You are on page 1of 5



Power Aware Routing Scheme in Ad Hoc

Divya Sharma, Ashwani Kush

Abstract— A recent trend in ad hoc network routing is the reactive on-demand philosophy where routes are established only
when required. Most of the protocols in this category, however, use single route and do not utilize multiple alternate paths. This
paper proposes a scheme to improve existing on-demand routing protocols by introducing the power aware virtual node scheme
in whole scenario. The scheme establishes the multi paths without transmitting any extra control message. It offers quick
adaptation to distributed processing, dynamic linking, low processing and memory overhead and loop freedom at all times. This
scheme uses the concept of Power awareness among route selection nodes by power states of each node in the topology
which insures fast selection of routes with minimal efforts and faster recovery. The scheme is incorporated with the Ad-hoc On-
Demand Distance Vector (AODV) protocol and the performance has been studied through simulation.

Index Terms— Ad Hoc networks, Alternate path, Routing Protocol, Power Consumption, Power Factor

——————————  ——————————


A n ad hoc network consists of hosts communicating

among themselves with portable radios. This net-
One distinguishing feature of Power aware ad hoc
routing protocol is its use of Power status for each route
work can be deployed without any wired base station or entry. Given the choice between two routes to a destina-
infrastructure support where routes are mainly multi-hop tion, a requesting node is required to select one with bet-
because of the limited radio propagation range. Topology ter power status and more active. The rest of the paper is
of the network changes frequently and unpredictably organized as follows. Section 2 illustrates the related
since its host moves randomly. Therefore, routing is an work which already has done on power aware routing
integral part of ad hoc communication, and has received protocols in Manet. New protocol description and system
interests from many researchers. In traditional “onde- methodology presented in section 3. Performance evalua-
mand” routing schemes like Ad Hoc On Demand Dis- tion via simulation is presented in Section 4 and conclu-
tance Vector Routing (AODV) scheme [3], when route sions are given in Section 5.
disconnects, nodes of the broken route simply drop data
packets because no alternate path to the destination is
available until a new route is established. When the net-
work traffic requires real time delivery (voice, for in-
stance), dropping data packets at the intermediate nodes
can be costly. The Power Aware Ad hoc routing Protocol Ad hoc networks have limited power capabilities mainly
algorithm enables dynamic, self starting, multi hop owing to the nature of the infrastructure they use. Power
routing between participating mobile nodes wishing to required by each mobile host can be classified into two
establish and maintain an ad hoc network. It allows mo- categories as (a) Communication-related power and (b)
bile nodes to maintain routes to destinations with more Non-communication-related power Three Layers are in-
stable route selection. This scheme responds to link brea- volved in communications as
kages and changes in network topology in a timely man- a) Physical layer Transmission power should be at a min-
imum level to maintain links. It should allow adapting to
ner and also takes care of nodes that does not have better
changes in transmission environment. Excessive trans-
power status and of Virtual nodes to participate in route mission power can cause interference to other hosts. b)
selection. Data Link Layer Energy conservation can be achieved by
using effective retransmission request schemes and sleep
mode operation. It is important to appropriately deter-
• Divya Sharma is with the Department of IT, ITM University Gurgaon,
mine when and at what power level a mobile host should
• Ashwani Kush is with the Department of Computer Science, University attempt retransmission. Node’s transceiver should be po-
college, Kurukshetra University, 132119. wered off when not in use. c) Network Layer In wireless
network it is important that the routing algorithm select
the best path from the viewpoint of power constraints as
part of route stability.Routing algorithms that can evenly

distribute packet-relaying loads to each nodes to prevent culated separately for the cases when transmit power is
nodes from being overused. fixed and when transmit power is varied dynamically.
2.1 Power Consumption States Variation is in terms of the distance that changes between
receiver and sender. For fixed case energy for each opera-
In a MANET, wireless communication involves usage of a
tion is given by [11]
transceiver at the source, intermediate, and destination
nodes. The transmitter sends control, route request and
E (packet) = b * packet_size + c
response, as well as data packets originating at or routed
through the transmitting node. The receiver is used to
Where b denotes packet size-dependent energy consump-
receive data and control packets -some of which are des- tion and c is a fixed cost for acquiring channel. For vary-
tined for the receiving node and some of which are for- ing transmit power Lin [12] proposed a local routing al-
warded. A wireless network interface has five possible gorithm. The authors assumed theα power needed for
energy consumption states (six including the off state). transmission is a linear function of d where d is distance
a) Transmit state for transmitting data, control and between two neighboring nodes and α is a parameter de-
routing packets pending on physical environment. This procedure re-
b) Receive state is for receiving data, control and routing quires GPS equipped systems, which in turn limits its
packets. usage and also it is based on least power cost routes
which in many cases lead to problems. In this case nodes
c) In the idle state, which is the default state for ad hoc
die soon and most of the cases those nodes die which are
environment, the interface can transmit or receive most needed to maintain network connectivity.
d) The sleep state has extremely low power consumption
as the interface can neither transmit nor receive in this
state. The previous case gave rise to number of battery cost
e) Lastly, a card can enter a reduced energy discard state aware routing algorithms as described below:
while the media carries uninteresting traffic. The decision (a) Minimum battery cost routing algorithm [13]: It mini-
to enter the reduced energy discard state is made by the mizes the total cost of the route. It minimizes the summa-
non-destination nodes in the range of the sender. This tion of inverse of remaining battery capacity for all nodes
decision is based on the packet size information in the on the routing path.
RTS (request to send) control packet that is exchanged (b) Min-Max battery cost algorithm [10] [13]: It is a mod-
between the sender and the receiver at the start of packet ification of minimum battery cost routing. It tries to avoid
transfer. The reduced energy state uses slightly less pow- the route with nodes having the least battery capacity
er than the idle state, but significantly more than that among all nodes in all possible routes.
used in the sleep state [3].
The objective of Power-aware Routing will be to extend
Main emphasis of research on routing protocols in Ad the useful service life of a MANET. This is highly desira-
Hoc networks has been delivery of packets and network ble in the network since death of certain nodes leads to a
performance. There has been very less amount of work possibility of network partitions, rendering other live
have done on power aware routing schemes, though it is nodes unreachable. Power aware source routing solves
very important aspect in route selection and performance the problem of finding a route π at route discovery time t
of protocol. Some study has been done in this context and such that the following cost function is minimized:
presented is a brief review of them.

3.1 Minimum Power Routing

Singh, Woo and Raghvendra [10] [11] proposed a routing Where

algorithm based on minimizing the amount of power re-
quired to send a packet from source to destination. The Pi is the transmit power of node i
equation is: Fi is the full charge battery capacity of node i
Ri remaining battery capacity of node i at time t
Min ∑ P(i,i +1) α is a positive weighting factor
i∋ path

Where P (i, i+1) denotes the power expended for transmit-

ting packet between node i and node i+1. Link cost is cal-

4 PROTOCOL CONCEPT comparing the simulation results of the AODV protocol

In this scheme main aim is to provide an efficient, more with and without applying this scheme and we also com-
stable and long lasting path from source to destination. pare this scheme with DSR and TORA. In this section, we
This routing scheme is designed for mobile ad hoc net- termed the AODV protocol that applied this scheme with
works with large number of nodes. It can handle low, alternate route and local repair scheme. We are using the
moderate, and relatively high mobility rates. It can handle network simulator NS-2 version 2.28. The Network Simu-
a variety of data traffic levels. This Protocol is designed lator NS-2 is a scalable simulation environment for wire-
for network topology in which the nodes can trust each less network systems. Simulation modeled a network of
other, and there are no malicious intruder nodes. Since 50 mobile hosts placed randomly within a 1000 meter *
the main aim is to improve the performance of existing 1000 meter area. Radio propagation range for each node
on-demand protocols (specifically AODV in this paper), was 300 meters and channel capacity was 2 Mb/s. Each
this protocol description is based on AODV. Modifica- run executed for 250 seconds of simulation time. Here we
tions to AODV for applying this scheme [14,15] are also are using random waypoint mobility model. Each node
introduced. There are three major operations in this randomly selects a position, and moves toward that loca-
scheme: RREQ (Route Request) phase, RERR (Route Er- tion with a speed between the minimum and the maxi-
rors) phase and local repair phase. As a major change to mum speed. Once it reaches that position, it becomes sta-
all existing protocols, we start Power related function tionary for a predefined pause time. After that pause
only when RREP (Route Reply) phase occurs as in the time, it selects another position and repeats the process.
start, when network is new, all nodes are fresh with ade- The minimum and the maximum speed were zero and 20
quate energy levels and can very well select shortest path m/s, respectively. The Metrics used for Performance com-
for data transmission. This in turn reduces overhead and parison are:
end to end delay.
Packet Delivery Ratio: The fraction of successfully re-
Route Construction: Route construction scheme can be ceived packets, which survive while finding their destina-
incorporated with reactive routing protocols. This scheme tion. This performance measure also determines the com-
does not require any modification to the AODV's RREQ pleteness and correctness of the routing protocol.
route request process. When a source wants to transmit a
data to a destination but does not have any route infor-
mation, it searches a route by broadcasting a ROUTE RE-
QUEST (RREQ) packet.
Successfully delivered packets, C is total number of flows,
f is id, R is packets received from f and T is transmitted
Route Error & Maintenance: In this scheme data trans-
from f.
mits continuously through the primary route unless there
is a route disconnection. When a node detects a link
End-to-End Delay: Average end-to-end delay is the delay
break, it performs a one hop data broadcast to its imme-
experienced by the successfully delivered packets in
diate neighbors. The node specifies in the data header
reaching their destinations. This is a good metric for
that the link is disconnected and thus the packet is candi-
comparing protocols. This denotes how efficient the un-
date for alternate routing. Upon receiving this packet
derlying routing algorithm is, because delay primarily
route maintenance phase starts by selecting alternate path
depends on optimality of path chosen.
and checking power status. Power factor is calculated
using status report based on three factors using a 2-D ma-
trix function. Three states are defined as Active, Critical
Average end-to-end Delay =
and Danger level. In this phase when route error message
has sent to previous neighbor of any intermediate node it
just reinitiates route construction phase by considering Routing Overhead: The number of routing packets sent
power status of all its virtual nodes. All this route main- by the routing protocol to deliver the data packets to des-
tenance occurs under local repair scheme. Local repair tination.
scheme tries to get rid off the regeneration of route re-
quest phase in the case of route break. Where S is no. of packets received successfully, ri is time
at which packet is received and si is time at which it is
sent, i is unique packet identifier.
To evaluate the performance made by the PAR, we are

nation. PAR delivers more data packets, and those pack-

ets that are delivered in PAR but not in AODV, take al-
ternate and possibly longer hop routes. PAR having long-
er delays than AODV does not represent its ineffective-
ness since these protocols use the same primary route.

A new scheme has been presented that utilizes power
states of each mobile node and alternate paths. This
scheme can be incorporated into any ad hoc on-demand
unicast routing protocol to improve reliable packet deli-
very in the face of node movements and route breaks.
Figure 1: Packet Delivery Ratio Alternate routes are utilized only when data packets can-
Figure 1 shows the throughput in packet delivery ratio. not be delivered through the primary route. As a case
We can see that this scheme improves the throughput study, this algorithm has been incorporated on AODV to
performance of AODV as well as this scheme performs check performance improvements. More simulations are
much better than DSR and TORA. in progress using speed and more pause time as func-
tions. Moreover system is under evaluation for sparse
medium as well. Simulation results have indicated that
our technique provides robustness to mobility and en-
hances protocol performance. However, this scheme may
not perform well under heavy traffic networks. We are
currently investigating ways to make this protocol robust
to traffic load. Additionally, we plan to further evaluate
this scheme by using more detailed and realistic channel
models with fading and obstacles in the simulation. The
performance of protocol is slightly costlier as compared to
AODV and its counterparts and causes slight overhead in
route selection initially.

[1] S.H. Bae, S.- J. Lee, W. Su and M. Gerla “The Design, Implementation
Figure2 : End-to-end delay and Performance Evaluation of On-Demand Multicast Routing Proto-
col in Multihop Wireless Networks “, IEEE Network,Special issue on Mul-
As the mobility increases (i.e., pause time gets shorter), ticasting Empowering the Next Generation Internet,Vol 14 no.1, pp. 70-77,
the performance gain by alternate routes becomes more 2000.
[2] A.Nasipuri, R. Castaneda, and S.R. Das,”Performance of Multi-
significant. Due to presence of neighboring which are in
Path routing for On-Demand Protocols in Mobile Ad Hoc Network”,
active state, nodes attempt to use multiple alternate paths
ACM/Baltzer Journal of Mobile Network and Applica-
for data delivery in the presence of route breaks, the pro- tions(MONET)
tocol is able to deliver more packets to the destination [3] C.E. Perkins, and E.M.Royer “Ad-Hoc On Demand Distance Vector
than AODV. AODV simply drops data packets when Routing,” Proc of 2nd IEEE Workshop on Mobile Computing Systems and
routes are disconnected. PAR also has some packet losses. Applications(WMCSA), New Orleans, pp. 90-100, 1999.
Alternate paths may be broken as well as the primary [4] J. Raju, and J.J.Garcia-Luna-Aceves, “A New approach to On-Demand
route because of mobility or power status of nodes, or be Loop Free Multi path Routing,” Proc of 8 th IEEE International conference
unavailable and not discovered during the route reply on computer communications and Networks(ICCCN), Boston, MA, pp. 522-
527, 1999.
phase. Moreover, packets can be lost because of collisions
[5] Z. Wang, and J.Crowcroft “Shortest Path first with emergency Exits,”
and contention problems.
Proc of ACM SIGCOMM Symposium on Communication Architectures and
protocols, Philadelphia, PA, pp. 166-176, 1990.
End-to-end delay is presented in Figure 2. As expected, [6] J. Moy, “Link state Routing,” In Routing in communication Networks,
PAR has longer delays than AODV. We can only measure edited by M.E. Steenstrup, Prentice Hall, pp. 135-157, 1995.
delays for data packets that survived to reach their desti- [7] V.D. park, and M.S.Corson, “A Highly AdaptivNew approach to On-

Demand Loe Distributed Routing Algorithm for Mobile wireless Net-

works,” Proc of IEEE International conference on computer communications
(INFOCOM), Kobe, Japan, pp. 1405-1413, 1997.
[8] N. Shancham, E. J. Craighill and A.A. Poggio, “Speech Trans-
port in Packet-Radio Networks with Mobile Nodes,” IEEE Jour-
nal on Selected Areas in communications,Vol. SAC-1,no.6, pp.
1084-1097, 1983.
[9] A Nasipuri, and S.R.Das, “On-Demand Multi path routing for Mobile
Ad Hoc Networks,” Proc of 8 th Annual IEEE International conference on
computer communications and Networks(ICCCN), Boston, MA, pp. 64-70,
[10] S. Singh, Mike Woo, C. S. Raghavendra, “Power Aware routing in
Mobile Ad Hoc Networks,” Proc of ACM MOBICOM 98, Dallas, 1998.
[11] S. Lindsay, K. Sivalingam, and C.S. Raghavendra, “Power
aware routing and MAC protocols for wireless and mobile
networks,” in Wiley Handbook on wireless Networks and Mo-
bile Computing; Ed., John Wiley & Sons, 2001.
[12] Stojimenovic, and X .Lin “Ppower aware localized routing in wireless
networks,” Proc of IEEE IPDPS, Cancum, Mexico, 2000.
[13] C.K. Toh, “Maximum Battery Life Routing to support Ubiquitous Mo-
bile Computing in wireless Ad Hoc Networks,,” IEEE Communication
Magazine, 2001.
[14] A.Kush, P. Gupta, "Power Aware Virtual Node routing Proto-
col for Ad Hoc Networks," International Journal of Ubiquitous
Computing and Communication(UBICC),Vol 2 No. 3, pp 55-62,
[15] A.Kush, P. Gupta and C. Hwang, "A Hybrid Stable Routing
Protocol for Mobile Ad Hoc Networks," Proc of the World Wire-
less Congress” WWC 2008,San Francisco,USA, pp 19-25,2008

Ms Divya Sharma is working as Assistant professor in IT Depart-

ment at ITM University, Gurgaon. She is pursuing PhD in computer
Science from Kurukshetra University. She has interest areas in Wir-
less networking and Power management. She is member of CSI,

Dr Ashwani Kush is working as Head of the Computer science

Department at University College Kurukshetra University, India. He
has done PhD in association with Indian Institute of Technology
Kanpur and Kurukshetra University. He has more than 60 research
papers to his credit in various International/National Journals and
conferences.He has authored 15 books in computer science for
undergraduate and school student. His research interests are Ad hoc
networks Routing, Security and Power awareness. He is member of