Tracking Mobile Sensor Nodes in Wildlife Francine Lalooses
Tracking Mobile Sensor Nodes in Wildlife Francine Lalooses Hengky Susanto EE 194 -Professor Chang
Outline n Recap n Tracking with Binary Sensor Network n Distributed Predictive Tracking Algorithm n Our Tracking Approach n Future Work n References
Recap n Sensors only monitor land animals n Animals are tagged n Sensors placed at certain location n Better understanding of region/animal relationship n Not specific to animal size or velocity
Tracking with Binary Sensor Network n Assumptions: n Sensors have only one bit of information n Sensors broadcast bit to base station (BS) n Proximity sensor requires one more bit n One bit of information gives accurate predictions about direction of motion Approaching (+) n Moving away (-) n n Simple broadcast protocol Tracking a Moving Object with a Binary Sensor Network, Dartmouth College.
Binary Sensor Network Geometry n Future position lies: n Inside plus sensor overlap (+) n Outside minus sensor (-) + - +
Tracking with a Proximity Bit n Simulation results n Estimated trajectory (star – dashed line) n Actual trajectory (triangle – line) n Plus sensors (squares) n Minus sensors (circles) n Object gets in range time 3 at
Binary Sensor Summary n Advantages: n Trajectory prediction error is low n Broadcasting single bits over network feasible n BS computation is fast n Disadvantages: n Only tracks one animal at a time n No consideration for energy efficiency n No failure recovery model
Tracking Algorithm: Low Duty Cycle What if. . .
Distributed Predictive Tracking Algorithm n No central point n Cluster based architecture n Assumptions: n Randomly distributed sensors n Default to normal beam n Hibernation mode n Predictive mechanism n Cluster head activates appropriate sensors before target arrives A Protocol for Tracking Mobile Targets using Sensor Networks, RPI
DPT: Target Descriptor Formulation n Target descriptor (TD) consists of: n Target’s identity n n n Target’s present location n n Sensor triplet triangulation Target’s next predicted location n Unique Created when target first detected Alerts CHs most likely approached Linear predictor Time stamp n Time TD created
DPT: Sensor Selection Algorithm CHi+1 CH CHi+2 CH = cluster head TD = target description upstream downstream n Prediction: n When CHi predicts target location, downstream CHi+1 receives message n CHi+1 has information of all sensors in cluster n Selection: n CHi+1 locally decides sensor-triplet to sense target and sends wake-up message n Each sensor sends location message to CHi+1 n CHi+1 formulates TDi+1
DPT: Failure Recovery n Failure: n If upstream CH does not receive confirmation from downstream CH, assumes downstream CH is not available and target lost n Target changes direction or speed and moves away from predicted location n Recovery: n Wake up all sensors within area n Calculated from target’s previous actual location
DPT: Energy Considerations n Sensor-hibernation method n Most sensors stay in hibernation mode n Only chosen sensors become active n Energy for obtaining TD of one location Etotal = (1 - pmiss) Esuccess + pmiss Efailure n Keep pmiss small to minimize energy consumed for recovery n Energy consumed in failure recovery Efailure = Esuccess + 3 EHBPHB + (1 -PHB)(3 EHB+C) n Failures cause extra communication between clusters and sensors
Our Tracking Approach Nature n Cluster based algorithm n Hierarchical approach Master CH CH CH d First level CH CH d n Variables: n distance = E[velocityrunning] * time
Our Tracking Algorithm n Calculation based on maximum hop and popularity Lost region 1 h 2 h 2 d 4 h n Variables: n h = CH hop count
Our Intuition n Accuracy finding lost target improves over time More information = better search boundary n Error handling wakes up all nodes in region of diameter 2*d n n Advantages over sweeping across region: n More energy efficient n Less network traffic
Sweeping Across the Region n Perform increment layer outward from last seen position until n n n lost target is found or reaches border layer Only notifies their neighbor at outer layer When successful, the founder takes over target When target is not found, border sensors report to node in charge Awake all nodes in region and flood network Running time is O(n) Example of sweeping: Sensor node layers
Future Work n Failure and recovery algorithm n Further develop our algorithm n Compare DPT with our algorithm n n n Performance Energy efficiency Error handling
References n A Protocol for Tracking Mobile Targets using Sensor Networks. H. Yang n n n and B. Sikdar. RPI. Tracking a Moving Object with a Binary Sensor Network. J. Aslam, Z. Butler, F. Constantin, V. Crespi, G. Cybenko, D. Rus. Dartmouth College, CSU Los Angeles. Rumor Routing Algorithm for Sensor Networks. D. Braginsky, D. Estrin. UCLA. The ACQUIRE Mechanism for Efficient Querying in sensor Networks. N Sadagopan, A Helmy. USC. Distribute Online localization in Sensor Networks Using a Moving Target. A Galstyan, K Lerman, S Pattem. USC. Distributed Target Classification and Tracking in Sensor Networks. R. R. Brooks, P Ramanathan, A Sayeed. Penn State University and University of Wisconsin. Detecting Moving Radioactive Source Using Sensor Networks. D Stephens, A Peurrung. Pacific National Laboratory.
Questions
- Slides: 20