1. 1 Sniper Detection Using Wireless Sensor Networks Joe Brassard Wing Siu EE-194WIR: Wireless Sensor Networks Presentation #4: April 14th, 2005

2. 2 Presentation #4 PinPtr In Action DFRF Review Improvements Spanning Tree Convergecast Power Management Issues Implicit Time Synchronization Hardware Improvements Questions

3. 3 PinPtr ScreenShots

4. 4 PinPtr Video

5. 5 DFRF Redux Each node contains an “engine” that determines what to do with received packets, according to a policy Policies determined by application Broadcast Gradient Spanning Tree

6. 6 DFRF Message Each message contains information about its data type, its rank, and the actual data it contains Rank indicates packet’s position towards its goal

7. 7 The DFRF State Machine Message handling can be represented in terms of a state machine Each message has a priority value (0 – 255, stored internally) Priority value indicates status Odd, keep but don’t transmit Even, transmit 0, new 255, old, discard Varies according to application

8. 8 Gradient Convergecast DFRF This method was tested out in PinPtr Data is forwarded to a root node Data packets transmitted up to three times Fast, robust, but can generate lots of overhead

9. 9 Fat Spanning Tree Convergecast (FAST) Instead of using the entire path to send message, flood a small “neighborhood” with data Neighborhood is defined using a spanning tree

10. 10 A FAST Node Knows the node ID of its parent, grandparent, and great- grandparent Assigns packets rank of grandparent Based on this information, received packets with sender ID [rank] of: Parent – this node is closer. Give higher priority Grandparent – same. Ignore. Great-grandparent – node is further away. Give lower priority. Other – not involved. Ignore. State diagram is similar to that of gradient convergecast

11. 11 Convergecast: Gradient vs. FAST

12. 12 Power Management Issues Keeping PinPtr on continuously drastically reduces lifespan of network How to decrease costs? Incorporate power management More efficient, cheaper hardware

13. 13 Power Management, I Time Synchronization is a very important step, but is continuous time synch. necessary? Implicit time synchronization can be realized during message routing Adding an “age” field to data packets adds little overhead but allows integrated time synch, or RITS

14. 14 Power Management, II Routing Integrated Time Synchronization

15. 15 Power Management, III RITS Advantages: Eliminates continuous, extraneous radio messages for time synch Enables power management (i.e., “light sleep” and “deep sleep” modes among motes) Can now use analog comparators to “wake” the mote when an acoustic event is recorded Very low overhead imposed on original message routing protocol

16. 16 Power Management, IV PinPtr system was designed and tested using the Mica2 mote Searched for other motes with lower operating power and/or cheaper cost Found Tmote Sky mote

17. 17 Power Management, V Mica2 Versus Tmote Sky

18. 18 Power Management, VI Mica2: COM (Serial) Interface TX max = 35mA Outdoor range ~500ft $200 per mote No standard encryption 16 pin expansion support Tmote Sky USB Interface TX max = 19.5mA Outdoor range ~420ft $130 per mote Hardware link-layer encryption 16 Pin expansion support Faster wake up time than Mica2 (<6us)

19. 19 References www.moteiv.com  Tmote Sky www.moteiv.com www.xbow.com  Mica2 www.xbow.com Ledeczi, “Pattern Oriented Composition and Synthesis of Middleware Services for NEST”, Vanderbilt Maroti, “The Directed Flood Routing Framework”, Vanderbilt Simon, et. Al., “Sensor Network Based Countersniper System”, Vanderbilt