ArticlePDF Available

Analysis of Cross-Layer Design of Quality-of-Service Forward Geographic Wireless Sensor Network Routing Strategies in Green Internet of Things

Authors:

Abstract

Wireless sensor networks suffer from some limitations such as energy constraints and the cooperative demands essential to perform multi-hop geographic routing for Internet of things (IoT) applications. Quality of Service (QoS) depends on a great extent on offering participating nodes an incentive for collaborating. This paper presents a mathematical model for a new generation by forwarding QoS routing determination that enables allocation of the optimal path to satisfy QoS parameters to support a wide range of communication-intensive IoT’s applications. The model is used to investigate the effects of multi-hop communication on a traffic system model designed with a Markov discrete time M=M=1 queuing model, applicable to green deployment of duty-cycle sensor nodes. We present an analytical formulation for the bit-error-rate, and a critical path-loss model is defined to the specified level of trust among the most frequently used nodes. Additionally, we address the degree of irregularity parameter for promoting adaptation to geographic switching with respect to two categories of transmission in distributed systems: hop-by- hop and end-to-end retransmission schemes. The simulations identified results for the average packet delay transmission, the energy consumption for transmission, and the throughput. The simulations offer insights into the impact of radio irregularity on the neighbor-discovery routing technique of both schemes. Based on the simulation results, the messages encoded with a non-return-to-zero (NRZ) have more green efficiency over multihop IoT (without loss of connectivity between nodes) than those encoded with the Manchester operation. The findings presented in this work are of great help to designers of IoT.
SPECIAL SECTION ON SURVIVABILITY STRATEGIES FOR EMERGING WIRELESS NETWORKS
Received February 13, 2018, accepted March 25, 2018, date of publication April 6, 2018, date of current version April 25, 2018.
Digital Object Identifier 10.1109/ACCESS.2018.2822551
Analysis of Cross-Layer Design of Quality-
of-Service Forward Geographic Wireless
Sensor Network Routing Strategies
in Green Internet of Things
MOHAMMED ZAKI HASAN 1, (Member, IEEE), FADI AL-TURJMAN 2, (Member, IEEE),
AND HUSSAIN AL-RIZZO1
1Systems Engineering Department, Donaghey College of Engineering and Information Technology, University of Arkansas at
Little Rock, Little Rock, AR 72204, USA
2Antalya Bilim University, 07190 Antalya, Turkey
Corresponding author: Mohammed Zaki Hasan (mohammed.z.hasan@ieee.org)
ABSTRACT Wireless sensor networks suffer from some limitations such as energy constraints and
the cooperative demands essential to perform multi-hop geographic routing for Internet of things (IoT)
applications. Quality of Service (QoS) depends to a great extent on offering participating nodes an incentive
for collaborating. This paper presents a mathematical model for a new-generation of forwarding QoS routing
determination that enables allocation of optimal path to satisfy QoS parameters to support a wide range
of communication-intensive IoT’s applications. The model is used to investigate the effects of multi-hop
communication on a traffic system model designed with a Markov discrete-time M/M/1 queuing model,
applicable to green deployment of duty-cycle sensor nodes. We present analytical formulation for the bit-
error-rate, and a critical path-loss model is defined to the specified level of trust among the most frequently
used nodes. Additionally, we address the degree of irregularity parameter for promoting adaptation to
geographic switching with respect to two categories of transmission in distributed systems: hop-by-hop
and end-to-end retransmission schemes. The simulations identified results for the average packet delay
transmission, the energy consumption for transmission, and the throughput. The simulations offer insights
into the impact of radio irregularity on the neighbor-discovery routing technique of both schemes. Based on
the simulation results, the messages en-coded with non-return-to-zero have more green efficiency over multi-
hop IoT (without loss of connectivity between nodes) than those encoded with the Manchester operation.
The findings presented in this paper are of great help to designers of IoT.
INDEX TERMS Wireless sensor networks, quality of Service, Internet of Things, geographical routing,
euclidean distance, network topology.
I. INTRODUCTION
Internet of Things (IoT) has been envisioned as one of
promised collaborative distributed networks that consisting
of numerous smart devices assembled for the purpose of
monitoring real environments [1]. These devices may be
scattered throughout an area of interest to observe a physical
phenomenon [2], [3]. For example, sensor nodes that are
deployed in hostile environments can track moving vehi-
cles on high-ways, monitor climate changes, and provide
early warnings for radiation hazards or an earthquake [4].
Moreover, the sensor nodes integrate with IoT as three
sub-systems: sensing, processing, and communicating [5].
In most IoT’s applications, they are placed above the floor
or within a liquid medium using an antenna that is a few
centimeters above the ground [1].
Several studies have demonstrated the communication
component of sensor nodes; however, the operation of ineffi-
cient design of the antenna exhausts most of the battery-life
of the sensor nodes, which might adversely affect the wireless
channel and lead to error-prone links and inefficient rout-
ing [6]. In another hand, traffic control access to the media
from the distributed sensor nodes should be strictly controlled
VOLUME 6, 2018
2169-3536 2018 IEEE. Translations and content mining are permitted for academic research only.
Personal use is also permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
20371
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
to avoid or reduce redundancy and collisions, which have a
dramatic impact on the lifetime of theses devices [7].
Hence, certain routing protocols in Wireless Sensor Net-
works (WSNs), such as geographic routing, require informa-
tion (for example, a self-configuring localization mechanism
or a-priori mechanism) that is adapted to the location of the
sensor nodes for the purpose of information delivery [8].
Thus, geographic routing protocols are considered more effi-
cient in IoT, since they minimize the size of sensor node
storage by storing only the information on direct neighbors
for forwarding the packet [9]. In this case, geographic routing
is usually considered to be a smart forwarding mechanism
whereby every node makes a decision to route the packet to
the closest neighbors or to the nodes that are closest to the
sink. Furthermore, such geographic routing protocols may be
efficient, low-over-head methods that have sufficient network
density, accurate localization, and significant link reliability
in the sensor networks [10]. However, unfortunately they
conserve more energy and bandwidth because the discovery
floods and state propagation are not required beyond a single
hop. Such energy conservation may generate unreliable links
for forwarding; these situation are referred to as weakest link
problems [11].
Unreliable links increase the rate of packet drops, delays,
and energy consumption because of continuous retransmis-
sion. Certain sensor nodes may decide which path to forward
the packets to based on the distance and number of hops and
on the loss characteristics [12]. However, sensor nodes in
IoT in most applications are distributed in an ad hoc fashion
broadly use two transmission techniques to communicate
among themselves or to send data to a sink: single-hop com-
munication (SHC) and multi-hop communication (MHC), as
shown in Fig. 1.
FIGURE 1. Single communication hop (SHC) and multi-hop
communication (MHC) routing algorithm in wireless sensor networks.
In single-hop communication, the sensor nodes can cover
the entire network and thus consume energy according to
the longest transmission range [8]. Therefore, because of
this fixed transmission range, extra energy is wasted even
when the sensor nodes are close to one another because of
retransmission caused by the unreliable links.
Meanwhile in MHC, the energy consumed for end-to-end
increases as a nonlinear function of both the number of relay-
ing nodes and the energy consumption for each individual
hop; the sensor nodes attempt to maximize the per-hop relia-
bility by forwarding to the nearest node with highly reliable
links. This results in increased energy expenditure because of
the increase in the number of hops until the final destination
is reached. Regardless of weakest link problems, the hop dis-
tance is mainly determined by the transmission range. More-
over, the energy consumed in each individual hop increases
with the transmission range set by the distance in relation to
the attenuation factor of the signal propagation. As a result,
differences between the number of hops and the transmission
range occur. For example, if the hop number is small and
the transmission range is large, the energy consumed for one
transmission increases nonlinearly. Alternatively, if the hop
distance is small for the same overall end-to-end transmission
range, the energy consumption is dominated by the energy
consumption of the transceivers; therefore, the total energy
increases in an almost linear manner as a function of the hop
number (in other word, with an increasing number of hops).
Apparently, there is a trade-off between the hop number,
transmission range, and link status of each hop in MHC to
achieve optimal green energy efficiency. More precisely, the
network designer must consider these three parameters that
govern energy consumption model in what is referred to as
the distance-hop-energy trade-off [13].
Increasingly, a large number of IoT’s applications require a
real-time approach combined with queuing theory to provide
a stochastic green model to guarantee the QoS parameters.
This green model and several other solutions proposed for
multimedia communications on the Internet and the wireless
environment cannot be directly applied to IoT because of the
influence of the wireless channel and the duty-cycle of the
sensor nodes on various design characteristics, in addition to
the resource allocation constraints of IoT applications [14].
Furthermore, the random nature of the wireless channel pro-
hibits guaranteeing deterministic QoS parameters in MHC
sensor networks. Consequently, an analysis of the stochastic
green model of QoS metrics is necessary to address the QoS
requirements in IoT’s applications [15].
In this paper, an analytical green model is presented for a
new-generation of forward QoS geographically routing deter-
mination that enables the allocation of the optimal path to
satisfy the required QoS parameters. This green model is an
extension the work reported [16] which intended to support a
wide range of communication-intensive real-time multimedia
for wireless sensor applications. Thus, the approach usually
involves all of the structural layers in the communication
protocol and the physical (PHY) and Media Access Con-
trol (MAC) layers of the IoT as depicts in Fig. 2.
The proposed model investigates the effects of MHC when
deriving the equation for the bit-error-rate (BER). A critical
path-loss model of the adaptive switching for two categories
of transmission schemes is designed when the topology of
the IoT matches the topology of the physical surroundings
20372 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
FIGURE 2. The three layers of IoT.
to govern the performance of energy consumption and QoS
parameters. This adaptive switching occurs in accordance
with the criterion embedded in the path-loss model parameter
called the degree of irregularity (DOI) which is a function
of the distance between two sensor nodes; this criterion is
used in deciding which path from the source node to the next
hop will be selected in forwarding the packet. Thus, effective
path-loss predication is considered significant green scheme
in the traffic system model, which is designed with a Markov
discrete-time M/M/1 queuing model and applied to duty-
cycled for each sensor nodes.
We describe the behavior of two categories of transmission
schemes to achieve a green model with applying the station-
ary distribution of the probability of the packet transmission
for both schemes for the purpose of investigating the per-
formance of the duty-cycle of the MAC layer in IoT. The
analysis of the energy consumption, delay, and throughput
of the this green model can be used to optimize the protocol
parameters essential to enable achieving the desired network
performance.
The main contribution of this paper is the development of a
framework for analyzing the optimal forwarding choices with
respect to QoS parameters, with the intention of quantifying
the impact of the relay of radio irregularity on MAC and
routing layers of two categories of transmission schemes in
IoT. We show that minimizing the energy consumption per
information along a selected path can be obtained by deriving
the end-to-end BER for more practical transmission schemes.
In summary, the contributions paper is summarizes as:
1) a mathematical green model to analyze the optimal
forwarding choice as a trade-off between distance-per
hop and overall hop counts for two categories of trans-
mission schemes;
2) a radio model to simulate these two categories of trans-
mission schemes under various channel conditions in
order to investigate the impact of the radio channel on
MAC and routing layers in IoT;
3) a mathematical model of channel access mecha-
nism based on discrete M/M/1 queuing modeling to
describe the behavior of duty-cycled MAC protocol in
order to analyze handling real-time traffic of a detected
event with acceptable transmission range under the
impact of radio irregularity.
The remainder of this paper is organized as follows.
In section II, an overview of previous analyses of
MHC schemes in WSNs is provided. The analysis and
traffic-system model are introduced in sections III and IV,
respectively. In section V, a simulation of the proposed
traffic system model is provided, and the results for each
site are presented. Finally, the conclusions are summarized
in section VI.
II. RELATED WORK
Many research groups are exploring issues related to the
design of lower-complexity nodes for deployment scheme
for green IoT as sensor hardware components [8]. There
are several projects such as smart cities, structural health
and smart lighting that have sought or are seeking ways to
integrate three functions: sensing, processing, and communi-
cating, into a single integrated circuit for various IoT’s appli-
cations with limited energy consumption [5]. Furthermore,
studies focusing on new techniques, such as cooperative
multilayer communication among nodes and network coding
for wireless communication using particle-sized sensor nodes
that are distributed for wide-area sensing [17]. However, the
increasing interest in IoT’s applications in multimedia sensor
networks have lead these studies to focus on increasing the
network performance by relying on an accurate link esti-
mation in order to ensure efficient use of energy resources
of the node [3], [18]. Cross-layer awareness is considered
as a potential solution to various issues and a means of
improving the performance in IoT because of the possibility
of involving both the PHY and MAC layers as shown in Fig. 2
to provide functions other than routing, such as the power
efficiency [19].
In [20] an investigation of the trade-off between two
schemes, Forward Error Correction (FEC) and Automatic
Repeat Request (ARQ), were proposed in terms of the
energy consumption, delay, and end-to-end BER. The authors
explored the improvement of the channel capacity by reduc-
ing the interference with the intention of transmitting power
through a channel-aware cross-layer design. FEC codes and
ARQ are considered as candidates for delay-sensitive traffic
in WSNs. Our approach and the system model proposed in
this paper depends on the analysis for cross-layer green IoT
through the derivation of the equation for the BER by defin-
ing a critical path-loss model of adaptive switching for the
QoS-sensitive traffic in the WSNs. Consequently, the effect
of BER on the QoS parameters of both hop-by-hop and end-
to-end retransmission schemes is studied. Furthermore, the
derivation of the equation for the BER has never been con-
sidered in the context of traffic system model in IoT before.
The technique referred to as the hop-length extension
assesses the performance of two routing protocol schemes.
An analysis of the hop-and-packet reception ratio (PRR) for-
warding strategies to maximize the lifetime and minimize the
energy consumption cost of sensor networks was performed
in [21] by introducing two new forwarding schemes, i.e. the
single-link energy efficient forwarding schemes (SEEF) and
multi-link energy-efficient forwarding (MEEF), based on a
realistic link loss model. The selected path for forwarding
was based on an energy efficiency metric, defined as the ratio
between the end-to-end delivery packet rate and the energy
VOLUME 6, 2018 20373
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
required for the transmission of the packets to reach the sink.
Therefore, our two proposed green schemes focused on the
trade-offs involved in minimizing the expected number of
retransmissions or the endeavor to increase the end-to-end
BER. However, this forwarding strategy does not maximize
the lifetime in terms of minimizing the energy cost. Because
those nodes are not addressed in the packet header as the
destination, they temporarily turn their radio unit to sleep
mode to save energy and achieve green model. As long as
some sensor nodes in this stochastic green model stay awake
to overhear the incoming packets, it might be more efficient
to prevent retransmission of the entire packet even if retrans-
mission requires more energy.
A new technique called cooperative communication
has been applied to Wireless Multimedia Sensor Net-
works (WMSNs) to improve network performance, and the
results are primarily classified in the literature as either coop-
erative ARQ protocols or transformation from single-hop to
multi-hop transmission. The Network-Coding-based Cooper-
ative ARQ MAC protocol for WSN (NCCARQ) proposed in
[22] focused on a centralized WSN topology to coordinate the
retransmission of channel access among a set of relay nodes
that operate in a promiscuous mode and provide bidirectional
connections between any pairs of sensor nodes. Each sensor
node stores a copy of the received data packet until a positive
acknowledgment from the sink is received; otherwise, the
error mechanism performs an error control operation on the
received messages. The protocol based on CSMA, therefore,
is compatible with the IEEE802.15.4 standard. This com-
patibility enables the NCCARQ to use the same structure in
the control packets and follows the same principles as the
standard with certain modifications to improve the efficiency
of the proposed protocol. It is similar to the model proposed
in this paper, which uses less control packets than ARQ-based
protocols and delivers increased energy efficiency while sat-
isfying QoS parameters.
A typical ad-hoc network operates according to a coopera-
tive multi-hop transmission approach, which achieves greater
power efficiency because it operates at a low signal-to-noise-
ratio (SNR) which is needed to cover the transmission range.
In [23], Feng and Cimini, Jr., adopted the linear multi-
hop transmission approach by considering quasi-static fading
without spatial reuse. This adaptation simplified the linear
multi-hop transmission approach by including Hybrid Auto-
matic Repeat Request (HARQ) retransmission protocols. The
authors focused on a design that provides the optimal number
of hops with a maximum delay along a linear multi-hop
network that achieves maximum end-to-end throughput. This
analytical framework allows the parameters to be set as an
optimization problem, which is solved using numerical meth-
ods. Likewise, Sikora et al. [24] considered an uncooperative
linear approach for multi-hop and single-hop network trans-
mission (in which the nodes do not cooperate and attempt to
access the channel simultaneously) to investigate the perfor-
mance of the distributed channel access capacity at the MAC
layer and the power channel at the PHY layer, especially in
delay and bandwidth-constrained scenarios. The analytical
framework provides the optimum number of hops under the
delay constraint using a sphere-packing bound. The authors
indicated that choosing the optimum number of hops using
Time Division Multi Access (TDMA) multi-hop transmission
resulted in asymptotic per-link spectral efficiency.
Our research in this paper is similar to previous stud-
ies [6] and [25] as far as fundamental construction of network
topologies is concerned but differs in the forwarding strategy,
in which each sensor node can be placed arbitrarily between
the source and the sink along a selected path. Additionally,
with respect to the energy consumption, our green model
incorporates all transmission operations with all of the circuit
processing energy consumption.
Many researchers have focused on a single-layer design
in centralized WSN topology to tackle energy dissipation
problems [4], [6], and [8]. The cross-layer design typically
focuses on multi-layers, such as MAC, routing, and transport,
but does not consider the PHY layer. The consideration of the
PHY layer was of paramount importance in the cross-layer
design in [26] to minimize energy consumption. The pro-
posed cross-layer works as follows. First, an estimation of the
channel gains between every node and the sink is performed.
Second, the MAC layer cooperates with this information to
design the time-slot lengths in a comfortable, energy-efficient
manner. Finally, the calculated time slot lengths and BER
determine the suitable modulation level for the PHY layer.
In [27] the physical layer was specifically considered, and
parameters such as the hop distance, transmission power,
and modulation schemes were regarded as open parameters
for the network designer. The authors proposed an approach
for minimizing the energy consumption and maximizing the
network lifetime by finding the optimal transmission distance
while controlling the transmit power for a given modulation
scheme and a given channel model. In the process, the authors
derived a new metric called energy per successfully received
bit, in which the probability of error was defined as a function
of a basic modulation scheme and depends on two param-
eters, the received signal energy and the noise level of the
channel.
Differently from existing approaches, our contributions
in this paper are based on real-time queuing theory [28]
which uses a simple stochastic M/M/1 queuing model to
investigate the performance of the duty-cycle MAC layer and
to simultaneously analyze its QoS parameters in terms of
average delivery delay, throughput, and energy consumption.
It should be noted that this paper does not focus on the
scheduling problem in real-time systems; instead, it is aimed
at providing an analytical tool to help designers to develop
new geographic routing communication solutions for green
IoT. For ease of understanding, we present an overview of
some existing SHC and MHC approaches in Table 1.
III. ANALYTICAL FORMULATION
The proposed framework is reviewed in this section by con-
sidering a certain number of sensor nodes that create a direct,
20374 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
TABLE 1. Some existing SHC and MCH approaches.
fully connected network topology. This topology consists of
sensor nodes distributed according to a 2-D matrix Poisson
distribution with a given density over a short interval. Thus,
the nodes making up a small density in an area are selected
as intermediate relays to construct the multi-hop network
topology and obtain a closed-form expression for the num-
ber of hops, which can be used to estimate the network
performance [29].
A framework is composed of two main components;
a channel access mechanism and a wireless link channel
access modeling.
FIGURE 3. The framework of the flow of packet forwarding mechanism.
To understand the impact of radio model on the QoS
requirements for multimedia data traffic, we take a close
look into the work flow on a MICA2 sensor device that a
sensor node forwards a packet towards the sink. As depicted
in Fig. 3, the flow begins with first stage at the physical
(PHY) (i.e. the wireless modeling) with a node perceiving
the physical wave that carries the packet from the source
node. This perceiving of node depends on the calculation
of corresponding DOI values that inform the variances of
SNR with incremental changes in directions. Each node’s
DOI value is calculated, and then we adjust the value of path
loss models based on a Kcoefficient in order to forward
direction from next hop to the nearest neighbor. This stage
will end when the node successfully receives ACK message
from next-hop intermediate node at the MAC (i.e. the channel
access mechanism). A packet is not accepted as long as any
bit of the packet is received with error. Finally, at the network
layer, the node is likely to cause mismatch between the path
estimation and the real forwarding in order to indicate a
reliable wireless channel from an unreliable one for the data
transmission.
A. CHANNEL ACCESS AND DUTY-CYCLE
NODE MECHANISM
The channel access mechanism is composed of two sub-
protocols: retransmission channel access and duty-cycle node
operations. The two sub-protocols are described below.
1) ANALYSIS OF RETRANSMISSION
CHANNEL ACCESS SCHEMES
The proposed framework enables a comprehensive com-
parison of two routing schemes in which medium access
to the channel is achieved through a request to send/clear
to send/data/acknowledgment RTS/CTS/DATA/ACK hand-
shaking to guarantee successful end-to-end retransmission or
multi-hop transmission, as illustrated in Fig. 4. The transmis-
sion schemes in distributed systems are categories as:-
FIGURE 4. Two categories of transmission in distributed systems.
1) Hop-by-hop retransmission routing scheme: in this
scheme at every next-hop, the intermediate node checks
the correctness of the data packet and requests for
retransmission with NACK packet until a correct data
packet arrives. After that, an ACK packet is transmitted
to the sender node indicating a successful transmis-
sion [30]. Figure 5depicts the mechanism of retrans-
mission, whereas the first data packet for example fails
between the nodes 2 and 3. Then node 3 sends an
NACK data packet asking node 2 for retransmission.
After that, node 2 retransmits the data packet, node 3
transmit an ACK data packet after successfully receiv-
ing the data packet.
2) End-to-end retransmission routing scheme: the inter-
mediate nodes simply forward received data packets to
VOLUME 6, 2018 20375
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
FIGURE 5. Hop-by-hop retransmission routing strategy.
FIGURE 6. End-to-end retransmission routing strategy.
the next hop and do not check the correctness of the data
packets until they arrive at the sink. Figure 6depicts the
mechanism of retransmission routing scheme, whereas
the received data packets forwarded to the next hop
do not check the correctness of the data packets until
they arrive at the sink. Moreover, the sink checks the
correctness of the data packets and retransmits with
an NACK packet to the source if the data packets are
incorrect [30].
2) MODELING DUTY-CYCLE NODE OPERATIONS
Each node has a finite queue size with heterogeneous ini-
tial energy and communication capabilities determined by
the transmission range. Moreover, a duty-cycle operation
is deployed in green IoT such that the sensor nodes enter
the sleep state when there is no ongoing transmission and
enter the wake-up state when transmission is required. The
Scheduling-Driven Sensing Traffic (SDST) application is
part of a very active field namely distributed tracking for
vehicles within smart cities or on high-ways with low sensor
rates but a higher reliability of messages transmission.
The tracking scenario raises a number of fundamental
information processing problems in distributed information
representation, discovery, storage and communication in col-
laborative processing, networking routing and aggregation,
data abstraction and query optimization, human-computer
interface interaction and finally in the software services [31].
Therefore, we focuses in this paper on how the information
processing aspect of the tracking problems such as how rout-
ing the collected information from the environment under
resources QoS constraints of IoT which may require speci-
fied multi-hop geographic routing which considerably differs
from other routing protocols.
Consequently, the research community has been searching
for appropriate IoT stack-layers that can provide suit-
able abstractions networking and the limitation of hard-
ware resources. While defining a unifying architecture that
involves all the structure layers in the communication pro-
tocol, PHY and MAC layers of the sensor network are
still active areas [31]. Moreover, the event-driven model
for traffic analysis and QoS constraints are considered the
most challenging among other models because it is contin-
uous, observer-initialized, and hybrid; thus, the occurrence
of events is completely unpredictable, resulting in arbitrary
traffic patterns [32]. We believe that an approximate solu-
tion of defining a unifying architecture is the principled
interaction or cooperation between IoT and sensor network
stack-layers.
To evaluate the performance of SDST application, we
consider a realistic channel model for dense distributed net-
worked sensors. This realistic channel model can improve
perceived SNR by decreasing average hop-distances to
improve a traffic-system queuing model for routing determi-
nation that enables the allocation of the path for reporting
events from the source to the sink in IoT.
Most recent empirical studies have based their assumptions
on the proposed mathematical model for a channel-aware sys-
tem derived from [33]; several studies have proposed new link
models, but these suffer from significant shortcomings [16].
Our channel-aware routing algorithm serves as a method
to evaluate QoS parameters by considering the relationship
between realistic hop-distance, number of hops, and link
quality of the path. More precisely, the determination of the
next hop for each node to transmit independently is based on
the probability of link quality at a distance from the sink. The
expected hop-distance is found by deriving the equation of
BER and defining the critical path-loss model parameters as
a function of the distance between two nodes to decide which
path from the source node to the next hop will be selected to
forward the packet.
IV. A MARKOV QUEUING MODEL FOR
DUTY-CYCLED NODE OPERATIONS
In order to satisfy the QoS requirements for IoT’s appli-
cations, a Markov discrete-time stochastic process M/M/1
queuing model for green model is proposed under a realistic
assumption of a finite queue size that can hold-up to m
packets for duty-cycling with schedule-driven operation for
sensor nodes. The arrival and departure of the data packets
are regulated under a realistic assumption of a finite queue
size. Therefore, the proposed model makes the following
assumptions:
20376 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
1. Event/packet arrivals denote a stochastic process
{A(τ)|τ0}that represents the total number of arrivals
that have occurred from time 0 to time τ; this procedure
creates an independent Poisson process at each node,
and the number of packet arrivals in any time slot is
distributed with a Poisson process with parameter λ.τ ,
for time of arrival τ τ 0.
2. Let πı(τ) denote the steady state of power for the node
at time τ; the inter-arrival δ0 times (that is, the
distribution of time at state ıbefore marking the tran-
sition) are independent and exponentially distributed
with the λ, where o(δ) is defined as a function of δsuch
that limδ0o(δ)
δ.
3. The queueing discipline of data packets is First-Come,
First-Served (FCFS).
4. The queueing system assumes equilibrium under the
condition that the probability of arrival is less than the
independent probability of transmitting the information
packet, or λ<β.
5. The processing and radio-transmission times are
independent and identical (i.i.d.) with an arbitrary
distribution.
6. Retransmission is supported.
7. When an event is sensed, the node processes it and
sends the information packet with a probability of
transmission per node per cycle, and every sensor node
in the network has an independent probability of trans-
mitting information packet βin the duty-cycle.
These assumptions are made based on [34] and [35] and are
similar to [36] and [37], which have been verified as valid
approximations of realistic scenarios. The proposed Markov
model shows that the power transition of each sensor node
in the network may be modeled by a discrete-time M/M/1
Markov chain, which represents a different predefined sta-
tus for a node for an event at the wake-up/sleep mode of
the duty cycle. Table 2lists all the notations used in the
model.
TABLE 2. Notations of the proposed model.
A node may exchange its status slot by slot, which cor-
responds to the transition from one state to another in the
Markov chain. Figure 7shows that the proposed Markov
model has limited queuing capacity with finite state slots
from left to right, which corresponds to 0 state for processing
packets in the queue and so on to mpackets in the queue
(full queue). Specifically, if a packet arrives and the queue
is full, then the packet is simply dropped; nevertheless, the
FIGURE 7. Discrete-time Markov chain for M/M/1 modeling of a traffic
queuing behavior system.
packets are removed from the queue when they are success-
fully transmitted.
FIGURE 8. Splitting of a Poisson process.
By contrast, when the queue is neither full nor empty,
then a node may obtain access to the channel to transmit
packets with an independent probability, as depicted in Fig. 8.
The analysis of the Markov discrete-time M/M/1 queu-
ing model offers insights into the traffic behavior of IoT
in general and points to an idea for a control algorithm.
The steady state probability and the transition probabilities
of moving from one state to another can be described as
follows:-
P0,ı=λıδ, ı=0,...,M,(1)
P0,M=λM,(2)
P=βλ0δ, ı=0,...,M,(3)
Pı,ı1=βδλı+1δ
+(1 βδ)λıδ, ı=1,...,M1,(4)
Pı,M=βδλMı+1δ
+(1 βδ)λMıδ, ı=1,...,M,(5)
Pı, =0,ı=2,...,M,  =0,...,ı2,(6)
If the event of interest is detected during the specified oper-
ations, then Equations 1and 2describe all transitions from
an empty-queue status to a non-empty status according to the
Poisson process probability of new packet arrival λ. The typi-
cal schedule-driven node operates with two timers: one for the
wake-up mode and another for sleep mode, for each node in
the network [30]. Therefore, if an abnormal event is detected
by a sensor node and needs to be transmitted to another node
or to the sink, the node stops the sleep-mode timer, turns
on its radio, and starts processing the event; otherwise, the
node remains in sleep mode. Equations 3and 6describe
the transition probability of the schedule-driven duty-cycle
node operation, including the processing and transmission
of information packets. Equations 4and 5also describe the
VOLUME 6, 2018 20377
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
non-transition probability state (i.e., the probability of having
a non-decreasing queue), which can be obtained from two
terms depending on the oldest information packets still in the
queue and winning the contention to access the channel (first
term) or otherwise (second term) [28], [31].
The proposed Markov model with a finite set of power
mode transition states for each is node defined in space S=
0,1,...,Mand transitional probabilities πm(τ) in matrix P
of being in state m(the system has mpackets), as illustrated in
Figure 7. The steady state equations for each schedule-driven
duty-cycle node operation are described as follows:
π0(τ)=π0(1 λδ)+π1βδ +o(δ) (7)
πM(τ)=πM1(τ)βδ +πM(τ)(1 β δ)+o(δ) (8)
πm(τ)=πm1(τ)βδ +πm(τ)
×(1 λδ βδ)+πm+1(τ)
×βδ +o(δ),m6= 0,m(9)
The proposed model is considered to be an irreducible, peri-
odic, and recurrent non-null Markov chain; therefore, the
model possesses the unique stationary probability 5m(τ)=
(50(τ), . . . , 5M(τ)), where 6M
m5m(τ)=1, 5m(τ)P=
5m(τ) which strictly provides that the mean rate of arrivals
per state λis less than the mean rate at which pack-
ets are obtained by the server per state β. Moreover, the
queue length process will become stable, and the number
of packets in the queue will be finite under this balanced
assumption. Thus, both the packet arrival information λand
the probability of successful transmission βfor a speci-
fied schedule-driven duty-cycle node operation in multi-hop
communication become variables in the transition matrix P.
Previously, 5m(τ) was considered to be a unique stationary
probability; hence, 5m(τ) can represent a function of both the
packet arrival information λand the probability of successful
transmission β.
Specifically, Equation 9is defined as a function that
describes the relationship among the steady state for a spec-
ified schedule-driven duty-cycle node operation, the packet
arrival information λ, and the probability of successful trans-
mission β, to assess the node in the network in varying to
access the media. The model depends on the mechanism of
protocol-specific channel access rules to obtain βfor each
node in the duty-cycle to win access to the channel, which
is obtained by the knowledge of the initial stationary steady
state matrix of the power mode for the node and the channel
access rules of the protocol. By solving the steady state equa-
tions, the stationary probabilities of the power mode and β
can be obtained. These values enable an analysis of the traffic
behavior of IoT in delay, throughput, and energy consumption
of the QoS parameters.
The model is used to investigate the behavior of two
routing categories, hop-by-hop and end-to-end transmission
schemes, shown in Figs. 5and 6to show how to optimize
protocol parameters to achieve the desired network perfor-
mance. The power transition mode of the node, described
as the wake-up period of the cycle, has a fixed length that
is determined by the MAC layer contention window size,
whereas the sleep period might be shorter or longer depend-
ing on the predefined duty-cycle, which is the ratio of the
wake-up period length to cycle length. Because there are
different protocols with different media access rules, the our
model uses RTS/CTS/DATA/ACK handshaking in addition
to ACK and message retransmission to guarantee successful
unicast multi-hop transmission.
FIGURE 9. The power profile of simplified sensor node model.
A. ENERGY CONSUMPTION MODELING
The proposed Markov chain evaluates the energy consump-
tion for multi-hop network communication by defining a
critical path-loss for considering the randomness of the hop-
distance between connected nodes. The main concepts in
terms of delay and throughput will be described below. The
proposed chain model assumes that every sensor node in the
network has the following four main power transition states
as depicted in Fig.9: transmit, receive, idle, and sleep, which
are denoted as PTX ,PRX ,Pidle, and Psleep , respectively. Fur-
thermore, Ptotal (n) is defined as the total power consumption
for these four power states during the transmission period via
nhops from the source to the sink along the selected route.
Thus,
P(n)=2Pstartup +PACKETLen
R((n1)PRX
+nPTX +2Pcir +Pidle +Pamp +Psleep) (10)
where 2Pstartup represents the power for the start-up
radio frequency (RF). The power amplifier is denoted as
Pamp =cγ(Dmax
n)α
η, where ηis the performance of the RF
power amplifier (which is widely cited by the WSN com-
munity to support a more realistic characterization of power
consumption [37]), γis defined as a sufficient SNR at the
received node, Rdenotes the transmission rate, cis a constant
value defined as proportional to the packet length, channel
attenuation, and non-linear effects of the power amplifier,
and Dmax is the total distance between the source and the
sink [39], [40].
Therefore, to transmit one DATA packet from the source
to the sink via n, it is necessary to provide to nunits
of transitional power. This operation sequentially occupies
nduty cycles; in each duty-cycle period, the total time
spent at each power state of the sensor node may be
20378 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
approximated as
lim
TimePeriod→∞ TimePeriod =TimePeriod5m(τ) (11)
The total energy consumption during the duty-cycle period
varies because the sensor node plays several roles. On one
hand, the node might be transmitting a DATA packet with a
successful transmission probability according to the steady
state of the Markov chain model. Meanwhile, the node might
be in receive mode, receiving a DATA packet successfully
with a different successful transmission probability because
each node selects its neighbors through the embedded crite-
ria in the path-loss model as opposed to random selection.
Because of a queue overflow, the node may fail to receive
a DATA packet; therefore, it goes into idle mode until it
enters sleep mode. The determination of the steady state
probabilities of the power mode depends on the regulation of
protocol for media access. Thus, the total energy consumed
at each node in the four power states is the sum of the
energy consumption of each power mode multiplied by the
corresponding steady state probability,
E(5m(τ)) =TimePeriod 5m(τ)
×(PRX +PTX +Pidle +Pamp +Psleep) (12)
Furthermore, energy consumption for a multi-hop communi-
cation network is expressed as
E(5m(τ)) =(n1)TimePeriod 5m(τ)PRX
+nTimePeriod 5m(τ)PTX
+TimePeriod 5m(τ)(Pidle +Psleep +Pamp )
(13)
The cost of the energy transition between these four states
may be obtained by multiplying the total cost of a single tran-
sition by the average number of transitions. However, the total
energy consumption at each power state and the energy cost
of transitioning from one state to another should not exceed
the total energy resource [41]. The energy consumption of
a successful hop-by-hop transmission from the source to the
sink with a four-way RTS-CTS-DATA-ACK handshake or a
three- way RTS-CTS-DATA handshake, is given as
Energyhbh =6n
(ı=1)
1
(1 PERı)E(5m(τ)) (14)
Where PER defines the packet error rate at hop ıof selected
path in specified direction as depicted in Fig. 5. Usually, the
definition model of PER assumes independence of packet
errors in different directions. This is called a poor wireless
link channel with respect to the function of distance between
two sensor nodes. However, the probability of a packet error
in any direction at hop ıis PER and no packet error is 1PER.
From the assumed independence of the errors, that is for
good wireless link channel, the probability for no error in the
n-hops from detect event toward the sink is (1 PER)n[16].
In the end-to-end retransmission scheme as depicted in
Fig. 6, the source node waits to receive either ACK or NACK
packets, which are sent only to the sink node; the interme-
diate nodes simply forward the DATA packets, and the total
expected number of transmissions may be evaluated by 1
Pst ,
where Pst =Qn
(ı=1)(1 PERı). Therefore, the total expected
energy consumption of transmitting a packet from the source
to the sink can obtained as
Energye2e=1
Pst
[
n
X
ı=1
E(5m(τ))] (15)
B. DELAY MODE LING
As a DATA packet travels from one node to a subsequent
node along the route toward the sink, it suffers from various
types of delays at each node along the selected path [42]. The
most important of these are the nodal processing, queuing,
transmission, contention, propagation, and switching delays.
The proposed model calculates the total delay by con-
sidering different delays types separately. Let dprco ,dqueue,
dtrans,dcont ,dprop, and dswit denote the processing, queuing,
transmission, contention, propagation, and switching delay,
respectively. The processing node delay depends on the net-
work data processing algorithm, but it can also include other
factors, such as the time needed to check for bit-level errors
in the packet that might have occurred in the process of
transmitting the packet bits from the upstream source node
to a neighboring node.
The queuing delay is defined as the time the DATA packet
spends in the queue as it waits for all the DATA packets ahead
of it to obtain access to the media. According to the station-
ary probability distribution 5m(t) of the proposed Markov
model, the mean queue delay of the DATA packet can be
calculated directly from Little’s result as [34].
dqueue =6
m=0m5m(τ)
λδ (16)
As previously mentioned in the model assumptions, the
queuing discipline of the data packets is first-come, first-
served (FCFS); moreover, a packet may be transmitted only
after all the packets that have arrived before they have been
transmitted. Thus, the transmission delay, which is also called
the store-and-forward delay, is defined as the amount of
time required to transmit the packets into the link, which is
given as
dtrans =S
(transmissionrateoflink)(17)
Most studies dismiss the contention delay in the calculation of
the total delay and even of the queuing delay. The contention
delay is defined as the interval from the time the DATA
packet reaches to the first queue slot to being successfully
transmitted and dropped from the queue. Moreover, for any
DATA packet newly joined the first queue slot, the discipline
will be to set a period in which it should start contending
for media access. The calculation of contention delay is thus
based on the probability of successful and/or unsuccessful
VOLUME 6, 2018 20379
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
transmission of the DATA packets during the cycle length,
dcont =T
X
ı=0
(ı+1)β(1 β)ı(18)
The propagation delay depends on two factors: the propaga-
tion speed of the physical medium of the link and the distance
between the two connecting nodes. Finally, the switching
delay is defined as the switching time between duty-cycle
node operations. Therefore, the total nodal delay is given by
Dtotal =dprco +dtrans+dqueue +dtrans +dcont +dprop +dswit
(19)
The nodal delay is given as
Dı=6n
ı=1dıprco +dıtrans +dıqueue +dıtrans +dıcont +dıprop +dıswit
(20)
For hop-by-hop transmission, the expected number of trans-
missions may be evaluated by Pı=1
1PERı; therefore, the
expected hop-by-hop delay in transmitting a DATA packet
from the source to the sink can evaluated by
Delayhbh =
n
X
ı=1
DıPı(21)
In the end-to-end retransmission scheme, the expected delay
may be calculated as
Delaye2e=1
Pst
Dı,(22)
C. THROUGHPUT MODELING
Throughput is defined as the amount of information suc-
cessfully delivered within a specified unit of time [37]. The
throughput is calculated during the duty-cycle of the node
operation on information packets within a given cycle time.
Successful transmission of information is described as the
probability of sending out information between two or more
connected nodes kthat are competing for media access in the
network which is a function of the current steady state of the
power transition mode of the nodes:
Pk(5m(τ)) =N1
k(1 5m(τ))k5m(τ)N1k(23)
In case of the use of RTS/CTS/DATA/ACK handshaking
between knodes that are competing for media access, the
probability of winning the contention and successfully send-
ing the RTS can obtained as follows:
pk=6Win
ı=1
1
win .(win ı+1
win )k,k=0,...,N1,(24)
The probability of successfully sending DATA packets is
calculated as
pk=6Win
ı=1
1
win .(win ı
win )k,k=0,...,N1,(25)
where win is the contention window size; therefore,
ps=
N1
X
k=0
Pk(5m(τ))psk (26)
and, finally, the throughput is given by a fraction of the length
of the node cycle time; that is,
Th =6
m=15m(τ)=150(τ).(27)
For multi-hop, communication, the number of nodes in the
network is N, the MAC layer DATA packets size is S, the
length of the cycle is T, and 5m(τ) is known; furthermore,
the only variable is the probability of successful DATA packet
transmission, which can be obtained according to the media
access protocol. Thus, the throughput of the network can be
determined as
ThMHC =N(1 5m(τ))ps
S
T(28)
D. WIRELESS LINK CHAN NEL MODELING
Because of the limited functionality of the sensors, power
consumption and reliability pose the greatest challenges
in IoT’s applications, such as how sensor nodes may be
deployed in varied locations with different deadlines using
more flexible and low-cost wireless links [43]. Different
sets of models, such as the link quality model, energy-
transmission model, and interference model, initially focused
on both designing and analyzing real-time routing protocols
for WSNs [44].
Several studies of wireless link channels partition the area
into the following three main regions: connected, transitional,
and disconnected regions. The transitional region is regarded
as the main region because: (1) it is quite significant in size
and generally characterized by asymmetric connectivity and
high-variance in reception packets rates [33], and (2) it has
contributed to the understanding of the realistic link layer
model for WSNs.
The definition of the transitional region results in dense
and random deployment of sensor devices for IoT to mon-
itor, sense, and control the environment, to perform local
processing, and to communicate the results with a base sta-
tion. It is important to understand how the channel can be
analyzed to determine the width of the transitional region
and how altering both can affect the extension of the transi-
tional region [45], which prospectively influences the traffic
behavior of MHC. Hence, the proposed green model for IoT
requires an abstraction to illustrate the log-normal path loss
channel and to present an approximate expression of the
variance in PER with respect to the function of distance;
with the Radio Irregularity Model (RIM), this abstraction also
provides the packet received ratio (PRR) as a function of the
SNR in MHC, as described in the subsequent sections.
E. THE LOG-NORMAL PATH-LOSS MODEL
Propagation phenomena are diverse and complex because of
the separation of the receiver and the transmitter, in addition
20380 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
FIGURE 10. The average RSSI reading as a function of the distance
(on logarithmic scale).
FIGURE 11. The SNR reading as a function of the distance (on logarithmic
scale) for MHC.
to having a random variation in power due to path shadowing,
which causes the signal strength to decay exponentially with
respect to the separation. Radio signal strength attenuates as
a function of distance as shown in Fig. 10. Basically, the
graph is composed of 1942 average received signal strength
indication (RSSI)-log distance which is expressed in meter
between sending and receiving sensor nodes. However, RSSI
is not only affected by the distance but also there are a few
other factors that affect the radio signal propagation and
hence the RSSI value that is perceived by a sensor node at an
equal distance [46]. Figure 11 shows the SNR as a function
of distance between all sending and receiving sensor nodes.
The measured SNR does not decrease monotonously with
distance, a SNR of 17dB is measured and at the shorter dis-
tance about of 18 meter. Let there be two links, both of them
affected by channel fading, and the measured SNR is 17dB
for one link and 25dB for the other. Thus, it could happen
that when the channel fading is affects both links in the same
way, the SNR of the former link falls to 5dB whereas the latter
still enjoys an SNR of 17dB, which is a more efficient link
quality than the former. Therefore, to compare the two links,
it is useful to account for their estimated SNR values, which
are considered to be very helpful in refining the link quality
judgment [16]. Furthermore, at the distance of 20 meter the
expected value of a RSSI is 95dBm. The measured signal
is much too strong, even at small distance. This effect occurs
in all experimental studies in industrial indoor environments
and cannot be explained by many theories [1], [47].
There are various propagation models for path loss, which
can be categorized into the following three types: (1) empir-
ical models, (2) deterministic models, and (3) stochastic
models. For the free space model a path loss proportional
to the square of the distance, for example the IEEE802.15.4
standard recalculates this path loss for MHC as
PL(dı)=PL (d0)+20ξlog10 (dı)
for dı<8meter,ı=1,2,...,n(29)
PL(dı)=PL (d0)+33ξlog10 (dı
8)
for dı>8meter,ı=1,2,...,n(30)
FIGURE 12. The path-loss versus the log-distance for MHC.
where PL(d0) is the path loss at a reference distance d0in
meter. Existing techniques either consider the path-loss is
known apriori by assuming the WSN environment is free
space, or obtain the path-loss through extensive channel
measurement and modeling by measuring both SNR and
distances in the same environment of WSN prior to system
deployment [48]. However, an accurate knowledge of the
path-loss is required in order to obtain an accurate esti-
mate of the inter-sensor distance from the corresponding
SNR measurement. Figure 12 shows the path-loss ver-
sus the log-distance for MHC approximated by linear
VOLUME 6, 2018 20381
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
regression [49]. Whenever the path-loss exponent ξis
increased, the node consumed energy with the increased node
distance and becomes more incomparable with the energy
received because of the exponential increasing. Meanwhile,
we defined the mean path loss as a function of the hop-
distance in relation to the power of the path-loss exponent ξ
as PL(dı)(dı
d0)ξ. Therefore, our model assumes that there
are multi-hop nodes between the source and the sink; thus, the
relationship between the path-loss and the multi-hop-distance
may be obtained by
PL(dı)=PL (d0)+20ξlog10 (dı
d0
)+ε(0, σ ),
ı=1,2,...,n(31)
Therefore, the received power PRX in dB is given by
PRX =PTX (dıPLdı) (32)
where PTX (dı) is defined as the output power, PL(dı) is the
power decay at a reference distance close to the transmitter,
d0TR can be 1 millimeter to 20 meters, ξis the path-
exponent, and ε(0, σ ) is defined as a zero-means Gaussian
distribution with standard deviation σ. Consequently, the
MHC path-loss derived by linear regression is given
as [1] and [49]
2(n, ξı)=6n
ı=1(PL(dı)PL (dn)
10ξılog10(dı
d0
))2,ı=1,...,n(33)
The number of hops nand the values of the path exponent
for each hop may be obtained through a linear regression
algorithm. The algorithm begins with the calculation 2(n, ξı)
and is repeated with an increasing number of hops. Finally,
the number of hops and the values of the path-exponent can
be obtained. We assume that the criterion of hop-distance
depends on the maximum power level of a specific output
power PTX , i.e., 0 PTX Pmax , therefore the transmission
range (TR) of any single-hop-distance is given as
TR =ξ
s(PRX +PTX )η
(1 21ξ),(34)
This indicates that there is a relationship between the
log-normal path-loss and transmission range, in which the
hop-distance can be attributed to the influence of the path-
loss because of the various propagation environments [1].
Figure 12 shows the exponential impact of the path loss on
the distance between the transmitter and the receiver.
With this knowledge of PTX(dı), the path-loss, and the
receiver sensitivity Sr, the PRX and SNR values can be
estimated to redefine the PER expression, which is derived
from equation (36) and corresponds to the Non-Coherent-
Frequency-Shift-Keying (NC-FSK) modulation used in early
WSN platforms, such as MICAZ [50]. Thus, SNR is given as
SNR =PTX PL(d0)10ξlog10 (dı
d0
)+ε(0, σ )Sr,
=PTX PL(dı)Sr,ı=1,...,n(35)
Certain companies supply the PER value for developers. For
example, the Zigbee implementation on the MC13213 chip
produced by FreeScale Semiconductor Inc. [51] has a basic
function called MLME Link Quality that developers can use
to obtain the current PER. The link quality of a hop from the
source to the sink is represented by the PER so the value of
the transmission of packets on each hop from the source to the
sink may be defined as an independent event that circumvents
the complexities of retransmission and requires no coding;
thus, the single BER leads to packet error. Therefore, the
successful delivery probability for one packet may be closely
approximated for MHC as [52].
PRR =(1 PER)n(36)
where
PER =(1 1
2exp(SNR
2
1
0.64 ))ρ.8.f.n(37)
For each correctly received packet, fdata frame length
is received over the time period f
bitrate , where bitrate =
numberofbitspersymbol.symbolrate, and ρis the encoding
ratio that used NRZ and Manchester encoding operations.
F. NON-ISOTROPIC TRANSMISSION MODELING
In addition to line-of-sight, signal propagates by the means
such as diffraction, reflection, scattering, etc., which exhibit
radio irregularity patterns that might influence the commu-
nication performance of WSNs in most environments. Radio
irregularity is considered to be a common and non-negligible
propagation phenomenon that arises because of several fac-
tors and results in irregularity in TR and diversity in packet
reception in various antenna directions; such irregularity has
significant direct or indirect influences on MAC because
of asymmetric radio collision or interference and on many
aspects of the upper layer performance because of asymmet-
ric paths [53], [54].
In generally, there are two factors that cause radio irregu-
larity: the device properties and the propagation media. The
device properties includes antenna characteristics, such as
antenna gains, radiation patterns, polarization, etc., the effects
of which Hasan et al. [1] investigated under three differ-
ent scenarios of antenna operations for path-loss prediction,
which has become a key issue in system design.
Unlike existing models, our model proposes that the radio
propagation model approximates the anisotropic property to
present the effects of irregularity on the localization tech-
nique for correct MAC in case of asymmetric radio collision
between two nodes, which occurs when a node is unable to
reserve the wireless channel, and on the routing performance
through paths asymmetrical to the traffic behavior in the two
transmission schemes.
Our model is defined as geo-directional-cast forwarding
that consists of a set of mathematical expressions that rep-
resent the non-isotropic log-normal path-loss with differ-
ences in transmit power levels of 0dB,3dB,5dB,7dB,
10dB,and 25dB as well as IEEE802.15.4 respectively
20382 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
FIGURE 13. SNR multipath fading for MHC at different power levels.
according to Equations 29 and 30 as shown in Fig. 13. The
SNR in function of distance on a semi-logarithmic scale
on multi-hop sensor nodes varies according to the propaga-
tion direction from the node to its neighbor. Figure 13 was
constructed with the average of decreasing SNR values of
packets sent at different power levels from sensor nodes to the
sinks. Figure 14 shows an example of a geo-directional-cast
mechanism where the source node forwards DATA packets
to only one-neighbor to reduce packet flooding and minimize
collision. Therefore, SNR is given as
SNR =PTX PL(dı)+fading,ı=1,...,n(38)
FIGURE 14. Geo-directional-cast forwarding for network topology.
The DOI is introduced to denote the irregularity of the
radio pattern. The DOI is based on the distance over which
one node can hear its neighbor. It is defined as ‘‘the maxi-
mum path loss percentage variation per unit degree change in
the direction of radio propagation’’ [53]. Figure 15 depicts
the radio irregularity pattern at various degrees. It can be
observed that when the DOI is zero, the transmission range
is considered as a perfect sphere, whereas any continued
increase in the DOI value causes the transmission range to
FIGURE 15. Degree of irregularity.
have an increasingly irregular radio pattern. To establish a
radio irregularity model, our propagation model relies on
real-data values (which have been repeatedly used in many
experiments on MICAZ sensor nodes in vehicle tracking sys-
tems) to approximate the radio irregularity by calculating the
corresponding DOI values [54]. To reflect the path-loss for
a specified angle toward an optimal forward direction from
the next hop to the nearest neighbor, a new coefficient Kis
defined; directional forwarding is used to adapt the bounds
of the DOI model value between an upper and a lower signal
propagation for the two categories of retransmission schemes
by adjusting the path-loss for forwarding to the next nodes
that have the best progress toward the sink in the specified
direction.
Beyond the upper bound, all neighbors are outside TR;
within the lower bound, all neighbors are guaranteed to be
within the inner TR, and the signal is strong enough to be
received correctly. Therefore, it becomes more critical to
carefully select the sensor nodes that participate in sensor
transmission range forwarding the information against its
resource consumption. Thus, the DOI modeling results are
given as
SNR =PTX DOIAdjustedPL(dı)+fading,(39)
where DOIAdjustedPL(dı)=PL (dı).KıSpecifically, Kis
defined as the adjustable th coefficient used to adjust the
path-loss value according to a specified direction, which is
calculated as
K=(1 if =0
K1±Rand.DOI if 0 <  < 360 N
(40)
where |K0K359| ≤ DOI .
Thus, minimizing the amount of energy consumption and
adjusting the transmission range as much as possible through
observed DOI value can significantly prolong the lifetime of
sensor network.
V. THE SENSOR NETWORK SCENARIO
AND SIMULATION RESULTS
This section analyzes the QoS parameters obtained from the
analysis of the effects of radio irregularity on a multi-hop net-
work routing protocol with the focus on realistic scenarios at
VOLUME 6, 2018 20383
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
low densities considering effects that might have a significant
influence on the performance of geographic routing in IoT.
In the following subsection, the average packet delay trans-
mission, the energy consumption for transmission, and the
throughput are analyzed in view of the impact of radio irreg-
ularity on the neighbor-discovery routing technique of both
the hop-by-hop and end-to-end retransmission schemes. The
parameters used in proposed model are presented in Table 3
MATLAB was used to construct a network topology consist-
ing of MICAZ sensor nodes in uniform Poisson distribution
over a short interval and implemented according to non-
coherent-frequency shift keying modulation with NRZ and
Manchester encoding.
TABLE 3. Experiments parameters.
A. THE SENSOR NETWORK SCE NARIO
Generally, there are two core issues for automated surveil-
lance : (1) object detection, and (2) tracking. Object detection
is an important capability for senor networks. Several events
that smart surveillance system in IoT has to detect in real-
time such as abandoned object alert, motion, object removal,
and observation of any other abnormal behavior. Supporting
a smart surveillance system has several additional charac-
teristics, challenges and factors which influence the design
of distributed smart devices in IoT. These characteristics,
challenges and factors are depend on the real-time multimedia
traffic data flow from surveillance cameras. Moreover, the
quality of the video depends on the lightning conditions.
Additionally, any suspicious activity or behavior should be
detected at the time of occurrence. Therefore, object detection
must be performed in real-time. But the strongly challenging
problem for detection in IoT is the optimization allocation of
sensing and communication resources to multiple competing
detection tasks spawned by emerging stimuli which is more
computationally complex and resource demanding.
This study provides a dynamic mean to define and form
a group of sensor nodes based on task requirements of
resources in response to external events and dynamic resource
allocation availability. The MICA2 sensor characteristics and
network topology scenario were defined with the following
assumptions under the SDST application behavior:
1) From a sensing and information processing, a vehicle
physical process model need to be designed which
correspondence to some real processes such as spatial
correlation of information, and variability over time.
The basic model is defined for each sensor node in
wireless network as a tuple Sn=ν, E,Pυ,PEwhere
νand Especify a network topology with its sensor
nodes νand link connectivity Eν.Pυis a set of
functions which characterizes the properties as a static
values of each sensor node in ν, including the sen-
sor location, computational capabilities, sensor output
type, and sensing modality.
PEspecifies the properties of each link such as the
quality of link. All sensor nodes have a common trans-
mission range TR, and are deployed with a uniform
distribution. The average number of nodes within an
area of TR conforms to a Poisson distribution with
parameter λwhere λ=5.TR2density. Thus, the
probability that νnodes cover an arbitrary geometric
point is λn
ν!expλ, ν =1,2, . . .. Furthermore, any pair
of sensor nodes in the network topology, for example,
(u, υ)ν, can obtain connectivity with another pair if
they are within TR from each other. The model assumes
that when a sensor node discovers a neighboring node,
a wireless link Eis established between them, and the
sensor can set the level of transmission power to be used
over that link. Furthermore, the model assumes that the
sensors can adjust their transmission range toward a
neighbor over time.
FIGURE 16. An object detection scenario showing moving vehicle sin
field of sensors.
2) The construction of the network topology as shown in
Fig. 16 depends on the location management, which
determines the localized information of uniformly
deployed sensor nodes. It assumes that all MICAZ
sensor nodes are in a fixed position and that the sink
is at the origin (0,0).
For multi-hop communication, there is at least a sin-
gle path connecting the source to the sink through
20384 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
intermediate sensor nodes, issuing network commands
such as ‘‘sleep,’’ ‘‘idle,’’ and ‘‘wake-up;’’ changing the
level of transmit power; and synchronizing the trans-
mission time. The object detection used to bring out
key of SDST application. As vehicles move along the
road with specific velocity, the information is stored at
a sensor node called leader node which are numbered
as 1,2,3,4,and 5.
These leader sensor nodes collect information from
their neighbor’s relevant intermediate sensor nodes and
forwarded toward direction of next hop of the nearest
leader sensor node. As the vehicle moves or environ-
mental conditions vary, the leaderships may change
hands among sensor nodes. Therefore, the movement
of leader between the sensor nodes may lead to design
localized communication, reducing overall commu-
nication and increasing the lifetime of the network.
Figure 17 shows with the average SNR values of the
leader sensor nodes that sent packets at different power
levels and the distance between the sensor nodes are
known.
3) Vehicles are simulated as sources of abstract sensor
readings that can be detected in sensing range of the
sensor node. Each reading is dispersed in the simulation
over at a certain point and at a certain time as
Y(p,t)=X
allvehiclesı
(Yı(τ))
(Kdı(τ)+1)a(41)
where Y(p, τ ) denotes the position of the vehicle phys-
ical process at a certain point pand at a certain time
τ, Yı(τ) denotes the position of ıth vehicle at time τ
and called snapshot of vehicle which help to defined
the maximum possible number of pickup of snapshots
for all vehicle s,dı(τ) denotes the distance of the
ıth vehicle from the sensor node at time τin the
sensing range.
The parameters Kand adefine multiplicative constant
and attenuation constant parameters that are set to 0.1
and 1 respectively and are used to determine the value
from a diffused vehicle. The road itself is specified in
the simulation space by two points given by their xand
ycoordinates. These points are set by the following four
parameters of a vehicle physical process model. Each
time a vehicle needs to enter in a highway, it follows a
Poisson Arrival process with a specific velocity along a
straight or non-straight route as shown in Fig.16. Each
sensor generates DATA packets of a constant size at a
given rate when the sensor node senses an anomalous
event; the first DATA packets are buffered while await-
ing their transmission to another node. As previously
mentioned, the queuing discipline of DATA packets is
stable and modeled as FCFS.
4) Each node tries to access the channel through a hand-
shaking technique, which is designed to resolve hidden
and exposed terminal problems. However, when the
source node generates DATA packets for transmission,
FIGURE 17. SNR versus logarithmic distance for MHC at different power
levels.
it sends a RTS message to its neighbor, which responds
with a CTS message. If the DATA packet transmission
is successful, then the source node receives ACK from
its neighbor in the hop-by-hop transmission scheme,
or an ACK message from the sink in the end-to-end
transmission scheme.
5) The proposed model depends on Lee’s model calcula-
tion [55] to estimate the path-loss since it is the most
commonly used in urban areas
The expected hop-distance can be attributed to the
influence of the path-loss due to the various propagation
directions, which is a function of SNR for different power
values PTX , as shown in equation (38). It can be observed
from Fig. 13 that for small values of SNR, the average hop-
distance value increases because nodes with a high-quality
channel may be chosen as the next hop in the specified direc-
tion, and other nodes may become the next hop. Therefore,
for the perfect route, the smallest number of hops to a sink
decreases for smaller SNR values.
B. THE IMPACT OF RADIO IRREGULAR ITY ON ENERGY
CONSUMPTION IN THE MULTI-HOP
ROUTING PROTOCOL
Generally, in most single-hop and multi-hop communication
scenarios, when TR Dmax the single-hop strategy is con-
sidered more energy efficient within the range of the radio.
Particularly, given the low-path-loss exponents because the
distance for one hop is close to the perfect value of SNR,
and the start-up power overhead makes the multi-hop strategy
inefficient for hop-distances of less than Dmax. Consequently,
increasing the amount of energy exhaustion of nodes that is
closer to the sink [56].
The supposed reduction in energy consumption is con-
sidered a central issue and should be analyzed in terms
of reducing the power overhead through a new strategy
for multi-hop communication when TR >Dmax ; thus,
multi-hop communication would become more attractive
VOLUME 6, 2018 20385
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
when the number of hops is denoted as the upper integer or
ceiling of TR
Dmax [16].
Consequently, the power overhead in retransmission
schemes must be properly analyzed. Figure 17 presents the
amount of energy consumed when the power overhead is
reduced for each individual node using a discrete M/M/1
Markov queuing model that synchronizes the power overhead
in sending and receiving control messages within a sensor
network. Results are shown for the hop-distances between
nodes for the two retransmission schemes with various PER.
The model is useful for most real-time IoT’s applica-
tions, particularly when the traffic load is heavy and changes
over time. When the PER is very low, the energy consump-
tion increases because the power amplifier must consume
more energy to guarantee a smaller PER as seen in equa-
tion (10). The average energy consumption in the hop-by-hop
retransmission scheme shows more energy efficiency than
the end-to-end retransmission scheme because the Markov
chain results in a more efficient synchronization overhead in
sending and receiving control messages in the hop-by-hop
retransmission scheme than in the end-to-end retransmission
scheme; it actually introduces more delay in the former than
in the latter scheme. Suppose a sensor camera detects an
abnormal event. Heavy traffic would be generated, and some
sensor nodes might have little or no chance to transit to the
sleep mode. In contrast, the sensor node has more sleep time
when there is little traffic. Therefore, increasing the traffic
sensor nodes results in fewer chances to go into sleep mode
and thus consumes more energy, whether in the hop-by-hop
or end-to-end scheme.
The effects of coding and channel access control on energy
consumption in various retransmission schemes are also com-
pared. The average energy consumption in the hop-by-hop
retransmission scheme is approximately 35.8%, whereas in
the end-to-end retransmission scheme, the value is approxi-
mately 65.9% because the DATA packet errors are not thrown
out until the packets are received by the sink in end-to-end
retransmission, which leads to higher energy wastage; such
wastage is avoided in hop-by-hop retransmission. Figure 18
depicts the energy efficiency of both retransmission schemes,
indicating that non-return-to-zero (NRZ) encoding is more
efficient than the Manchester encoding operation because
the word error probability in encoding DATA packets with
the Manchester coding operation is larger when the number
of hops is increased by virtue of duplication of the DATA
packets.
C. THE IMPACT OF RADIO IRREGULAR ITY ON AVERAGE
ENERGY CONSUMPTION IN THE MULTI-HOP
ROUTING PROTOCOL
Figure 19 shows the average delay versus the hop-distance
from the source to the sink with various PER for the
two retransmission schemes. The results indicate that the
delay first hover near zero and then increases as the hop-
distance increases; the maximum delay in hop-by-hop trans-
mission with NRZ coding is larger than that in end-to-end
FIGURE 18. The energy consumption of both retransmission schemes.
FIGURE 19. The average delay in retransmission schemes.
retransmission with the same coding operation. When the
number of hops increases to more than two, the average
maximum delay in the end-to-end retransmission scheme
becomes approximately 31.9% which is less than that in
hop-by-hop retransmission. Whenever the number of hops is
increased, the maximum delay in end-to-end retransmission
improves to approximately 41.8% (compared to approxi-
mately 47.9% in hop-by-hop retransmission) because every
intermediate node must transmit ACK/NACK packets in hop-
by-hop retransmission, which leads to more traffic pattern
transmissions and the increased delay compared to the end-
to-end retransmission scheme. Compared to NRZ encod-
ing, Manchester encoding causes a larger increase in delay
in both retransmission schemes because the packet length
is double-decoded in the operation; therefore, the reliabil-
ity of the Manchester-coded packet length mis given as
(1 PER)m[57]. The M/M/1 queuing model can deliver
20386 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
incoming DATA packets as soon as they arrive at the network;
thus, only a few DATA packets are accumulated in the queue
for sending in every duty cycle (as illustrated in Fig. 7), and
the DATA packet delay begins with a small value. Therefore,
when transmitting DATA packets with specific probability β
and with stationary distribution 5m, increasing the number
of hops might cause an increase in the contention delay for
each sensor node in the network before the associated RTS
message is sent. Consequently, the average queue length,
which cannot be longer than the queue capacity m, increases
in response to the dropping of DATA packets from the queue;
otherwise, the retransmission is defined indefinitely as .
Therefore, the average delay increases according to equa-
tion (18) as the number of hops increases because βand 5m
are more sensitive to an increase in the number of hops; this
can be attributed to the influence of the path-loss as a result
of the various propagation directions.
D. THE IMPACT OF RADIO IRREGULAR ITY ON
THROUGHPUT IN THE MULTI-HOP
ROUTING PROTOCOL
Figure 20 depicts the amount of DATA throughput versus
the hop-distance from the source to the sink with various
PER for the two retransmission schemes. The results illustrate
the changes in the throughput trend, where the hop-distance
between the source and the sink increases with larger SNR,
and the data rate becomes limited by the number of hops; the
throughput bounds for each retransmission scheme decrease.
In turn, the MAC delivers all DATA packets as soon as
they arrive at the network; therefore, the throughput starts
with an increasing trend. When the MAC reaches its delivery
limit and can no longer deliver more incoming DATA packets
from other nodes in the network, the DATA packets become
backlogged in the queue, which may eventually overflow, as
shown in Fig. 8. This consequently causes βto decreases
with the increasing number of hops toward the sink, and more
packets are ultimately dropped because of queue overflow or
collision during retransmission.
The DATA packets sent over the multi-hop network
are encoded with Manchester operation. The throughput is
approximately 63.3kbps in end-to-end retransmission and
41.3kbps in hop-by-hop retransmission, which represents a
larger decrease compared to that of the NRZ encoding oper-
ation (approximately 163kbps in end-to-end retransmission
and 139.6kbps in simple hop-by-hop retransmission). This
results because the duplication of DATA packets might have
a significant effect on the efficiency of the system and the
specified network capacity.
As previously mentioned, the proposed model uses the
probability of successful transmission; therefore, the dupli-
cation of DATA packet lengths decreases the probability of
successfully receiving DATA packets and increases the prob-
ability of error rates. Thus, the transmission power varies for
each link between connecting nodes, (1 PER)n, where the
value of PER depends on the SNR and on the modulation and
encoding method used [56]. However, this is only true if the
FIGURE 20. The average throughput in both retransmission schemes.
packet overhead is not taken into consideration; otherwise,
the throughput of the system approaches zero per hop. Thus,
the optimal packet length must be considered to obtain the
highest energy efficiency and network capacity. However, if
TR were to decrease too much, network connectivity would
be compromised, and the average per-node throughput would
drop considerably, as shown in Fig. 20. Therefore, setting the
common power transmission level to the minimum value to
achieve full network connectivity is the optimal choice for
increasing network throughput.
VI. CONCLUSIONS
In order to eliminate the impact of the single-hop routing
strategy in IoT with uniform sensor nodes distribution. A fun-
damental question raised is whether it is advantageous to
route over many short hops or over a smallernumber of longer
hops? Multi-hop geographic routing green schemes propose
that transmit as far as possible and outperform nearest-
neighbor routing models. Clearly, the benefits of short-hop
routing include the SNR gain to choose the furthest neighbor
that can be reached with sufficient reliability. Therefore, this
paper provides a comprehensive study for the hypothesis that
the effective anisotropic radio properties inside the deploy-
ment scheme for green IoT that have an influence on the
components of energy consumption and on the traffic system
model.
A traffic system model is developed with a Markov
discrete-time M/M/1 queuing model for duty-cycle nodes
to describe the green behavior of two categories of transmis-
sion schemes and to derive the stationary distribution of the
probability of packet transmission for both schemes to inves-
tigate the performance of the duty cycle of the MAC layer
in analyzing the delay, throughput, and energy consumption
simultaneously.
The results are compared with hop-by-hop and end-to-end
retransmission with NRZ and Manchester encoding opera-
tions. Based on the results, the messages encoded with NRZ
VOLUME 6, 2018 20387
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
have more efficiency over multi-hop networks (without loss
of connectivity between nodes) than those encoded with the
Manchester operation. The findings presented in this work are
of great help to designers of WSNs. The implications of this
study for future works are as follows:
It is important to match the optimal modulation scheme
and encoding operation with the expected hop- distance
and channel-noise model to guarantee the efficient usage
of limited sensor residual energy and to achieve a long
network lifetime. This match should be obtained using
heuristic algorithms, which are considered to be very
promising and effective in designing multi-hop geo-
graphic routing schemes.
The use of heuristic algorithms is recommended to
find the optimal angle for the nodes chosen around
the optimum distance with some probability; the opti-
mum transmit energy would likely change according
to the geographic routing schemes chosen for the relay
sensor nodes to meet QoS requirements. The proposed
model should also be applied to actual hardware devices
and implemented in realistic scenarios (such as Motes,
MICAZ, or Libelium waspmote) and the results should
be evaluated.
REFERENCES
[1] M. Z. Hasan, H. Al-Rizzo, and F. Al-Turjman, ‘‘A survey on multi-
path routing protocols for QoS assurances in real-time wireless multi-
media sensor networks,’IEEE Commun. Surveys Tuts., vol. 19, no. 3,
pp. 1424–1456, 3rd Quart., 2017.
[2] C. Zhu, V. C. M. Leung, L. Shu, and E. C.-H. Ngai, ‘‘Green Internet of
Things for smart world,’IEEE Access, vol. 3, pp. 2151–2162, Nov. 2015.
[3] Z. H. Mir and Y.-B. Ko, ‘‘Collaborative topology control for many-to-
one communications in wireless sensor networks,’IEEE Access, vol. 5,
pp. 15927–15941, 2017.
[4] M. Z. Hasan and F. Al-Turjman, ‘‘Optimizing multipath routing with
guaranteed fault tolerance in Internet of Things,’IEEE Sensors J., vol. 17,
no. 19, pp. 6463–6473, Oct. 2017.
[5] T. Soyata, L. Copeland, and W. Heinzelman, ‘‘RF energy harvesting for
embedded systems: A survey of tradeoffs and methodology,’’ IEEE Cir-
cuits Syst. Mag., vol. 16, no. 1, pp. 22–57, 1st Quart., 2016.
[6] M. Z. Hasan and T.-C. Wan, ‘‘Optimized quality of service for real-time
wireless sensor networks using a partitioning multipath routing approach,’
J. Comput. Netw. Commun., vol. 2013, Apr. 2013, Art. no. 497157.
[7] F. M. Al-Turjman, A. E. Al-Fagih, and H. S. Hassanein, ‘‘A novel cost-
effective architecture and deployment strategy for integrated RFID and
WSN systems,’’ in Proc. IEEE Int. Conf. Comput., Netw. Commun. (ICNC),
Feb. 2012, pp. 835–839.
[8] G. Singh and F. Al-Turjman, ‘‘A data delivery framework for cognitive
information-centric sensor networks in smart outdoor monitoring,’Com-
put. Commun. J., vol. 74, no. 1, pp. 38–51, 2016.
[9] F. M. Al-Turjman and H. S. Hassanein, ‘‘Enhanced data delivery frame-
work for dynamic information-centric networks (ICNs),’’ in Proc. IEEE
38th Conf. Local Comput. Netw. (LCN), Oct. 2013, pp. 810–817.
[10] M. Z. Zamalloa, K. Seada, B. Krishnamachari, and A. Helmy, ‘‘Efficient
geographic routing over lossy links in wireless sensor networks,’’ ACM
Trans. Sensor Netw., vol. 4, no. 3, pp. 1–33, 2008.
[11] F. Al-Turjman, ‘‘Energy-aware data deliveryframework for safety-oriented
mobile IoT,’’ IEEE Sensors J., vol. 18, no. 1, pp. 470–478, Jan. 2018.
[12] K. Almi’ani, A. Viglas, and L. Libman, ‘‘Tour and path planning methods
for efficient data gathering using mobile elements,’’ Int. J. Ad Hoc Ubiq-
uitous Comput., vol. 21, no. 1, pp. 11–25, 2016.
[13] K. Zeng, W. Lou, and M. Li, Multihop Wireless Networks: Opportunistic
Routing, vol. 25. Hoboken, NJ, USA: Wiley, 2011.
[14] L. Shu, ‘‘Geographic routing in duty-cycled industrial wireless sensor
networks with radio irregularity,’’ IEEE Access, vol. 4, pp. 9043–9052,
2016.
[15] Y. Wang, M. Vuran, and S. Goddard, ‘‘Cross-layer analysis of the end-
to-end delay distribution in wireless sensor networks,’’ IEEE/ACM Trans.
Netw., vol. 20, no. 1, pp. 305–318, Feb. 2012.
[16] F. M. Al-Turjman, H. S. Hassanein, W. M. Alsalih, and M. Ibnkahla,
‘‘Optimized relay placement for wireless sensor networks federation in
environmental applications,’’ Wireless Commun. Mobile Comput., vol. 11,
no. 12, pp. 1677–1688, 2011.
[17] Libelium Comunicaciones Distribuidas S.L.. Accessed: Sep. 2013.
[Online]. Available: http://www.libelium.com
[18] J. Long, ‘‘Reliability guaranteed efficient data gathering in wireless sensor
networks,’IEEE Access, vol. 3, pp. 430–444, 2015.
[19] S. Ehsan and B. Hamdaoui, ‘‘A survey on energy-efficient routing
techniques with QoS assurances for wireless multimedia sensor net-
works,’IEEE Commun. Surveys Tuts., vol. 14, no. 2, pp. 265–278, 2nd
Quart., 2012.
[20] M. C. Vuran and I. F. Akyildiz, ‘‘Cross-layer analysis of error control in
wireless sensor networks,’’ in Proc. SECON 3rd Annu. IEEE Commun. Soc.
Sensor Ad Hoc Commun. Netw., Sep. 2006, pp. 585–594.
[21] M. Busse, T. Haenselmann, and W. Effelsberg, ‘‘Energy-efficient forward-
ing schemes for wireless sensor networks,’’ presented at the Int. Symp.
World Wireless, Mobile Multimedia Netw., 2006.
[22] A. Antonopoulos and C. Verikoukis, ‘‘Network-coding-based cooperative
ARQ medium access control protocol for wireless sensor networks,’Int.
J. Distrib. Sensor Netw., vol. 8, no. 1, p. 601321, 2011.
[23] H. Feng and L. J. Cimini, Jr., ‘‘On the optimum number of hops in a multi-
hop linear network with randomly located nodes,’’ in Proc. ICC, 2012,
pp. 2329–2333.
[24] M. Sikora, J. N. Laneman, M. Haenggi, D. J. Costello, Jr., and T. Fuja,
‘‘On the optimum number of hops in linear wireless networks,’’ in Proc.
IEEE Inf. Theory Workshop, Oct. 2004, pp. 165–169.
[25] M. Z. Hasan, F. Al-Turjman, and H. Al-Rizzo, ‘‘Optimized multi-
constrained quality-of-service multipath routing approach for multime-
dia sensor networks,’IEEE Sensors J., vol. 17, no. 7, pp. 2298–2309,
Apr. 2017.
[26] M. Holland, T. Wang, B. Tavli, A. Seyedi, and W. Heinzelman, ‘‘Optimiz-
ing physical-layer parameters for wireless sensor networks,’ACM Trans.
Sen. Netw., vol. 7, no. 4, pp. 1–28, 2011.
[27] J. Abouei, K. N. Plataniotis, and S. Pasupathy, ‘‘Green modulations in
energy-constrained wireless sensor networks,’’ IET Commun., vol.5, no. 2,
pp. 240–251, 2011.
[28] J. P. Lehoczky, ‘‘Real-time queueing network theory,’’ in Proc. 18th IEEE
Real-Time Syst. Symp., Dec. 1997, pp. 58–67.
[29] D. Hu and J. P. Wu Fan, ‘‘Optimum number of hops of linear multi-hop
networks with equidistant relays,’’ in Proc. Int. Workshop High Mobility
Wireless Commun. (HMWC), 2014, pp. 174–177.
[30] S. Huimin, L. Zhonghai, A. Jantsch, Z. Dian, and Z. Li-Rong, ‘‘Analytical
evaluation of retransmission schemes in wireless sensor networks,’’ in
Proc. IEEE 69th Veh. Technol. Conf. (VTC Spring), Apr. 2009, pp. 1–5.
[31] F. Zhao, J. Liu, J. Liu, L. Guibas, and J. Reich, ‘‘Collaborative signal and
information processing: An information-directed approach,’Proc. IEEE,
vol. 91, no. 8, pp. 1199–1209, Aug. 2003.
[32] H. S. Mahdi and I. J. N. Y. Mohammed Baithoon, ‘‘Low transmission delay
based on LEACH and PEGASIS protocol in WSN,’’ Int. J. Comput. Sci.
Mobile Comput., vol. 4, no. 5, pp. 209–217, 2015.
[33] M. Zuniga and B. Krishnamachari, ‘‘Analyzing the transitional region
in low power wireless links,’’ in Proc. IEEE SECON 1st Annu. IEEE
Commun. Soc. Conf. Sensor Ad Hoc Commun. Netw., Oct. 2004,
pp. 517–526.
[34] M. E. Woodward, Communication and Computer Networks: Modelling
With Discrete-Time Queues. Los Alamitos, CA, USA: IEEE Computer
Society Press, 1994, pp. 79–81, ch. 4.
[35] M. Z. Hasan and F. Al-Turjman, ‘‘Evaluation of a duty-cycled asyn-
chronous X-MAC protocol for vehicular sensor networks,’’ EURASIP J.
Wireless Commun. Netw., vol. 2017, no. 1, p. 95, 2017.
[36] M. Z. Hasan, H. M. Al-Rizzo, M. Günay, ‘‘Lifetime maximization by
partitioning approach in wireless sensor networks,’EURASIP J. Wireless
Commun. Netw., vol. 2017, no. 1, p. 15, 2017.
[37] Y. Ou and W. Heinzelman, ‘‘Modeling and performance analysis for duty-
cycled MAC protocols with applications to S-MAC and X-MAC,’’ IEEE
Trans. Mobile Comput., vol. 11, no. 6, pp. 905–921, Jun. 2012.
[38] S. Cui, A. J. Goldsmith, and A. Bahai, ‘‘Energy-constrained modu-
lation optimization,’IEEE Trans. Wireless Commun., vol. 4, no. 5,
pp. 2349–2360, Sep. 2005.
20388 VOLUME 6, 2018
M. Z. Hasan et al.: Analysis of Cross-Layer Design of QoS Forward Geographic Wireless Sensor Network Routing Strategies
[39] W. Qin, M. Hempstead, and W. Yang, ‘‘A realistic power consumption
model for wireless sensor network devices,’’ in Proc. 3rd Annu. IEEE
Commun. Soc. Sensor Ad Hoc Commun. Netw. (SECON), Sep. 2006,
pp. 286–295.
[40] C. Taddia and G. Mazzini, ‘‘On the retransmission methods in wireless
sensor networks,’’ in Proc. IEEE 60th Veh. Technol. Conf. (VTC-Fall),
Sep. 2004, pp. 4573–4577.
[41] D. Jung, T. Teixeira, and A. Savvides, ‘‘Sensor node lifetime analysis:
Models and tools,’ACM Trans. Sensors Netw., vol. 5, no. 1, pp. 1–33,
2009.
[42] M. Z. Hasan and F. H. Al-Turjman Al-Rizzo, ‘‘Evaluation of a duty-
cycled protocol for TDMA-based Wireless Sensor Networks,’’ in Proc.
Int. Wireless Commun. Mobile Comput. Conf. (IWCMC), Sep. 2016,
pp. 964–969.
[43] S. Kurt and B. Tavli, ‘‘Path-loss modeling for wireless sensor networks:
A review of models and comparativeevaluations,’’IEEE Antennas Propag.
Mag., vol. 59, no. 1, pp. 18–37, Feb. 2016.
[44] B. Krishnamachari, Networking Wireless Sensors. Cambridge, U.K.:
Cambridge Univ. Press, 2005.
[45] M. Z. Zamalloa and B. Krishnamachari, ‘‘An analysis of unreliability and
asymmetry in low-power wireless links,’’ ACM Trans. Sensor Netw., vol. 3,
no. 2, 2007, Art. no. 7.
[46] R. Pahtma, J. Preden, R. Agar, and P. Pikk, ‘‘Utilization of received signal
strength indication by embedded nodes,’Elektron. Elektrotechn., vol. 93,
no. 5, pp. 39–42, 2009.
[47] F. Vanheel, J. Verhaevert, E. Laermans, I. Moerman, and P. Demeester,
‘‘Automated linear regression tools improve RSSI WSN localization in
multipath indoor environment,’’ EURASIP J. Wireless Commun. Netw.,
vol. 2011, no. 1, pp. 1–27, 2011.
[48] G. Mao, B. D. O. Anderson, and B. Fidan, ‘‘Path loss exponent estimation
for wireless sensor network localization,’Comput. Netw., vol. 51, no. 10,
pp. 2467–2483, Jul. 2007.
[49] F. Al-Turjman, ‘‘Cognitive routing protocol for disaster-inspired Internet
of Things,’Future Generat. Comput. Syst., vol. 3, no. 14, pp. 1–13, 2017.
[50] I. F. Akyildiz and M. C. Vuran, Wireless Sensor Networks, vol. 4. Hoboken,
NJ, USA: Wiley, 2010.
[51] F. S. Presentation, ‘‘An introduction to Motorola’s RCF (reconfigurable
compute fabric) technology,’’ Freescale Semiconduct., Austin, TX, USA,
Tech. Rep., 2004.
[52] A. A. Ahmed and N. Fisal, ‘‘A real-time routing protocol with load distri-
bution in wireless sensor networks,’’ Comput. Commun., vol. 31, no. 14,
pp. 3190–3203, 2008.
[53] G. Zhou, T. He, S. Krishnamurthy, and J. A. Stankovic, ‘‘Impact of radio
irregularity on wireless sensor networks,’’ in Proc. 2nd Int. Conf. Mobile
Syst., Appl., Services, 2004, pp. 125–138.
[54] G. Zhou, T. He, S. Krishnamurthy, and J. A. Stankovic, ‘‘Models and
solutions for radio irregularity in wireless sensor networks,’ACM Trans.
Sensor Netw., vol. 2, no. 2, pp. 221–262, 2006.
[55] W. C. Y. Lee, ‘‘Lee’s model [cellular radio path loss prediction],’’ in Proc.
IEEE 42nd Veh. Technol. Conf., vol. 1. May 1992, pp. 343–348.
[56] A. W. Khan, A. H. Abdullah, M. H. Anisi, and J. I. Bangash,
‘‘A comprehensive study of data collection schemes using mobile sinks in
wireless sensor networks,’Sensors, vol. 14, no. 2, pp. 2510–2548, 2014.
[57] H. Karvonen, Z. Shelby, and C. Pomalaza-Raez, ‘‘Coding for energy
efficient wireless embedded networks,’’ in Proc. Int. Workshop Wireless
Ad-Hoc Netw., 2004, pp. 300–304.
MOHAMMED ZAKI HASAN received the Ph.D.
degree in computer science from the Network
Research Group, School of Computer Sciences,
Universiti Sains Malaysia, in 2014. He is cur-
rently a Visiting Faculty with the Systems Engi-
neering Department, University of Arkansas at
Little Rock. His is involved in software testing for
software defined network to allow a centralized
management and control of networking devices,
programmability, and increased network reliabil-
ity. Moreover, he is working in the area of Internet of Things and wireless
multimedia sensor networks routing design architecture, deployment, and
performance evaluation.
FADI AL-TURJMAN received the Ph.D. degree
in computing science from Queen’s University,
Canada, in 2011. He is a Professor with the Com-
puter Engineering Department, Antalya Bilim
University, Turkey. He is also a leading authority in
the areas of smart/cognitive, wireless and mobile
networks’ architectures, protocols, deployments,
and performance evaluation. His record spans over
170 publications in journals, conferences, patents,
books, and book chapters, in addition to numerous
keynotes and plenary talks at flagship venues. He is serving as the Lead Guest
Editor in several journals including the IET Wireless Sensor Systems, MDPI
Sensors and Wiley. He is also the Publication Chair for the IEEE International
Conference on Local Computer Networks in 2018. He is the sole author for
three recently published books about cognition and wireless sensor networks’
deployments in smart environments with Taylor and Francis, CRC New York
(a top tier publisher in the area).
HUSSAIN AL-RIZZO received the B.Sc. degree
(Hons.) in electronics and communications, the
Postgraduate Diploma degree (Hons.) in electron-
ics and communications, the M.Sc. degree (Hons.)
in microwave communication systems from the
University of Mosul, Mosul, Iraq, in 1979, 1981,
and 1983, respectively, and the Ph.D. degree in
computational electromagnetics, wireless commu-
nications, and the global positioning system from
the Radiating Systems Research Laboratory, Elec-
trical and Computer Engineering Department, University of New Brunswick,
Fredericton, NB, Canada in 1992. From 1983 to 1987 he was with the Elec-
tromagnetic Wave Propagation Department, Space and Astronomy Research
Center, Scientific Research Council, Baghdad, Iraq. In 1987, he joined the
Radiating Systems Research Laboratory, Electrical and Computer Engineer-
ing Department, University of New Brunswick, Fredericton, NB, Canada.
Since 2000, he has been with the Systems Engineering Department, Uni-
versity of Arkansas at Little Rock, where he is currently a Professor of
telecommunication systems engineering. He has published over 200 peer-
reviewed journal papers and presentations, book chapters, and two patents.
His research areas include implantable antennas and wireless systems, smart
antennas, 4G LTE-A, WLAN/MIMO deployment and load balancing, elec-
tromagnetic wave scattering by complex objects, design, modeling and test-
ing of high-power microwave applicators, precipitation effects on terrestrial
and satellite frequency re-use communication systems, field operation of
NAVSTAR GPS receivers, data processing, and accuracy assessment, effects
of the ionosphere, troposphere and multipath on code and carrier-beat phase
GPS observations and the development of novel hybrid Cartesian/cylindrical
FDTD models for passive microwave components. For his various academic
achievements he received the Best Doctoral Graduate Award in science and
engineering from the University of New Brunswick.
VOLUME 6, 2018 20389
... In our revised definition of hop-by-hop reliability, sensors that forward data directly to the central node must perform retransmissions. Reference [15] addresses the degree of irregularity parameters to facilitate adaptation to geographic switching for two types of transmission in distributed systems: hop-by-hop and end-to-end retransmission schemes. e simulations determined results for average packet delay transmission, transmission energy consumption, and throughput. ...
Article
Full-text available
Multipath data transmission is a key problem that needs to be solved urgently in wireless sensor networks. In this paper, sensor node failure, link failure, energy exhaustion, and external interference affect the stability and reliability of network data transmission. A multipath transmission strategy for wireless sensor networks based on improved shuffled frog leaping algorithm is proposed. A mathematical model of multipath transmission in wireless sensor networks is established. In the shuffled frog leaping algorithm, combined with the transition probability in the particle swarm optimization algorithm, random individuals in the subgroup are introduced to assist the search when updating the frog individual position, which improves the algorithm's ability to jump out of the local optimum and improves the quality of the optimization algorithm solution. The model is applied to multipath transmission in wireless sensor networks. Then, the shuffled frog leaping algorithm is used to update, divide, and reorganize the sensor nodes to select the optimal node to establish the optimal transmission path and improve the stability and reliability of the network. Simulation experiments show that the algorithm in this paper can ensure the reliability of data transmission, reduce the network packet loss rate and network energy consumption, and reduce the average delay of data transmission.
... Robustness and scalability are two main factors that need to be considered in implementing packet routing in WSN. Due to each sensor node in WSN having limited transmission range, packets can only be forwarded from the source node to the destination node through other sensor nodes using multi-hop technique [2]. ...
Article
Full-text available
Routing packets in Wireless Sensor Network (WSN) is challenging due to the distribution of sensor nodes with different ability. Inefficient routing may lead to higher failure rate, higher latency and higher energy consumption. One of the common approaches to solve this problem is by using bio-inspired routing algorithms due to their abilities to adapt with dynamic environment. This paper proposed an improved ant colony system for packing routing in WSN that focuses on exploration and exploitation techniques. In the proposed routing algorithm, the best path to be used for packet transmission will be determined by considering the remaining energy of each sensor node to reduce the hotspot problem. Local pheromone update and global pheromone update are used with the aim to prevent imbalanced energy depletion of sensor nodes and to balance the packet distribution. The proposed routing algorithm was validated against several bio-inspired routing algorithms in medium and large sized networks. The results suggested that it has outperformed in terms of success rate, packet loss rate, latency and energy efficiency.
... Hasan et al. [21] proposed formulation using analytics for the error rate and a decisive path-loss structure is defined using specified level. There is also trust among nodes that are most frequently used. ...
Article
Full-text available
The research work presents, constrained network coding technique to ensure the successful data transmission based composite channel cmos technology using dielectric properties. The charge fragmentation and charge splitting are two components of the filtered switch domino (FSD) technique. Further behavior of selected switching is achieved using generator called conditional pulse generator which is employed in Multi Dynamic Node Domino (MDND) technique. Both FSD and MDND technique need wide area compared to existing single node-keeper domino technique. The aim of this research is to minimize dissipation of power and to achieve less consumption of power. The proposed research, works by introducing the method namely Interference and throughput aware Optimized Multicast Routing Protocol (IT-OMRP). The main goal of this proposed research method is to introduce the system which can forward the data packets towards the destination securely and successfully. To achieve the bandwidth and throughput in optimized data transmission, proposed multicast tree is selected by Particle Swarm Optimization which will select the most optimal host node as the branches of multi cast tree. Here node selection is done by considering the objectives residual energy, residual bandwidth and throughput. After node selection multi cast routing is done with the concern of interference to ensure the reliable and successful data transmission. In case of transmission range size is higher than the coverage sense range, successful routing is ensured by selecting secondary host forwarders as a backup which will act as intermediate relay forwarders. The NS2 simulator is used to evaluate research outcome from which it is proved that the proposed technique tends to have increased packet delivery ratio than the existing work. 1 Introduction Environmental monitoring system in home based applications, wearable devices like sensors are integrated with applications which interact and accessed remotely to the people [1]. A recent developmental application like smart city, smart medical system, smart transports etc uses IoT for its deployment [2]. IoT is used in high end sensing applications like information sensing, data transmission, cloud distributed system for storage and process [3]. Other applications are E-healthcare, Internet of medical things, digital security, smart watch so on. These major applications are interconnected using 4G and 5G architectures [4].
... In wireless networks, the cross-layer approach overcomes all pitfalls of the conventional model. In the cross-layer model, each layer can communicate with other layers via layer boundaries [17,19,22,31,36,39]. Each layer sends feedback to another layer to optimize the entire network's performances in a cross-layer-based protocol stack. ...
Article
Full-text available
The rapid proliferation of wireless networks increases the demand for video transmission in surveillance, online gaming, video streaming, and other applications. Still, assurance of Quality of Experience (QoE) in video transmission is restricted by multiple constraints, including wireless medium characteristics, multipath routing limitations, and so on. QoE is measured at the source in many prior works, which is not suitable for QoE assurance. This paper addresses all these issues in video transmission over wireless networks with a novel cross-layer design. The proposed cross-layer approach initially measures video quality at a destination based on video quality score (VQS) which is obtained from past destination by application layer, and the feedback is given to the source to guarantee QoE. For improving the quality of video transmission, Quality-based Adaptive Scalable Video Coding (QA-SVC) based video coding is performed in the source node to increase the transmission efficiency. The encoded video is transmitted over multiple paths which are scheduled using the Enriched Particle Swarm Optimization with Multiple Solutions (EPSO-MS) algorithm by considering numerous metrics to reduce the data loss. Improved Artificial Neural Network (IANN) and Deficit Weighted Round Robin (DWRR) are jointly used to schedule the video in intermediate nodes based on priority level which reduces the waiting delay with efficient video quality for transmitting the video before deadline. Video scheduling and priority classification are supported by the Modified Real-time Transport protocol (M-RTP) for fast priority provisioning. QoE of the video is assured and enhanced by performing video quality estimation based on VQS and it is carried out at the destination node using Type-2 Fuzzy Logic (T2FL). Finally, the proposed cross-layer design is modeled in NS-3.26 and evaluated based on throughput (2mbps (high)), jitter (20 ms (low)), goodput (1 mbps(high)), delay (25 ms (low)), Peak Signal-to-Noise Ratio (PSNR) (30 dB (high)), packet drop (7% (low)), bandwidth utilization (20%(high)), and mean opinion score (MOS) (2 (high)).
Article
This paper demonstrates, network-level performance analysis and implementation of smart city Internet of Things (IoT) system with Infrastructure as a Service (IaaS) level cloud computing architecture. The smart city IoT network topology performance is analyzed at the simulation level using the NS3 simulator by extracting most of the performance-deciding parameters. The performance-enhanced smart city topology is practically implemented in IaaS level architecture. The intended smart city IoT system can monitor the principal parameters like video surveillance with a thermal camera (to identify the virus-like COVID-19 infected people), transport, water quality, solar radiation, sound pollution, air quality (O3, NO2, CO, Particles), parking zones, iconic places, E-suggestions, PRO information over low power wide area network in 61.88 km × 61.88 km range. Primarily we have addressed the IoT network-level routing and quality of service (QoS) challenges and implementation level security challenges. The simulation level network topology analysis is performed to improve the routing and QoS. Blockchain technology-based decentralization is adopted to enrich the IoT system performance in terms of security.
Preprint
Full-text available
Information may be accessed from a distance thanks to computer networks. Wireless or wired networks are also possible. Due to recent developments in wireless infrastructure, wireless sensor networks (WSNs) were developed. Activities or events occurring in the environment are monitored, recorded, and managed by WSN. Through a variety of routing techniques, data relaying is done in these systems. The fourth industrial revolution, or Industry 4.0, is defined as the integration of complex physical automation systems made up of machinery and devices connected by sensors and managed by software. This is done to boost the efficiency and reliability of operations. Industry 4.0 is viewed as a possibility because of industrial IoT, the concept of leveraging IoT technology in manufacturing. delivering, in an industrial setting, a means of connecting engines, power grids, and sensors to the cloud. In this essay, we'll try to comprehend how the Internet of Things (IoT) works in wireless sensor networks and how it might be used in various situations.
Article
Full-text available
The vast number of smart cloud applications that communicate with the “things” over a variety of physical networks and communication protocols contribute to the rise of complexity in Internet of Things (IoT) systems. The diversity of end‐user requirements related to the volume of generated data, its variety, and the velocity of its transmission makes quality of service (QoS) provisioning even more crucial and challenging in IoT. This paper provides a comprehensive and up‐to‐date survey of QoS support in IoT networks and communication protocols. An analysis of the QoS‐aware IoT architectures, layer‐dependent QoS metrics, and network resource optimization methods utilized in IoT systems are given. The limitations of the current state‐of‐the‐art studies for efficient delivery of QoS metrics are discussed. The paper concludes with future research directions on end‐to‐end QoS provisioning in IoT. This paper provides a comprehensive and up‐to‐date survey of QoS support in IoT networks and communication protocols. An analysis of the QoS‐aware IoT architectures, layer‐dependent QoS metrics, and network resource optimization methods utilized in IoT systems are given. The limitations of the current state of the art studies for efficient delivery of QoS metrics are discussed. The paper concludes with future research directions on end‐to‐end QoS provisioning in IoT.
Chapter
This study investigates the challenges and opportunities particularly in current transportation policies that may arise to change existing mobility systems with the help of intelligent autonomous vehicle (IAV) technologies. India is already one of the highest levels of population, so all this leading to a massive increase in road traffic which leads to serious issues such as permanent traffic jams increased air pollution at harmful levels, accidents, or even human losses. Hence, technology and transportation infrastructure providers are needed to provide safer, large scalable, better flexible, and low-cost effective solutions to all these problems. This paper provides a comprehensive review of the relevant literature and explores intelligent transportation systems (ITS) which is one of the best solutions, and this offers traffic monitoring, guidance, or alerting by communication capabilities in between vehicles to avoid any accidents in terms of safety elements (Litman in Autonomous vehicle implementation predictions implications for transport planning, 2018; Pino et al. in IEEE Access 6:17527–17532, 2018; Bagloee et al. in J Mod Transp 24(4):284–303, 2016; Aazam et al. in IEEE Commun Mag, 2018 (in press); Gong et al. in IEEE Trans Intell Transp Syst 19:390–401, 2017; Miettinen in Evolutionary algorithms in engineering and computer science: recent advances in genetic algorithms, evolution strategies, evolutionary programming, GE. Wiley, New York, 1999; Making Innov E-J, makingofinnovation.com, Oct 2015; Campioni et al. in IEEE Trans Veh Technol J Latex Class Files 14(8), 2015; El Zoghby et al. in IEEE 68(2) ITVTAB, 2019; IEEE International symposium on technology in society (ISTAS) proceedings, 2015; NCSL—National conference of state legislatures, database on autonomous vehicles legislation; Cai et al. in An empirical air-to-ground channel model based on passive measurements in LTE, 1140; Qin et al. in IEEE Trans Veh Technol; Meng et al. in IEEE Trans Control Syst Technol 25:1480–1487, 2017; Zeng et al. in Joint communication and control for wireless autonomous vehicular platoon systems. CoRR, 2018) [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15]. So it is important to do the improvement in new vehicular automation methods. It must have its organizations, specialists, and foundations to the center for their safe and successive journey under enhancing street security and traveling comfort which is the prime need of nowadays (Lian et al. in Channel models a non-stationary 3-D wideband GBSM for HAP-MIMO communication systems, 1128; Du et al. in Resource allocation in vehicular networks based on dual-side cost minimization, 1079; Gonzalez-Martín et al. in Analytical models of the performance of C-V2X mode 4 vehicular communications, 1155; Wei et al. in IEEE Trans Veh Technol; Wang et al. in Optimizing content dissemination for real-time traffic management in large-scale internet of vehicle systems, 1093; Jo et al. in Image-to-image learning to predict traffic speeds by considering area-wide spatio-temporal dependencies; Vehicle automation IEEE 2019: collision avoidance intelligent vehicles system vehicle-to-everything; Du et al. in Design and assessment of an electric vehicle power train model based on real-world driving and charging cycles, 1178; Peng et al. in Connected vehicle series vehicular communications: a network layer perspective, 1064; Wei et al. in An integrated longitudinal and lateral vehicle following control system with radar and vehicle-to-vehicle communication, 1116) [16,17,18,19,20,21,22,23,24,25]. All over the world, currently automobile manufacturers are mainly focusing on developing, exhibiting, producing, and promoting new vehicle features with advance controlled strategies that could make possible the exchange of information with self-organized robotics and try to interfaces with new algorithms into the smart automobile world (Self-driving cars: the next revolution kpmg.com | cargroup.org CAR (car centre for automotive research); Hasan et al. in IEEE Access 6:20371–20389, 2018; Guerrero-ibanez et al. in IEEE Wirel Commun 22:122–128, 2015; Li et al. in IEEE Trans Veh Technol, 2017; Abdelhamid et al. in IEEE Trans Intell Transp Syst 99:1–14, 2017; Jo et al. in Image-to-image learning to predict traffic speeds by considering area-wide spatio-temporal dependencies, 1188; Tan and Hu in IEEE Trans Veh Technol; Garcia-Garcia et al. in Appl Soft Comput 70:41–65, 2018; Xiao et al. in Inf Sci 432:543–558, 2018; Favarò et al. in Presented at PSAM Topical 2017 on human reliability, quantitative human factors, and risk management, Munich, Germany, 7th–9th June 2017) [26,27,28,29,30,31,32,33,34,35]. The vision of this paper is to provide a multimodal transportation system that could overview all the recent research work in the field of interconnected transportation environment giving help to millions of vehicles that are facing traffic problems and safety issues. To this end, we propose smart automated connected vehicles with the help of intelligent transport systems (ITS) and can provide various application services to improve the safety, efficiency, reliability, and comfort of the driving system (Autonomous cars: past, present and future—a review of the developments in the last century, the present scenario and the expected future of autonomous vehicle technology, Jan 2015 with 15,903; Autonomous vehicles UF: automated guided vehicles autonomous cars autonomous driving autonomous trucks unmanned autonomous vehicle BT: autonomous systems intelligent vehicles RT: artificial intelligence mechatronics multi-agent systems vehicular automation NT: unmanned autonomous vehicles, by The Institute of Electrical and Electronics Engineers (IEEE), 2019; Int J Inf Commun Comput Technol, Jagan Institute of Management Studies, New Delhi Article, Jan 2016; Proceedings of 2014 RAECS UIET, Panjab University Chandigarh, 06–08 Mar 2014; Future Internet J, Published: 24 Jan 2019; Martin-Vega et al. in IEEE Trans Veh Technol 67:3069–3084, 2018; Bazzi et al. in IEEE Trans Wirel Commun 17:2402–2416, 2018; Noor-A-Rahim et al. in IEEE Access 6:23786–23799, 2018;Technical report, IEEE, Piscataway, NJ, USA, 2006; Bazzi et al. in IEEE Access 6:71685–71698, 2018) [36,37,38,39,40,41,42,43,44,45].KeywordsAnt coloniesBee’s coloniesKilobitSwarm intelligenceVolvo carsTesla autopilotGoogle self-driving carArtificial intelligenceIoTPractical actual accidentsITSV2VDSRCVA NETWireless sensor networksAccident preventionPost-accident investigation
Article
Over the years cloud computing has seen significant evolution in terms of improvement in infrastructure and resource provisioning. However the continuous emergence of new applications such as the Internet of Things (IoTs) with thousands of users put a significant load on cloud infrastructure. Load balancing of resource allocation in cloud-oriented IoT is a critical factor that has a significant impact on the smooth operation of cloud services and customer satisfaction. Several load balancing strategies for cloud environment have been proposed in the past. However the existing approaches mostly consider only a few parameters and ignore many critical factors having a pivotal role in load balancing leading to less optimized resource allocation. Load balancing is a challenging problem and therefore the research community has recently focused towards employing machine learning-based metaheuristic approaches for load balancing in the cloud. In this paper we propose a metaheuristics-based scheme Data Format Classification using Support Vector Machine (DFC-SVM), to deal with the load balancing problem. The proposed scheme aims to reduce the online load balancing complexity by offline-based pre-classification of raw-data from diverse sources (such as IoT) into different formats e.g. text images media etc. SVM is utilized to classify “n” types of data formats featuring audio video text digital images and maps etc. A one-to-many classification approach has been developed so that data formats from the cloud are initially classified into their respective classes and assigned to virtual machines through the proposed modified version of Particle Swarm Optimization (PSO) which schedules the data of a particular class efficiently. The experimental results compared with the baselines have shown a significant improvement in the performance of the proposed approach. Overall an average of 94% classification accuracy is achieved along with 11.82% less energy 16% less response time and 16.08% fewer SLA violations are observed.
Chapter
With the advent of the Internet of Things (IoT), machine to machine (M2M) communication, and the related ecosystem, recently a new paradigm of Industrial IoT (IIoT) has emerged.
Article
Full-text available
An asynchronous medium access control (MAC) duty-cycled protocols have higher energy efficiency and lower packet latency than synchronized ones due to reduced idle listening. Moreover, they provide efficient utilization of energy supplied to mobile sensors. They are considered very important in MAC protocols due to the adverse effects of hidden terminals which causes energy consumption in sensor networks. Therefore, in this paper, the impact of hidden terminals on the performance of an asynchronous duty-cycled MAC protocol X-MAC for vehicle-base sensor is investigated via analysis and simulations. We propose a Markov model to analyze the quality-of-service (QoS) parameters in terms of energy consumption, delay, and throughput. Our analytical model provides QoS parameter values that closely match the simulation results under various network conditions. Our model is more computationally efficient and provides accurate results quickly compared with simulations. More importantly, our model enables the designers to obtain a better understanding of the effects of different numbers of mobile sensor nodes and data arrival rates on the performance of an asynchronous MAC duty-cycled protocol.
Article
Full-text available
Wireless Sensor Networks (WSNs) are used to collect and send various kinds of messages to a base station, for instance, WSNs are used in many important applications such as health care, military and monitoring buildings. Routing protocol (such as LEACH and PEGASIS) is one of the important issues in WSN, when applying routing protocol on sensor nodes causes a delay in data transmission between sensor nodes and cluster head. This paper proposed a protocol named Enhanced PEGASIS protocol (EPEGASIS) based on clustering mechanism with k-means algorithm to compute the distances between sensor nodes and their own cluster head in each cluster. The EPEGASIS protocol achieves low transmission delay in wireless sensor networks. The simulation result shows that EPEGASIS protocol reduced the transmission delay between sensor nodes by about 27% compared with LEACH and by about 36% compared with PEGASIS protocols. Keywords – Wireless Sensor Network (WSN), PEGASIS protocol, LEACH protocol, Transmission delay, K-means algorithm. Introduction A wireless sensor network is an active research area with numerous workshops and conferences arranged each year. A Wireless Sensor Networks (WSN) are a set of hundreds or thousands of micro sensor nodes that have capabilities of sensing, establishing wireless communication between each other and doing computational and processing operations [1]. The clustering protocol as a hierarchical protocol is more extensive than the other types of protocols, and the clustering protocols generally contain two steps: one is forming the cluster and the other is transmitting the data [2]. In general, classification of a WSN routing methodology can be done into two main categories; based on network structure or based on the protocol operation. Depending on the network structure, different routing schemes fall into this category. A sensor network can be hierarchical or flat in the sense that every sensor has the same role and function. Therefore the connections between the nodes are set in short distance to establish the radio communication. Alternatively, the network can be hierarchical, where the network is divided into clusters comprising of number of nodes. Cluster head, which is master node, within each respective cluster is responsible for routing the information to other cluster head [1]. The sensed data are transmitted to the base station based on the assumed transmission range by the sensor nodes. In the transmission range, the energy levels of the
Article
Full-text available
The vision of Wireless Multimedia Sensor Networks (WMSNs) is to provide real-time multimedia applications using wireless sensors deployed for long-term usage. Quality of Service (QoS) assurances for both best effort data and real-time multimedia applications introduced new challenges in prioritizing multipath routing protocols in WMSNs. Multipath routing approaches with multiple constraints have received considerable research interest. In this paper, a comprehensive survey of both best effort data and real-time multipath routing protocols for WMSNs is presented. Results of a preliminary investigation into design issues affecting the development of strategic multipath routing protocols that support multimedia data in WMSNs are also presented and discussed from the network application perspective.
Chapter
A sensor network is designed to perform a set of high-level information processing tasks, such as detection, tracking, or classification. Measures of performance for these tasks are well defined, including detection, false alarms or misses, classification errors, and track quality. Commercial and military applications include environmental monitoring (e.g. traffic, habitat, security), industrial sensing and diagnostics (e.g. factory, appliances), infrastructure protection (e.g. power grid, water distributions), and battlefield awareness (e.g. multi-target tracking).
Article
The proliferation of wireless multimedia sensor networks (WMSNs) has given rise to intelligent transportation systems (ITS) as a mobile data-sharing model. This vision can be extended under the umbrella of the mobile Internet of Things (IoT) to include versatile resources such as smartphones, Radio Frequency Identification (RFID) tags and sensors on roads that can be utilized in emergency situations. The facilitation of such a vision faces key challenges in terms of interoperability, resource management and energy consumption. In this work, we propose agile data delivery framework that caters for service-based applications in smart cities where multimedia data is heavily exchanged. Optimized routing approach that operates with limited resources in highly dynamic topologies is investigated and recommended. This approach assists in specifying which path a data packet should follow in order to determine the optimal usage of the available resources while satisfying QoS constraints for a wide range of real-time multimedia applications in safety and security fields. Simulation results, which have been validated via solid analytical analysis, are used to assess and outline the efficiency of the proposed approach in terms of system throughput, energy consumption, and average end-to-end delay against other similar approaches in the literature.
Article
Topology control is relevant in wireless sensor network because of two reasons, namely minimal sensor coverage and power constraints. The former condition is typically satisfied by high-density deployment, whereas the latter mainly concerns with the control protocol design that is adaptable. Controlling communication topology is at the center of the efforts to optimize network performance while improving energy conservation. A dense topology often results in high interference and lower spatial reuse thus reduced capacity, while sparse topology is susceptible to network partitioning and sub-optimal path selection from the routing layer. Topology control has been extensively studied in both flat and hierarchical network by mean of power adjustment and clustering, respectively. Despite a common goal of making the topology less complex both techniques differ in their approach. While the focus of clustering is to form a connected backbone which consists of a minimum subset of nodes, i.e., dominating set, power adjustment focus on minimizing energy consumption. Combining both approaches remains a relatively lesser explored area. We proposed a hybrid framework called Collaborative Topology Control Protocol (also CTCP), which combines dominating set based clustering and transmission power adjustment. The protocol operates in two stages. During the first stage, a parameterized Minimum Virtual Connected Dominating Set (MVCDS) algorithm is executed to obtain clusters of various desirable properties. In the second stage, each cluster-head executes a distributed power adjustment algorithm. The simulation results show that the proposed topology control framework is capable of versatile performance in terms of transmission range/energy cost, the number of neighbors, edges and hop distance. Moreover, the topology construction process uses the locally available information only with minimal communication overhead.
Article
In this paper, we propose a framework for data delivery in large-scale networks for disaster management, where numerous wireless sensors are distributed over city traffic-infrastructures, shopping-malls’ parking areas, airports’ facilities, etc. In general, our framework caters for energy-efficient applications in the Internet of Things (IoT) where data is propagated via relays from diverse sensor-nodes towards a gateway connected to a large-scale network such as the Internet. We consider the entire network energy while choosing the next hop for the routed packets in the targeted wireless sensor network. Our delivery approach considers resource limitations in terms of hop count, and remaining-energy levels. Extensive simulations are performed and achieved results confirm the effectiveness of the proposed approach in comparison to other baseline energy-aware routing protocols in the literature.
Article
Modern multimedia sensor networks impose strict constraints on both the delay and energy consumption when time-critical data must be reported to the sink within a limited bandwidth without any loss. Failure to transmit an event to the sink occurs for many reasons, including inherence limitations of sensors, power consumption, and reliability. We propose a mathematical model for a novel Quality of Service (QoS) routing-determination method. The proposed scheme enables determining the optimal path to provide appropriate shared radio satisfying the QoS for a wide range of real-time intensive media. The mathematical model is based on the Lagrangian Relaxation method (LR), to control adaptive switching of hopby- hop QoS routing protocols. The embedded criteria for each objective function are used to decide which path from source to sink will be selected. Simulation results show that, compared with existing routing protocols, the approach proposed in this paper significantly improves the packet received ratio, energy consumption, and average end-to-end delay of the sensor node.