proposed by authors in [71] to improve network stability
environmental monitoring protocols. Using energy harvest- and efficiency. Their approach includes the selection of the
ing in environmental monitoring hampers the destruction cluster head for each cluster based on the node’s energy
of environmental pollution from the disposal of batteries. level, the amount of energy harvested, and the number of
Energy harvesting is increasingly becoming important in its neighbours. The leach-based clustering protocol uses a
IoT implementations due to the massive number of sensor lower threshold of 0.1 J and a higher threshold of 1 J to com-
nodes deployed for some applications, primarily working pete in the cluster head selection. Other protocols imple-
for long periods without human interferences. menting prediction models that may be difficult in WSN
Energy challenges like traditional WSN applications may are made possible due to energy harvesting. Weather
not constrain energy harvesting due to energy availability. forecasting-based applications include implementations in
Some identified EH sources in environmental monitoring IproEnergy [72], autoregressive (AR) models [73], ARIMA
include sunlight, vibration, sound, wind, thermal, electro- models [74], and reinforcement learning applications using
magnetic waves, and body heat and movement. Hence, some Q-learning [75].
applications apply architectures such as mobile, stationary, Mechanical kinetic energy harvesting converts mechani-
or hybrid. They may also be classified as single-tier, multi- cal motions and vibration harnessed from the environment
tier, or homogeneous [68]. Sources like solar energy, com- into electrical energy. This energy source is self-power sens-
monly used in many applications, use the photovoltaic ing, convenient, energy-saving, sustainable, and eco-friendly.
energy harvesting approach. In photovoltaic energy harvest- It is applied in aerospace, biomedical engineering, and mili-
ing, solar light energy is converted into electrical energy to tary and environmental monitoring applications [76]. How-
recharge the batteries of WSN nodes. The highest energy ever, mechanical energy has few implementations in WSN
conversion for outdoor solar energy harvesting was due to low conversion efficiency, low power output, and con-
15mW/cm2 and an efficiency of about 30%. It is the pre- version being time-dependent and may even cause damage
ferred renewable energy source. It is primarily available in to the device. It requires unique modulation of the energy
the environment where sensor nodes are deployed and have source to harness the energy from the ambient environment.
the highest energy conversion efficiency. Solar energy har- Applications depending on mechanical include structural
vesting does not pollute the environment but requires little condition monitoring, smart devices and cities, biomedical
preservation and may be stored for several years. Solar har- and wearable devices, and machine monitoring. Biomass,
vesters may be deployed in applications that include agricul- which converts organic materials from plants and animals,
ture (farms), forest monitoring, greenhouse monitoring, and has seen applications in WSNs where devices deployed in
animal monitoring. Recently, commercial applications have unreachable locations harvest energy from organic ambient
been deployed by Crossbow Inc. USA, using Mote View sources. Examples of sources include corn, soyabeans,
2.0 for measuring parameters such as temperature, humid- woody plants, paper, cotton, food, wood wastes, and animal
ity, pressure, acceleration, and light. The application uses and human sewage. Authors in [77] harvested energy from
the Zigbee protocol for communication with an expandable switchgrass, while authors in [78] mentioned the extraction
distance of 100m to 1.5 km. An example of implementation of Xenopus oocytes from female frogs to power capacitors
was discussed in the survey by authors in [69] which in their application.
reviewed the literature on solar energy harvesting WSN
(SHE-WSN). 5. Environmental Monitoring Application
A smart agricultural application simulated by the Design Requirements
authors in [70] is intended to extend the network lifetime
using solar energy harvesting. Using similar tools described Environmental monitoring applications are different in
by Sharma et al., the lifetime of nodes deployed on a farm many ways. Some of the applications are dynamically
was extended from 5.75 days when energy harvesting was deployed, and others are statically deployed. In dynamic
not implemented to 115.75 days. With an increase in deployments, the nodes are primarily mobile, and in static
throughput of 160 kbits/s from 100 kbits/s, the duty cycle deployments, the nodes are positioned at various points in
of these sensor nodes could be adjusted upwards to more the environment. In each of these deployments, the design
than 25% (which is a message transmitted every 4 s) since requirements may differ. Also, environmental monitoring
energy is no longer the main challenge of these networks. applications are characterised by energy efficiency, network
12 Journal of Sensors
complexity, scalability, data transmission, bandwidth, and WSN topologies such as basic peer-to-peer, linear, star,
processing storage. These characteristics are critical design tree, a cluster tree, or mesh are used in EMAs to monitor
criteria considered in EMAs. Different topologies associated and set up the sensor network. Sensor nodes may be
with the various applications determine the routing and deployed remotely at an area of interest to acquire data wire-
MAC protocols employed for the energy-constrained sensor lessly using one WSN topology to monitor agriculture, the
node. Efficient routing protocols schedule routes efficiently environment, or water resources. Topologies can be built
to minimise the amount of energy consumed by the nodes either statically or dynamically. Several sensor nodes can
to prolong the lifetime of the sensor network. With efficient freely move in dynamic topologies in some application
routing, gains are made in data communication. MAC layer domains (for example, water quality monitoring and ocean-
protocol requirements such as duty cycling, slot scheduling, ography). The topology can self-organize when individual
time synchronisation, node prioritisation, and efficient sensor nodes fail or deplete their energy. In dynamic topol-
channel utilisation improve performance and increase net- ogies, when new nodes are added to the network after some
work lifespan. MAC protocol implementation in WMAs nodes fail, the identity of the new node enters the network
achieves high reliability, effective scheduling, and efficient without changing the topology [81].
time synchronisation in EMAs [79]. In WSNs for EMAs, the transmission power of the nodes
determines the network topology, which directly impacts
network performance [82]. The network topology in EMAs
5.1. Quality-of-Service Requirements. Quality of service can be designed to accommodate mobile or static sensor
(QoS) guarantees that the network provides the expected nodes and a sink. Mobile nodes in a sensor network are typ-
results. In WSNs for EMAs, there are essential parameters ically designed to deal with the dynamic network topology
that the application may be designed to achieve. These required for node mobility. Because node mobility in WSNs
QoS parameters include throughput, delay, packet delivery makes the topology dynamic, the communication protocol
ratio, and energy consumption. For example, in forest fire becomes more complicated, necessitating more processing
monitoring, the nodes are deployed to access real-time data resources and energy. The network topology may generate
and remotely acquire data from forest zones for decision- minimal heavy traffic load depending on the network size
makers to make decisions based on the data received from and application domain (animal tracking, water quality,
the sensor nodes. The sensor nodes report any unusual tem- oceanography, fire monitoring). The sensor node in an
perature, smoke, oxygen levels, and humidity that may be EMA must be able to reconfigure itself for different network
collected to mitigate forest fires. Collecting and studying topologies (such as star and tree topology). All sensor nodes
event data from the forest employ sensors that may trigger directly communicate with the sink node in the star topol-
alarms, calculate and track humidity levels and temperature ogy. Most sensor nodes in a tree or mesh communicate with
variations, and detect smoke patterns. From such sensor net- neighbours to maintain connectivity with the sink node [84].
works, throughput and data reliability are essential. Also, EMAs’ network topology requirements should consider
measuring delay is crucial because the different sensors the number of nodes and the distance between neighbour
may collect the data in specified periods. This means that nodes to determine node placement density, network diam-
delays in sensor data reporting may cause forest fires. The eter, and coverage. The minimum number of data relays
sensor nodes’ energy efficiency will optimise network perfor- between sensor nodes is used when packets are transmitted
mance and uptime. For example, in forest fires, sensor nodes between nodes in the network [85]. Researchers developing
must operate in the forest for a long time. Therefore, effi- applications to monitor the environment must consider net-
cient energy consumption or low energy consumption of work topology requirements because network functionality
the sensor nodes’ energy will make the node live longer. and stability are critical to meeting application objectives.
Energy-aware sensor network architecture (SNA) techniques To determine the type of topology, it is also necessary to
are paramount in such networks [80]. Therefore, protocols consider the traffic load density. Network topologies should
designed at the various layers must be energy-aware to be well-designed to be fault-tolerant, reconfigurable, energy-
enhance the overall network lifetime. efficient, and scalable [86].
5.2. Topology Requirements. WSN network topology is the 6. WSN Security Approaches for EMAs
physical or logical placement or arrangement of sensor
nodes in an observed area of interest. It also includes how WSN application security is a complex problem to solve. It
sensor nodes communicate within the network to collect requires finding the best method to maximize network per-
data from the environment and transmit it to a base station formance while dealing with complex restrictions. The goal
via a sink node [81]. Topology requirements in WSN for is to find a reasonable balance between efficiency, energy
EMAs vary depending on the environment and application. efficiency, and routing protocol design. Wireless sensor net-
Network topologies are typically application-specific, and works (WSNs) are being utilised for various applications,
the structure serves as an essential foundation in EMAs including environmental monitoring applications (EMAs),
[82]. Network topologies must be designed to balance the like a new computer and network infrastructure platform,
energy consumed by sensor nodes while the network lifetime and operational security is a significant problem. Security
is maximized. The topology of a network can directly impact issues in WSN for EMAs are more concerned with the reli-
its performance [83]. ability of the network, positioning of the nodes
Journal of Sensors 13
Table 4: Security attacks at the layer(s) on the OSI model.
Network layers Security attacks Effects of network/node Defense mechanism
Message corruption, DoS, disrupt Increases packet reception time
Application Firewalls and the use of antiviruses
or intercept confidential data Reduces data reliability
Degradation of energy
Fake packets are injected Provision of authentication
Transport Session hijacking, DoS
Data integrity, availability, and Reducing packet response rate
authenticity are affected
Continuous request to send packets
Wormhole, blackhole, sinkhole, Encryption
floods the network
Sybil, selective forwarding, spoofing, Authorisation
Network Generation of false messages
altered or replayed routing Probing
Creation of routing loops
information, internet smurf Monitoring firewalls
Causes selective forwarding
Retransmission of data
Traffic analysis, HELLO flood, Error correction
The decreased energy level of the sensor node
Data link monitoring, channel exhaustion, Use of virtual private networks
Nodes may miss the transmission
802.11 disruptions (MAC) Reprogramming senor devices
Degradation in network performance
Corrupts or sends a large number of packets
Jamming, node malfunctioning, Hiding
Sensor node physically tampered
node destruction, tampering Region mapping
Physical Addition of other sensor nodes
(direct node attack), DoS, radio Spread-spectrum techniques
Network services get stopped (data collection)
interference, interception such as DSSS and FHSS
Network energy is exploited
(localization), and the topological characteristics that may The network layer handles packet routing from one node to
affect data collection, data processing, and increased delay another. Several attacks at this layer in wireless sensor net-
in the network [18, 87]. works take advantage of the routing mechanism. In WSNs,
In WSNs, a security attack is defined as any attempt to the transport layer ensures that the entire message is deliv-
expose, steal, manipulate, modify, or obtain unauthorised ered in the exact sequence. At this level, some attacks can
access to information in the sensor network [87]. A wireless be made. New connection requests are produced repeatedly
sensor network is highly vulnerable to attacks because the in floating till the resources reach their maximum capacity.
sensor nodes are physically unprotected. In WSN, there are To develop secure WSNs, application designers must
two types of attacks: active and passive. The attackers just consider security objectives (integrity, availability, authorisa-
monitor the communication channel in a passive attack, tion, authentication, and confidentiality). Data integrity
but they modify the data stream in an active attack. Passive ensures that data is not tampered with by unauthorised
attacks include eavesdropping, node dysfunction, node parties. Data availability ensures that authorised system
destruction, and traffic analysis. Active attacks occur when stakeholders have immediate and unrestricted access to the
an adversary attempts to disrupt the operation of the net- system’s and network’s resources. The most critical aspect
work under attack. Active attacks include denial-of-service of security is availability. The authority to access the data
(DoS), sinkhole attacks, flooding, and Sybil. At various tiers is primarily concerned with maintaining the confidentiality
of the network, many attacks exist. The attackers aim to of the information. Authentication requires genuine access
exploit the nodes directly at the physical layer. Jamming is to sensor nodes and the network by application designers
a denial-of-service attack in which the victim’s computer’s throughout implementation. The sensor nodes must provide
functions are disrupted. There are various attacks at various authorised data to stakeholders while operating the network
layers of the network. Table 4 presents security attacks at the [5]. Encryption techniques are used to protect the privacy of
layers on the OSI model. At the physical layer, attackers system resources and operations. As a result, data confiden-
attempt to exploit the nodes physically. Physical layer threats tiality is conditional on a certain level of information [88].
put data availability, integrity, and confidentiality at risk.
The data link layer is responsible for framing, addressing, 6.1. Types of Security Threats in EMAs. Security targets in
error correction, and flow control. At this layer, a variety WSNs for EMAs apply to the wireless medium between sen-
of attacks may occur. When two separate nodes send data sor nodes, and the sensor devices are susceptible to attacks
on the same frequency, the packet’s data varies by a small and vulnerabilities. The environmental monitoring applica-
amount. As a result, the packet becomes unusable, and the tions are designed to achieve specific objectives. Some appli-
data is deleted. An attacker may cause these collisions [5, cations such as water quality, air quality, animal tracking,
87]. and forest fire are life-threatening, endangering the lives of
For communication, many networks employ a two-way humans and animals. They require essential security mea-
handshake. An attacker can make a constant request to send sures to ensure that the nodes remain operational 100% of
the packet. The network link may be flooded as a result. the time. Security attacks of the sensor nodes may cause
Interrogation is the term used to describe this type of attack. the nodes to malfunction, disrupt the network, and
14 Journal of Sensors
Table 5: Description of various security threats.
Attack Description
This attack overflows the network with traffic, utilises more bandwidth, and causes services or
resources to become unavailable to the user. This type of attack is most common in all WSN
Denial of service (DoS)
applications. It is aimed at interrupting the system, making it unavailable or unusable, which attacks
its availability and efficiency.
In this situation, the attacker installs a malicious node which masquerades as a genuine node, which
Selective forwarding may refuse to forward or simply discard specific messages. This jeopardises the availability and
integrity of system data.
This exploits the network by adding a node that collects all data as though it were the base station.
Sinkhole
This threatens the confidentiality of the system where applicable.
In this case, the compromised node assumes several identities (creating the illusion of being present
Sybil attack in multiple locations) to connect with a large number of nodes. This poses a risk to the network’s
confidentiality.
The attack records the information to a different location before transmitting it or a portion of it.
Wormhole
This is a threat to data integrity.
The HELLO packet is transmitted to the nodes, and the attacked device may be misidentified as a
HELLO flood neighbour seeking to communicate with it. Its objective is to use network resources. The availability
of network resources is jeopardised as a result.
This is a more direct attack whereby the attacker can wreak havoc on the network by sending falsified
Spoofed, altered, or replayed
error messages or establishing routing loops. This compromises the network’s integrity and
routing information
availability.
With this attack, the targeted node is forced to transmit the reply to the malicious node by providing
Blackhole
fake route information. This disrupts availability.
This attack seeks to either disable the node so that a compromised node with the same identifier may
Node destruction
take its place or prohibit it from gathering data.
Monitor and eavesdropping The goal of this exploit is to obtain network information. This jeopardises confidentiality.
The goal is to capture and analyse messages to derive information from communication patterns. Its
Traffic analysis threat stems from its capacity to function even when data are encrypted and compromised
confidentiality.
Node replication This attack duplicates nodes and uses them to set up multiple attacks.
This attack takes three significant steps: it accepts a message, alters it to make it incomprehensible,
Message corruption and finally sends it to its intended recipient—the integrity and availability of data are being
compromised.
Jamming disrupts the sensor nodes’ RF signals, rendering them inoperable and compromising
Jamming
availability.
Node malfunctioning creates erroneous data that might jeopardise the integrity of the cluster heads’
Node malfunctioning
data-aggregation process.
introduce delays in data communication, halting the moni- imal operational energy, limited memory capacity, and con-
toring process and degrading the overall quality of service. strained computing abilities. These nodes can sense, record,
Table 5 illustrates the attacks and descriptions of the various and monitor environmental conditions. The sensor node
security threats. and the sensor network have a lot of practical uses, but they
Also, when the sensor nodes and network are attacked, are also challenged with several deployment problems of
the attackers may reprogram the nodes to transmit false data which security is paramount. The node is deployed in hostile
readings, which may endanger the lives of humans and environments, making them physically vulnerable to attacks
animals. (adversaries and natural disasters). When sensor nodes are
Attackers may also cause the sensor nodes to continu- deployed in hostile environments, the kind of topology
ously send packets until they are exhausted, drain their formed because of the node depleting its energy or being
energy, and die. Table 6 depicts some common types of damaged by animals is unclear. The section examines differ-
attacks and the environmental monitoring application they ent security approaches suitable for specific operations in
affect, which an attacker may exploit to render the sensor WSNs for EMAs. Some data gathered from sensor nodes
network inefficient affecting its intended use. in EMAs may be sensitive. Hence, there is a need to safe-
guard the node, network and sensed data to prevent any
6.2. Node and Network Security Approaches in attack or tampering. The data obtained from the sensor
Environmental Monitoring Applications. This paper has nodes and transferred to a base station require levels of
established that wireless sensor networks (WSNs) have min- authorisation and authentication to access it. These security
Journal of Sensors 15
Table 6: Threats in environmental applications.
Animal River/ocean Forest fire Air Precision Earth/
Threat Active Passive
tracking monitoring detection quality agriculture landslide
Denial of service × × × × × × ×
Selective forwarding × × × × × ×
Sinkhole × × × × × ×
Sybil × × × ×
Wormhole × × × × × ×
HELLO flood × × × × × × ×
Spoofed, altered, or replayed × × × ×
routing information
Blackhole × × × × ×
Node destruction × × × × × × ×
Monitor and eavesdropping × × × × ×
Traffic analysis × × ×
Node replication × × × × × ×
Message corruption × × × × × × ×
Jamming × × × × × × ×
Node malfunction × × × × × × ×
Data
aggregator
Base
station
Sensor
nodes
Target region
Figure 9: Data aggregation in a wireless sensor network (adapted from [92]).
levels will prevent unauthorised data from getting into the prevent false alarms. Data can be aggregated within a node,
hands of the wrong users [89]. across a network, at the sink, or the base station. All of these
data collection points require some level of security. Where
sensitive data is being collected, as in military deployments,
6.2.1. Secure Data Aggregation. The sensor nodes in EMAs encryption may be needed to prevent data from being read
collect enormous amounts of data. The sensed data must and compromised by adversaries [93].
be aggregated to avoid overload at the base station. As illus-
trated in Figure 9, data aggregation is how sensed data is
processed and combined en route by intermediary sensor 6.2.2. Access Control. Access control is a security strategy
nodes. In a typical WSN, many sensor nodes gather that governs who has access to and uses resources in a com-
application-specific data from the environment and transfer puting environment. It is a fundamental security concept
it to a centralized base station for processing, analysing, and used to mitigate risk. The two types of access control are
use by the application [90]. The basic strategy is to analyse physical and logical access control. Physical access control
data from multiple sensor nodes as transmitted collectively. is used to restrict access to physical IT assets. Logical access
The plain text sensed by the deployable environment’s nodes control governs all communication systems, files, and data
may be encrypted before being transferred to the base sta- connections. Nodes are fixed or mounted on endangered
tion. For example, water quality data such as pH or dissolved species (for example, turtles) in animal tracking applications
oxygen sensed by nodes deployed in a river is encrypted after to track their movements. Physical access to the nodes is
sensing and transmitted to the base station. Before data impossible while the animals are in motion. Still, logically,
aggregation operations can be performed on the data, the the algorithms or protocols are accessible to researchers
encrypted water quality data is decrypted at the base station who track the position of the animals and all data residing
[91, 92]. In real-time applications, data aggregation occurs to on the node. This could be applied to the battlefield, forest
16 Journal of Sensors
fire, and water resource monitoring. Farmers can have phys- keys and (2) the protection of the keys through key manage-
ical and logical control over agricultural monitoring and use ment (secure key generation, storage, distribution, use, and
the necessary security at each level to provide the restrictions destruction). Poor algorithms embedded in a robust key
required for the device and the data [94]. management framework are just as likely as good algorithms
embedded in a weak key management framework to fail.
6.2.3. Secure Routing. Nodes in WSNs for environmental Data encryption, authentication, and digital signatures are
monitoring communicate to transfer data from one node tasks that require using cryptographic algorithms [96].
to another. Data exchange between nodes necessitates a
secure pathway to avoid data compromise. Because nodes 6.3. Challenges of WSNs for EMAs. Wireless sensor net-
in a WSN route packets over multiple paths, secure routing works’ intense constraints and demanding application envi-
techniques must be used to secure the route. Secure routing ronments make computer security for such systems more
is a set of transport and network security controls that apply complicated than traditional networks. Some challenges, like
to routing protocols and individual nodes. However, to most systems, have a negative impact on the system’s oper-
establish the network architecture (reactive, proactive, or ations and resources. WSNs are not immune to challenges,
hybrid), nodes must communicate with their peers to use particularly in terms of security. Wireless communication
one of the routing protocols. Safe routing and secure data is less secure due to its distributed nature, allowing easy
forwarding are two approaches to routing security used in interception. An adversary can easily intercept, change, or
EMAs. Secure routing necessitates node collaboration to rerun any message. An attacker can easily intercept legal
share accurate routing information and keep the network packets and inject malicious ones [97].
connected. Secure data forwarding requires the protection
of data packets from corruption, dropping, and manipula- 6.3.1. Constrained Resources. A constrained resource is one
tion by an untrusted source. in which you have a finite supply. Sensor nodes in WSNs
are subject to resource constraints such as limited power,
6.2.4. Secure Localization. The positional information of restricted communication bandwidth, limited processing
nodes mounted on animals and floating in water sources is capability, and limited storage capacity due to their varying
critical in mobile-based sensor deployments such as animal sizes and cost. Wireless channels in the network are typically
tracking and water quality monitoring. For example, in the unstable when sensor data is shared among nodes or trans-
case of animal tracking, the location of endangered species ferred for data processing, resulting in unpredictable trans-
such as turtles, elephants, and others will provide informa- mission delays and packet losses. Energy constraints, for
tion. Localization computes the location or position of sen- example, require strategies to save energy to enable nodes
sor nodes in WSNs. Because of the dynamic nature of the to operate for a longer time since sensor nodes mainly rely
applications, WSN deployment has shifted from static to on batteries. The deployment type also affects the communi-
dynamic (mobile). Because such networks are dynamic, cation and connectivity between nodes in the network. This
node localization is also variable, making the process critical challenge also requires optimal node deployment schemes to
in WSNs. Knowledge of a network entity’s physical location ensure effective communication among the nodes in the net-
is useful in various applications. The primary factor in loca- work in EMAs.
tion determination is a group of particular nodes known as
anchor/beacon nodes, which are resource privileged and 6.3.2. Node Failure. Sensor nodes are the main components
have more excellent storage and computing capacity. Other of the wireless sensor network for environmental monitor-
unknown nodes calculate their position in various ways ing applications. They operate unattended and are capable
based on the location of anchor/beacon nodes. As a result, of adapting to their deployable environment. Due to its size,
malicious anchor/beacon nodes must be prevented from it comes with stringent energy requirements because, on
providing false location information, as unknown nodes rely most occasions, the nodes are not inaccessible to humans.
entirely on them to compute their position [95]. Hence, its battery and other hardware cannot fail to operate
since monitoring will not be possible. Sensor nodes comprise
6.2.5. Cryptographic Algorithms. The communication infra- sensing, energy, transceiver, and processing units. The node
structure (radios in the network) in WSNs for EMAs may fails to operate if any of these components fail, affecting the
also be compromised. If the radio is compromised, it will overall network’s operation. In EMAs, node failure means
no longer be able to securely communicate the data obtained no sensing can occur, and there will be an optimal commu-
from the sensor nodes. Furthermore, the data obtained from nication medium [93].
the sensor nodes may necessitate additional security at the
base station and backup systems if the base station is 6.3.3. Network Converge and Fault Tolerance. Typically, sen-
attacked. As a result, using cryptographic algorithms to pro- sor nodes fail because of a lack of power, physical harm, or
vide the desired security is critical. Cryptographic algorithms environmental interferences, and such failures should not
are methods and procedures for securing a system linked impact the WSN’s intended mission. Fault tolerance refers
with keys. The crypto security of an environmental monitor- to a WSN’s capacity to maintain sensor functionality in the
ing application against attacks and malicious infiltration can face of node failures. Fault-tolerance protocols and algo-
be defined by two factors: (1) the strength of the keys and the rithms must be designed to handle their specified perfor-
efficiency of procedures and protocols associated with the mance levels. For example, tolerance can be minimal in a
Journal of Sensors 17
house, but tolerance levels must be quite high for environ- 7. Conclusion
mental monitoring to maintain operation [98].
The applications of WSN technology for environmental
monitoring have been discussed in this paper. The paper
6.3.4. Data Privacy. Data privacy is the ability to control reviewed vital protocols used in WSNs for EMAs, focusing
when, how, and to what extent data or information is shared on some protocols in the WSN protocol stack. Given the
or divulged to others. WSNs in EMAs, like most data sys- importance of security in environmental monitoring appli-
tems, collect sensitive information, necessitating the need cations, this review discussed wireless sensor network rout-
to protect its privacy, which in turn protects its confidential- ing protocols, security implementations, the types of
ity. However, the data collected by the sensors can be security threats in environmental monitoring applications,
obtained by an attacker using a powerful receiver, resulting node and network security practices, and some suggested
in a privacy breach [99]. Furthermore, a hacker can quickly WSN challenges for EMAs. Researchers interested in design-
introduce harmful messages into the network by using wire- ing protocols at the physical, data link, and network layers
less communication, allowing easy eavesdropping and should consider the type of applications and the associated
potentially breaching data privacy. peculiarities. The paper also examines some essential proto-
cols researchers use to track animals, water quality, and for-
est health.
6.3.5. Location and Positioning of Nodes. Localization tech-
niques are critical in applications such as animal tracking,
where the location of the animals at a given time is required, Data Availability
particularly in the case of endangered species. Localization is No data is available.
a technique used in WSNs for EMAs to determine the loca-
tion of sensor nodes. WSNs are thousands of tiny nodes, Conflicts of Interest
making GPS installation on each sensor node impractical.
Furthermore, even when GPS is installed on sensor nodes, The authors declare that there is no conflict of interest
it does not provide ideal localization results. Establishing regarding the publication of this paper.
location references on each sensor node in a dense network
is practically impossible [100]. It thus creates a problematic Acknowledgments
situation in which the sensor nodes must define their current
position without the support of any special equipment such This work was funded in part by the BANGA-AFRICA, Uni-
as a global positioning system. versity of Ghana, under the Seed Research Grant UG-BA/
SRG-001/2022.
6.3.6. Propagation Delay. The nodes in EMAs are strategi-
cally placed throughout the deployment area to achieve References
detection and identification probabilities close to 100% for [1] K. S. Adu-Manu, C. Tapparello, W. Heinzelman, F. A. Kats-
specified parameters [101]. However, even when the nodes riku, and J.-D. Abdulai, “Water quality monitoring using
detect the required parameters, propagation delays affect wireless sensor networks,” ACM Transactions on Sensor Net-
their transmission during in-node or in-network communi- works, vol. 13, no. 1, pp. 1–41, 2017.
cation and a node sink communication. The propagation [2] W. Honghui, T. Xianguo, L. Yan et al., “Research of the hard-
delay is the time it takes packets to travel through the trans- ware architecture of the geohazards monitoring and early
mission medium and is limited by the speed of light. For warning system based on the IoT,” Procedia Computer Sci-
example, if the source and destination are both in the same ence, vol. 107, pp. 111–116, 2017.
deployment area or region but are 100 meters apart, the [3] B. Jiang, G. Huang, T. Wang, J. Gui, and X. Zhu, “Trust based
propagation delay can be estimated to be 1 microsecond. If energy efficient data collection with unmanned aerial vehicle
they are separated by 500 meters in a different deployment in edge network,” Transactions on Emerging Telecommunica-
environment, the delay can be estimated to be 0.1 seconds tions Technologies, vol. 33, no. 6, article 3942, 2020.
[102]. Factors influencing node propagation with underwa- [4] S. Bader, Enabling autonomous environmental measurement
ter nodes differ from those influencing node propagation systems with low-power wireless sensor networks terms with
with tracking animals on land in environmental monitoring low-power wireless sensor networks, [Ph.D. thesis], Mid Swe-
applications. As a result, the number of packets transmitted den University, Faculty of Science, Technology and Media,
Department of Information Technology and Media, Sweden,
between the source and destination nodes may vary depend- 2011.
ing on the environment and channel conditions. Delays in
[5] M. Keerthika and D. Shanmugapriya, “Wireless sensor net-
packet transmission have serious consequences for WSNs works: active and passive attacks - vulnerabilities and coun-
used in environmental monitoring applications that require termeasures,” Global Transitions Proceedings, vol. 2, no. 2,
a response time of fewer than 200 milliseconds. When the pp. 362–367, 2021.
propagation delay exceeds the threshold in such applications [6] M. Almazaideh and J. Levendovszky, “Novel reliable and
as forest fire monitoring, battlefield, and freshwater quality, energy-efficient routing protocols for wireless sensor net-
it may affect stakeholders’ decisions to avoid the effects if works,” Journal of Sensor and Actuator Networks, vol. 9,
no such delays occur. no. 1, 2020.
18 Journal of Sensors
[7] F. Engmann, K. S. Adu-Manu, J.-D. Abdulai, and F. A. Kats- [24] U. Draz, A. Ali, M. Bilal et al., “Energy efficient proactive
riku, Applications of prediction approaches in wireless sensor routing scheme for enabling reliable communication in
networks, IntechOpen, London, 2021. underwater Internet of Things,” IEEE Transactions on Net-
[8] H. M. A. Fahmy, “Energy harvesting projects for WSNs,” in work Science and Engineering, vol. 8, no. 4, pp. 2934–2945,
Wireless Sensor Network, Springer, Cham, 2020. 2021.
[9] M. Y. Arafat, M. A. Habib, and S. Moh, “Routing protocols [25] D. Kothandaraman, S. Naik Korra, A. Balasundaram, and
for UAV-aided wireless sensor networks,” Applied Sciences, S. Magesh Kumar, “Sequence number based secure routing
vol. 10, no. 12, p. 4077, 2020. algorithm for IoT networks,” Materials Today: Proceedings,
[10] I. Daanoune, B. Abdennaceur, and A. Ballouk, “A compre- 2021.
hensive survey on LEACH-based clustering routing protocols [26] F. Tabbana, “Performance analysis of AODVDSDV and ZRP
in wireless sensor networks,” Ad Hoc Networks, vol. 114, arti- routing protocols,”Wireless Sensor Networks using NS2 Tool,
cle 102409, 2021. vol. 10, pp. 279–297, 2020.
[11] Y. Jia, “LoRa-based WSNs construction and low-power data [27] K. S. Adu-Manu, F. Katsriku, J.-D. Abdulai, J. M. Gómez, and
collection strategy for wetland environmental monitoring,” W. Heinzelmann, “Network lifetime maximization with
Wireless Personal Communications, vol. 114, no. 2, adjustable node transmission range,” in Smart Cities/Smart
pp. 1533–1555, 2020. Regions–Technische, wirtschaftliche und gesellschaftliche
[12] O. I. Khalaf and G. M. Abdulsahib, “Energy efficient routing Innovationen, pp. 693–707, Springer Vieweg, Wiesbaden,
and reliable data transmission protocol in WSN,” Interna- 2019.
tional Journal of Advances in Soft Computing and Its Applica- [28] O. A. Osanaiye, O. O. Ogundile, and F. Aina, “Evaluating
tions, vol. 12, no. 3, pp. 45–53, 2020. DoS jamming attack on reactive routing protocol in wireless
[13] V. Pandiyaraju, R. Logambigai, S. Ganapathy, and sensor networks,” African Journal of Science, Technology,
A. Kannan, “An energy efficient routing algorithm for WSNs Innovation and Development, pp. 1–8, 2021.
using intelligent fuzzy rules in precision agriculture,”Wireless [29] P. Sarao, “Ad hoc on-demand multipath distance vector
Personal Communications, vol. 112, no. 1, pp. 243–259, 2020. based routing in ad-hoc networks,” The Journal of Communi-
[14] K. S. Adu-Manu, N. Adam, C. Tapparello, H. Ayatollahi, and cation, vol. 14, no. 4, pp. 2953–2953, 2019.
W. Heinzelman, “Energy-harvesting wireless sensor net- [30] L. Chan, K. Gomez Chavez, H. Rudolph, and A. Hourani,
works (EH-WSNs),” ACM Transactions on Sensor Networks “Hierarchical routing protocols for wireless sensor network:
(TOSN), vol. 14, no. 2, pp. 1–50, 2018. a compressive survey,” Wireless Networks, vol. 26, no. 5,
[15] F. Engmann, F. A. Katsriku, J.-D. Abdulai, and K. S. Adu- pp. 3291–3314, 2020.
Manu, “Reducing the energy budget in WSN using time [31] A. K. Singh, A. Bhalla, P. Kumar, and M. Kaushik, “Hierar-
series models,” Wireless Communications and Mobile Com- chical routing protocols in WSN: a brief survey,” in 2017
puting, vol. 2020, Article ID 8893064, 15 pages, 2020. 3rd International Conference on Advances in Computing,
[16] A. Kochhar, P. Kaur, P. Singh, and S. Sharma, “Protocols for Communication & Automation (ICACCA)(Fall), pp. 1–6,
wireless sensor networks: a survey,” Journal of Telecommuni- Dehradun, India, September 2017.
cations and Information Technology, vol. 1, no. 2018, pp. 77– [32] K. Haseeb, N. Islam, A. Almogren, I. U. Din, H. N. Almajed,
87, 2018. and N. Guizani, “Secret sharing-based energy-aware and
[17] K. Sohraby, D. Minoli, and T. Znati,Wireless sensor networks: multi-hop routing protocol for IoT based WSNs,” IEEE
technology, protocols, and applications, John wiley & Sons, Access, vol. 7, pp. 79980–79988, 2019.
United States, 2007. [33] A. Ghaffari and H. R. Bannaeian, “QoS-based routing proto-
[18] M. Elhoseny, A. E. Hassanien, M. Elhoseny, and A. E. Hassa- cols for wireless sensor networks: a survey,” World Applied
nien, “Secure data transmission in WSN: an overview,” Stud- Sciences Journal, vol. 14, no. 6, pp. 866–875, 2011.
ies in Systems, Decision and Control, vol. 165, pp. 115–143, [34] R. Zagrouba and A. Kardi, “Comparative study of energy effi-
2019. cient routing techniques in wireless sensor networks,” Infor-
[19] J. Yang,Design and implementation of large-scale wireless sen- mation, vol. 12, no. 1, p. 42, 2021.
sor networks for environmental monitoring applications, Uni- [35] J. N. Al-Karaki and A. E. Kamal, “A taxonomy of routing
versity of North Texas, United States, 2010. techniques in wireless sensor networks,” inHandbook of Sen-
[20] A. Sarkar and T. Senthil Murugan, “Routing protocols for sor Networks: Compact Wireless and Wired Sensing Systems,
wireless sensor networks: what the literature says?,” Alex- pp. 116–139, CRC Press, 2004.
andria Engineering Journal, vol. 55, no. 4, pp. 3173–3183, [36] F. Lakrami, N. Elkamoun, and M. E. Kamili, “Advances in
2016. ubiquitous networking,” Lecture Notes in Electrical Engineer-
[21] B. Abidi, A. Jilbab, and M. E. Haziti, “Routing protocols for ing, vol. 366, pp. 287–300, 2016.
wireless sensor networks: a survey,” in Advances in Ubiqui- [37] B. Bhushan and G. Sahoo, “Routing protocols in wireless sen-
tous Computing, pp. 3–15, Academic Press, United States, sor networks,” Studies in Computational Intelligence, vol. 776,
2020. 2019.
[22] N. Shabbir and S. R. Hassan, “Routing protocols for wireless [38] J. J. Cancela, M. Fandiño, B. J. Rey, and E. M. Martínez,
sensor networks (WSNs),” in Wireless Sensor Networks - “Automatic irrigation system based on dual crop coefficient,
Insights and Innovations, Intecopen Science, 2017. soil and plant water status for Vitis vinifera (cv Godello and
[23] S. Arjunan and P. Sujatha, “Lifetime maximization of wireless cv Mencía),” Agricultural Water Management, vol. 151,
sensor network using fuzzy based unequal clustering and pp. 52–63, 2015.
ACO based routing hybrid protocol,” Applied Intelligence, [39] M. Usha Rani and S. Kamalesh, “Energy efficient fault toler-
vol. 48, no. 8, pp. 2229–2246, 2018. ant topology scheme for precision agriculture using wireless
Journal of Sensors 19
sensor network,” in 2014 IEEE International Conference on International Conference on Development and Application
Advanced Communications, Control and Computing Tech- Systems, DAS 2020 - Proceedings, pp. 69–72, Suceava, Roma-
nologies, pp. 1208–1211, Ramanathapuram, India, May 2014. nia, 2020.
[40] M. Gameil and T. Gaber, “Wireless sensor networks-based [54] A. Chehri and R. Saadane, “Zigbee-based remote environ-
solutions for cattle health monitoring: a survey,” in Interna- mental monitoring for smart industrial mining,” in
tional Conference on Advanced Intelligent Systems and Infor- SCA2019: The Fourth International Conference on Smart City
matics, pp. 779–788, Springer, Cham, 2020. Applications, pp. 2–7, Casablanca Morocco, 2019.
[41] B. Sharma and D. Koundal, “Cattle health monitoring system [55] M. Dangana, S. Ansari, Q. H. Abbasi, S. Hussain, and M. A.
using wireless sensor network: a survey from innovation per- Imran, “Suitability of NB-IoT for indoor industrial environ-
spective,” IET Wireless Sensor Systems, vol. 8, no. 4, pp. 143– ment: a survey and insights,” Sensors, vol. 21, no. 16, p. 5284,
151, 2018. 2021.
[42] Y. Liu and C. Bi, “The design of greenhouse monitoring sys- [56] S. Duangsuwan, A. Takarn, R. Nujankaew, and
tem based on ZigBee WSNs,” in 2017 IEEE international con- P. Jamjareegulgarn, “A study of air pollution smart sensors
ference on computational science and engineering (CSE) and LPWAN via NB-IoT for Thailand smart cities 4.0,” 2018
IEEE international conference on embedded and ubiquitous 10th International Conference on Knowledge and Smart Tech-
computing (EUC), pp. 430–433, Guangzhou, China, 2017. nology: Cybernetics in the Next Decades, KST, 2018, pp. 206–
[43] Z. Yiming, Y. Xianglong, G. Xishan, Z. Mingang, and 209, Chiang Mai, Thailand, 2018.
W. Liren, “A design of greenhouse monitoring & control sys- [57] M. Ibrahim, A. Elgamri, S. Babiker, and A. Mohamed, “Inter-
tem based on ZigBee wireless sensor network,” in 2007 Inter- net of Things based smart environmental monitoring using
national Conference on wireless communications, networking the Raspberry-Pi computer,” 2015 5th International Confer-
and Mobile computing, pp. 2563–2567, Shanghai, China, Sep- ence on Digital Information Processing and Communications,
tember 2007. ICDIPC, 2015, pp. 159–164, Sierre, Switzerland, 2015.
[44] T. Hidayat, R. Mahardiko, and S. T. D. Franky, “Method of [58] Y. Liu, A. Liu, N. Zhang, X. Liu, M. Ma, and Y. Hu, “DDC:
systematic literature review for Internet of Things in ZigBee dynamic duty cycle for improving delay and energy efficiency
smart agriculture,” 2020 8th International Conference on in wireless sensor networks,” Journal of Network and Com-
Information and Communication Technology, ICoICT, 2020, puter Applications, vol. 131, pp. 16–27, 2019.
pp. 20–23, Yogyakarta, Indonesia, 2020. [59] K. Lokesh Krishna, J. Madhuri, and K. Anuradha, “A ZigBee
[45] S. Cecchi, S. Spinsante, A. Terenzi, and S. Orcioni, “A smart based energy efficient environmental monitoring alerting and
sensor-based measurement system for advanced bee hive controlling system,” 2016 international conference on infor-
monitoring,” Sensors, vol. 20, no. 9, p. 2726, 2020. mation communication and embedded systems, ICICES,
[46] L. Joris, F. Dupont, P. Laurent, P. Bellier, S. Stoukatch, and 2016, Chennai, India, 2016, 2016.
J. M. Redoute, “An autonomous Sigfox wireless sensor node [60] J. A. Lossio-Ventura and H. Alatrista-Salas, “Information
for environmental monitoring,” IEEE Sensors Letters, vol. 3, management and big data: SIMBig overview,” in 6th Interna-
no. 7, pp. 01–04, 2019. tional Conference, SIMBig 2019, Lima, Peru, 2016.
[47] S. Gil-Lebrero, F. J. Quiles-Latorre, M. Ortiz-López, [61] S. Zafar, G. Miraj, R. Baloch, D. Murtaza, and K. Arshad,
V. Sánchez-Ruiz, V. Gámiz-López, and J. J. Luna-Rodríguez, “An IoT based real-time environmental monitoring system
“Honey bee colonies remote monitoring system,” Sensors, using Arduino and cloud service,” Engineering, Technology
vol. 17, no. 1, 2016. & Applied Science Research, vol. 8, no. 4, pp. 3238–3242,
[48] S. S. Reka, B. K. Chezian, and S. S. Chandra, “A novel 2018.
approach of IoT-based smart greenhouse farming system,” [62] G. G. Xie and J. H. Gibson, “A network layer protocol for
in Green Buildings and Sustainable Engineering, Springer, UANs to address propagation delay induced performance
2019. limitations,” Oceans Conference Record (IEEE), vol. 4, no. -
[49] M. Y. Cheng, K. C. Chiu, Y. M. Hsieh, I. T. Yang, J. S. Chou, November, pp. 2087–2094, 2001.
and Y. W. Wu, “BIM integrated smart monitoring technique [63] H. Yan, Z. J. Shi, and J. H. Cui, “DBR: depth-based routing for
for building fire prevention and disaster relief,” Automation underwater sensor networks,” Networking 2008 Ad Hoc and
in Construction, vol. 84, pp. 14–30, 2017. Sensor Networks, Wireless Networks, Next Generation Inter-
[50] S. Kaushik and C. Singh, “Monitoring and controlling in food net, vol. 4982, pp. 72–86, 2008.
storage system using wireless sensor networks based on Zig- [64] M. Zorzi, P. Casari, N. Baldo, and A. F. Harris, “Energy-effi-
bee & Bluetooth modules,” International Journal of Multidis- cient routing schemes for underwater acoustic networks,”
ciplinary in Cryptology and Information Security, vol. 2, no. 3, IEEE Journal on Selected Areas in Communications, vol. 26,
pp. 7–10, 2013, http://warse.org/pdfs/ijmcis01232013.pdf. no. 9, pp. 1754–1766, 2008.
[51] S. Aram, A. Troiano, F. Rugiano, and E. Pasero, “Low power [65] A. A. Ahmed and N. Fisal, “A real-time routing protocol with
and Bluetooth-based wireless sensor network for environ- load distribution in wireless sensor networks,” Computer
mental sensing using smartphones,” in Artificial Intelligence Communications, vol. 31, no. 14, pp. 3190–3203, 2008.
Applications and Innovations, pp. 332–340, Springer, 2012. [66] A. Ali Ahmed, “An enhanced real-time routing protocol with
[52] P. Di Gennaro, D. Lofú, D. Vitanio, P. Tedeschi, and load distribution for mobile wireless sensor networks,” Com-
P. Boccadoro, “WaterS: a Sigfox-compliant prototype for puter Networks, vol. 57, no. 6, pp. 1459–1473, 2013.
water monitoring,” Internet Technology Letters, vol. 2, no. 1, [67] G. Sakya and V. Sharma, “ADMC-MAC: energy efficient
article e74, 2019. adaptive MAC protocol for mission critical applications in
[53] F. Pitu and N. C. Gaitan, “Surveillance of SigFox technology WSN,” Sustainable Computing: Informatics and Systems,
integrated with environmental monitoring,” in 2020 15th vol. 23, pp. 21–28, 2019.
20 Journal of Sensors
[68] D. K. Sah and T. Amgoth, “Renewable energy harvesting review,” International Journal of Computer Applications,
schemes in wireless sensor networks: a survey,” Information vol. 179, no. 28, pp. 42–47, 1970.
Fusion, vol. 63, pp. 223–247, 2020. [83] J. Hou and Y. Zhang, “Mobile-service based approach for
[69] H. Sharma, A. Haque, and Z. A. Jaffery, “Solar energy har- topology control of wireless sensor networks,” Wireless Per-
vesting wireless sensor network nodes: a survey,” Journal of sonal Communications, vol. 102, no. 2, pp. 1839–1851, 2018.
Renewable and Sustainable Energy, vol. 10, no. 2, 2018. [84] G. M. E. Rahman and A. W. Khan, “LDAP: lightweight
[70] H. Sharma, A. Haque, and Z. A. Jaffery, “Maximization of dynamic auto-reconfigurable protocol in an IoT-enabled
wireless sensor network lifetime using solar energy harvesting WSN for wide-area remote monitoring,” Remote Sensing,
for smart agriculture monitoring,” Ad Hoc Networks, vol. 94, vol. 12, no. 19, p. 3131, 2020.
article 101966, 2019. [85] A. Sobchuk, Y. Kravchenko, M. Tyshchenko, P. Gawliczek,
[71] S. M. Bozorgi, A. Shokouhi Rostami, A. A. R. Hosseinabadi, and O. Afanasyeva, “Analytical aspects of providing a fea-
and V. E. Balas, “A new clustering protocol for energy ture of the functional stability according to the choice of
harvesting-wireless sensor networks,” Computers and Electri- technology for construction of wireless sensor networks,”
cal Engineering, vol. 64, pp. 233–247, 2017. in 2019 IEEE International Conference on Advanced Trends
[72] H. K. Qureshi, U. Saleem, M. Saleem, A. Pitsillides, and in Information Theory (ATIT), pp. 102–106, Kyiv, Ukraine,
M. Lestas, “Harvested energy prediction schemes for wireless 2019.
sensor networks: performance evaluation and enhance- [86] M. Dong, H. Li, Y. Li, Y. Deng, and R. Yin, “Fault-tolerant
ments,” Wireless Communications and Mobile Computing, topology with lifetime optimization for underwater wireless
vol. 2017, 14 pages, 2017. sensor networks,” Sādhanā, vol. 45, no. 1, 2020.
[73] T. Xie, G. Zhang, H. Liu, F. Liu, and P. Du, “A hybrid fore- [87] R. Jadhav and V. Vatsala, “Security issues and solutions in
casting method for solar output power based on variational wireless sensor networks,” International Journal of Computer
mode decomposition, deep belief networks and auto- Applications, vol. 162, no. 2, pp. 14–19, 2017.
regressive moving average,” Applied Sciences, vol. 8, no. 10, [88] V. Ekong and U. Ekong, “A survey of security vulnerabilities
p. 1901, 2018. in wireless sensor networks,” Nigerian Journal of Technology,
[74] A. G. Salman and B. Kanigoro, “Visibility forecasting using vol. 35, no. 2, p. 392, 2016.
autoregressive integrated moving average (ARIMA) models,” [89] J. Bhola, S. Soni, and G. K. Cheema, “Recent trends for secu-
Procedia Computer Science, vol. 179, no. 2019, pp. 252–259, rity applications in wireless sensor networks - a technical
2021. review,” in Proceedings of the 2019 6th International Confer-
[75] M. Prauzek, N. R. A. Mourcet, J. Hlavica, and P. Musilek, “Q- ence on Computing for Sustainable Global Development,
learning algorithm for energy management in solar powered INDIACom 2019, pp. 707–712, New Delhi, India, 2019.
embedded monitoring systems,” in 2018 IEEE Congress on [90] S. Roy, M. Conti, S. Setia, and S. Jajodia, “Secure data aggre-
Evolutionary Computation, CEC 2018 - Proceedings, Rio de gation in wireless sensor networks,” Networks, vol. 7, no. 3,
Janeiro, Brazil, 2018. pp. 1040–1052, 2012.
[76] H. X. Zou, L. C. Zhao, Q. H. Gao et al., “Mechanical modula- [91] A. Aseeri and R. Zhang, “Secure data aggregation in wireless
tions for enhancing energy harvesting: principles, methods sensor networks: enumeration attack and countermeasure,”
and applications,” Applied Energy, vol. 255, article 113871, in ICC 2019 - 2019 IEEE International Conference on Com-
2019. munications (ICC), Shanghai, China, 2019.
[77] V. Larnaudie, M. D. Ferrari, and C. Lareo, “Switchgrass as an [92] S. Ozdemir and Y. Xiao, “Secure data aggregation in wireless
alternative biomass for ethanol production in a biorefinery: sensor networks: a comprehensive overview,” Computer Net-
perspectives on technology, economics and environmental works, vol. 53, no. 12, pp. 2022–2037, 2009.
sustainability,” Renewable and Sustainable Energy Reviews, [93] A. J. Watt, M. R. Phillips, C. E. Campbell, I. Wells, and
vol. 158, article 112115, 2022. S. Hole, “Wireless sensor networks for monitoring underwa-
[78] J. Singh, R. Kaur, and D. Singh, “Energy harvesting in wireless ter sediment transport,” Science of the Total Environment,
sensor networks: a taxonomic survey,” International Journal vol. 667, pp. 160–165, 2019.
of Energy Research, vol. 45, no. 1, pp. 118–140, 2021. [94] A. Razaque and S. S. Rizvi, “Secure data aggregation using
[79] B. Paharia and K. Bhushan, “A comprehensive review of dis- access control and authentication for wireless sensor net-
tributed denial of service (DDoS) attacks in fog computing works,” Computers & Security, vol. 70, pp. 532–545, 2017.
environment,” in Handbook of Computer Networks and [95] M. B. Shanthi and D. K. Anvekar, “Secure localization for
Cyber Security: Principles and Paradigms, Springer, 2019. underwater wireless sensor networks based on probabilistic
[80] K. S. Adu-manu, A study into lifetime maximization of wire- approach,” in Proceedings of 2018 2nd International Confer-
less sensor networks for water quality monitoring, [Ph.D. the- ence on Advances in Electronics, Computers and Communica-
sis], University of Ghana, 2020. tions, ICAECC 2018, pp. 1–6, Bangalore, India, 2018.
[81] W. Tiberti, F. Caruso, L. Pomante, M. Pugliese, M. Santic, [96] E. Barker and Q. Dang, Recommendation for key manage-
and F. Santucci, “Development of an extended topology- ment – part 3: application-specific key management guidance,
based lightweight cryptographic scheme for IEEE 802.15.4 vol. 800-57, NIST Special Publication, 2007.
wireless sensor networks,” International Journal of Distrib- [97] V. Kumar, A. Jain, and P. N. Barwal, “Wireless sensor net-
uted Sensor Networks, vol. 16, no. 10, Article ID works: security issues, challenges and solutions,” Interna-
155014772095167, 2020. tional Journal of Information & Computation Technology,
[82] C. Mallick and S. Satpathy, “Challenges and design goals of vol. 4, no. 8, pp. 859–868, 2014, http://www.irphouse.com.
wireless sensor networks: a sate-of-the-art challenges and [98] C. S. Kingsly and J. G. C. Chandran, “Critical study on con-
design goals of wireless sensor networks: a sate-of-the-art straints in wireless sensor network applications,”
Journal of Sensors 21
International Journal of Engineering Research & Technology,
vol. 2, no. 7, pp. 1311–1316, 2013.
[99] N. Sharma and R. Bhatt, “Privacy preservation in WSN for
healthcare application,” Procedia Computer Science,
vol. 132, pp. 1243–1252, 2018.
[100] J. Kuriakose, S. Joshi, R. Vikram Raju, and A. Kilaru, “A
review on localization in wireless sensor networks,” Adv.
Intell. Syst. Comput., vol. 264, pp. 599–610, 2014.
[101] A. J. Garcia-Sanchez, F. Garcia-Sanchez, F. Losilla et al.,
“Wireless sensor network deployment for monitoring wildlife
passages,” Sensors, vol. 10, no. 8, pp. 7236–7262, 2010.
[102] S. Khanvilkar, F. Bashir, D. Schonfeld, and A. Khokhar,
“Multimedia networks and communication,” in The Electri-
cal Engineering Handbook, pp. 401–425, University of Illi-
nois, Chicago, 2005.