Download Target Tracking in Sensor Networks

Target Tracking in Sensor Networks
Choong Seon Hong
Kyung Hee University
[email protected]
Fall 2006
Tracking
 One of the most important applications of sensors is target
tracking.
 Each node can sense in multiple modalities such as acoustic,
seismic and infrared.
 The type of signals to be sensed are determined by the
objects to be tracked
 Given a sensor network, use the sensors to determine the
motion of one or more targets
 Canonical domain for DSNs - much of what we have seen so
far is applicable
 data routing, query propagation, wireless protocols
 Typically requires more cooperation among entities than
other examples we have seen
 Compare: “is there an elephant out there?” vs. “where has that
particular elephant been?”
Fall 2006
Objectives to be satisfied
 Collaborative Signal Processing (CSP)
 Distributive processing
 Goal oriented, on-demand processing
 Information fusion
 Multi-resolution processing
Fall 2006
Collaborative Signal Processing (CSP)
 To facilitate detection, identification and
tracking of targets, global information in both
time and space must be collected and analyzed
over a specified space-time region
 However individual nodes provide spatially
local information only
 CSP provides data representation and control
mechanisms to collaboratively process and
store sensor information, respond to external
events and report results
Fall 2006
Distributive processing
Raw signals are sampled and processed
at individual nodes but are not directly
communicated over the wireless channel
Instead each node extracts relevant
summary statistics from the raw signal,
which are typically smaller in size
The summary statistics are stored locally
in individual nodes and may be
transmitted to other nodes upon request.
Fall 2006
Goal oriented, on-demand processing
To conserve energy, each node should
perform signal processing tasks that are
relevant to the current query
In the absence of a query, each node
should retreat into a standby mode to
minimize energy consumption
A sensor node should not automatically
publish extracted information, but should
forward information only when needed
Fall 2006
Information fusion
To infer global information over a certain
space-time region, CSP must facilitate
efficient hierarchical information fusion
High bandwidth time series data must be
shared between neighboring nodes for
classification purposes
Lower bandwidth data may be
exchanged between more distant nodes
for tracking purposes.
Fall 2006
Multi-resolution processing
 Depending on the nature of the query, some CSP
tasks may require higher spatial resolution involving a
finer sampling of sensor nodes, or higher temporal
resolution involving higher sampling rates
 Example: Reliable detection is achievable with
relatively coarse space-time resolution, whereas
classification typically requires higher resolution
Fall 2006
Tracking Challenges
 Data dissemination and storage
 Resource allocation and control
 Operating under uncertainty
 Real-time constraints
 Data fusion (measurement interpretation)
 Multiple target disambiguation
 Track modeling, continuity and prediction
 Target identification and classification
Fall 2006
Tracking Domains
 Appropriate strategy depends on the sensors’
capabilities, domain goals and environment
Requires multiple measurements?
Bounded communication?
Target movement characteristics?
No single solution for all problems
 For example…
Limited bandwidth encourages local processing
Limited sensors requires cooperation
Fall 2006
Why Not Centralized?
 Scale!
 Data processing combinatorics
 Resource bottleneck (communication,
processing)
 Single point of failure
 Ignores benefits of locality
Fall 2006
Why Not (fully) Distributed?
 Redundant information and computation
 Can increase uncertainty
 Lack of unified view
 High communication costs
Fall 2006
Different Approaches of Tracking
 Tree-Based
 Cluster-Based
 Prediction-Based
Fall 2006
Scalable Tracking Using Networked Sensors (STUN)
 Tree based
- H. T. Kung and D. Vlah. “Efficient Location
Tracking Using Sensor Networks.” WCNC,
March 2003.
- Chih-Yu Lin and Yu-Chee Tseng “Structures
for In-Network Moving Object Tracking in
Wireless Sensor Networks” BROADNETS’04
Fall 2006
STUN (cont’d)
 The method will need to
handle a large number
of moving objects at
once
 This method uses a
hierarchy to connect the
sensors
 The leaves are sensors
 The querying point as the
root
 The other nodes are
communication nodes
Z
X
1
Y
2
3
4
 Example of message pruning hierarchy.
Consider the direction messages from
sensors that detect the arrival of a car.
Sensor 1’s message will update the
detected sets of all its ancestors. The
message from sensor 2 and 4 do not
update the detected sets of their parents
and thus will be pruned there. The
message from sensor 3 updates only its
parent Z and thus will be pruned at X
Fall 2006
STUN (cont’d) -- Example
Fall 2006
STUN (cont’d)
 Advantage
Message pruning
Routing
 Disadvantage
Building the tree (the structures of the tree)
Fall 2006
Dynamic Convoy Tree-Based Collaboration (DCTC)
 Wensheng Zhang and Guohong Cao, “DCTC:
Dynamic Convoy Tree-Based Collaboration for Target
Tracking in Sensor Networks” IEEE TRANSACTIONS
ON WIRELESS COMMUNICATIONS, VOL. 3, NO. 5,
SEPTEMBER 2004
 Wensheng Zhang and Guohong Cao, “Optimizing
Tree Reconfiguration for Mobile Target Tracking in
Sensor Networks” INFOCOM 2004
Fall 2006
DCTC (Cont’d) - Introduction
 DCTC relies on a tree structure called
“convoy tree”
 The tree is dynamically configured to add
some nodes and prune some nodes as the
target moves.
Fall 2006
DCTC (cont’d) – Basic Idea
When a target shows up for the first
time, an initial convoy tree is constructed
The root collects data from nodes
surrounding the target, and process data
When the target moves, the
membership of the tree is changed
The structure of the tree is reconfigured
if necessary
Fall 2006
Cluster-Based
 Wei-Peng Chen, Jennifer C. Hou, and Lui
Sha, Fellow, IEEE “Dynamic Clustering for
Acoustic Target Tracking in Wireless Sensor
Networks” IEEE TRANSACTIONS ON
MOBILE COMPUTING, VOL. 3, NO. 3,
JULYSEPTEMBER 2004
Fall 2006
Dynamic Clustering for Acoustic Target Tracking
 A CH volunteers to become active
 When it detects that the strength of a received acoustic
signal exceeds a predetermined threshold
 The signal matches one of the signal patterns which the
system intends to track.
 The tasks of an active CH
 Broadcasting a packet that contains the energy and the
extracted signature of the detected signal to sensors
 Receiving replies from sensors
 Estimating the location of the target based on replies
 Sending the result to subscriber(s).
 Energy-Based Localization
 The fundamental principle applied in the energy-based
approaches is that the signal strength (i.e., energy) of a
received signal decreases exponentially with the
propagation distance
Fall 2006
Dynamic Cluster – The Continuous Objects
 Continuously distributed
across a region
 Occupy a large area
 Trend to diffuse, increase
in size, change in sharp,
split into multiple relatively
smaller continuous objects
Fall 2006
Prediction-Based
 Please go through the following papers
Yingqi Xu Winter, J. Wang-Chien Lee “Predictionbased strategies for energy saving in object
tracking sensor networks” 2004 IEEE International
Conference on Mobile Data Management, 2004.
Xu, Y.; Winter, J.; Lee, W.-C. “Dual prediction
based reporting for object tracking sensor
networks” MOBIQUITOUS 2004
Fall 2006
Organization-Based Tracking
 Use structure, roles to control data and
action flow
 Can be static, or dynamically evolved
 Maintain an organizational hierarchy for
achieving energy efficient tracking solution
Fall 2006
Distributed Target Classification and Tracking in
Sensor Networks --Proceedings of the IEEE, vol. 91,
no. 8, pp. 1163-1171, 2003.
Fall 2006
Problem Domain
 Single target
 Fixed, acoustic sensors
 Requires multiple measurements
 Limited ad-hoc wireless network
 Track and classify target
(classification, which uses a supervised learning
technique, is not discussed here)
Fall 2006
Location-Centric Tracking
Control and data flow at each node:
 Initialization: disseminate sensor information
 Receive candidates: describe approaching
targets
 Local detections: gather measurements
 Merge detections: form track, compare
candidates
 Determine confidence: estimate uncertainty
 Estimate track: predict future target location
 Transmit track: notify relevant sensors
Fall 2006
Location-Centric Tracking
 “Closest point of approach” (CPA)
measurements
 Target detection causes cell formation
Cells formed around the target’s estimated
location
Intended to include relevant sensors
 Manager is selected
Node with greatest signal strength
 Manager collects local CPA’s
Linear regression over CPA
node locations
Fall 2006
Location-Centric-Tracking
 Estimated location compared to prior tracks
Projections from candidate tracks
 Cell created for tracking in new area
Size is a function of target velocity
Tracking information propagated to cell
 Tracking repeats…
Fall 2006
Location-Centric Advantages
 Avoids combinatorial explosion of track
association
Centralized: n targets, n candidate locations =
n2
Distributed: 1 target, n candidate locations = n
 Reduces communication costs (multi-hop ad hoc)
Saves energy
Fall 2006
Using and Maintaining Organization in a
Large-Scale Sensor Network
Bryan Horling, Roger Mailler,
Mark Sims and Victor Lesser
Multi-Agent Systems Lab
University of Massachusetts
Fall 2006
Problem Domain
 Fixed doppler radars
 Requires multiple,
coordinated
measurements
 Multiple targets
 Shared 8-channel RF
communication
Fall 2006
Sensor Characteristics
 Hardware
Fixed location,
orientation
Three 120° radar heads
Agent controller
 Doppler radar
Amplitude and
frequency data
One (asynchronous)
measurement at a time
Fall 2006
Organizational Control
 Use organization to address scaling issues
 Environment is partitioned
Constrains information propagation
Reduces information load
Exploits locality
 Agents take on one or more roles
Limits sources of information
Facilitates data retrieval
 Other techniques also built into negotiation
protocol and individual role behaviors
Fall 2006
Typical Node Layout
• Nodes are arranged or scattered, and have varied orientations.
• One agent is assigned to each node.
Fall 2006
Partitioning of Nodes
• The environment is first partitioned into sectors.
• Sector managers are then assigned.
Fall 2006
Competition for Sensor Agents
• Sector members send their capabilities to their managers.
• Each manager then generates and disseminates a scan schedule.
Fall 2006
Track Manager Selection
• Nodes in the scan schedule perform scanning actions.
• Detections reported to manager, and a track manager selected.
Fall 2006
Managing Conflicted Resources
• Track manager discovers and coordinates with tracking nodes.
• New tracking tasks may conflict with existing tasks at the node.
Fall 2006
Data Fusion (Track Generation)
• Tracking data sent to an agent which performs the fusion.
• Results sent back to track manager for path prediction.
Fall 2006
Sector Size
 This one parameter affects many things…
 Sector manager load
Smaller sector –› smaller manager directory
Larger sector –› better sector coverage
 Track manager actions
Smaller sector –› fewer update messages
Larger sector –› larger manager directory
 Communication distance, agent activity,
 Empirical evaluation of varying these
parameters
Fall 2006
Recommended Reading










Efficient in-network moving object tracking in wireless sensor networks Chih-Yu Lin; Wen-Chih Peng;
Yu-Chee Tseng;Mobile Computing, IEEE Transactions on Volume 5, Issue 8, Aug. 2006 Page(s):1044 –
1056
Self-organizing sensor networks for integrated target surveillance Biswas, P.K.; Phoha, S.; Computers,
IEEE Transactions on Volume 55, Issue 8, Aug. 2006 Page(s):1033 – 1047
CollECT: Collaborative Event deteCtion and Tracking in Wireless Heterogeneous Sensor Networks
Kuei-Ping Shih; Sheng-Shih Wang; Pao-Hwa Yang; Chau-Chieh Chang; Computers and Communications,
2006. ISCC '06. Proceedings. 11th IEEE Symposium on 26-29 June 2006 Page(s):935 – 940
Wireless sensor network based model for secure railway operations Aboelela, E.; Edberg, W.;
Papakonstantinou, C.; Vokkarane, V.; Performance, Computing, and Communications Conference, 2006.
IPCCC 2006. 25th IEEE International 10-12 April 2006
Adaptive tracking in distributed wireless sensor networks Lizhi Yang; Chuan Feng; Rozenblit, J.W.;
Haiyan Qiao; Engineering of Computer Based Systems, 2006. ECBS 2006. 13th Annual IEEE International
Symposium and Workshop on 27-30 March 2006
A Monte Carlo Method for Joint Node Location and Maneuvering Target Tracking in a Sensor
Network Miguez, J.; Artes-Rodriguez., A.; Acoustics, Speech and Signal Processing, 2006. ICASSP 2006
Proceedings. 2006 IEEE International Conference on Volume 4, 2006
Target Tracking in a Two-Tiered Hierarchical Sensor Network Vemula, M.; Bugallo, M.F.; Djuric, P.M.;
Acoustics, Speech and Signal Processing, 2006. ICASSP 2006 Proceedings. 2006 IEEE International
Conference on Volume 4, 2006
Localization and Tracking in Sensor Systems Manley, E.D.; Al Nahas, H.; Deogun, J.S.; Sensor
Networks, Ubiquitous, and Trustworthy Computing, 2006. IEEE International Conference on Volume
2, 2006 Page(s):237 – 242
Efficient Online State Tracking Using Sensor Networks
Halkidi, M.; Kalogeraki, V.; Gunopulos, D.; Papadopoulos, D.; Zeinalipour-Yazti, D.; Vlachos, M.; Mobile
Data Management, 2006. MDM 2006. 7th International Conference on 10-12 May 2006 Page(s):24 – 24
Achieving Real-Time Target Tracking UsingWireless Sensor Networks
Tian He; Vicaire, P.; Ting Yan; Liqian Luo; Lin Gu; Gang Zhou; Stoleru, R.; Qing Cao; Stankovic, J.A.;
Abdelzaher, T.; Real-Time and Embedded Technology and Applications Symposium, 2006. Proceedings of
the 12th IEEE 04-07 April 2006 Page(s):37 - 48
Fall 2006
Thanks !
Fall 2006