EL

Sensor Networks– Part III

PT Dr. Sudip Misra
Associate Professor
Department of Computer Science and Engineering
IIT KHARAGPUR
N
Email: [email protected]
Website: http://cse.iitkgp.ac.in/~smisra/

Introduction to Internet of Things 1
 Target Tracking

EL
Fig a: Push‐based formulation: Nodes
compute the position of the target and
periodically notify the sink node. A cluster
structure is commonly used in this case
PT Fig b: Poll‐based formulation: Nodes register the
presence of the target to permit a low‐cost query.
Data reports are sent toward the sink only when
Fig c: Guided formulation: Some nodes
(beacon nodes) define a trajectory to the
target. The tracker follows this trail to
N
there is a query to be answered. Tree structure is
intercept the target. Face structure is often
often used in this case
used in this case
Source: Éfren L. Souza, Eduardo F. Nakamura, and Richard W. Pazzi. 2016. Target Tracking for Sensor Networks: A Survey. ACM Computing Survey, 49, 2, 2016

Introduction to Internet of Things 2
 Target Tracking (contd.)

EL
PT
N
Source: Éfren L. Souza, Eduardo F. Nakamura, and Richard W. Pazzi. 2016. Target Tracking for Sensor Networks: A Survey. ACM Computing Survey, 49, 2, 2016

Introduction to Internet of Things 3
 WSNs in Agriculture
 AID: A Prototype for Agricultural Intrusion Detection Using Wireless 

EL
Sensor Network
 A set of sensor nodes are deployed over an agricultural field
 Each of the board are enabled with two type of sensors:
a) Passive Infrared (PIR)
b) Ultrasonic
PT
 When an intruder enters into the field through the boundary (perimeter)
of the field, the PIR sensor detects the object.
N
 The ultrasonic sensor senses the distance at which the object is located
Source: Sanku Kumar Roy, Arijit Roy, Sudip Misra, Narendra S Raghuwanshi, Mohammad S Obaidat, AID: A Prototype for Agricultural Intrusion Detection Using Wireless 
Sensor Network,  IEEE International Conference on Communications (ICC), 2015

Introduction to Internet of Things 4
 WSNs in Agriculture (contd.)

EL
PT
N
Source: Sanku Kumar Roy, Arijit Roy, Sudip Misra, Narendra S Raghuwanshi, Mohammad S Obaidat, AID: A Prototype for Agricultural Intrusion Detection Using Wireless 
Sensor Network,  IEEE International Conference on Communications (ICC), 2015

Introduction to Internet of Things 5
 Wireless Multimedia Sensor Networks (WMSNs)
 Incorporation of low cost camera (typically CMOS ) to wireless 
sensor nodes 

EL
 Camera sensor (CS) nodes
 capture multimedia (video, audio, and the scalar) data, expensive and 
resource hungry, directional sensing range
 Scalar sensor (SS) nodes
PT
 sense scalar data (temperature, light, vibration, and so on), omni‐
directional sensing range , and low cost
 WMSNs consist of less number of CS nodes and large number of SS 
nodes
N
Source: S. Misra, G. Mali, A. Mondal, "Distributed Topology Management for Wireless Multimedia Sensor Networks: Exploiting Connectivity and Cooperation", International 
Journal of Communication Systems (Wiley), 2014

Introduction to Internet of Things 6
 Wireless Multimedia Sensor Networks (WMSNs)

 WMSNs Application

EL
 In security surveillance, wild‐habitat monitoring, environmental
monitoring,  SS nodes cannot provide precise information 
 CS nodes replace SS nodes to get precise information 

PT
 Deployment of both CS and SS nodes can provide better sensing and 
prolong network lifetime
N
Source: S. Misra, G. Mali, A. Mondal, "Distributed Topology Management for Wireless Multimedia Sensor Networks: Exploiting Connectivity and Cooperation", International 
Journal of Communication Systems (Wiley), 2014

Introduction to Internet of Things 7
 Topology Management in WMSNs
• Video data are larger in size (e.g., 1024 bytes) which require larger bandwidth 
and consume high battery power

EL
• Coverage of the event should be provided as soon as the event occurs
• Connectivity is another important metric that should be provided during video 
data transfer from the event area to the control center 

PT
• Therefore, Misra et al. proposed the distributed topology management of the 
WMSNs considering coverage, connectivity, and network lifetime
• Coverage of the event is provided by using Coalition Formation Game between 
the CS and SS nodes 
N
Source: S. Misra, G. Mali, A. Mondal, "Distributed Topology Management for Wireless Multimedia Sensor Networks: Exploiting Connectivity and Cooperation", International 
Journal of Communication Systems (Wiley), 2014

Introduction to Internet of Things 8
Nanonetworks
 Nanodevice has components of sizes in the order nano‐meters.

EL
 Communication options among nanodevices
 Electromagnetic
 Molecular

PT
N
Introduction to Internet of Things 9
 EL
PT
N
Source: Akyildiz and Jornet, “Electromagnetic Wireless Nanosensor Networks”, Nano Communication Networks, 2010

Introduction to Internet of Things 10
 Molecular Communication

 Molecule used as information

EL
 Information packed into vesicles
 Gap junction works as mediator between cells and vesicles

PT
Information exchange between communication entities using
molecules
 Performed at NTT, Japan lab
N
Sources:
Jornet and Akyildiz, “Graphene‐based plasmonic nano‐antenna for terahertz band communication in nanonetworks”, IEEE JSAC, 2013
S. Hiyama, Y. Masitani, T. Suda, “Molecular transport system in molecular communication”, NTT Documo Technical Journal, Vol. 10, No. 3

Introduction to Internet of Things 11
 Electromagnetic-based Communication

 Surface Plasmonic Polarition

EL
(SPP) generated upon   
electromagnetic beam
 EM communication for 
Nanonetworks centers around 
0.1‐10 Terahertz channel 
PT
N
Sources:
Jornet and Akyildiz, “Graphene‐based plasmonic nano‐antenna for terahertz band communication in nanonetworks”, IEEE JSAC, 2013
S. Hiyama, Y. Masitani, T. Suda, “Molecular transport system in molecular communication”, NTT Documo Technical Journal, Vol. 10, No. 3

Introduction to Internet of Things 12
 Underwater Acoustic Sensor Networks
• In a layered shallow oceanic region, the inclusion of the effect of
internal solitons on the performance of the network is important.

EL
• Based on various observations, it is proved that non-linear internal
waves, i.e., Solitons are one of the major scatters of underwater
sound.

PT
• If sensor nodes are deployed in such type of environment, inter-node
communication is affected due to the interaction of wireless acoustic
signal with these solitons, as a result of which network performance
N
is greatly affected.
Source: A. Mandal, S. Misra, M. K. Dash, T. Ojha, "Performance Analysis of Distributed Underwater Wireless Acoustic Sensor Networks in the
Presence of Internal Solitons", International Journal of Communication Systems (Wiley)

Introduction to Internet of Things 13
 Oceanic forces and their impact
• The performance analysis of UWASNs renders meaningful insights
with the inclusion of a mobility model which represents realistic

EL
oceanic scenarios.
• The existing works on performance analysis of UWASNs lack the
consideration of major dominating forces, which offer impetus for a
node’s mobility.

PT
• The existing works are limited to only shallow depths and coastal
areas. Therefore, in this paper, Mandal et al. used a physical
mobility model, named oceanic forces mobility model (OFMM), by
incorporating important realistic oceanic forces imparted on nodes.
N
In this model, nodes move in 3D ocean column.
Source: A. K. Mandal, S. Misra, T. Ojha, M. K. Dash, M. S. Obaidat, "Oceanic Forces and their Impact on the Performance of Mobile
Underwater Acoustic Sensor Networks",International Journal of Communication Systems (Wiley)

Introduction to Internet of Things 14
 3-Dimensional Localization in USNA

• Silent & energy-efficient scheme for mobile UWSNs

EL
• Iterative approach
 Less initiators nodes (anchors) required

• Mobility prediction
 Enhanced accuracy
PT
• Only 3 surface anchor nodes required
N
Source: T. Ojha and S. Misra, "MobiL: A 3-Dimensional Localization Scheme for Mobile Underwater Sensor Networks", Proceedings of the
19th Annual National Conference on Communications (NCC 2013), IIT Delhi, New Delhi, India, Feb. 15-17, 2013.

Introduction to Internet of Things 15
 HASL: High-Speed AUV-Based Silent Localization for
Underwater Sensor Networks
• Get GPS Beacon

EL
• Start Dead- Reception
reckoning • Silent • Trilateration
• Broadcast beacon listening • Get z from
at const. interval • Receive pressure
Beacon
Sending PT ‘Effective’
set of beacon
message
sensor
Location
Estimation
N
Source: T. Ojha and S. Misra, “HASL: High-Speed AUV-Based Silent Localization for Underwater Sensor Networks”, Proceedings of the
9th International Conference on Heterogeneous Networking for Quality, Reliability, Security and Robustness(Qshine 2013), Springer, Greater
Noida, India, January 2013.

Introduction to Internet of Things 16
 Opportunistic localization
• Objective

EL
• Unlocalized nodes: to
localize with minimum
localization delay.

• Localized nodes: select a

localized with min. energy

consumption.
PT
transmission power level such
that max. no. of nodes can be
Perspective of
unlocalized node
Perspective of
localized node
N
Source: S. Misra, T. Ojha, A. Mondal, “Game-theoretic Topology Control for Opportunistic Localization in Sparse Underwater Sensor
Networks”, IEEE Transactions on Mobile Computing, vol. 14, no. 5, pp. 990-1003, 2014.

Introduction to Internet of Things 17
 A Self-Organizing Virtual Architecture

 Tic-tac-toe-arch: A self-organizing

EL
virtual architecture for underwater
sensor networks.
Calculating the duration of connectivity

PT
between the underwater nodes
A self-organizing network architecture by
utilizing the dynamic formation of virtual
N
topology
Source: T. Ojha, M. Khatua and S. Misra, "Tic-Tac-Toe-Arch: A Self-organizing Virtual Architecture for Underwater Sensor Networks", IET
Wireless Sensor Systems, Vol. 3, No. 4, December 2013, pp. 307-316.

Introduction to Internet of Things 18
 Virtual Topology Formation

EL
• Broadcast “REQ”
• Receive “RPLY”
Best Neighbour 
Neighbour 
Finding
PT Selection
• Calculate td
• Select Neighbour 
• Set the selected 
node in ‘Active’ 
mode for td time
with Max. td Set Duty 
N
Cycle

Introduction to Internet of Things 19
 EL
PT
N
Introduction to Internet of Things 20
 EL
Sensor Networks – Part IV
PT Dr. Sudip Misra
Associate Professor
Department of Computer Science and Engineering
IIT KHARAGPUR
N
Email: [email protected]
Website: http://cse.iitkgp.ac.in/~smisra/

Introduction to Internet of Things 1
WSN Coverage
 Coverage – area‐of‐interest is covered satisfactorily

EL
 Connectivity – all the nodes are connected in the network, so that 
sensed data can reach to sink node
 Sensor Coverage studies how to deploy or activate sensors to cover 
the monitoring area
 Sensor placement
 Density control
 Two modes
PT
 Static sensors
N
 Mobile sensors

Introduction to Internet of Things 2
 Definitions:
 Sensing range rs
 Transmission range  rt

EL
 Relationship between coverage and connectivity rs
 If transmission range  2 * sensing range, 
PT
 coverage implies connectivity
 Most sensors satisfy the condition!
 Coverage is the main issue
P
D
N
S rt

Introduction to Internet of Things 3
Coverage
 Determine how well the sensing field is monitored or tracked by 

EL
sensors
 To determine, with respect to application‐specific performance 
criteria, 

PT
 in case of static sensors, where to deploy and/or activate them 
 in case of (a subset of) the sensors are mobile, how to plan the trajectory 
of the mobile sensors. 
 These two cases are collectively termed as the coverage problem in 
N
wireless sensor networks.

Introduction to Internet of Things 4
Coverage (contd.)
 The purpose of deploying a WSN is to collect relevant data for 

EL
processing or reporting 
 Two types of reporting 
 event driven

 on demand   
PT
 e.g. forest fire monitoring

 e.g. inventory control system
 Objective is to use a minimum number of sensors and 
N
maximize the network lifetime 

Introduction to Internet of Things 5
Coverage (contd.)
 The coverage algorithm proposed are either centralized or distributed and 
localized

EL
 Distributed: Nodes compute their position by communicating with their neighbors only.
 Centralized: Data collected at central point and global map computed.
 Localized: Localized algorithms are a special type of distributed algorithms where only a 
subset of nodes in the WASN participate in sensing, communication, and computation. 

 Sensor deployment methods

PT
 Deterministic versus random

 Sensing and communication ranges
 Objective of the problem: maximize network lifetime or minimum number of 
N
sensors.

Introduction to Internet of Things 6
Coverage Problems in Static WSNs

 Most problems can be classified as 

EL
 Area coverage 
 Point coverage 
 Barrier coverage

PT
N
Introduction to Internet of Things 7
 Area Coverage
 Energy‐efficient random coverage

EL
 Connected random coverage 
 A network is connected if any active node can communicate with 
any other active node 

PT
 Zhang and Hou proved that if the communication range Rc is at 
least twice the sensing range Rs, then coverage implies 
connectivity
N
Source: Zhang and Hou, “Maintaining Sensing Coverage and Connectivity in Large Sensor Networks”, Ad Hoc & Sensor Wireless Networks, Vol. 1, pp. 89‐124, 2005.

Introduction to Internet of Things 8
Area Coverage (contd.)

 An important observation is that an area is completely 

EL
covered if there are at least two disks that intersect and all 
crossing are covered 
 Based on these they proposed a distributed, localized 
PT
algorithm called optimal geographical density control
N
Introduction to Internet of Things 9
Point Covergae

 Objective is to cover a set of points

EL
 Random point coverage – Distribute sensors randomly, so that every 
point must be covered by at least one sensor at all times
 Deterministic point coverage – Do the same in a deterministic 
manner.
PT
N
Introduction to Internet of Things 10
 Barrier Coverage
 1‐barrier coverage – covered by at least 1 sensor

EL
 2‐barrier coverage – covered by at least 2 sensors
 K‐barrier coverage – covered by at least k sensors

PT
N
Introduction to Internet of Things 11
Barrier Coverage (contd.)

EL
Weak Coverage
PT Strong Coverage
N
Introduction to Internet of Things 12
 Coverage Maintenance Crossings

 A continuous region R is covered if 

EL
 Exist crossings in R
 Every crossing in R is covered

 Crossings: intersection points between

disk boundaries or between monitored
space boundary and disk boundaries
 A crossing is covered if it is in the
interior region of at least one node’s
PT
N
Not covered
coverage disk
Source: Zhang and Hou, “Maintaining Sensing Coverage and Connectivity in Large Sensor Networks”, Ad Hoc & Sensor Wireless Networks, Vol. 1, pp. 89‐124, 2005.

Introduction to Internet of Things 13
 Optimality Conditions
 Optimality conditions for  P

EL
minimizing overlap while covering crossings B
 If nodes A and B are fixed,  A
node C should be placed such that
OR = OQ O Q

PT
 If nodes A, B, and C all can 
change their locations, then
OP = OR = OQ
 If all nodes have the same sensing range,
R
C
N
the distance between them is 3  rs

Introduction to Internet of Things 14
Optimal Geographical Density Control (OGDC)
Algorithm
 A node (A) volunteers as a starting node

EL
 Broadcasts a message containing
 Ideal direction ( randomly selected )
 Another node (B) closest to the ideal distance and angle becomes 
active

active PT
 A node (C) covering P and closest to the optimal location becomes 

 Repeatedly cover uncovered crossings with nodes that incur 
minimum overlap. 
N
 A node sleeps if its coverage area is completely covered

Introduction to Internet of Things 15
 EL
PT
N
Source: https://www.researchgate.net/figure/269392166_fig2_Fig‐2‐Optimal‐Geographical‐Density‐Control‐OGDC‐algorithm

Introduction to Internet of Things 16
Optimal Geographical Density Control (OGDC)
Algorithm (contd.)
 Select a starting node

EL
 Each node voluntarily participates with 
probability p
 Chooses a back‐off time randomly

PT
 If it does not hear anything from its 
neighbors, declares itself as starting 
node
 Declares its position and preferred 
A
N
direction

Introduction to Internet of Things 17
Optimal Geographical Density Control (OGDC)
Algorithm (contd.)
 On receiving message from a starting node

EL
 Each node computes the deviation from 
desired position (based on distance and 
angle)

message.
PT
 Chooses a back‐off time randomly
 When back‐off expires, it sends power ON 

 Then, it declares its position and preferred 
A
N
direction

Introduction to Internet of Things 18
Optimal Geographical Density Control (OGDC)
Algorithm (contd.)
 The process continues 

EL
until the entire area is  Q
covered B
A 
 The nodes already 
PT
covered go to sleep mode
P
O C
N
Introduction to Internet of Things 19
Optimal Geographical Density Control (OGDC)
Algorithm: Highlights
 A node initiates the process with desired distance and angle

EL
 Other nodes calculates the deviation, and the optimal one is 
chosen
 The process continues for all nodes
PT
 All covered nodes go to sleep mode
 This process is continued at each round
N
Introduction to Internet of Things 20
 EL
PT
N
Introduction to Internet of Things 21
 EL
Sensor Networks– Part V
PT Dr. Sudip Misra
Associate Professor
Department of Computer Science and Engineering
IIT KHARAGPUR
N
Email: [email protected]
Website: http://cse.iitkgp.ac.in/~smisra/

Introduction to Internet of Things 1
Stationary Wireless Sensor Networks

 Sensor nodes are static

EL
 Advantages:
 Easy deployment
 Node can be placed in an optimized distance—Reduce the total 
number of nodes

 Disadvantages:
PT
 Easy topology maintenance

 Node failure may results in partition of networks
N
 Topology cannot be change automatically 

Introduction to Internet of Things 2
Stationary Wireless Sensor Networks (Contd.)

EL
PT
N
Introduction to Internet of Things 3
Stationary Wireless Sensor Networks (Contd.)

EL
PT Failure Failure
N
Introduction to Internet of Things 4
Stationary Wireless Sensor Networks (Contd.)

EL
Failure Failure

PT Split of networks
N
Introduction to Internet of Things 5
 Stationary Wireless Sensor Networks (Contd.)

EL
Solution? PT
To mobilize the sensor nodes 
N
Mobile Wireless Sensor Networks (MWSN)

Introduction to Internet of Things 6
Mobile Wireless Sensor Networks
 MWSN is Mobile Ad hoc Network (MANET)

EL
 Let us remember from previous lectures:‐
 MANET‐Infrastructure less network of mobile devices connected wirelessly which 
follow the self‐CHOP properties
 Self‐Configure
 Self‐Heal




Self‐Optimize
Self‐Protect
PT
Wireless Sensor Networks‐
 Consists of a large number of sensor nodes, densely deployed over an area.
N
 Sensor nodes are capable of collaborating with one another and measuring the condition of their
surrounding environments (i.e. Light, temperature, sound, vibration).

Introduction to Internet of Things 7
Mobile Wireless Sensor Networks (contd.)

EL
MANET

PT WSN
N
MWSN

Introduction to Internet of Things 8
Components of MWSN

 Mobile Sensor Nodes

EL
Sink

PT
N
Introduction to Internet of Things 9
Components of MWSN

 Mobile Sensor 

EL
Nodes
 Sense physical 


parameters
from the 
environment
PT Sink
N
Introduction to Internet of Things 10
Components of MWSN

 Mobile Sensor Nodes

EL
 Sense physical parameters 
 from the environment
Sink


PT
 When these nodes come
 in close proximity of sink,
deliver data
N
Introduction to Internet of Things 11
Components of MWSN

 Mobile Sensor Nodes

EL
 Sense physical parameters 
from the environment
 When these nodes come in  Sink

PT
close proximity of sink, 
deliver data
N
Introduction to Internet of Things 12
Components of MWSN

 Mobile Sink

EL
 Moves in order to 
collect data from 
sensor nodes Sink

PT
 Based on some 
algorithm sink moves 
to different nodes in 
N
the networks

Introduction to Internet of Things 13
Components of MWSN

 Mobile Sink

EL
 Moves in order to 
collect data from 
sensor nodes
PT
 Based on some 
algorithm sink moves 
to different nodes in 
Sink
N
the networks

Introduction to Internet of Things 14
Components of MWSN

 Data Mules

EL
 A mobile entity
 Collects the data from 
sensor nodes Sink


PT
 Goes to the sink and 
delivers the collected
data from different
N
 sensor nodes

Introduction to Internet of Things 15
Underwater MWSNs
 Senses different parameters under the 

EL
sea or water levels
 Can be linked with Autonomous 

PT
Underwater Vehicles (AUVs)
 Applications: Monitoring‐marine life, 
water quality etc.
N
Introduction to Internet of Things 16
Terrestrial MWSNs
 Sensor nodes typically deployed over land 

EL
surface
 Can be linked with Unmanned Aerial Vehicles 
(UAVs)
PT
 Applications: Wildlife monitoring, surveillance, 
object tracking
N
Introduction to Internet of Things 17
Aerial MWSNs
 Nodes fly on the air and sense data (physical 

EL
phenomena or multimedia data)
 Typical example is Unmanned Aerial Vehicles 
(UAVs)
PT
 Applications: Surveillance, Multimedia data 
gathering
N
Introduction to Internet of Things 18
Possible Entity as Mobile Nodes in Daily-life
 Human

EL
 Mobility can not be predict
 Cell phone can gather information and deliver data to an access point
 Vehicles
 Sensor equipped on it

PT
 Sense data from different geographical locations and transmit to 
road side unit (RSU)
 Mobile Robot
 Controllable sensor node
N
 Collect data by predefined instructions
 Deliver the data to a specific unit

Introduction to Internet of Things 19
Human-centric Sensing
 Today, smartphones and PDAs are equipped with several sensors, e.g., 

EL
accelerometer and gyroscope
 Miniaturization & proliferation of such devices give rise to new sensing 
paradigms such as,
 Participatory sensing

PT
 People‐centric sensing
 Opportunistic sensing
 Basic idea:
 Humans carry their devices and move around
N
 Sensors embedded within the devices record readings
 Sensory readings are then transmitted

Introduction to Internet of Things 20
 Human-centric Sensing (contd.)
 Three distinct roles (not necessarily mutually exclusive) played by 
humans 

EL
 Sensing targets: Humans themselves are sensed, e.g., personal health monitoring
 Sensor operators: Humans use sensors and applications in smartphones & PDAs to sense 
surroundings
 Data source: Humans disseminate & collect data without actually using any sensor, e.g., 

 Challenges in human‐centric sensing:
 Energy of devices
 Participant selection
PT
updates posted in social networking sites
N
 Privacy of users
Source: M. Srivastava, T. Abdelzaher, and B. Szymanski, “Human‐centric sensing,” Philosophical Transactions of the Royal Society A: Mathematical, 
Physical and Engineering Sciences, vol. 370, no. 1958, pp. 176–197, Jan. 2012.

Introduction to Internet of Things 21
 Participatory Sensing
 Proposed by Burke et al., 2006

EL
 Distributed sensing by devices carried by humans
 Goal: Not just collect data, but allow common people to 
access data and share knowledge
 Collected data provides: PT
 Quantitative information, e.g., CO2 level
 Endorsement of authenticity, e.g., via geo‐tagged location & timestamp
N
Source: J. Burke, D. Estrin, M. Hansen, A. Parker, N. Ramanathan, S. Reddy, and M. B. Srivastava, “Participatory sensing,” in Workshop on World‐Sensor‐
Web (WSW’06): Mobile Device Centric Sensor Networks and Applications, Boulder, Colorado, USA, 2006, pp. 117–134.

Introduction to Internet of Things 22
Delay Tolerant Networks
 Lack of end‐to‐end communication paths

EL
 High latency
 Asymmetric data rates; erroneous channels
 WSN and MWSN:

PT
 Typically assume the availability of end‐to‐end path 
between any sensor node and BS
 We saw data MULEs earlier
 Such WSNs, in general, belong to the category of delay 
N
tolerant wireless sensor networks (DT‐WSNs)

Introduction to Internet of Things 23
 EL
PT
N
Introduction to Internet of Things 24
 EL
UAV Networks
PT Dr. Sudip Misra
Associate Professor
Department of Computer Science and Engineering
IIT KHARAGPUR
N
Email: [email protected]
Website: http://cse.iitkgp.ac.in/~smisra/

Introduction to Internet of Things 1
Features of UAV Networks
 Mesh or Star networks.

EL
 Flexible deployment and management of new services using SDN.
 Routing protocol should be adaptive in nature.
 Contribute towards greening of the network.


Multi‐tasking. PT
Large coverage area.

N
Easily reconfigurable for varying missions.

Introduction to Internet of Things 2
Key Issues
 Frequently change in network topology.

EL
 Relative position of UAV may change.

PT
 Malfunctioning of UAVs

 Intermittent link nature.
N
 Lack of suitable routing algorithm.

Introduction to Internet of Things 3
 Considerations in UAV Networks
Feature Single UAV System Multi‐UAV System
Failures High Low

EL
Scalability Limited High
Survivability Poor High
Speed of Mission Slow Fast
Cost
Bandwidth required
Antenna
Complexity of Control
PT Medium
High
Omni‐directional
Low
High
Medium
Directional
High
N
Failure to coordinate Low Present
Source: Lav Gupta, Raj Jain, and Gabor Vaszkun. "Survey of important Issues in UAV communication networks." IEEE Communications Surveys & Tutorials 18.2 (2015): 1123‐
1152.

Introduction to Internet of Things 4
UAV Network Constraints
Frequent link breakages

EL
Prone to malfunction

PT
Huge power requirements

Very complex
N
Physically prone to environmental effects: winds, rain, etc.

Introduction to Internet of Things 5
UAV Network Advantages
High Reliability

EL
High Survivability
Single Malfunction Proof

PT
Cost Effective
Efficient
N
Speeded up missions

Introduction to Internet of Things 6
UAV Network Topology: Star
 Typically  two types –

EL
 Star Configuration, 
 Multi‐star Configuration .
 In Star Configuration, UAV is directly 
connected to the ground station.
 In Multi‐star Configuration, UAVs 



High latency.
PT
form multiple star topology. One 
node from each group connects to 
the ground station.

Highly dependent on ground 
N
station.
Star Configuration Multi‐star 
Configuration

Introduction to Internet of Things 7
UAV Network Topology: Mesh
 Typically two types –

EL
 Flat Mesh Network, 
 Hierarchical Mesh 
Network.
 Flexible


Reliable
Nodes are 
PT
interconnected
N
 More secure   Flat Mesh  Hierarchical Mesh
Configuration Configuration

Introduction to Internet of Things 8
 UAV Topology Comparison
Star Network Mesh Network

EL
Point‐to‐point Multi‐point to multi‐point
Central control point present Infrastructure based may have a control center, Ad hoc has no central 
control center
Infrastructure based Infrastructure based or Ad hoc
Not self configuring
Single hop from node to central
point
Devices cannot move freely
PT Self configuring
Multi‐hop communication

In ad hoc devices are autonomous and free to move. In infrastructure 
based movement is restricted around the control center
N
Links between nodes and central Inter node links are intermittent
points are configured
Nodes communicated through central controller Nodes relay traffic for other nodes

Introduction to Internet of Things 9
FANETs: Flying Ad Hoc Networks
 Network formation using UAVs which 

EL
ensures longer range, clearer line of sight 
propagation and environment‐resilient 
communication.
 UAVs may be in same plane or organized at 
varying altitudes.

PT
 Besides self‐control, each UAV must be 
aware of the other flying nodes of the 
FANET to avoid collision.
 Popular for disaster‐time and post‐disaster 
N
emergency network establishment.

Introduction to Internet of Things 10
FANETs: Flying Ad Hoc Networks (contd.)

 Features:

EL
 FANET Inter‐plane communication
 FANET Intra‐plane communication
 FANET‐ Ground Station communication


PT
FANET‐ Ground Sensor communication
FANET‐VANET communication
N
Introduction to Internet of Things 11
Ad-Hoc FANETs

 A2A links for data delivery among UAVs.

EL
 Heterogeneous radio interfaces can be 
considered in A2A links, such as XBee‐PRO 
(IEEE 802.15.4) and Wi‐Fi (IEEE 802.11). 
 Ground networks may be stationary WSNs or 

PT 


VANETS or Control stations.
UAV‐WSN link‐up may be used for 
collaborative sensing as well as data‐muling.
UAV‐VANETS link‐up may be used for visual  
N
guidance, data‐muling and coverage 
enhancement.

Introduction to Internet of Things 12
 Gateway Selection in FANETs
 Main communication requirements of UAV 
networks are:

EL
 Sending back the sensor data.
 Receiving the control commands.
 Cooperative trajectory planning.
 Dynamic task assignments.

PT  Number of UAV‐ground remote connections 
should be controlled to avoid interference. 
 Reduced nodes in the UAV network should act as 
gateways, to allow communication between all 
N
UAV and the ground
Source: F. Luo et al., "A Distributed Gateway Selection Algorithm for UAV Networks," in IEEE Transactions on Emerging Topics in Computing, vol. 3, no. 1, pp. 22‐33, March 
2015.

Introduction to Internet of Things 13
 Gateway Selection in FANETs (contd.)
 Entire UAV network coverage area divided into 

EL
sub‐areas.
 Sub‐areas collectively cover the entire 
communication area.
 Size of sub‐area to be controlled and adjusted 

PT 


dynamically.
Adjustments based on UAV‐interconnections and 
derived metrics.
The derived metrics are optimized for several 
N
iterations till optimum state is achieved.
Source: F. Luo et al., "A Distributed Gateway Selection Algorithm for UAV Networks," in IEEE Transactions on Emerging Topics in Computing, vol. 3, no. 1, pp. 22‐33, March 
2015.

Introduction to Internet of Things 14
 Gateway Selection in FANETs (contd.)

 Gateway selection initiated by selection of the 

EL
most stable node in the sub‐area.
 Consecutively, the partition parameters are 
optimized according to topology.
 Each UAV acquires the information of all UAVs 
within its 2 hops. PT
N
Source: F. Luo et al., "A Distributed Gateway Selection Algorithm for UAV Networks," in IEEE Transactions on Emerging Topics in Computing, vol. 3, no. 1, pp. 22‐33, March 
2015.

Introduction to Internet of Things 15
 Layered Gateway in FANETs
 Multi‐layered UAV topologies select one 

EL
gateway.
 The gateways from each layer communicate to 
forward information between layers, as well 
as from ground control.

PT
 Will increase the delay between ground 
control and higher layers.
 Not suitable for time‐critical relaying tasks.
N
Introduction to Internet of Things 16
 FANETs & VANETs

EL
PT
N
Source: Y. Zhou, N. Cheng, N. Lu and X. S. Shen, "Multi‐UAV‐Aided Networks: Aerial‐Ground Cooperative Vehicular Networking Architecture," in IEEE Vehicular Technology 
Magazine, vol. 10, no. 4, pp. 36‐44, Dec. 2015.

Introduction to Internet of Things 17
 FANETs & VANETs (contd.)

EL
PT
N
Source: Y. Zhou, N. Cheng, N. Lu and X. S. Shen, "Multi‐UAV‐Aided Networks: Aerial‐Ground Cooperative Vehicular Networking Architecture," in IEEE Vehicular Technology 
Magazine, vol. 10, no. 4, pp. 36‐44, Dec. 2015.

Introduction to Internet of Things 18
 Trajectory Control for Increasing Throughput
 UAVs with queue occupancy above a 

EL
threshold experience congestion resulting 
in communication delay.
 Control station instructs UAVs to change 
centers of trajectory.

PT 
 Command given based on traffic at “busy” 
communication link.
To provide enhanced coverage, UAVs may 
be commanded to change radius of their 
N
trajectories.
Source: Fadlullah, Zubair Md, et al. "A dynamic trajectory control algorithm for improving the communication throughput and delay in UAV‐aided networks." 
IEEE Network 30.1 (2016): 100‐105.

Introduction to Internet of Things 19
 Trajectory Control for Increasing Throughput (contd.)

EL
PT
N
Source: Fadlullah, Zubair Md, et al. "A dynamic trajectory control algorithm for improving the communication throughput and delay in UAV‐aided networks." 
IEEE Network 30.1 (2016): 100‐105.

Introduction to Internet of Things 20
 EL
PT
N
Introduction to Internet of Things 21
 EL
Machine to Machine Communication
PT Dr. Sudip Misra
Associate Professor
Department of Computer Science and Engineering
IIT KHARAGPUR
N
Email: [email protected]
Website: http://cse.iitkgp.ac.in/~smisra/

Introduction to Internet of Things 1
Introduction
 Communication between machines or devices with computing 

EL
and communication facilities.
 Free of any human intervention.
 Similar to industrial supervisory control and data acquisition 
systems (SCADA).
PT
 SCADA is designed for isolated systems using proprietary 
solutions, whereas M2M is designed for cross‐platform 
N
integration. 

Introduction to Internet of Things 2
 EL
PT
N
Introduction to Internet of Things 3
M2M Overview
Sensors

EL
Network

Information 
Extraction

PT Processing

Actuation
N
Introduction to Internet of Things 4
M2M Applications
 Environmental monitoring
 Civil protection and public safety

EL
 Supply Chain Management (SCM)
 Energy & utility distribution industry (smart grid)
 Intelligent Transport Systems (ITSs)




Healthcare
PT
Automation of building
Military applications
Agriculture
N
 Home networks

Introduction to Internet of Things 5
 M2M Features
 Large number of nodes or devices.

EL
 Low cost.
 Energy efficient.
 Small traffic per machine/device.



PT
Large quantity of collective data.
M2M communication free from human intervention.
Human intervention required for operational stability and 
N
sustainability.
Source: Kim, Jaewoo, et al. "M2M Service Platforms: Survey, Issues, and Enabling Technologies." IEEE Communications Surveys and Tutorials
16.1 (2014): 61‐76.

Introduction to Internet of Things 6
 M2M Node Types

EL
M2M

M2M

M2M
Low  Mid  High 
PT
End End End
N
Source: Kim, Jaewoo, et al. "M2M Service Platforms: Survey, Issues, and Enabling Technologies." IEEE Communications Surveys and Tutorials
16.1 (2014): 61‐76.

Introduction to Internet of Things 7
 Low-end Sensor Nodes
 Cheap, and have low capabilities.
 Static, energy efficient and simple.

EL
 Deployment has high density in order to increase network 
lifetime and survivability.

PT
 Resource constrained, and no IP support.
 Basic functionalities such as, data aggregation, auto 
configuration, and power saving.
 Generally used for environment monitoring applications.
N
Source: Kim, Jaewoo, et al. "M2M Service Platforms: Survey, Issues, and Enabling Technologies." IEEE Communications Surveys and Tutorials
16.1 (2014): 61‐76.

Introduction to Internet of Things 8
 Mid-end Sensor Nodes
 More expensive than low‐end sensor nodes.

EL
 Nodes may have mobility.
 Fewer constraints with respect to complexity and energy efficiency.
 Additional functionalities such as localization, Quality of Service 

intelligence. PT
(QoS) support, TCP/IP support, power control or traffic control, and 

 Typical application includes home networks, SCM, asset 
N
management, and industrial automation.
Source: Kim, Jaewoo, et al. "M2M Service Platforms: Survey, Issues, and Enabling Technologies." IEEE Communications Surveys and Tutorials
16.1 (2014): 61‐76.

Introduction to Internet of Things 9
 High-end Sensor Nodes
 Low density deployment.

EL
 Able to handle multimedia data (video) with QoS
requirements.
 Mobility is essential.
 Example: smartphones.
PT
 Generally applied to ITS and military or bio/medical 
applications.
N
Source: Kim, Jaewoo, et al. "M2M Service Platforms: Survey, Issues, and Enabling Technologies." IEEE Communications Surveys and Tutorials
16.1 (2014): 61‐76.

Introduction to Internet of Things 10
 M2M Ecosystem
Device Providers

EL
Internet Service Providers (ISPs)

Platform Providers

PT
Service Providers

Service Users
N
Source: Kim, Jaewoo, et al. "M2M Service Platforms: Survey, Issues, and Enabling Technologies." IEEE Communications Surveys and Tutorials
16.1 (2014): 61‐76.

Introduction to Internet of Things 11
 EL
PT
N
Introduction to Internet of Things 12
 M2M Service Platform (M2SP)

EL
PT
N
Source: Kim, Jaewoo, et al. "M2M Service Platforms: Survey, Issues, and Enabling Technologies." IEEE Communications Surveys and Tutorials
16.1 (2014): 61‐76.

Introduction to Internet of Things 13
M2M Device Platform
 Enables access to objects or devices connected to the Internet 

EL
anywhere and at any time.
 Registered devices create a database of objects from which 
managers, users and services can easily access information.
 Manages device profiles, such as location, device type, address, and 
description.
PT
 Provides authentication and authorization key management 
functionalities.
 Monitors the status of devices and M2M area networks, and 
N
controls them based on their status.

Introduction to Internet of Things 14
M2M User Platform
 Manages M2M service user profiles and provides functionalities such as, 

EL
 User registration
 Modification
 Charging
 Inquiry.

PT
 Interoperates with the Device‐platform, and manages user access 
restrictions to devices, object networks, or services.
 Service providers and device managers have administrative privileges on 
their devices or networks.
 Administrators can manage the devices through device monitoring and 
N
control.

Introduction to Internet of Things 15
M2M Application Platform
 Provides integrated services based on device collected data‐

EL
sets.
 Heterogeneous data merging from various devices used for 
creating new services.

PT
 Collects control processing log data for the management of 
the devices by working with the Device‐platform.
 Connection management with the appropriate network is 
N
provided for seamless services.

Introduction to Internet of Things 16
M2M Access Platform
 Provides app or web access environment to users.

EL
 Apps and links redirect to service providers.
 Services actually provided through this platform to M2M devices.
 Provides App management for smart device apps.
 PT
App management manages app registration by developers and 
provides a mapping relationship between apps and devices.
 Mapping function provides an app list for appropriate devices.
N
Introduction to Internet of Things 17
Non-IP based M2M Network

EL
PT
N
Introduction to Internet of Things 18
IP-based M2M Network

EL
PT
N
Introduction to Internet of Things 19
 M2M Area Network Management Features
 Fault tolerant
 Scalable

EL
 Low cost, low complexity
 Energy efficient
 Dynamic configuration capabilities


Minimized management traffic
Application dependence: 
 Data‐centric application, 
 Emergency application,
PT
N
 Real‐time application
Source: Kim, Jaewoo, et al. "M2M Service Platforms: Survey, Issues, and Enabling Technologies." IEEE Communications Surveys and Tutorials
16.1 (2014): 61‐76.

Introduction to Internet of Things 20
 EL
PT
N
Introduction to Internet of Things 21