-
An efficient void resolution method for geographic routing
in wireless sensor networks
The Journal of Systems and Software 82 (2009) 963–973
Young Il Ko a,*, Chang-Sup Park b, In Chul Song a, Myoung Ho Kim
a a Division of Computer Science, Korea Advanced Institute of
Science and Technology,
373-1, Guseong-Dong, Yuseong-Gu, Daejeon 305-701, Republic of
Korea b Department of Computer Science, Dongduk Women’s University,
Republic of Korea
Speaker : TSUNG-JU LEE Data : 2013.5.21
-
Outline
Motivation Introduction Basic proposed method Other
considerations Experimental results Conclusion
The Journal of Systems and Software
-
Motivation
The Journal of Systems and Software
• Geographic routing is an attractive choice for routing data in
wireless sensor networks because of lightweight and scalable
characteristics
• Conventional geographic routing methods commonly use the
right-hand rule for void resolution
• However, the detour path produced by the right-hand rule is
not energy-efficient in many cases
• The right-hand rule often produces a path containing many
redundant nodes
3
Fig. 1. A routing path reduction
-
Introduction(1/3)
• Wireless sensor networks consist of many sensor nodes which
have sensing, computing, and wireless communication
capabilities
• Energy efficiency should be considered a primary factor in
developing sensor network applications
• An efficient technique is required to transport data across a
sensor network because energy consumption in data communication is
much greater than that of local computation
The Journal of Systems and Software 4
-
Introduction(2/3)
• Geographic routing is a promising approach for message
transmission in large-scale location-aware sensor networks
• Most of the previous geographic routing approaches use a
combination of greedy forwarding and a certain variation of face
routing
• However, greedy forwarding may route a message to a stuck node
– Stuck node
• A node that has no neighbor whose distance to the destination
is shorter than that of the current node
• When greedy forwarding is not possible due to a void area, we
need a void resolution mechanism
The Journal of Systems and Software 5
-
Introduction(3/3)
• When a message gets stuck at a node in the course of greedy
forwarding, the routing mode of GPSR is changed into perimeter
forwarding
• In the perimeter forwarding, a message is routed by the
right-hand rule along the faces in the planar graph of the sensor
network until either it can be routed by greedy forwarding again or
it returns to the node where it entered the perimeter mode
• However, the methods based on the right-hand rule may produce
inefficient routing paths
The Journal of Systems and Software 6
-
Outline
Motivation Introduction Basic proposed method Other
considerations Experimental results Conclusion
The Journal of Systems and Software
-
The Journal of Systems and Software
Basic proposed method(1/10) • Azimuth angle and distance
– In this paper, It uses the 2D polar coordinate system to
locate the sensor nodes in a sensor network
8
-
Basic proposed method(2/10)
• Quadrant division – Quadrant 1 is the area that has a shorter
distance and a larger azimuth angle than the node n – Quadrant 2 is
the area that has a longer distance and a larger azimuth angle than
the node n – Quadrant 3 is the area that has a longer distance and
a smaller azimuth angle than the node n – Quadrant 4 is the area
that has a shorter distance and a smaller azimuth angle than the
node n
The Journal of Systems and Software 9
-
Basic proposed method(3/10)
• Overall message transmission policy – Greedy forwarding
(1) If there are neighbors whose distances to the destination
are shorter than that of the current node, forward a message to the
neighbor whose distance is the shortest among them (2) If the
destination is reached, stop (3) If a stuck node is reached, start
void resolution-forwarding
– Void resolution (VR)-forwarding (1) Determine a quadrant of
the current node to send a message from the current node based on
the quadrant selection scheme (2) Select a successor node ns among
the nodes in the quadrant determined in step 1, based on the node
selection heuristics, and send the routing message to it .Now ns
becomes the current node
The Journal of Systems and Software 10
-
Basic proposed method(4/10)
• Overall message transmission policy(cont.) – Choose one of the
following cases
(1) If ns satisfies the condition of greedy forwarding, i.e.,
the distance of ns to the destination is shorter than that of the
stuck node, resume greedy forwarding at ns (2) Otherwise, go to
step 1(Determine a quadrant of the current node)
The Journal of Systems and Software 11
-
Basic proposed method(5/10)
• Void resolution-forwarding – Determination of the routing
quadrant
The Journal of Systems and Software 12
-
Basic proposed method(6/10)
• Void resolution-forwarding(cont.) – Selection of the successor
node
• The underlying motivation of all the heuristics is to select a
node that can result in a routing path going out of the quadrant
more quickly and thus having a smaller number of hops to the
destination than a path generated by the conventional right-hand
rule
The Journal of Systems and Software 13
-
Basic proposed method(7/10) • Void
resolution-forwarding(cont.)
– Selection of the successor node(cont.) • Example
– Node R is selected by the proposed heuristic for quadrant 1,
which has the shortest distance in quadrant 1
The Journal of Systems and Software 14
Fig. 6. Node selection in quadrant 1.
-
Basic proposed method(8/10)
• Void resolution-forwarding(cont.) – Selection of the successor
node(cont.)
• In any quadrant, multiple nodes can tie on the node selection
criterion for each quadrant
• It uses additional heuristics to break the tie based on the
values of the other polar coordinate of the nodes
• This strategy can make a routing path which is more close to
the faces of the void
The Journal of Systems and Software 15
-
Basic proposed method(9/10)
• Void resolution-forwarding(cont.) – Routing termination
during
VR-forwarding • First, It aims to detect a
ring-type void surrounding the destination
• It sums up the amount of differences between the azimuth
angles of a receiver and a sender at every message hop starting
from the stuck node
The Journal of Systems and Software 16
-
Basic proposed method(10/10)
• Void resolution-forwarding(cont.) – Routing termination
during
VR -forwarding(cont.) • As another case, Fig. (b)
shows a ring-type void surrounding the source node
• It measures and accumulate the size of the angle between two
rays originating from the source
The Journal of Systems and Software 17
-
Outline
Motivation Introduction Basic proposed method Other
considerations Experimental results Conclusion
The Journal of Systems and Software
-
The Journal of Systems and Software
Other considerations(1/4) • The quadrant-level right-hand rule
first examines the Quadrant 1 of the
current node, it can miss a routing path passing through a node
in the backward quadrant
19
-
Other considerations(2/4)
• To find a direct node, the current node needs to investigate
its backward quadrant first prior other quadrants by the
quadrant-level right hand rule
• The qualification for a direct node satisfied the condition
that depends on the current direction as described in Table 3
The Journal of Systems and Software 20
-
Other considerations(3/4)
• Void resolution using direct nodes
The Journal of Systems and Software 21
Fig. 18. An efficient detour path through a direct node (when
the current routing direction is F-direction)
-
Other considerations(4/4)
• Switchover to GPSR – The proposed VR-forwarding void
resolution scheme has a possibility
to fail to find a routing path from the source to the
destination by entering into a loop even when there exists a path
in a given sensor network
– It suggests a strategy to exploit the GPSR method when
VR-forwarding cannot find a detour path
– It changes the routing method from VR-forwarding to GPSR when
the following condition is satisfied:
– The right-hand side of the above inequality formula means the
number of message hops in a routing path along the boundary of the
entire sensor network
The Journal of Systems and Software 22
Where w and h denote the width and height of the given sensor
network area and r is the radius of the radio range of the sensor
nodes
-
Outline
Motivation Introduction Basic proposed method Other
considerations Experimental results Conclusion
The Journal of Systems and Software
-
Experimental results(1/6)
• Simulation setting – In the experiments, It has generated
sensor networks having 1500 to 5000
nodes which are randomly scattered over the flat area of 600 by
600 square meters
– It uses a T-shape void whose aspect ratio (i.e. the ratio of
width to height) is 5:4
– It selects a source and destination node near the lower left
corner and upper right corner of the sensor network,
respectively
The Journal of Systems and Software 24
-
Experimental results(2/6) • Energy efficiency
– Fig. 11 presents the results of experiments with various
numbers of fixed-size voids in sensor networks
– Average hop stretches • The ratio of the number of hops in a
routing path to the number of hops in the shortest path
from the source to the destination
The Journal of Systems and Software 25
-
Experimental results(3/6)
• Energy efficiency(cont.)
The Journal of Systems and Software 26
-
Experimental results(4/6)
• Energy efficiency(cont.)
The Journal of Systems and Software 27
-
Experimental results(5/6) • Message delivery success rate
The Journal of Systems and Software 28
Fig. 14. Delivery success rate of the proposed method
-
Experimental results(6/6)
• Effect of the hybrid approach
The Journal of Systems and Software 29
-
Conclusion • In this paper, It has proposed a new geographic
routing scheme called VR-
forwarding which can efficiently detour void areas
• Based on the polar coordinate system to locate sensor
nodes
• It provides the quadrant-level right-hand rule and node
selection heuristics to generate an energy-efficient routing path
when the greedy forwarding fails due to various types of voids
• The experimental results also present that it can produce more
efficient routing paths than the previous GPSR geographic routing
protocol based on the conventional right-hand rule
The Journal of Systems and Software 30
-
THANK YOU FOR LISTENING
An efficient void resolution method for geographic routing in
wireless sensor
networksOutlineMotivationIntroduction(1/3)Introduction(2/3)Introduction(3/3)OutlineBasic
proposed method(1/10)Basic proposed method(2/10)Basic proposed
method(3/10)Basic proposed method(4/10)Basic proposed
method(5/10)Basic proposed method(6/10)Basic proposed
method(7/10)Basic proposed method(8/10)Basic proposed
method(9/10)Basic proposed method(10/10)OutlineOther
considerations(1/4)Other considerations(2/4)Other
considerations(3/4)Other considerations(4/4)OutlineExperimental
results(1/6)Experimental results(2/6)Experimental
results(3/6)Experimental results(4/6)Experimental
results(5/6)Experimental results(6/6)ConclusionThank you for
listening