college of engineering non-uniform grid- based coordinated routing priyanka kadiyala major advisor:...
TRANSCRIPT
College of Engineering
Non-uniform Grid-based Coordinated
Routing
Non-uniform Grid-based Coordinated
RoutingPriyanka Kadiyala
Major Advisor: Dr. Robert Akl
Department of Computer Science and Engineering
Priyanka Kadiyala
Major Advisor: Dr. Robert Akl
Department of Computer Science and Engineering
OutlineOutline
• Research objective
• Overview of sensor networks
• Related work
• Motivation
• Non-uniform grid-based coordinated routing protocol
• Simulations and results
• Research objective
• Overview of sensor networks
• Related work
• Motivation
• Non-uniform grid-based coordinated routing protocol
• Simulations and results
04/19/23
Overview of Sensor NetworksOverview of Sensor Networks
• Ad hoc networks of tiny battery powered sensor nodes capable of sensing, processing and communicating data.
• Applications - Video surveillance, traffic monitoring, environmental monitoring, structure and system health monitoring in buildings and aircraft interiors.
• The main source of energy is battery, no external supply of power - major constraint is energy available.
• Ad hoc networks of tiny battery powered sensor nodes capable of sensing, processing and communicating data.
• Applications - Video surveillance, traffic monitoring, environmental monitoring, structure and system health monitoring in buildings and aircraft interiors.
• The main source of energy is battery, no external supply of power - major constraint is energy available.
04/19/23
Research ObjectiveResearch Objective
To increase the lifetime of the sensor network by using non-uniform grid based routing for the case of random node deployment.
To increase the lifetime of the sensor network by using non-uniform grid based routing for the case of random node deployment.
04/19/23
Overview of Sensor NetworksOverview of Sensor Networks
04/19/23
Overview of Sensor NetworksOverview of Sensor Networks
Protocols for WSNs
• Flooding, Gossiping, SPIN, LEACH, PEGASIS, Directed Diffusion, and GEAR
• Energy efficient protocols that allow nodes to be put to sleep are GAF, SPAN, STEM, ASCENT, CEC, AFECA and GBCR.
Protocols for WSNs
• Flooding, Gossiping, SPIN, LEACH, PEGASIS, Directed Diffusion, and GEAR
• Energy efficient protocols that allow nodes to be put to sleep are GAF, SPAN, STEM, ASCENT, CEC, AFECA and GBCR.
04/19/23
Related WorkRelated Work
Flooding :
• In flooding every node that receives a packet broadcasts it to its neighbors. If the node receives the packet for the first time, it is stored in the buffer. If it is a redundant packet, it is discarded.
Flooding :
• In flooding every node that receives a packet broadcasts it to its neighbors. If the node receives the packet for the first time, it is stored in the buffer. If it is a redundant packet, it is discarded.
04/19/23
Flooding SimulationFlooding Simulation
04/19/23
Related WorkRelated Work
Geographic Adaptive Fidelity:
• A virtual grid is proposed with only one node active at a time in each grid.
• Other nodes save energy by turning their radios off, or by entering sleep mode.
• Each node in GAF has three states: sleeping, discovery and active states respectively.
Geographic Adaptive Fidelity:
• A virtual grid is proposed with only one node active at a time in each grid.
• Other nodes save energy by turning their radios off, or by entering sleep mode.
• Each node in GAF has three states: sleeping, discovery and active states respectively.
04/19/23
Related WorkRelated Work
Span:
• Forms a backbone network of active nodes that participate in routing.
• A node in Span can only be in two states: coordinator and a non-coordinator.
• A node volunteers to be the coordinator if two of its neighbors fail to communicate with each other, either directly or through another coordinator.
Span:
• Forms a backbone network of active nodes that participate in routing.
• A node in Span can only be in two states: coordinator and a non-coordinator.
• A node volunteers to be the coordinator if two of its neighbors fail to communicate with each other, either directly or through another coordinator.
04/19/23
Related WorkRelated Work
Grid-based Coordinated Routing:
• Combines flooding, GAF and Span.
• Network is partitioned into square shaped grids.
• In each grid, one node participates in routing while other nodes are put to sleep to conserve energy.
Grid-based Coordinated Routing:
• Combines flooding, GAF and Span.
• Network is partitioned into square shaped grids.
• In each grid, one node participates in routing while other nodes are put to sleep to conserve energy.
04/19/23
04/19/23
0 200 400 600 800 10000
200
400
600
800
1000
0 200 400 600 800 10000
200
400
600
800
1000 Source
Source 1 2 3 4 5 6 7 8 9 10
11 12 13 14 15 16 17 18 19 20
21 22 23 24 25 26 27 28 29 30
31 32 33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48 49 50
51 52 53 54 55 56 57 58 59 60
61 62 63 64 65 66 67 68 69 70
71 72 73 74 75 76 77 78 79 80
81 82 83 84 85 86 87 88 89 90
91 92 93 94 95 96 97 98 99 100
04/19/23
04/19/23
MotivationMotivation
• To save energy by radio range adjustment, dividing the network into sections of different grid sizes based on a range-traffic relationship has been proposed.
• Our work is motivated from the concept of non-uniform grid sizes across the network using coordinated routing.
• To save energy by radio range adjustment, dividing the network into sections of different grid sizes based on a range-traffic relationship has been proposed.
• Our work is motivated from the concept of non-uniform grid sizes across the network using coordinated routing.
04/19/23
Non-uniform Grid-based Coordinated Routing Protocol
Non-uniform Grid-based Coordinated Routing Protocol
• The entire test area is divided into grids.
• Estimate the grid size to ensure proper connectivity between two coordinator nodes in adjacent grids.
• A coordinator node is elected in each grid to participate in routing.
• Energy depletion of nodes is taken into account for load balancing in the network.
• The entire test area is divided into grids.
• Estimate the grid size to ensure proper connectivity between two coordinator nodes in adjacent grids.
• A coordinator node is elected in each grid to participate in routing.
• Energy depletion of nodes is taken into account for load balancing in the network.
04/19/23
Estimating the Grid SizeEstimating the Grid Size
• To ensure connectivity and efficient usage of node energy, the grid size should neither be too large nor too small.
• To ensure connectivity and efficient usage of node energy, the grid size should neither be too large nor too small.
04/19/23
Estimating the Grid Size (Contd.)
Estimating the Grid Size (Contd.)
• The amount of energy that is required to establish a link between two nodes is proportional to the distance between the two nodes raised to a constant power, called the path loss exponent, n .
• If S is the receiver sensitivity, the communication link between the two nodes leads to a successful transmission between the nodes if the power of the received signal is greater than S.
• The amount of energy that is required to establish a link between two nodes is proportional to the distance between the two nodes raised to a constant power, called the path loss exponent, n .
• If S is the receiver sensitivity, the communication link between the two nodes leads to a successful transmission between the nodes if the power of the received signal is greater than S.
nt dP /Pr
04/19/23
Rnr
2r
.5/R r n
Estimating the Grid Size (Contd.)
Estimating the Grid Size (Contd.)
We define an upper bound on the grid size as 200 m and consider a lower bound of 100 m.
04/19/23
Non-uniform grid structuresNon-uniform grid structures
Designing the grid structures:
• Areas of high node density can be used efficiently for a grid size of 200 m.
• Areas of low node density require a grid size of 100 m.
• Random node placement implies sparsely and densely populated areas, therefore requiring a non-uniform distribution of grid size.
Designing the grid structures:
• Areas of high node density can be used efficiently for a grid size of 200 m.
• Areas of low node density require a grid size of 100 m.
• Random node placement implies sparsely and densely populated areas, therefore requiring a non-uniform distribution of grid size.04/19/23
Types of non-uniform gridsTypes of non-uniform grids
• Source non-uniform grid structure : suitable for low density around the source node and high density around sink node.
• Sink non-uniform grid structure : suitable for high density around the source node and low density around the sink node.
• Alternating non-uniform grid structure : suitable for random node placement across the network.
• Source non-uniform grid structure : suitable for low density around the source node and high density around sink node.
• Sink non-uniform grid structure : suitable for high density around the source node and low density around the sink node.
• Alternating non-uniform grid structure : suitable for random node placement across the network.04/19/23
Source non-uniform grid structure
Source non-uniform grid structure
0 100 200 300 400 600 800 10000
200
400
600
700
800
900
1000
0 100 200 300 400 600 800 10000
200
400
600
700
800
900
1000
04/19/23
Sink non-uniform grid structureSink non-uniform grid structure
0 200 400 600 700 800 900 10000
100
200
300
400
600
800
1000
0 200 400 600 700 800 900 10000
100
200
300
400
600
800
1000
04/19/23
Alternating non-uniform grid structure
Alternating non-uniform grid structure
0 200 300 500 600 800 900 10000
200
300
500
600
800
900
1000
0 200 300 500 600 800 900 10000
200
300
500
600
800
900
1000
04/19/23
Coordinator node electionCoordinator node election
• Each node has a randomly assigned ID.
• From each grid, the node with maximum node ID is the coordinator node .
• To distribute load across the coordinator nodes in a fair manner, load balancing is employed.
• Each node has a randomly assigned ID.
• From each grid, the node with maximum node ID is the coordinator node .
• To distribute load across the coordinator nodes in a fair manner, load balancing is employed.
04/19/23
Load BalancingLoad Balancing
• If coordinator node energy > 25% of battery life,
node rank = node rank +1
• If the energy < 25% of battery life,
node rank = node rank + 2
• For each grid, the current coordinators are replaced with lower ranked nodes.
• If coordinator node energy > 25% of battery life,
node rank = node rank +1
• If the energy < 25% of battery life,
node rank = node rank + 2
• For each grid, the current coordinators are replaced with lower ranked nodes.
04/19/23
04/19/23
04/19/23
Simulations and resultsSimulations and results
Assumptions:
• Energy consumption by nodes is assumed as Idle:transmit:receive = 1:2:1.5
• Test area is assumed to be replicating an actual sensor field of size 1000 m in the x-direction and 1000 m in the y-direction.
• Position of nodes deployed is assumed to be the same for all grid structures.
Assumptions:
• Energy consumption by nodes is assumed as Idle:transmit:receive = 1:2:1.5
• Test area is assumed to be replicating an actual sensor field of size 1000 m in the x-direction and 1000 m in the y-direction.
• Position of nodes deployed is assumed to be the same for all grid structures.
04/19/23
04/19/23
04/19/23
Network Parameters Network Parameters
04/19/23
ResultsResults
Metrics:
• Normalized energy
• Network lifetime
Graphs:
• Network lifetime graph.
• Energy depletion graph.
Metrics:
• Normalized energy
• Network lifetime
Graphs:
• Network lifetime graph.
• Energy depletion graph.
04/19/23
Network Lifetime GraphNetwork Lifetime Graph
04/19/23
0 1 2 3 4 5 0
100
200
300
400
500
600
700
800
900
1000Comparison of netw ork alive time for 100 nodes
Uniform Alternate Source non-uniform Sink non-uniform
Net
wor
k al
ive
time
in s
econ
ds
04/19/23
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 50
200
400
600
800
1000
1200
1400
1600
1800Comparison of netw ork alive time for 200 nodes
Uniform Alternate Source non-uniform Sink non-uniform
Net
work
aliv
e t
ime in s
econ
ds
04/19/23
0 0.5 1 1.5 2 2.5 30
500
1000
1500
2000
2500
3000
3500
4000
4500
5000Comparison of netw ork alive time for 1000 nodes
Uniform Alternate
Net
wor
k al
ive
time
in s
econ
ds
Energy Depletion GraphEnergy Depletion Graph
04/19/23
0 100 200 300 400 500 600 700 800 900 10000
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1Comparison of rate of energy depletion for 100 nodes
Time (seconds)
Ene
rgy
alternate grid structure
uniform grid structuresink non-uniform grid structure
source non-uniform grid structure
04/19/23
0 200 400 600 800 1000 1200 14000
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1Comparison of rate of energy depletion for 200 nodes
Time (units)
Ene
rgy
Alternate grid structure
Uniform grid structureSink non-uniform grid structure
Source non-uniform grid structure
0 500 1000 1500 2000 2500 3000 3500 4000 4500 50000
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1Comparison of rate of energy depletion for 800 nodes
Time (units)
Ene
rgy
Alternately non-uniform grid
Uniform grid
04/19/23
04/19/23
Node
density
Lifetime for Uniform
grid structure (seconds)
Lifetime for Alternating
non – uniform grid structure
(seconds)
Lifetime for Source non-uniform grid
(seconds)
Lifetime for Sink non-
uniform grid (seconds)
100 nodes 600 880 550 800
200 nodes 680
1250
780 1170
400 nodes 1800 1840 1660 1250
1000 nodes
3000 4500 4000 4300
ConclusionConclusion
• Different non-uniform grid structures provide different levels of energy savings and network lifetime.
• For random node deployment, using a non-uniform grid structure of alternating small and large grid size improves network lifetime over a uniform grid structure.
• Different non-uniform grid structures provide different levels of energy savings and network lifetime.
• For random node deployment, using a non-uniform grid structure of alternating small and large grid size improves network lifetime over a uniform grid structure.
04/19/23
0 200 300 500 600 800 900 10000
200
300
500
600
800
900
1000
04/19/23
Future WorkFuture Work
• Implementation on actual motes.
• Mobility of nodes.
• Irregular distribution of nodes.
• Implementation on actual motes.
• Mobility of nodes.
• Irregular distribution of nodes.
04/19/23
Thank youThank you
Questions ? Questions ?
04/19/23