Collaborative and privacy-aware sensing for observing urban movement patterns Nelson Collaborative and privacy-aware sensing for Gon¸ calves, Rui Jos´ e, Carlos Baquero observing urban movement patterns Universidade do Minho and INESC TEC, PT Nelson Gon¸ calves, Rui Jos´ e, Carlos Baquero Universidade do Minho and INESC TEC, PT Esorics DPM, September 2013
Urban Movement Patterns Collaborative and privacy-aware sensing for observing urban movement patterns Nelson Gon¸ calves, Rui Jos´ e, Carlos Macro-level detection of aggregated urban movement can assist Baquero Universidade do infrastructure management. Minho and INESC TEC, PT In a tourism office: “Are the individuals in this art gallery likely to have visited a given art museum first?” In a shopping mall: “Which shops are visited most likely after the movie theater? and before the theater?”
Individual Movement Patterns Collaborative and privacy-aware sensing for observing urban movement patterns Nelson Gon¸ calves, Rui Jos´ e, Carlos Baquero Universidade do Minho and INESC TEC, PT Individuals often carry devices than can be detected Local detections can be shared and allow movement tracking 02:27:e4:f2:cd:0a W.Foyer 11:Sep:2013:19:12:33 MAC pseudonyms can be correlated to individuals
Research Questions Collaborative and privacy-aware sensing for observing urban movement patterns Nelson Gon¸ calves, Rui Jos´ e, Carlos Baquero Universidade do Minho and INESC TEC, PT 1 Can we design a mechanism that preserves privacy while allowing limited accuracy tracking of movement patterns? 2 Can higher accuracy collective movement result from lower accuracy individual tracking?
Precedence Filters Collaborative and privacy-aware sensing for observing urban movement patterns Our approach, Precedence Filters , builds heavily on: Nelson Bloom Filters (for probabilistic set membership) and on, Gon¸ calves, Rui Jos´ e, Carlos Baquero Vector Clocks (for distributed causality tracking). Universidade do Minho and The goal is to present a probabilist trace of past user locations, when INESC TEC, PT at a given location. @ Subway Bank → Market → Subway And collectively collect common routes.
Tools: Bloom Filters Collaborative and Bloom filter for set { x , y , z } with 3 hash functions. privacy-aware sensing for observing urban movement hash_fun 1 patterns hash_fun 2 Nelson hash_fun 3 Gon¸ calves, Rui { x,y,z } Jos´ e, Carlos Baquero Universidade do Minho and INESC TEC, PT 1 0 0 1 0 0 1 0 1 0 1 0 0 1 1 0 0 1 0 1 0 w Querying for element w yields a false positive. Larger filters depict larger precision for the same stored set size.
Tools: Vector Clocks Collaborative and privacy-aware sensing for observing urban movement Captures causality ( happens before ) relations without wall clocks patterns Nelson [1,0,0] [2,0,0] [3,0,0] [4,3,3] Gon¸ calves, Rui Jos´ e, Carlos P1 Baquero Universidade do Minho and INESC TEC, PT [2,2,0] [2,3,3] P2 [2,1,0] P3 [0,0,1] [2,2,2] [2,2,3] [2 , 2 , 0] → [2 , 2 , 3] → [4 , 3 , 4]
System Model and Design Collaborative and privacy-aware sensing for observing urban movement patterns Nelson Gon¸ calves, Rui Jos´ e, Carlos Network of local sensing devices (e.g. WiFi Hotspots) Baquero Universidade do MAC/Pseudonyms cannot leave the local sensing device Minho and INESC TEC, PT Tracking can exhibit false routes (plausible deniability) No network communication failures Network communication is faster than user movement A node holds a filter and caches cells from other filters
Precedence Filters: Algorithm Collaborative and privacy-aware sensing for observing urban movement patterns All filters have cells at 0 and they can take natural numbers Nelson Gon¸ calves, Rui A MAC address a is sensed in scanner node X Jos´ e, Carlos Baquero Using hashes X calculates to which cells item a is mapped Universidade do Minho and INESC TEC, PT Each other node sends to X the value on those cells Node X updates the caches of node’s filters on those cells In X filter, on those cells, it stores the maximum known value, plus one. This creates a fingerprint for a that is after all other sightings. From this information a node can construct its probabilistic view of the sequence of visits of a sensed device.
Mobility Traces Collaborative and privacy-aware sensing for observing urban movement patterns Trace with recurrent visits Nelson Gon¸ calves, Rui Jos´ e, Carlos Baquero Universidade do Minho and Subway → Market → Bookshop → Bank → Market → Subway INESC TEC, PT Precedence filters only capture the last of recurring visits Trace with more recent visits Bookshop → Bank → Market → Subway
Metrics and Data Sets Collaborative and privacy-aware sensing for A data set of Bluetooth sightings by static nodes was used from observing urban Leguay at all, from 2006, where 18 static nodes tracked 9244 distinct movement patterns users. This trace was replayed and complemented by a derived Nelson Gon¸ calves, Rui synthetic trace that expands the trace length and number of users. Jos´ e, Carlos Baquero Universidade do Minho and INESC TEC, PT Precedence Filters false positives create fictitious transitions. For evaluation we observe the relative proportion of these transitions. A value of 0 . 5 means that 50% of transitions are false.
Data Set: Location visits Collaborative and privacy-aware sensing for observing urban movement patterns Total and Distinct number of sightings for each fixed scanner Total and Distinct number of sightings for each fixed scanner 3000 3000 Nelson Total devices Total sightings Gon¸ calves, Rui Distinct devices Distinct sightings Jos´ e, Carlos 2500 2500 Baquero Universidade do 2000 2000 Minho and Number of sightings Number of sightings INESC TEC, PT 1500 1500 1000 1000 500 500 0 0 38 39 40 42 52 37 53 44 43 46 48 50 45 49 47 41 54 51 P1 P2 P3 P4 P5 P6 P7 P8 P9 P10 P11 P12 P13 P14 P15 P16 P17 P18 Scanner name Scanner name Distribution of detections in locations on real and synthetic traces
Inaccuracy vs False Positive Probability Collaborative and privacy-aware sensing for observing urban movement 1.0 1.0 patterns Individual Global Nelson 0.8 0.8 Gon¸ calves, Rui Jos´ e, Carlos Baquero Inaccuracy Inaccuracy 0.6 0.6 Universidade do Minho and INESC TEC, PT 0.4 0.4 0.2 0.2 0.0 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 CBFs' False Positive Probability CBFs' False Positive Probability Real and synthetic traces for the same trace length and users Global measures quality of aggregated transition prevalence
Extended synthetic trace Collaborative and Effects of increased trace size (100) and tracked users (100000) privacy-aware sensing for observing urban movement 1.0 patterns Nelson Gon¸ calves, Rui 0.8 Jos´ e, Carlos Baquero Universidade do Minho and Inaccuracy 0.6 INESC TEC, PT 0.4 0.2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 CBFs' False Positive Probability For longer runs higher quality aggregated data can be extracted from low quality (higher privacy) individual movement tracking.
Take home message Collaborative and privacy-aware sensing for observing urban New technique, Precedence Filters , joins Bloom Filters and VCs movement patterns Controlling filter size WRT number of devices, dictates accuracy Nelson Gon¸ calves, Rui False positives translate to fictitious visits to locations Jos´ e, Carlos Baquero Proportion of fictitious visits supports plausible deniability Universidade do Minho and INESC TEC, PT 50% user inaccuracy can support aggregated 10% inaccuracy
Photos Collaborative and privacy-aware sensing for observing urban movement patterns Nelson Gon¸ calves, Rui Jos´ e, Carlos Baquero Attribution under Creative Commons. Universidade do Minho and INESC TEC, PT http://www.flickr.com/photos/skyjuice/ http://www.flickr.com/photos/unknowndomain/ http://www.flickr.com/photos/library of congress/
Recommend
More recommend
