Multihop Relay
Techniques for Communication
Range Extension in Near-Field Magnetic Induction
Communication Systems
Mehrnoush
Masihpour
, Daniel Franklin and Mehran Abolhasan
University of Technology Sydney
Sydney, Australia
mehrnoush.masihpour@student.uts.edu.au; daniel.franklin@uts.edu.au; mehran.abolhasan@uts.edu.au
Abstract
munications and
near field magnetic induction communication
(NFMIC) is discussed. Three multihop relay strategies for
NFMIC are proposed: Non Line of Sight Magnetic Induction
Relay (NLoS-MI Relay), Non Line of Sight Master/Assistant
Magnetic Induction Relay1 (NLoS-MAMI Relay1) and Non Line
of Sight Master/Assistant Magnetic Induction Relay2 (NLoS-
MAMI Relay2). In the first approach only one node contributes
to the communication, while in the other two techniques (which
are based on a master-assistant strategy), two relaying nodes
are employed. This paper shows that these three techniques
can be used to overcome the problem of dead spots within
a body area network and extend the communication range
without increasing the transmission power and the antenna size
or decreasing receiver sensitivity. The impact of the separation
distance between the nodes on the achievable RSS and channel
data rate is evaluated for the three techniques. It is demonstrated
that the technique which is most effective depends on the specific
network topology. Optimum selection of nodes as relay master
and assistant based on the location of the nodes is discussed. The
paper also studies the impact of the quality factor on achievable
data rate. It is shown that to obtain the highest data rate, the
optimum quality factor needs to be determined for each proposed
cooperative communication method.
Index Terms—NFMIC; propagation model; relay; cooperative
communication; MI-Relay; magneto inductive waveguide; multi-
hop communications; body area networks; range extension
I. I
NTRODUCTION
The number
of applications for wireless communication
technologies continue to grow rapidly [1]–[4]. However, the
availability of frequency spectrum is limited. In many sit-
uations, multiple users and/or networks need to share the
same spectrum, leading to increased interference. In many
communication networks, such as in public safety communica-
tions, different frequency-hopping and other spread-spectrum
methods have been adopted to mitigate interference due to
the spectral overlaps, and to make the existence of radio
transmissions less obvious [5].
Most existing wireless devices use radiative electromagnetic
(EM) waves for data transmission between personal electronic
devices. While EM-RF based systems are well suited to long
range data exchange, they are not the best possible solution
for communications over very short distances (such as personal
area networks). EM waves are capable of traveling very long
distances, and received power decays with the square of
communication distance [5], [6]. Therefore, the transmitted
signal can be received at distances far away from the source.
Although this characteristic of the EM waves is beneficial
for long range communications, it may be problematic for
communications over very short distances. For instance, a
transmitted signal which conveys confidential information
within a battlefield may be detected by unauthorised parties.
Even if the information cannot be decrypted, the detection
of the transmitted signal may reveal the location of the
transmitter.
Recently, a new technology called Near Field Magnetic In-
duction Communications (NFMIC) has emerged as a promis-
ing solution for short range communications [6]. While con-
ventional radio communication systems use an antenna to
propagate EM waves into free space for data transmission,
NFMIC communications occurs through the magnetic cou-
pling of two compact coils [5]–[13]. The resulting magnetic
field does not propagate far into free space, which allows
the communication to be established and retained within
short distances. This class of transmission is known as near
field communications, while communication using radiating
electromagnetic waves may be referred to as far field commu-
nications.
The boundary between the near field and far field, i.e. the
maximum possible communication range in near field, is a
function of frequency. The distance from the source into which
the magnetic field is radiated into free space is generally
considered to equal λ/2π [6]. This point in space is considered
the end of near field region and the beginning of the far
field. Therefore, to maintain NFMIC, the distance between
the source and destination needs to be less than λ/2π.
NFMIC offers advantages over conventional EM-RF com-
munications when it is used for short proximity communica-
tions. It can provide better signal quality since its behaviour
is much more predictable than RF [5], [6], [10], [12], [13].
RF communications often suffer from frequency spectrum
contention, reflection, shadowing and fading resulting from
the surrounding environment and the presence of objects
such as vehicles, buildings and the human body. By contrast,
NFMIC is mainly affected by the magnetic permeability of
the channel and is more robust to reflection, shadowing and
diffraction. Therefore, it can be an appropriate physical layer
for Body Area Networks (BAN). A BAN refers to the low
—In this
paper, multihop relaying in RF-based com-
JOURNAL OF NETWORKS, VOL. 8, NO. 5, MAY 2013
999
© 2013 ACADEMY PUBLISHER
doi:10.4304/jnw.8.5.999-1011
power communications between smart sensors, within close
proximity to the human body, and its potential applications
may be categorised as follows:
Medical and health care: A BAN can be used for different
purposes at hospitals such as automatic medical diagnosis,
treatment and dosing to improve the quality of treatment and
management efficiency in hospitals [14]–[16]. It is also useful
for remote patient monitoring. In medical ICT (MICT), the
main purpose of using a BAN is to collect the vital information
regarding a patient’s condition such as blood pressure, body
temperature, glucose level, heartbeat and brain or cardiac
signals and transmit the data to a command unit (action unit)
or a central controller, which can be a smart device located in
hospital and controlled by a doctor or nurse. It also may be a
digital device controlled by the patients themselves [14]–[16].
BAN devices used for medical purposes are often in the form
of implants and need to be located inside the human body.
In such an environment, the transmitted EM signal is highly
attenuated by the body tissue since communication channel is
in fact the human body and contains body tissues and water.
Assisting people with disabilities: In this usage model,
BAN devices may be used for object detection such as detect-
ing stairs and vacant seat in trains also to provide guidance for
routing and positioning [14]. As an example, a BAN can be
used to assist a speech-impaired person [14], in which sensors
may be located on the person’s fingers to collect information
such as the movement of fingers and relative position of
fingers in respect to each other and also to the hand and
communicate the gathered information to a central node to
be further interpreted as vocal language [14].
Entertainment: A BAN may be used by a person for
entertainment purposes such as gaming, music and video
playing and so on [14], [16]. The typical devices in such
networks are mobile phones, laptop computers, music players
and headsets [16]. This usage scenario requires the highest
data rate among all the applications discussed here, since the
real time video streams require data rates in range of 384 kbit/s
up to 20 Mbit/s [16]. Since the cost and power consumption
needs to be minimised, it is very challenging to achieve
required data rates for this category of BAN application.
Personal fitness monitoring: BAN for fitness monitoring
typically consists of a music player and some sensors collect-
ing the information relevant to the exercise, such as sensors to
monitor heart rate, speed, body temperature, oxygen level and
rate of glucose consumption [16]. The collected information
may be further sent through a gateway, to a central data base
or to a coach, monitoring the athlete [16]. This can highly
improve the training of professional athletes.
Public Safety: A BAN may be used by firefighters, police,
ambulance officers, emergency service or military personel for
public safety purposes. Vital information from individuals and
the ambient environment may be collected in order to detect
an emergency situations which may require quick actions
from outside [16]. Information such as the level of toxic
gas in the air and the temperature can be collected and the
sensor may warn the person or the action unit [15], [16].
One example of BAN usage model in military is a U.S Army
program known as warfighter physiological status monitoring
(WPSM) [17]. This programs aims to address two issues.
Firstly, to reduce injuries caused by environmental factors
such as high temperature and altitude sickness [17]. Authors
of [17] discuss that having access to WPSA data enables
the commanders at different levels to effectively have access
to their troops and enhance their performance. According to
[17], the second purpose of WPSM program is to increase the
chance of survival for casualties. WPSM information can help
the combat medic to quickly access the wounded person.
To improve the reliability of RF communication systems,
higher transmission power may be used. However, increas-
ing the transmission power may lead to interference, inter-
system frequency contention and higher power consumption.
Increasing the transmission power to achieve higher signal to
noise ratio also results in security risks. By increasing the
power, the chance of the signal being detected by unauthorised
parties increases. By contrast, NFMIC not only achieves higher
reliability but also reduces the required power consumption.
This is due to the inherent properties of near field MI waves. A
MI signal attenuates with the sixth power of distance, or about
-60 dB per decade of distance [6], [9], [18]. Although this
property of MI makes it unsuitable for transmission over long
distances, it allows efficient communication over a short range.
It also results in less interference with other communication
systems and reduces frequency spectrum contention [5]–[7],
[10], [12], [13], [18].
Due to its low power consumption, reliability and the in-
herent difficulty of long-range detection, NFMIC is considered
to be a good solution for short range military communication
applications [5], [6], [18]. NFMIC can also be used in a wide
range of non-military applications such as contactless payment
cards, medical implants and monitoring devices, personal wire-
less electronics and so on. NFMIC is also a promising solution
for underwater and underground communications in which
signal transmission is difficult, inefficient or impossible [19]–
[21]. While EM waves are severely attenuated by soil, water,
body tissues and rocks, MI waves are capable of penetrating
more deeply in such environments [19]–[21]. These benefits
are countered by the limited data rate achievable through MI
communications systems.
The contribution of this paper is to study the application
of cooperative communications to NFMIC systems in order to
extend the achievable communication range and enhance the
channel capacity. In this paper, three cooperative communica-
tion techniques are proposed to enhance system performance
where there is no line of sight (NLoS) between the source and
the final destination. Methods whereby idle NFMIC devices
can be utilised as cooperative relay nodes to ensure good signal
quality at the final receiver are discussed. The propagation
model in such scenarios is evaluated for the three different
multihop relaying techniques.
The reminder of this paper is structured as follows: Section
II discusses related works on the topic, Section III presents the
proposed relaying strategies, in Section IV simulation results
are discussed, and finally a summary of contributions is given
in Section V.
1000
JOURNAL OF NETWORKS, VOL. 8, NO. 5, MAY 2013
© 2013 ACADEMY PUBLISHER
II. BACKGROUND
To improve achievable communication range and to enhance
capacity without increasing transmission power or receiver
sensitivity, multihop relaying has been added to wireless
communication system such as cellular networks [22]–[26],
UWB [27], ZigBee [28] and many more [29]–[31]. In general,
multihop relaying refers to a communication technique in
which data is routed to the destination through one or more
intermediate nodes located between the source and destination.
Multihop relaying can achieve higher capacity or provide
extended coverage and consequently higher reliability and
throughput with lower cost and less complexity compared to
conventional peer-to-peer communication systems [3], [4].
A number of different types of multihop networks have been
proposed. The first is multihop infrastructure-based systems
[1]–[4], [22], which consist of one or more fixed relaying
station that are used along with the main base station to relay
the data between a source (which could be a base station, user
station or another relay) and a destination (base station, user
station or another relay) [1]–[4]. This type of multihop relay
is appropriate for long range cellular networks to cover dead
spots (areas that are out of direct communication range of a
base station) or to enhance network capacity in highly crowded
area such as cities, shopping malls and amusement parks.
Multihop ad hoc is another multihop method which is
suitable for both short range and long range communications
[22], [25], [29], [32], [33]. In multihop ad hoc, there is no
need for fixed infrastructure. Electronic devices such as mobile
phones and laptops can be connected in a peer-to-peer fashion
and relay the transmitted data from a source node to other
nodes until the destination is reached. Multihop ad hoc can be
used for inter and intra vehicle communications, personal area
networks, local area networks, underground communications
as well as communications in the battlefield. Multihop ad hoc
is also useful in the event of natural disasters such as floods
and storms, where fixed infrastructure may be damaged or
destroyed as a result of the disaster [22], [25], [29], [32], [33].
Where multihop ad hoc networks are used in combination
with fixed infrastructure networks, the resulting network is
known as multihop hybrid [25], [34]. In such systems, traffic
can be relayed by other devices to allow communication with
a user far away from the source and without the need to hop
through a single base station. This can be useful in busy and
populated areas, where the base station is heavily loaded by
data traffic. It also can enhance system coverage when a user
is located outside the coverage range of a base station (for
example, in dead spots). In this paper, only the multihop ad
hoc technique is considered since it is the most suitable for
short range communication systems and body area networks
in particular.
Ad hoc networks are classified into two categories, based on
the architecture of the network; centralised (cluster-based) and
decentralised (distributed) networks [22], [30]. A centralised
network consists of a number of nodes and only one cluster
head, which is periodically elected by the other nodes in
the network. The cluster head is in possession of all of the
information about the entire network and should be located in
the best-connected position amongst all other nodes [30]. By
contrast, in distributed ad hoc networks, all nodes have the
same amount of information about the network.
While centralised networks have complex architectures and
limited flexibility, distributed networks are simpler to imple-
ment [30]. However, distributed networks suffer from larger
end-to-end delay and higher rates of packet collision. Dis-
tributed networks are less prone to network failure, because if a
node fails, there are connections to other nodes which can pro-
vide alternate paths to a destination [30]. Therefore distributed
networks are suitable for multihop communications. Since they
are more robust to network failure, decentralised multihop
ad hoc networks work well for military communications and
disaster recovery applications, since robustness is a critical
factor in such scenarios [30].
Another factor that makes distributed networks more suit-
able for military applications is their lower transmission power
requirements. Since each node is not required to transmit the
traffic through a central controller, the individual transmission
power can be lower. Each node can communicate with a
destination through its neighbours; therefore, communication
is performed via multiple shorter links instead of one link
with higher transmission power. High transmission power in
military communications poses security risks through location
disclosure [30]. Thus low transmission power is highly desir-
able for military communications.
Multihop ad hoc has been considered for range extension
and increased robustness in different short range communica-
tions systems such as wireless local area networks (WLANs)
[22], ZigBee [28] and ultra wide band (UWB) [27]. In [28], the
authors have developed a prototype system for home security
and automation which uses ZigBee-based multihop sensor net-
works. Authors of [28] claim that it can theoretically achieve
unlimited coverage range. Achieving a large coverage area
through single hop peer to peer networks for such applications
requires long range devices, which are often expensive and
power-hungry [28].
In UWB networks, the coverage range is also limited and
high data rates may not be achievable through a conventional
single hop method. In [27], a simulation environment is
proposed which can simulate both the physical and MAC
(medium access control) layers of OFDM-based UWB multi-
hop network. Using this simulation environment, the authors
have evaluated the performance of a multihop relay UWB
network to determine whether it improves system performance
measures such as end to end delay and packet loss [27].
It is concluded that the IEEE 802.15.3 TDMA MAC layer
can perform adequately in multihop UWB networks if proper
scheduling and routing methods are precisely defined and
implemented. However, further study is required into more
efficient scheduling schemes such as Self-organised Time
Division Multiple access (S-TDMA), to enhance the capacity
and frequency reuse in such communication systems [27].
Although extensive studies have been conducted in multihop
RF communication systems, this concept has not been widely
investigated for near field magnetic induction communication
systems. As mentioned earlier, NFMIC is limited to very short
communication distances. Different techniques used in RF
JOURNAL OF NETWORKS, VOL. 8, NO. 5, MAY 2013
1001
© 2013 ACADEMY PUBLISHER
Fig. 1. Magnetic waveguide and circuit model (adopted from [19])
communications for range extension can be used in NFMIC
to overcome the limited communication range. However, since
the nature of signal transmission in NFMIC differs from
RF communications, it is important to study range extension
methods which are most applicable to NFMIC. In this paper,
three different multihop methods to be applied in a NFMIC
system are proposed.
The magneto-inductive waveguide method has been studied
as a possible solution for multihop communications in NFMIC
[19]–[21], [35]–[39]. A magneto-inductive waveguide commu-
nication system consists of a number of NFMIC nodes, where
the transmitter sends the data to a receiver via multihop relay.
Each node receives the data from its nearest neighbour on one
side and transmits to the next neighbour on its other side via
magnetic field coupling. Multihopping is performed until the
data is delivered to the final destination. A typical waveguide
system model can be seen in Fig. 1. As can be seen from
the figure, all the cooperative nodes are passively powered
and there is no need for an individual power source at each
relaying node.
In [19]–[21], the magneto-inductive waveguide approach is
studied for underground communications, where RF systems
perform poorly due to the adverse channel conditions. In
such an environment, the communication channel consists of
rock and soil, possibly containing water and organic matter.
Underground RF communications suffer from three major
problems: high path loss, large antenna size and dynamic and
unpredictable channel conditions. The authors of [19]–[21]
suggest that by using NFMIC, the problems of large antenna
size and dynamic channel condition may be mitigated. MI
waves are not significantly affected by humidity, soil and rock
since they all have nearly the same magnetic permeability as
air [19]–[21]. However, the high path loss is still a problem
and leads to limited coverage.
To overcome the limited range, authors in [19], [20] have
investigated how a magneto inductive waveguide can be used
to extend the communication distance. The performance of
the improved magneto-inductive model is compared with the
conventional MI and EM communication techniques. The
authors conclude that by implementing a waveguide system,
lower path loss can be achieved regardless of the level of water
content in the soil [19], [20].
In [21], Triangle Centroid (TC) deployment algorithms
for underground MI sensor networks are proposed. In this
algorithm, a Voronoi diagram is used to partition the network
into non-overlapping triangular cells and a three pointed
star topology in each trianglular cell is used to obtain a k-
connected network (k>3) [21]. The authors show that this
algorithm is more robust to network failure than the Minimum
Spanning Tree (MST) algorithm which is only 1-connected.
The MST algorithm connects the entire network together with
the optimum number of relaying nodes; however, nodes have
only one connection, therefore the network is not robust to
node failure [21]. Although this topology is well suited for
underground communications, it is not realistic for a body
area network. In a body area network, nodes may be randomly
located and might frequently change their location. Therefore,
in this paper different multihop methods in a three dimensional
environment are proposed which are more applicable to a body
area network.
III. PROPOSED NFMIC COOPERATIVE RELAY
ALGORITHMS (NL
OS)
A. Network Model
In this section, cooperative communication methods ap-
plicable to a personal area network are proposed. Three
different relaying methods will be evaluated using a simple
network model to show how the idle intermediate nodes can
be used to extend the coverage range. The three techniques
are denoted NLoS-MI Relay, NLoS-MAMI Relay1 and NLoS-
MAMI Relay2. The network consists of a number of wireless
nodes; however, for simplicity it is assumed that only 4 nodes
contribute to the communication: a transmitter (source), a re-
ceiver (destination) and two intermediate nodes which function
as cooperative relay nodes. The source and destination are
separated from each other by a distance d. However, there
is no direct link between them; the target receiver is out of
the communication range of the transmitter. It is assumed
that there are two idle devices between the source and the
sink, which can be utilised to assist the communication by
providing an indirect path from the transmitter to the receiver
over which information may be relayed. The transmitter is
separated from the relay 1 and 2 by distance (x-component of
the distance) x
Tx,R1
and x
Tx,R2
respectively, and the receiver
is located at a distance x
R1,Rx
and x
R2,Rx
from relay 1
and 2 respectively. Relay R1 is assumed to be closer to the
transmitter and R2 is located closer to the receiver such that
any distance-dependent differences in performance may easily
be evaluated. Both the source and sink have a direct link with
R1 and R2. To avoid spectrum contention, the network uses
Time Division Multiple Access (TDMA). As is usual in single-
channel wireless systems half duplex transmission is used,
meaning that a node can either transmit or receive data during
a specific time slot, but cannot do both simultaneously. A relay
node receives the signal from the transmitter, amplifies it and
then forwards it to the next hop, which could either be the final
receiver or another relay node (this is known as the Amplify
and Forward cooperative relaying technique) [40].
1002
JOURNAL OF NETWORKS, VOL. 8, NO. 5, MAY 2013
© 2013 ACADEMY PUBLISHER
B. Relay node selection metric
Different relay selection criteria can be considered to choose
either of intermediate nodes as relay node such as signal
to interference and noise ratio, angle of arrival (AoA), time
difference of arrival (TDoA) and separation distance between
the nodes. In NFMIC, communication distance has a critical
impact on the received signal strength and on the achiev-
able data rate. Since signal attenuation is proportional to
the sixth power of distance rather than the square as in
the case of RF communications, it is the dominant factor
in determining achievable system performance. Furthermore,
since in an NFMIC personal area network, communication
occurs over very short distances, shadowing and multipath
effects are not as critical as in RF communications. Hence, the
separation distance between the nodes is the most appropriate
criterion for optimum performance achievement. This paper
studies the impacts of distance of relaying nodes with respect
to transmitter/receiver on the system performance, in order
to show the optimum selection of the relaying nodes. The
performance is measured according to the received signal
strength at the target receiver as well as the maximum end
to end throughput capacity.
C. Physical Channel Model
In this section, a peer to peer communication model is de-
scribed. Fig. 3 illustrates an ideal near field magnetic induction
communication system, in which there is no angular or lateral
misalignment between the transmitting and receiving antenna
coils. The system consists of a transmitter and a receiver
separated from each other by distance d. The circuit model
of such a system is also shown in Fig. 3.
According to [41], the power transfer function for this
scenario is:
P
Rx
P
Rx
=
μ
2
0
N
2
T
N
2
R
A
2
R
ω
2
16π
2
R
Tx
R
Rx
H
2
INT
(1)
where the magnetic field strength is [41]:
H
INT
=
π
0
dI
Tx
× x
x
3
=
r
4
T
π
2
(r
2
T
+ d
2
)
3
(2)
The cross sectional area of the receiving coil is:
A
R
=2· π · r
2
R
(3)
The total resistances of the receiving and transmitting circuits
are:
R
Rx
=(2· π · r
R
· N
R
· R
0
)+R (4)
R
Tx
=(2· π · r
T
· N
T
· R
0
)+R
S
(5)
Fig. 2. NLoS-MI Relay
Fig. 3. Ideal transmitting and receiving coil configuration and the circuit
model (adapted from [41])
Fig. 4. Lateral Misalignment (adapted from [41])
whereR
L
and R
S
are the resistance of load and source respec-
tively and r
R
,r
T
, N
R
and N
T
are the radius and number
of turns of the receiving and transmitting circuit respectively.
R
0
is the per unit resistance of the wires used to build the
coils (copper wire in this case). Therefore the power transfer
function for the ideal communication link becomes [41]:
P
Rx
R
Tx
=
μ
2
0
· N
2
T
· N
2
R
· r
4
R
· ω
2
· r
4
T
16.R
Tx
· R
Rx
· (r
2
T
+ d
2
)
3
(6)
According to [42], [43] the power transfer function can be
also expressed as:
P
Rx
R
Tx
= Q
T
Q
R
k
2
(7)
k is the coupling coefficient and Q
T
and Q
R
denote the quality
factor of transmitting and receiving antennas [43]:
Q
T
=
ωL
T
R
Tx
=
ω
μ
0
πN
2
T
r
2
T
l
T
R
Tx
(8)
Q
R
=
ωL
R
R
Rx
=
ω
μ
0
πN
2
R
r
2
R
l
R
R
Rx
(9)
L
T
and L
R
are the inductance of the transmitting and
receiving coils respectively and l
T
and l
R
are the length of
JOURNAL OF NETWORKS, VOL. 8, NO. 5, MAY 2013
1003
© 2013 ACADEMY PUBLISHER
