College of Engineering Non-uniform Grid- based Coordinated Routing Priyanka Kadiyala Major Advisor:...

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