Wireless Sensor Networks Research issues and Addressing
approaches BY Eman Shaaban, PhD Associate Professor Computer
Systems Dept. Faculty of computer and information science Ain-Shams
university, Cairo, Egypt [email protected] Slide 2
Outline Part 1: Introduction - Mote research lab at FCIS - Common
design issues in WSNs Part 2: Related Researches - Relative
RSS-Based GSM Localization - Enhancing S-LEACH security for WSNs -
Mobility-Aware MAC Protocol for Delay-Sensitive WSN - Efficient
routing protocol for VANET Part 3: Related Works In Progress - A
scalable sensor localization scheme for WSN - Connectivity
Restoration in WSNs Through Node Repositioning Slide 3 A sensor
network is composed of a large number of sensor nodes that are
densely deployed inside or very close to the phenomenon:
Deterministic or Random deployment Self-organizing capabilities
Wireless Sensor Networks Slide 4 Sensor readings are transmitted
over a wireless channel to a running application that makes
decisions. Slide 5 Main Constraints Limited processing and power
capabilities. Scarce network resources Weak security Intermittent
connectivity and frequent node failures Heterogeneous systems
(Hardware, O/S, Applications) Densely deployment Frequent topology
changes Broadcast communication paradigm Possible absence of unique
global ID Slide 6 Classification of WSN Applications Event
Detection and Reporting: The significant design problem is that of
routing the event report to the sink, once the event is detected
Slide 7 Classification of WSN Applications Data Gathering and
Periodic Reporting: each sensor is expected to constantly produce
some amount of data which has to be sent to the sink. The sink
might not be directly interested in the individual measurements,
but could require a distributed computation of some function of the
sensor readings. Slide 8 Classification of WSN Applications
Sink-initiated Querying: This enables the sink to extract
information from different regions in space. For the underlying
communication protocols, we need effective means to address and
route data to and from dynamic sets of sensors. Slide 9
Classification of WSN Applications Tracking-based Applications:
when the target is detected, the sink needs to be notified
promptly. Then, the sink may initiate queries to receive
time-stamped location estimates of the target, so that it can
calculate the trajectory and keep querying the appropriate sets of
sensors. Slide 10 MoteWorks WSN Research Lab at FCIS End-to-End
enabling platform for the creation of WSN Processor- Radio-Data
Logger Mote Sensor cluster or interface card Crossbows Focus Slide
11 MoteWorks WSN Research Lab at FCIS Slide 12 Hardware Platform:
30 mote of type MICAz 2.4 GHz: Microprocessor: ATmega128L Radio:
CC2420 IEEE 802.15.4 compliant, ZigBee ready radio frequency
transceiver integrated with an Atmega128L External Serial Flash:
AT45DB041 512 KB 51-Pin Expansion connector Slide 13 MoteWorks WSN
Research Lab at FCIS 20 MDA100 ( Mote Data Acquisition board)
51-Pin Expansion connector 10-bit analog input Light sensor
Temperature sensor Prototyping area supports connection to eight
channels of the Motes analog to digital converter (ADC07). USART
and the I2C digital communications bus. The prototyping area has 45
unconnected holes used for breadboard of circuitry Slide 14
MoteWorks WSN Research Lab at FCIS 10 MIB520 (Mote Interface
Boards) for Programming, gateway, and base station. Protocol
characteristics and performance can be validated through
experimentations using a real sensor network environment. Slide 15
Research Classification Slide 16 Component level: improving the
sensing, communication, storage, and computation capabilities of an
individual sensor device. Communication level: mechanism of
networking and coordinating several sensor devices in an
energy-efficient and scalable fashion. Service level: are developed
to enhance the application, system performance and network
efficiency Slide 17 Research at Communication Level Optimize the
communication protocols, to best satisfy the application level
objectives as each WSN application imposes a unique set of goals
and produces a different type of data traffic. For each application
class, different problems associated with designing the
communication protocols. application specific design are
recommended. Cross-layer Collaboration between all the layers to
make the protocols lightweight and energy- efficient. Slide 18
Event-to-Sink reliability (reliable event detection at sink) rather
than end-to-end reliability (reliable delivery of individual
packets from a source to destination) Sink is only interested in
the collective information of sensor nodes within the event radius
and not their individual data Event radius sink Transport Layer
Slide 19 Common Design Issues in WSNs Service Level Compression and
aggregation Localization Synchronization Topology control Security
Slide 20 Compression and Aggregation Data-compression: Compressing
data before transmission to base station Data aggregation: Data is
collected from multiple sensors and Combined together to transmit
to base station Reduces the energy consumption of packet
transmissions, lowers the traffic load and therefore reduces the
contentions and collisions Slide 21 Compression and Aggregation
Data fusion can also be integrated with data- centric routing aims
to locate routes that lead to the largest degree of data
aggregation. Power efficient in-network distributed data processing
algorithms to implement efficient query processing and data
management. Most of data-gathering and fusion mechanisms reside in
or below the network layer Slide 22 Address-centric routing vs
Data-centric routing Slide 23 The network is divided into clusters,
and each cluster has associated a cluster head node. Each node in
the cluster sends its sensed data to the cluster head where it
belongs. The cluster head aggregates the data packets received into
a single packet, which is transmitted to the sink Clustering
improves resource utilization and prolong network lifetime and also
provide load balancing if appropriately configured The salient
advantage of using clusters in a sensor network comes from
in-network data aggregation. Clustering Slide 24 Clustering Slide
25 Localization Localization has become the hot research spot of
wireless sensor network. Localization is critical for data
stamping, clustering, topology control, location-based information
querying and geographical routing. Localization algorithms: secure,
minimum cost and localization errors. Efficient and accurate
localization for sensors in WSNs of arbitrary sizeis an essential
requirement for tomorrows wireless sensor networks to provide their
intended services Slide 26 Localization Range based schemes: use
absolute point to point distance estimates (range) or angle
estimates in location calculation. Range-based methods require
additional equipment. Range free schemes: use only connectivity
information to locate the entire sensor network. For example
hop-counting techniques in which each anchor computes an average
size for one hop. Slide 27 Localization Techniques The coordinates
of nodes in a network can be calculated using: Geometrical
techniques: triangulation, trilateration and multitrilateration.
Multidimensional scaling: convert distance information into the
coordinate vector. Algorithms for convex and nonconvex
optimization: formulating the localization problem as a nonlinear,
nonconvex optimization task solved by global optimization solvers.
Hybrid schemes that use two different techniques. Slide 28
Hierarchical cluster-based location systems Hierarchical
cluster-based solutions are often proposed to improve scalability
and efciency of the location system. For position estimation of
cluster heads usually more complex but accurate protocols are used.
The remaining nodes can use a simpler but less accurate method with
cluster heads as reference nodes. Slide 29 Localization: boundary
detection Localized edge detection Centralized edge determination
More Challenges to estimate a boundary using mobile sensing nodes.
Recent research effort to achieve promising accuracy of boundary
estimation in the future. Slide 30 Synchronization Time
synchronization: adjusting sensors local clocks to a common time
scale for: Data global time-stamp Cooperation Scheduling sleep-wake
patterns of the nodes in power-saving algorithms. Both localization
and synchronization have communication overheads, and these
tradeoff issues have been addressed in several works Slide 31
Topology Control Coverage topology control: maximize a reliable
sensing area while consuming less power. sensing coverage of the
entire region of interest is assured. Degree of coverage is
application dependent. Impacts on energy conservation Connectivity
topology control: concern more about network connectivity and
emphasizes the message retrieve and delivery in the network. Slide
32 Coverage topology control: Given a fixed number of static and
mobile nodes how should they be deployed in a monitoring region so
that area coverage is maximum? Coverage increased Slide 33
Connectivity Based Topology Control Nodes transmit at max power
levels Nodes transmit at min power levels High energy consumption
High interference Low throughput Network may partition Slide 34
Global connectivity Low energy consumption Low interference High
throughput Mechanisms to maintain an efficient sensor connectivity
topology: controlling the radio power level to achieve optimized
connectivity topology. Power Management that maintains a good
wake/sleep schedule. Mobile relays to link disjoint batches of
nodes. Connectivity Based Topology Control Slide 35 Topology
Control for Tolerating node failures: Forming k-connected WSN
Forming k-node disjoint communication paths between pairs of nodes
in the network to tolerate the failure of up to k-1 consecutive
node failures without suffering partitioning, and achieve that
using the least number of redundant nodes. Such optimization is a
very challenging problem that has been proven to be NP-hard for
most of the formulations of sensor deployment, even for k=1.
Several heuristics have been proposed to find suboptimal solutions.
Slide 36 2 Vertex disjoint paths bet. 1&11 Slide 37 Place the
minimum number of relay nodes such that each sensor is connected to
at least two relays and the inter-relay network is 2- connected.
Topology Control for Tolerating node failures: Forming k-connected
WSN Slide 38 The backup nodes can be simply passive spares among
redundant nodes, or among the active nodes. When failure of
critical nodes is detected, these active spares will have to quit
what they are doing and relocate to substitute failed nodes.
Assigned k distinct spares will enable the recovery to take place
even if k-1 of these spares. Topology Control for Tolerating node
failures: Designate backups for critical nodes Slide 39 A backup
node that would cause the least coverage degradation is favored. In
addition, low node degree would make a node an attractive backup.
Other solutions opt to localize the scope of the recovery by
picking backups within the 2-hop neighborhood of a failed critical
node Af. If not found, the search widens to include more distant
nodes. Upon detecting the failure of Af, the designated spare will
travel to replace Af or a series of cascaded relocation on the
shortest route between Af and the selected backup will be triggered
to split the travel load on multiple nodes Slide 40 Real-time
connectivity restoration implements a recovery procedure when a
node failure is detected. Such a reactive methodology better suits
dynamic WSNs. The idea is to utilize existing alive nodes which can
move and reposition them to the appropriate locations. Effectively,
the network topology is restructured to regain strong connectivity.
Topology Control for Tolerating node failures: Reactive
Connectivity Restoration Schemes ` Slide 41 Topology Control for
Tolerating node failures: Reactive Connectivity Restoration Schemes
Slide 42 Outline Part1: Introduction - Mote research lab at FCIS -
Common design issues in WSNs Part 2: Related Researches - Relative
RSS-Based GSM Localization - Enhancing S-LEACH security for WSNs -
Mobility-Aware MAC Protocol for Delay-Sensitive WSN -Efficient
routing protocol for VANET Part 3: Related Works In Progress - A
scalable sensor localization scheme for WSN - Connectivity
Restoration in WSNs Through Node Repositioning Slide 43 Slide 44
Relative RSS-Based GSM Localization This study has proposed and
implemented a robust technique for localization using Relative
Received Signal Strength (RRSS) of GSM. The study has been tested
and analyzed in Egypt roads using realistic data and Android smart
phones. The performance evaluation showed good results compared
with other similar environment fingerprint positioning techniques.
Slide 45 Database Correlation-based localization Store the
location-dependent network parameters seen by MS (fingerprints),
for the whole coverage area of the location system in a database
that is used by a location estimation algorithm. When MS needs to
be located, the necessary measurements are performed and
transmitted to the location server. The location server calculates
the MS location by comparing the transmitted information and the
fingerprints of the database. Slide 46 Slide 47 Fingerprinting
Taking RSS at the MS as the fingerprint variable. A single
fingerprint in the database consists of: Location coordinates RSS
from serving base station and other neighboring base stations in
that location. With GSM/GPS it is possible to obtain RSS from the
serving cell and maximum of five neighboring cells. Slide 48
Relative RSS-Based GSM Localization RSSs of the BSs are often
influenced by some static factors such as the distances,
obstructions and the transmission powers of the Aps. RSS at a
particular location would vary time to time due to multi-path
propagation and fading. Although RSSs received from a BS at a
location change over time, the relative RSSs which refer to the
relations of the RSSs received from the different base stations are
more stable than the absolute RSSs. Hence we define rules on RRSSs
for every location in a space and infer a users location by
matching the rules observed at the user and the rules obtained for
the space. Slide 49 Relative RSS based GSM localization Offline
phase: the data base of fingerprint is built where RRSS for each
antenna at a particular position/sample. Online or matching phase:
calculate the most probable sample which has maximum correlation
between rules. Slide 50 Offline phase Collecting fingerprint at
normal speed of roads is more practical than stopping at each
training point. Building an average RSS table to record the average
RSS from all cells for each location over a day. For each location,
rules are generated automatically for each pair of BSs based on the
average RSSs of the BSs at that location. The accuracy of the
location estimate is highly dependent on the density of the set of
collected fingerprints Slide 51 Offline phase Each training point
has n(n-1)/2 rules where n is number of valid detected cells. Rule
pattern for GSM: RSS1 (LAC1, CID1) {>,
