Module-1

Introduction to Sensor Networks,

Localization, and Time Synchronization
 Wireless Sensor Networks (WSN)
The “many-tiny” principle: wireless networks of
thousands of inexpensive miniature devices capable
of computation, communication, and sensing

Their use throughout society “could well dwarf

previous milestones in the information revolution”: U.S.
National Research Council Report, 2001.

Berkeley Mote (MICAz MPR 2400 Series)

 Timeline
1970’s: Wired sensors connected to a central location
1980’s: Distributed wired sensor networks
1993: LWIM project at UCLA
1999-2003: DARPA SensIT project: UC Berkeley, USC,
Cornell etc.
2001: Intel Research Lab at Berkeley focused on WSN
2002: NSF Center for Embedded Networked Sensing
2001-2002: Emergence of sensor networks industry;
startup companies including Sensoria, Crossbow,
Ember Corp, Sensi Cast plus established ones: Intel,
Bosch, Motorola, General Electric, Samsung.
2003-2004: IEEE 802.15.4 standard, Zigbee Alliance.
 Wireless Sensor Networks (WSN)
Provide a bridge between the real physical and
virtual worlds

Allow the ability to observe the previously

unobservable at a fine resolution over large spatio-
temporal scales

Have a wide range of potential applications to industry,

science, transportation, civil infrastructure, and security.
 Some Sample Applications
Habitat and Ecosystem Monitoring

Seismic Monitoring

Civil Structural Health Monitoring

Monitoring Groundwater Contamination

Rapid Emergency Response

Industrial Process Monitoring

Perimeter Security and Surveillance

Automated Building Climate Control

5
Basic Components of a WSN Node
 Challenges

Energy Efficiency

Responsiveness

Robustness

Self-Configuration and Adaptation

 Challenges (contd.)

Scalability

Heterogeneity

Systematic Design

Privacy and Security

Deployment

9
 Deployment Issues

Structured versus Randomized Deployment

Does the network involve
(a) structured placement, either by hand or via autonomous
robotic nodes, or
(b) randomly scattered deployment?

Overdeployed versus Incremental Deployment

Should the network be deployed a priori with redundant
nodes, or can nodes be added or replaced incrementally
when the need arises? In the former case, sleep scheduling
is desirable to extend network lifetime
 Deployment Issues
Network topology
Is the network topology going to be a simple star topology, a grid,
an arbitrary multi-hop mesh, or a two-level cluster hierarchy?

Homogeneous versus heterogeneous deployment

Are all sensor nodes of the same type or is there a mix of high-
and low-capability devices? In case of heterogeneous
deployments, there may be multiple gateway/sink devices

Coverage metrics
What is the kind of sensor information desired from the
environment and how is the coverage measured? This could be
on the basis of detection and false alarm probabilities or whether
every event can be sensed by K distinct nodes
 Network Topologies

(a) a star-connected single-hop topology, (b) flat multi-

hop mesh, (c) structured grid, and (d) two-tier hierarchical
cluster topology
Localization
 Localization Issues
Location information necessary/useful for many functions,
including measurement stamps, coherent signal
processing, cluster formation, efficient querying and
routing.

Key Questions:
▪ What to localize?
▪ When to localize?
▪ How well to localize?
▪ How to localize?
 Localization

Each individual sensor observation can be

characterized essentially as a tuple ofbthe form < S,T,
M >, where S is the spatial location of the
measurement, T the time of the measurement, and
M the measurement itself.

The location information of nodes in the network is

fundamental for a number of reasons:
1. To provide location stamps for individual sensor
measurements that are being gathered.
2. To locate and track point objects in the
environment.
3. To monitor the spatial evolution of a diffuse
phenomenon
4. To determine the quality of coverage.
 Localization Contd…
5. To achieve load balancing
6. To form clusters.
7. To facilitate routing
8. To perform efficient spatial querying.
 Localization Contd…

1. What to localize?
This refers to identifying which nodes have a priori
known locations (called reference nodes) and which
nodes do not (called unknown nodes). There are a
number of possibilities. The number and fraction of
reference nodes in a network of n nodes may vary all
the way from 0 to n − 1.
2. When to localize?
In most cases, the location information is needed
for all unknown nodes at the very beginning of
network operation. In static environments, network
localization may thus be a one-shot process. In other
cases, it may be necessary to provide localization on-
the-fly
 Localization Contd…
3. How well to localize?
This pertains to the resolution of location
information desired. Depending on the application, it
may be required for the localization technique to
provide absolute x y z coordinates, or perhaps it will
suffice to provide relative coordinates (e.g. “south of
node 24 and east of node 22”); or symbolic locations (e.
g. “in room A”, “in sector 23”, “near node 21”).
4. Where to localize?
The actual location computation can be performed
at several different points in the network: at a central
location once all component information such as inter-
node range estimates is collected; in a distributed
iterative manner within reference nodes in the
network; or in a distributed manner within unknown
nodes.
 Localization Contd…

5. How to localize?
Finally, different signal measurements can be used
as inputs to different localization techniques. The
signals used can vary from narrowband radio signal
strength readings or packet-loss statistics, UWB RF
signals, acoustic/ultrasound signals, infrared. The
signals may be emitted and measured by the reference
nodes, by the unknown nodes, or both.
 Localization approaches
Coarse-grained localization using minimal
information:
These typically use a small set of discrete
measurements, such as the information used to compute
location. Minimal information could include binary
proximity (can two nodes hear each other or not?),
near–far information (which of two nodes is closer to a
given third node?), or cardinal direction information (is
one node in the north, east, west, or south sector of the
other given node?).

Fine-grained localization using detailed

information:
These are typically based on measurements, such as RF
power, signal waveform, time stamps, etc., that are either
real-valued or discrete with a large number of
 Coarse Grained Node Localization
Several techniques provide approximate solutions for
node localization based on the use of minimal
information:

▪ Binary Proximity
▪ Centroids Calculations
▪ Geometric Constraints
▪ Approximate point in triangle APIT
Geometric Constraints
Reference
node
Unknown
node
Constrained
location
region
Dis Secto
c r

Quadra Annul
nt us
(Doherty, Pister and
Ghaoui ‘01)
Approximate Point in Triangle (APIT)

(He, Huang, et al.

‘03)
 Fine-Grained Node Localization
Basic Approach: Ranging
– ranging using radio signal strengths (m-level
accuracy)
– ranging using time difference of arrival (cm-
level accuracy over short distances)

Position estimation is then an MMSE problem:

E j = Ri,j - √ ((xi-xj)2+(yi-yj)2) Find (xi,yi) to minimize Σ (Ej)2

Angle of arrival techniques are particularly useful in

conjunction with ranging
Radio signal-based distance estimation
(RSS)

RSS-based ranging may perform much better

in situations where the fading effects can be
combatted by diversity techniques
 Time Difference of Arrival

To
Transmitter

RF Acousti
c

Receiv
er Tr Ts

Distance ≅ (T r – Ts) .
Vs

One minor limitation of acoustic ranging is that it

generally requires the nodes to be in fairly close
proximity to each other (within a few meters) and
preferably in line of sight.
 Angle of arrival (AOA)
Angles can potentially be estimated by using rotating
directional
beacons, or by using nodes equipped with a phased array of
RF or ultrasonic receivers.
A very simple localization technique, involving three
rotating reference beacons at the boundary of a
sensor network providing localization for all interior
nodes
 Fine-Grained Node Localization (contd.
)Pattern matching techniques such as RADAR (Bahl and
Padmanabhan, ‘00) require pre-training of signal
strengths at different locations in the environment.

Ecolocation (Yedavalli et al. ‘04) is based on sequence

decoding.
– Record the received signal strengths at different reference
nodes from a given unknown node, and order these into a
sequence
– Return as the unknown node’s location the location that “best
matches” the measured sequence

© Bhaskar Krishnamachari 2005 28

 Network Localization
• Different from node localization. Few reference nodes
and several networked unknown nodes.

• Several approaches:
– Constraint satisfaction/optimization (centralized)
– Joint estimation using ranging estimates (centralized)
– Multihop distance estimation (distributed)
– Iterative localization (distributed)
– Potential fields (distributed)

© Bhaskar Krishnamachari 2005 29

 RSS-based joint
estimation
If radio signal strengths can be measured between all
pairs of nodes in the network that are within the
detection range, then a joint maximum likelihood
estimation (MLE) technique can be used to determine
the location of unknown nodes in a network
First an expression is derived for the likelihood that the
obtained matrix of power measurements would be
received given a particular location set for all nodes;
the objective is then to find the location set that
maximizes this likelihood.

© Bhaskar Krishnamachari 2005 30

 Iterative Localization
The iterative multilateration technique is applicable
whenever inter-node distance information is available
between all neighboring nodes

One begins by determining the location of an unknown

node that has the most reference nodes in its
neighborhood. In a distributed version, the location of
any node with sufficient references in its neighborhood
may be calculated as the initial step

Centroid calculation could be used at each iterative step

by the unknown nodes, instead of using distance-based
triangulation if only connectivity information is available.

© Bhaskar Krishnamachari 2005 31

 Iterative Localization

(0,10
) C D (10,10
F
)
(5,5)
E I
G L J

K
A
H B
(0,0 (10,0
)
)

© Bhaskar Krishnamachari 2005 32

 Iterative Localization

(0,10 1. (5,
) C 25/3) D (10,10
F )
(5,5)
3 G 2 E 6
J I 7
L
K
A 4
H B
(0,0 5 (10,0
)
)

© Bhaskar Krishnamachari 2005 33

Multi-hop distance-estimation approaches
Estimates of distances to reference nodes that may be
several hops away. These distances are propagated
from reference nodes to unknown nodes using a basic
distance-vector technique.
There are three variants of this approach:
1. DV-hop: Each unknown node determines its distance
from
various reference nodes by multiplying the least number
of hops by the reference nodes with an estimated
average distance per hop
2. DV distance: If inter-node distance estimates are
directly available for each link in the graph
3. Euclidean propagation: Geometric relations can be
used in addition to distance estimates to determine more 34
accurate estimates to reference nodes
 Reference-less Localization
• What if there are no reference nodes with known
locations?
• Three-step solution:
1. If all boundary nodes have known locations, use
iterative centroid calculations
2. If boundary nodes do not have known locations,
use pairwise hop-hop counts to get approximate
locations and apply step 1.
3. If nodes are not aware of the boundary, use a flood
to identify boundary nodes and apply step 2.
• The solution provides only a relative map, useful for
geographic routing
© Bhaskar Krishnamachari 2005 35
 Illustration of Reference-less
Localization

Localization assuming
Correct
only known
locations
boundary nodes
© Bhaskar Krishnamachari 2005 36
 Rigidity and Unique Localization

The network is said to be uniquely localizable if there

exists a unique assignment of (x,y) coordinates to all
nodes that is consistent with the available positions
and distance estimates in a network.

Let GN be the grounded graph, i.e. the network graph

augmented with distance-labeled edges between all
pairs of reference nodes.

Theorem: A network is uniquely localizable if its

grounded graph is globally rigid.

© Bhaskar Krishnamachari 2005 37

 Global Rigidity
A
Not Rigid
E B
B E
A
D
C D C

(a) (b)

A
Rigid

E B E B
A

D C D C

(c)
© Bhaskar Krishnamachari 2005 (d) 38
 Time Synchronization

© Bhaskar Krishnamachari 2005 39

 Reasons for Time Synchronization

For time-stamping measurements

For in-network signal processing
For localization
For cooperative communication
For medium-access
For sleep scheduling
For coordinated actuation

© Bhaskar Krishnamachari 2005 40

 Key issues
The clock at each node consists of timer
circuitry, often based on quartz crystal
Oscillators.
The simplest approach to time
synchronization in a distributed system is
through periodic broadcasts of a consistent
global clock
The requirements for time synchronization
can vary greatly from application to
application.
In some applications, it will be necessary to
keep
© Bhaskar all nodes
Krishnamachari 2005 synchronized globally to an
41

external reference
 Traditional approaches
A fundamental technique for two-node clock
synchronization is known as Cristian’s algorithm.

© Bhaskar Krishnamachari 2005 42

 Traditional approaches
Send time – which includes any processing time and
time taken to assemble and move the message to the
link layer.
Access time – which includes random delays while
the message is buffered at the link layer due to
contention and collisions.
Propagation time – which is the time taken for point-
to-point message travel. While negligible for a single
link, this may be a dominant term over multiple hops
if there is network congestion.
Receive time – which is the time taken to process the
message and record its arrival.
© Bhaskar Krishnamachari 2005 43
 Fine-grained clock
synchronization
1. Reference broadcast
synchronization

The reference broadcast synchronization (RBS)

© Bhaskar Krishnamachari 2005 technique 44
 Fine-grained clock
synchronization
2. Pair-wise sender-receiver synchronization
(TPSN): Time-sys protocol for Wireless
Network

© Bhaskar Krishnamachari 2005 45

 Fine-grained clock
synchronization
3. Flooding time synchronization protocol (FTSP):
▪ Interrupt handling time: This is the delay in
waiting for the processor to complete its
current instruction before transferring the
message in parts to the radio.
▪ Modulation/encoding time: This is the time
taken by the radio to perform modulation and
encoding at the transmitter, and the
corresponding demodulation and decoding at
the receiver.

© Bhaskar Krishnamachari 2005 46

 Fine-grained clock
3. Flooding timesynchronization
synchronization protocol (FTSP):
FTSP has two main components:
Multiple time measurements:
▪ The sender takes several time stamp measurements during
transmission, one at each byte boundary
▪ These measurements are normalized by subtracting an appropriate
multiple of the byte transmission time, and only the minimum of these
multiple measurements is embedded into the message
Flooded messaging:
▪ First, a single uniquely identifiable node in the network provides the
global clock
▪ The reception of each broadcast message allows a receiver to
accumulate a reference synchronization point
▪ When a receiver accumulates several reference points, it becomes
synchronized with itself
© Bhaskar Krishnamachari 2005 47
 Fine-grained clock
synchronization
4. Predictive time synchronization:

In the real world, clock drift can vary over time quite drastically
due to environmental temperature and humidity changes.
This reassessment is to re-synchronize nodes periodically at
the same interval.
The frequency of inter-node time sampling is adaptively
adjusted.
Depending on the time sampling period S, a window of T/S
prior sample measurements is used.
This technique not only predicts the clock drift (through linear
regression), but also estimates the error in the prediction

© Bhaskar Krishnamachari 2005 48

 Coarse-grained data
The
synchronization
Wisden system presents an excellent lightweight
alternative to clock synchronization approaches that are
suitable for data-gathering applications.
The approach is to collect and record latency
measurements within each packet in a special
residence time field as the packet propagates through
the network.
In this approach, only the base station is required to
have an accurate reference clock.
Say the time spent by packet i at the kth intermediate
node on an n+1 hop-path to the destination is

© Bhaskar Krishnamachari 2005 49

 Coarse-grained data
synchronization
let the time at the base station d when the packet is
received be then the packet’s original contents are
time-stamped to have been generated at source s at

© Bhaskar Krishnamachari 2005 50