IEEE Sponsored 9th International Conference on Intelligent Systems and Control (ISCO)2015

A Simple Routing Algorithm for Multi-hop

Wireless Network

Ishita Chakraborty Upasana Sarmah

CSE Dept. Royal Group ofInstitutions CSE Dept. Royal Group ofInstitutions
Affiliated to Gauhati University Affiliated to Gauhati University
Guwahati-35, India Guwahati-35, India
ishitachakrabortyl8@gmail.com

Abstract- Routing plays a critical role in determining the length hop. These networks do not require a sophisticated
performance of a multi-hop wireless mesh network. When there infrastructure. Therefore, they can be rapidly deployed in
are a large number of nodes available the task of communication remote, hostile, or hard to reach areas. [3]. A localized routing
becomes quite challenging. Our main concern is energy-efficient algorithm must know the coordinates of the user, its neighbors
routing. Energy conservation is a critical issue in wireless
as well as those of the destination, Base Station (BS) at each
networks for node and network lifetime, as the nodes are
point of time. It is not required to have full knowledge of the
powered by batteries only. One major approach for energy
topology of the entire network. Upon arrival to an IM, the
conservation is to establish a communication session which
consumes the lowest total transmission energy. This optimization local information of the IM regarding its neighbors can be
problem is referred as minimum-energy routing. used and the edges connecting IM to them. Using this plus the
Our work, we have considered a simple routing information that is stored in the user's local memory, an edge
algorithm which can be implemented in both centralized and incident to IM is chosen, and traversed until the next node is
distributed manner. Energy consumed in each of the route is reached, unless that is IM=BS, in which case it stops.[4].
calculated. The route in which energy consumption is the less is Wireless networks need efficient power management, since
chosen for the communication.
transmission power is a precious resource. Power failure of a
mobile node not only affects the node itself but also its ability
Index Terms- Wireless Mesh Networks, Routing, Energy
to forward packets on behalf of others and thus the overall
Consumption, Throughput, Minimum Angle, Minimum Distance.
network lifetime. This optimization problem is referred as
minimum-energy routing discussed in [5].Problem related to
multiple access is discussed in [6]. The authors in [7]
I. INTRODUCTION discussed various categories of energy efficient routing

Wireless Mesh Network (WMNs) is a communication protocol.

network. A WMN contains dozens or hundreds of nodes

11. METHODOLOGY
called as Wireless Mesh Nodes. These nodes are small radio
transmitters and are organized in a manner of mesh topology. Routing in multi-hop wireless networks is a critical process.
These radio transmitters are functionally analogous to wireless Routing schemes should be carefully considered in order to

routers. The network connection is spread out among these maintain the connection between the communicating nodes. It
nodes. Even for a large area, these nodes can communicate is the sole responsibility of the routing protocol to establish a

among each other or share the network connection. WMNs communication link between the source and the destination.
can connect big cItles using eXIstmg technology Routing protocols can depend on some Metric in order to

inexpensively, easily and effectively. In order to communicate choose the best path for forwarding of packets [8 - 12]. The

wirelessly with users and also with each other, Wireless Mesh metric that we have focused is the Transmission energy.
Nodes can use the Wi-Fi standards: 802.11a, and g. [1 -2]. A For performance evaluation of the routing algorithms we

communication path is established between two parties either have considered, four network scenarios. The network

through a single-hop transmission in which the parties are scenarios contain a BS at the center and end users who want to
within transmission range, or through relaying by intermediate establish links with it. There are one or more IMs which help
nodes (IMs). In Multi-hop Wireless Networks, one or more in forwarding packets through the links for communication to
IMs are present between the two communicating parties along take place between the user U and the BS.

the route. These IMs receive and forward the packets via A. Minimum Distance Routing Algorithm
wireless links. Over the last few years, multi-hop wireless
In this routing algorithm, an end user node U, wants to
networks have gained considerable amount of attention. This
communicate with the BS. For constructing the route, U first
might be due to the fact that the required transmission energy
fmds the IMs within a given transmission range. The IM with
for multiple hops of shorter length is less than a single longer

978-1-4799-6480-2/15/$31.00 © 2015 IEEE

 IEEE Sponsored 9th International Conference on Intelligent Systems and Control (ISCO)2015
which U has the minimum distance is selected and a link is For each node, information about its neighbors and
established. This IM node represents the next hop and acts as base station is sufficient.
Minimum-Distance forwarders (or sometimes called as ii. The effect of user's mobility is not considered in this
relays). [n the same manner, the role of U is transferred to the study.
chosen [M node and finally the BS is reached. Here, an Ill. There is always a node available within a
important point to be noted is that if the user U initiates the transmission range for connection based on
communication it is said to be a distributed routing (OR). But Minimum Angle (MA) and Minimum Distance (MD)
if the communication is initiated by the BS, it is called a criteria.
centralized routing (CR). Thus, Minimum Distance routing IV. Network lifetime IS sustainable for all
initiated by U is called Minimum Distance Distributed communication to take place successfully. We define
Routing (MD-DR) and the Minimum Distance routing the network lifetime as duration of time until the first
initiated by the BS is called Minimum Distance Centralized node in a network fails due to battery exhaustion.
Routing (MD-CR). The routes obtained for both MD-DR and [16]
MD-CR may be different. Also the, end-to-end throughput v. Link errors might influence the transmission of
performance may differ [13].Adaptive routing mechanism for packets. But we have not considered this in our study
on demand routing protocol is analyzed in [14]. and we assume that as long as there is a possible
route from U to BS, the communication is successful.
VI. Throughout this paper, the links indicated by green
B. Minimum Angle Routing Algorithm
color are selected on basis of MD criteria and those
Here, the user U in order to construct the route firstly draws a indicated by blue color are selected by MA criteria.
reference line to the BS which is the destination. Within a VII. We assume that every user has a 15Mb file to
given transmission range, U now chooses the IM node which transfer.
makes the minimum angle with the reference line drawn. The viii. The energy consumption is proportional to the square
role of U is now transferred to the chosen IM node and the of the distance between two nodes.[17]
same procedure is repeated until the BS is reached. It is We also assume that the nodes obtain different bit rates over
significant to note that at each step the reference line is drawn various transmission distances, as shown in Table 11.
between the sender node which can be the user U or the [M
node and the BS. Hence, the route is constructed hop by hop TABLE [I. BIT RATES AND TRANSMISSION RANGES
and the [Ms are called Minimum-Angle forwarders [13]. The
Minimum Angle routing initiated by U is called Minimum Modulation Coding Transmission Bit
Angle Distributed Routing (MA-DR) and the Minimum Angle Rate Range (Km) Rate
routing initiated by the BS is called Minimum Angle (Mbps)
Centralized Routing (MA-CR). [13] and [15] reports MA-DR QPSK Yz 5 2
and MA-CR based routing and scheduling results for three
[6-QAM Yz 3.5 5.5
example networks.
64-QAM % 2 11
Ill. PERFORMANCE EVALUATION

For performance evaluation of our algorithm, we consider four A. Results and Discussion
network scenarios of medium size, and find the routes for
[n the first network topology, shown in Fig 1 the nodes are
various numbers of users to the BS. Table I shows the
placed around the BS which is placed at the center. There are
particulars of the network scenario.
30 nodes including a BS and 3 users which are denoted as Ul,
U2, and U3, respectively. The transmission range here is 3.5
TABLE I. PARTICULARS OF FOUR NETWORK SCENARIO
km. and the routes established by the users based on MD
algorithm and MA algorithm.
Network No. of No. No. No. Transmission Energy spent in transmission depends upon the distance
Scenario Nodes of of of Range travelled by the signal. [[8]
Users IMs BS (Km) E[] = Kd

1 30 3 26 1 3.5 Where, K = constant, (since in this study the main aim is the
comparison of energy consumptions, so for simplicity we have
2 50 4 45 1 3
taken K=[).
3 [00 4 95 1 4.5
d = distance between nodes i and j.
4 20 3 16 1 4 The total energy consumption for transmission of a packet on
a specific route (from a source node to its corresponding
The following cases are considered for every network scenario destination) can be determined as follows:

I. To route a packet to the base station the nodes do not

need to know the exact positions of all other nodes.
 IEEE Sponsored 9th International Conference on Intelligent Systems and Control (ISCO)2015

1 0 00
Where p represents a route and link (i,j) I: p. Selecting the path .p wP �
91]0 .u ....,
with a minimum energy requirement improves the energy .tf1 �
efficiency of the network.Therefore, based on energy 800 ;fI9�
�1 � oo lili1
consumed by nodes present in both routes, the route with .73 .'
mo
�
minimum energy consumption is chosen for communication
!I 000
and is shown in the Fig. I (nodes present in the energy j
efficient route are indicated by +). 12
Cl
�o

2(]0
8,.. -lOO .' 9
!is
3(]O ". .7
1
fP6
• OIl' ... .;n
160
�
.7 2(J0 � ,.0
140
100 ,g� ,.
rr
"" 120 .. ' ..
:!. 0
g mo
.. . .0 0 100 :!OO 3DO 500 600 700 BOO ·000 1000
8... x co-,ordinate
8[)
Fig. 3. Routes by MA and MD for Network Scenario 3
6[)

4[)
80
Ii'

70
BO 100 120 140 160 lBO 20 0
x co---onfI1 ate-

60
Fig. 1. Routes by MA and MD for Network Scenario 1

50
The second network scenario consists of 50 nodes including
the BS at the center. The transmission range is 3 km. Fig. 2,
40
shows the routes constructed for users Ul, U2, U3, and U4 by
both MA and MD routing algorithms along with the energy 30
efficient path.
20
2110

lao 10
0 10 20 30 40 50 60 70 80 90

1110
Fig. 4. Routes by MA and MD for Network Scenario 4
1-10

i 120 In Fig. 1, the routes found by applying both MA and MD

... algorithms are shown. The power supply of wireless nodes is
'e lOO
�
,., 80
very limited, and the transmission
proportional to the square of distance between the two nodes.
energy is directly

!lO [19], [20].

Since we know the relation, D 2 Di + n;=
40
Where, D = Di + D , it is ideal to send information in smaller
20 hops rather than one transmission over a longer hop distance.
O L-�__�__-L��__���������� For user UI, the path created by MA Algorithm considers
o 2!l -10 60 8D lOD 1 211 180 200
• OOr-ORfin ate
the intermediate nodes with smaller hops compared to longer
ones in MD path. So, the path created by MA consumes less
Fig. 2. Routes by MA and MD for Network Scenario 2 energy than the other path. For user U2 and U3, MD
Algorithm provides the least energy consumed path in
The third network consists of 100 nodes with four users U I, comparison to MA path shown in Fig. 5. For UI-BS pair, the
U2, U3, and U4. The transmission range is 4.5 Km. The routes least energy consumed path also provides better throughput.
established by both MA and MD routing algorithms and the But, it is seen that it is not always the case that the path with
energy efficient path shown.in the Fig. 3. least energy consumption provides high throughput. As for
The fourth scenario, shown in Fig. 4 consists of 20 U3, from Fig. 6, we can see that the throughput by MA path is
nodes including the BS and users U1, U2, and U3. Here, more but since it consumes higher energy for transmission, the
the transmission range for all nodes is 4 km. MD path is chosen with moderate throughput.
 IEEE Sponsored 9th International Conference on Intelligent Systems and Control (ISCO)2015

10000 For the third network scenario, the energy consumption for

'"
� 9000 _Minimum Angle each path is shown in Fig. 9 and throughput for each user is
-5 8000 _Minimum Distance shown in Fig. 10. For U1, U2 and U4, the MA path consumes
� 7000 less energy and also provides better throughput. For user U3
:gE 6000 the MD route provides better results
� 5000
g 4000
o
>- 3000 1.6 .Minimum Angle
Cl
� 2000 1.4
c
_Minimum Distance
•
w 1000 � 12
o .<>

15 26 29 �
•
Subscriber station id �
..c:
0.8
g> 0.6 • •
Fig. 5. Energy Consumed for each User (Network Scenario 1) e •
r= 0.4
• •
02
•
+Minimum Angle o +-------,---,
1 o 20 40 60
0.9 .Minimum Distance •• Subscriber Station ID
'[ 08
� 07 •
� 0.6 Fig. 8. Throughput (Network Scenario 2)
� 05 • •
� 04 60000
::J .Minimum Angle
e 0.3 fi)
Q)
� 0.2 .Minimum Distance
:; 50000
o
0.1 ...,
0+------,---,--, § 40000
o 10 20 30 40 �
Subscriber Station Id
§'" 30000
c
o
Fig. 6. Throughput (Network Scenario 1) (,) 20000
>-
Cl

In Fig. 2. U2 instead of one longer hop from U2 to

� 10000
w
intermediate node 26, four smaller hops are taken which
o
reduces the energy consumption. But too many smaller hops 6 88 72 62
are not always acceptable. As for user U1, instead of single Subscriber Station Id
longer hop from node 39 to node 42, the MD routing considers
8 smaller hops in between these two nodes. This leads to Fig. 9. Energy Consumed for each User (Network Scenario3).
overall high energy consumption. So, the MA path is
considered. We can also see that too many hops also affect the 0.6
• Minimum Angle
End-to-End Throughput from the Fig. 8. The throughput is •
0.5 • Minimum Distance
also less in the MD path. For U3 and U4, the least energy
fi)
consumed path also provides better throughput. .8- 0.4
:lE
:s • •
a. 0.3 •
10000 ..c: • •
Cl
.Minimum Angle ::J
fi) 9000 e 0.2 •
-S 8000
o .Minimum Distance
..c:
I-
2. 7000 0.1
� 6000 0
iil 5000
0 20 40 60 80 100
§ 4000
o Subscriber Station Id
>- 3000
� 2000 Fig. 10. Throughput (Network Scenario 3).
.lJ 1000
o
For the fourth network scenario, the energy consumption for
8 19 45 49
each path is shown in Fig. 11 and throughput for each user is
Subscriber station id
shown in Fig. 12.

Fig. 7. Energy Consumed for each User (Network Scenario2)

 IEEE Sponsored 9th International Conference on Intelligent Systems and Control (ISCO)2015
For UI and U2, the MA path consumes less energy and also [4] Evangelos Kranakis, Harvinder Singh, and Jorge Urrutia.
"Compass Routing on Geometric Networks". In ProceedinQs of
provides better throughput. For user U3 the MD route
the 11th Canadian Conference on Computational Geometry
provides better results. (CCCG '99) , 1999.
[5] P.-J. Wan G. Calinescu X.-Y. Li_ O. Frieder, "Minimum-Energy
1400 Broadcast RoutinQ in Static Ad Hoc Wireless Networks".
INFOCOM 2001 Twentieth Annual Joint Conference of the
-.. 1200 IEEE Computer and Communications Societies. Proc.
•
:; IEEE (Volume:2 ), pp. 1162 - 1171, Apr. 2001
.!i 1000
:;- [6] Tamer El Batt and Anthony Ephremides, "Joint Scheduling and
•
E 800 Power Control for Wireless Ad Hoc Networks", IEEE
� Transactions on Wireless Communications, Vol. 3, NO. 1, Jan.
0 600 • Minimum Angle
0
u 2004
>- 400 • Minimum Distance
!." [7] Nikolaos A. Pantazis, Stefanos A. Nikolidakis and Dimitrios D.
•
0
... 200 Vergados, "Energy-Efficient Routing Protocols in Wireless
Sensor Networks: A Survey", IEEE Communications Surveys &
Tutorials, Vo!. 15, No. 2, Second Quarter 2013
11 20
[8] T. Hou and V.O.K. Li, "Transmission Range Control in
Subscriber station Id
Multihop Packet Radio Networks", IEEE Trans. Comm., Vol.
34, no. 1, pp. 38-44, 1986.
[9] T. Melodia, D. Pompiliand T.F. Akyildiz, "Optimal Local
Fig. 11. Energy Consumed for each User (Network Scenario4) Topology Knowledge for Energy Efficient Geographical Routing
in Sensor Networks," Proc. IEEE INFOCOM, Vo!. 3, pp. 1705-
1.2
1716, Mar. 2004

1 • [10] K. Kar, M. Kodialam, T.V. Lakshman, and L. Tassiulas,

� "Routing for Network Capacity Maximization in Energy­
Co
! 0.8 • Constrained Ad-Hoc Networks." Proc. IEEE INFOCOM. Vol. L
..
pp. 673-681, Apr.2003.
[ 0.6 • Minimum Angle
[11] L. Lin, N.B. Shroft: and R. Srikant, "Asymptotically Optimal
..c
t>J)
5 0.4 • Power-Aware Routing for Multihop Wireless Networks with
Renewable Energy Sources," Proc. IEEE INFOCOM, Vo!. 2,
•
pp. 1262-1272, Mar. 2005
� 0.2 • Minimum Distance
[12] Sungoh Kwon, Member, IEEE, and Ness B. Shroff, Fellow,
0 IEEE, "Energy-Efficient SINR-Based Routing for Multihop
Wireless Networks", IEEE Transactions on Mobile Computing,
0 10 20 30
VOL. 8, NO. 5, MAY 2009
Subscriber station Id
[13] Ishita Chakraborty, and Md. Anwar Hussain, "A Simple Joint
Fig. 12. Throughput (Network Scenario 4) Routing and Scheduling Algorithm for a Multihop Wireless
Network," ICCS1ll2, AUS, Dubai, December 2012.
IV. CONCLUSION [14] Lamia Romdhani and Christian Bonnet, "Energy Consumption
Rate-Based Routing for Mobile Ad Hoc Networks", International
In our paper, a simple routing algorithm is taken into account.
Journal Of Wireless and Mobile Computing.
For performance evaluation, each routing algorithm is
[15] Md. Anwar Hussain, Md. Faiz Iqbal Faiz, and Ishita
simulated under different scenarios consisting different Chakraborty, "Simple Routing Algorithm for Multi-hop Wireless
number of nodes. Our simulations mainly concentrate on Network", Journal of Emerging Trends in Computing and
distributed route construction. The route construction Information Sciences, Vol. 4.Special Issue ICCSll , ISSN 2079-
8407 Pg 58-65
algorithms are simple and require only the (x-y coordinate)
[16] Intae Kang and Radha Poovendran, "On Lifetime Extension and
information of the nodes and the route so constructed is
Route Stabilization of Energy-Ef
f icient Broadcast Routing over
energy efficient. In our further work we show the MANET', Department of Electrical Engineering University of
performances of the network by mixing both MA and MD for Washington, Seattle, WA. 98195-2500, U.S.A
optimal routing and compare them for both centralized and [17] Backhyun Kim, and Iksoo Kim, "Energy Aware Routing
Protocol in Wireless Sensor Networks", TJCSNS International
distributed route.
Journal of Computer Science and Network Security, VOL.6
No.l, January 2006
REFERENCES
[18] M. M. Chandane, S. G. Bhirud, S.V. Bonde, "Distributed
[1] Mihail L. Sichitiu, "Wireless Mesh Networks: Opportunities and Energy Aware Routing Protocol for Wireless Sensor Network",
Challenges", Dept. of Electrical and Computer Engineering, International Journal of Computer Applications(0975-
North Carolina State University Raleigh, NC 27695, USA. 8887)Volume 34-No.3, November 2011 Pg- 6 to 11.
[2] V. Lakshmi Praba,A. Mercy Rani, "Efficient Neighbor Routing [19] Ivan Stojmenovic and XU Lin, "Power-Aware Localized Routing
in Wireless Mesh Networks", International Journal of Computer in Wireless Networks", IEEE Transactions on Parallel and
Science and Business Informatics, ISSN: 1694-2108 Vol. 1, No. Distributed Systems, Vo!. 12, No. 10, Oct. 2001.
1. MAY 2013
[20] R Ramanathan, R Hain, "Topology Control of Multihop
[3] Torsten Braun, Andreas Kassler, Maria Kihl, Veselin Wireless Networks Using Transmit Power Adjustment".
Rakocevic, Vasilios Siris, Geert Heijenk, "Traffic and QoS In Proc. Nineteenth Annual Joint Conference of the IEEE
Management in Wireless Multimedia Networks", pp 201-265, ComDuter and Communications Societies (lNFOCOM'OO),
doi: 10.1007/978-0-387-85573-8 5, 2009, ISBN 978-0-387- pages 404--413, March 2000.
85572-1, Vol: 31, Publisher: Springer US