Optimizing Mixing in Pervasive Networks: A Graph-Theoretic Perspective

Slides:



Advertisements
Similar presentations
Impact of Interference on Multi-hop Wireless Network Performance
Advertisements

On the Optimal Placement of Mix Zones Julien Freudiger, Reza Shokri and Jean-Pierre Hubaux PETS, 2009.
Impact of Interference on Multi-hop Wireless Network Performance Kamal Jain, Jitu Padhye, Venkat Padmanabhan and Lili Qiu Microsoft Research Redmond.
1/22 Worst and Best-Case Coverage in Sensor Networks Seapahn Meguerdichian, Farinaz Koushanfar, Miodrag Potkonjak, and Mani Srivastava IEEE TRANSACTIONS.
Winter 2004 UCSC CMPE252B1 CMPE 257: Wireless and Mobile Networking SET 3f: Medium Access Control Protocols.
Minimizing Seed Set for Viral Marketing Cheng Long & Raymond Chi-Wing Wong Presented by: Cheng Long 20-August-2011.
Self-Organized Anonymous Authentication in Mobile Ad Hoc Networks Julien Freudiger, Maxim Raya and Jean-Pierre Hubaux SECURECOMM, 2009.
1 Advancing Supercomputer Performance Through Interconnection Topology Synthesis Yi Zhu, Michael Taylor, Scott B. Baden and Chung-Kuan Cheng Department.
Yu Stephanie Sun 1, Lei Xie 1, Qi Alfred Chen 2, Sanglu Lu 1, Daoxu Chen 1 1 State Key Laboratory for Novel Software Technology, Nanjing University, China.
Quantifying Location Privacy: The Case of Sporadic Location Exposure Reza Shokri George Theodorakopoulos George Danezis Jean-Pierre Hubaux Jean-Yves Le.
Haiming Jin, He Huang, Lu Su and Klara Nahrstedt University of Illinois at Urbana-Champaign State University of New York at Buffalo October 22, 2014 Cost-minimizing.
Mini-Project 2007 On Location Privacy in Vehicular Mix-Networks Julien Freudiger IC-29 Self-Organised Wireless and Sensor Networks Tutors: Maxim Raya Márk.
Distributed Algorithms for Secure Multipath Routing
1 A Distortion-based Metric for Location Privacy Workshop on Privacy in the Electronic Society (WPES), Chicago, IL, USA - November 9, 2009 Reza Shokri.
1 Delay-efficient Data Gathering in Sensor Networks Bin Tang, Xianjin Zhu and Deng Pan.
An Efficient and Spontaneous Privacy-Preserving Protocol for Secure Vehicular Communications Hu Xiong, Konstantin Beznosov, Zhiguang Qin, Matei Ripeanu.
Dynamic Hypercube Topology Stefan Schmid URAW 2005 Upper Rhine Algorithms Workshop University of Tübingen, Germany.
ICNP'061 Benefit-based Data Caching in Ad Hoc Networks Bin Tang, Himanshu Gupta and Samir Das Computer Science Department Stony Brook University.
1 Worst and Best-Case Coverage in Sensor Networks Seapahn Meguerdichian, Farinaz Koushanfar, Miodrag Potkonjak, Mani Srivastava IEEE TRANSACTIONS ON MOBILE.
Zoë Abrams, Ashish Goel, Serge Plotkin Stanford University Set K-Cover Algorithms for Energy Efficient Monitoring in Wireless Sensor Networks.
Online Data Gathering for Maximizing Network Lifetime in Sensor Networks IEEE transactions on Mobile Computing Weifa Liang, YuZhen Liu.
Maximum Network lifetime in Wireless Sensor Networks with Adjustable Sensing Ranges Mihaela Cardei, Jie Wu, Mingming Lu, and Mohammad O. Pervaiz Department.
Distributed Combinatorial Optimization
Department of Computer Engineering Koc University, Istanbul, Turkey
Approximation Algorithms Motivation and Definitions TSP Vertex Cover Scheduling.
Exposure In Wireless Ad-Hoc Sensor Networks S. Megerian, F. Koushanfar, G. Qu, G. Veltri, M. Potkonjak ACM SIG MOBILE 2001 (Mobicom) Journal version: S.
1 Algorithms for Bandwidth Efficient Multicast Routing in Multi-channel Multi-radio Wireless Mesh Networks Hoang Lan Nguyen and Uyen Trang Nguyen Presenter:
A Fair Scheduling for Wireless Mesh Networks Naouel Ben Salem and Jean-Pierre Hubaux Laboratory of Computer Communications and Applications (LCA) EPFL.
Toward Prevention of Traffic Analysis Fengfeng Tu 11/26/01.
PIC: Practical Internet Coordinates for Distance Estimation Manuel Costa joint work with Miguel Castro, Ant Rowstron, Peter Key Microsoft Research Cambridge.
Network Aware Resource Allocation in Distributed Clouds.
WMNL Sensors Deployment Enhancement by a Mobile Robot in Wireless Sensor Networks Ridha Soua, Leila Saidane, Pascale Minet 2010 IEEE Ninth International.
DARP: Distance-Aware Relay Placement in WiMAX Mesh Networks Weiyi Zhang *, Shi Bai *, Guoliang Xue §, Jian Tang †, Chonggang Wang ‡ * Department of Computer.
Network Survivability Against Region Failure Signal Processing, Communications and Computing (ICSPCC), 2011 IEEE International Conference on Ran Li, Xiaoliang.
Chi-Cheng Lin, Winona State University CS 313 Introduction to Computer Networking & Telecommunication Chapter 5 Network Layer.
Maximum Network Lifetime in Wireless Sensor Networks with Adjustable Sensing Ranges Cardei, M.; Jie Wu; Mingming Lu; Pervaiz, M.O.; Wireless And Mobile.
On the Age of Pseudonyms in Mobile Ad Hoc Networks Julien Freudiger, Mohammad Hossein Manshaei, Jean-Yves Le Boudec and Jean-Pierre Hubaux Infocom 2010.
Downlink Scheduling With Economic Considerations to Future Wireless Networks Bader Al-Manthari, Nidal Nasser, and Hossam Hassanein IEEE Transactions on.
Architectures and Algorithms for Future Wireless Local Area Networks  1 Chapter Architectures and Algorithms for Future Wireless Local Area.
Preserving Location Privacy in Wireless LANs Jiang, Wang and Hu MobiSys 2007 Presenter: Bibudh Lahiri.
Ahmed Osama Research Assistant. Presentation Outline Winc- Nile University- Privacy Preserving Over Network Coding 2  Introduction  Network coding 
GameSec 2010 November 22, Berlin Mathias Humbert, Mohammad Hossein Manshaei, Julien Freudiger and Jean-Pierre Hubaux EPFL - Laboratory for Computer communications.
On the Topology of Wireless Sensor Networks Sen Yang, Xinbing Wang, Luoyi Fu Department of Electronic Engineering, Shanghai Jiao Tong University, China.
On Non-Cooperative Location Privacy: A Game-theoreticAnalysis
1 - CS7701 – Fall 2004 Review of: Detecting Network Intrusions via Sampling: A Game Theoretic Approach Paper by: – Murali Kodialam (Bell Labs) – T.V. Lakshman.
Virtual Trip Lines for Distributed Privacy- Preserving Traffic Monitoring Baik Hoh et al. MobiSys08 Slides based on Dr. Hoh’s MobiSys presentation.
1 G-REMiT: An Algorithm for Building Energy Efficient Multicast Trees in Wireless Ad Hoc Networks Bin Wang and Sandeep K. S. Gupta Computer Science and.
Efficient Computing k-Coverage Paths in Multihop Wireless Sensor Networks XuFei Mao, ShaoJie Tang, and Xiang-Yang Li Dept. of Computer Science, Illinois.
Privacy Preserving Payments in Credit Networks By: Moreno-Sanchez et al from Saarland University Presented By: Cody Watson Some Slides Borrowed From NDSS’15.
1 Covert Communication based Privacy Preservation in Mobile Vehicular Networks Rasheed Hussain*, Donghyun Kim**, Alade O. Tokuta**, Hayk M. Melikyan**,
Reliable Multicast Routing for Software-Defined Networks.
Efficient Resource Allocation for Wireless Multicast De-Nian Yang, Member, IEEE Ming-Syan Chen, Fellow, IEEE IEEE Transactions on Mobile Computing, April.
Mix networks with restricted routes PET 2003 Mix Networks with Restricted Routes George Danezis University of Cambridge Computer Laboratory Privacy Enhancing.
1 Vehicular Networks Slides are integrated from researchers at EPFL.
Coverage Problems in Wireless Ad-hoc Sensor Networks Seapahn Meguerdichian 1 Farinaz Koushanfar 2 Miodrag Potkonjak 1 Mani Srivastava 2 University of California,
NOTE: To change the image on this slide, select the picture and delete it. Then click the Pictures icon in the placeholder to insert your own image. Fast.
March 9, Broadcasting with Bounded Number of Redundant Transmissions Majid Khabbazian.
TU/e Algorithms (2IL15) – Lecture 12 1 Linear Programming.
Efficient Point Coverage in Wireless Sensor Networks Jie Wang and Ning Zhong Department of Computer Science University of Massachusetts Journal of Combinatorial.
Track Me If You Can: On the Effectiveness of Context-based Identifier Changes in Deployed Mobile Networks. Authors: Laurent Bindschaedler, Murtuza Jadliwala,
SYNERGY: A Game-Theoretical Approach for Cooperative Key Generation in Wireless Networks Jingchao Sun, Xu Chen, Jinxue Zhang, Yanchao Zhang, and Junshan.
T H E O H I O S T A T E U N I V E R S I T Y Computer Science and Engineering 1 1 Sriram Chellappan, Xiaole Bai, Bin Ma ‡ and Dong Xuan Presented by Sriram.
Privacy Vulnerability of Published Anonymous Mobility Traces Chris Y. T. Ma, David K. Y. Yau, Nung Kwan Yip (Purdue University) Nageswara S. V. Rao (Oak.
Slide 1 Toward Optimal Sniffer-Channel Assignment for Reliable Monitoring in Multi-Channel Wireless Networks Donghoon Shin, Saurabh Bagchi and Chih-Chun.
Impact of Interference on Multi-hop Wireless Network Performance
Prof. Yu-Chee Tseng Department of Computer Science
A Study of Group-Tree Matching in Large Scale Group Communications
Presented By Siddartha Ailuri Graduate Student, EECS 04/07/17
Multi-Objective Optimization for Topology Control in Hybrid FSO/RF Networks Jaime Llorca December 8, 2004.
Dhruv Gupta EEC 273 class project Prof. Chen-Nee Chuah
Presentation transcript:

Optimizing Mixing in Pervasive Networks: A Graph-Theoretic Perspective Murtuza Jadliwala, Igor Bilogrevic and Jean-Pierre Hubaux ESORICS, 2011

Wireless Trends Always on Background apps Smart Phones Vehicles Watches Cameras Passports

Peer-to-Peer Wireless Networks 1 2 Peer-to-Peer wireless network Vehicular networks, delay tolerant networks, mobile social networks WiFi, Bluetooth Location privacy problem Third party can track location of nodes by monitoring identifiers Obtain location traces MAC address, authentication credentials Message Identifier

Examples VANETs Social networks Urban Sensing networks Nokia Instant Community Urban Sensing networks Delay tolerant networks Peer-to-peer file exchange

Location Privacy Problem Monitor identifiers used in peer-to-peer communications a Easy mass surveillance of location (not by network operator, but by anyone with WiFi sniffer) b c

Location Privacy Attacks Pseudonym Message Identifier Pseudonymous location traces Home/work location pairs are unique [1] Re-identification of traces through data analysis [2,4,3,5] Attack: Spatio-Temporal correlation of traces Linkability breaks anonymity. Need spatial and temporal decorrelation of traces => Filtering based on tracking model [1] P. Golle and K. Partridge. On the Anonymity of Home/Work Location Pairs. Pervasive Computing, 2009 [2] A. Beresford and F. Stajano. Location Privacy in Pervasive Computing. IEEE Pervasive Computing, 2003 [3] B. Hoh et al. Enhancing Security & Privacy in Traffic Monitoring Systems. Pervasive Computing, 2006 [4] B. Hoh and M. Gruteser. Protecting location privacy through path confusion. SECURECOMM, 2005 [5] J. Krumm. Inference Attacks on Location Tracks. Pervasive Computing, 2007

Location Privacy with Mix Zones Prevent long term tracking b ? 1 a 2 Traditional solution Spatial and temporal decorrelation of location traces Mix zones Mix zone Change identifier in mix zones [6,7] Key used to sign messages is changed MAC address is changed [6] A. Beresford and F. Stajano. Mix Zones: User Privacy in Location-aware Services. Pervasive Computing and Communications Workshop, 2004 [7] M. Gruteser and D. Grunwald. Enhancing location privacy in wireless LAN through disposable interface identifiers: a quantitative analysis. Mobile Networks and Applications, 2005

Mix-zone Placement in Road Networks Mix zone placement most effective at intersections [8] Enables mixing (covers) at roads leading in and out of the intersection Mix-zones incur cost Communication loss Routing delays Cost vary from intersection to intersection How to place mix-zones? All roads are covered Overall cost is minimized Mix Cover problem Communication loss due to silent period Identifier change causes routing delays [8] L. Buttyan, T. Holczer, and I. Vajda. On the effectiveness of changing pseudonyms to provide location privacy in VANETs. ESAS 2007

Previous Work on Mix zone Placement Optimization Approach [9] Mixing effectiveness using a flow-based metric Given upper bound on mix zones, max. distance between them and cost, where to place mix zones that maximizes mixing effectiveness Do not address the coverage problem Game-theoretic Approach [10,11] Game-theoretic model of optimal attack and defense strategies Only consider local, and not network-wide, intersection characteristics – – [9] J. Freudiger, R. Shokri, and J-P. Hubaux. On the optimal placement of mix zones. PETS 2009 [10] M. Humbert, M. H. Manshaei, J. Freudiger, and J-P. Hubaux. Tracking games in mobile networks. GameSec 2010 [11] T. Alpcan and S. Buchegger. Security games for vehicular networks. IEEE Transactions on Mobile Computing, 2011

Outline Mix Cover (MC) Problem Algorithms Evaluation and Results What are mix zones formally? And what was done in the past with mix zones.

Graph-Theoretic Model 𝐺≡ 𝑉,𝐸,𝑤,𝑑 Intersections  Vertices (V) Roads  Edges (E) Mixing cost at intersection  Vertex weight (w) Node intensity on road or demand  Edge weight (d) One for each direction, for 𝑒≡ 𝑢,𝑣 , 𝑑 𝑒 = ( 𝑑 𝑢 𝑒 , 𝑑 𝑣 𝑒 ) 3 7 9 2 6 4 8 6 3 4 2 2 10 8 8 7 2 6 2 3 7 6 3 2 12 9 2 2 5 2 9 1 8 1 2 2 1 9 4 Computations offline Assume knows mobility profiles 2 5 4 4 1

Mix Cover (MC) Problem 𝐺≡ 𝑉,𝐸,𝑤,𝑑 Determine a subset 𝑉 𝑀𝐶 ⊆𝑉 and a capacity 𝑐 𝑣 , ∀𝑣∈ 𝑉 𝑀𝐶 s.t. ∀𝑒≡ 𝑢,𝑣 , at least one of 𝑢 or 𝑣∈ 𝑉 𝑀𝐶 ∀𝑣∈ 𝑉 𝑀𝐶 , 𝑐 𝑣 ≥max ( 𝑑 𝑣 𝑒 ) for all 𝑒 covered by 𝑣 (capacity indicates the largest demand the intersection can handle) Total weighted cost 𝑥∈ 𝑉 𝑀𝐶 𝑐 𝑥 . 𝑤 𝑥 is minimized 3 10 6 7 9 2 6 4 8 6 3 4 2 2 10 8 8 7 7 2 6 2 3 7 6 3 2 12 9 2 2 5 2 9 9 1 8 1 2 2 1 9 4 Computations offline Assume knows mobility profiles 2 2 5 4 4 1 4 6x6 + 2x5 + 7x12+ 10x8 + 4x1 + 9x9 = 295

Why Mix Cover? A mix cover provides both these! Mix zone deployment that provides two guarantees: Privacy guarantee All roads are covered at least at one end Nodes go without mixing over at most one intersection Cost guarantee Minimum network-wide mixing cost A mix cover provides both these! Privacy: Nodes go without mixing at at most one intersection

Combinatorial Properties Generalization of Weighted Vertex Cover (WVC) problem Different from the Facility Terminal Cover (FTC) [13] generalization of WVC In FTC, each edge has only a single demand Result 1: Mix Cover problem is NP-hard No efficient algorithm for finding optimal solution, even finding a good approximation seems hard Proof by polynomial-time reduction from WVC Privacy: Nodes go without mixing at at most one intersection [13] G. Xu, Y. Yang, and J. Xu. Linear Time Algorithms for Approximating the Facility Terminal Cover Problem. Networks 2007

Outline Mix Cover (MC) Problem Algorithms Evaluation and Results What are the possible strategies and constraints to take into account?

Three Algorithms Optimization using Linear Programming “Divide and Conquer” approach Largest Demand First Smallest Demand First 1 2 3

Integer Program Formulation Cost guarantee Privacy guarantee Capacity requirement where 𝑤 𝑣 mixing cost at vertex 𝑣 𝑥 𝑣 decision variable indicating selected capacity of vertex 𝑣 Constraint Integer programming : Makes sense to guarantee minimal privacy. Might not have solution. We can relax this and consider instead Linear programming + heuristics Uses our metric to compute mix zone effectiveness a priori 𝑧 𝑣 𝑒 decision variable for vertex 𝑣 covering edge 𝑒 Result 2: LP relaxation of the above IP can guarantee a polynomial-time 2-approximation for the Mix Cover problem

Largest Demand First (LDF) For each edge, replace smaller demand with larger demand Round off the demands to the closest power of 2 Divide into subgraphs 𝐺 𝑘 based on the rounded edge demands 2 𝑘 Obtain 𝑆 𝐺 𝑘 =WVC−2Approx( 𝐺 𝑘 ) for each 𝐺 𝑘 For all 𝑣∈𝑆 𝐺 𝑘 , 𝑆 𝑀𝐶 =(𝑣,𝑐 𝑣 ) , where 𝑐 𝑣 = max{ 2 𝑘 |∀𝑘 s.t. 𝑣∈𝑆 𝐺 𝑘 } Output 𝑆 𝑀𝐶 𝐺≡ 𝑉,𝐸.𝑤.𝑑 𝐺′≡ 𝑉,𝐸.𝑤.𝑑′ Intuition for 1: The intuition behind such a transformation is that if a vertex is able to cover the larger demand, then it will definitely be able to cover any demand smaller or equal to the larger demand

LDF – Combinatorial Results A solution to MC problem on 𝐺′ is also a solution for 𝐺 Result 3: 𝑂𝑃𝑇 𝐺 ′ ≤2𝛼𝑂𝑃𝑇(𝐺), where 𝑂𝑃𝑇 is the optimal solution and 𝛼=max{ 𝑑 𝑢 𝑒 − 𝑑 𝑣 𝑒 ,∀𝑒∈𝐺} Result 4: LDF is a linear time 4𝛼𝛽-approximation algorithm for mix cover where 𝛽 is approximation ratio of WVC−2Approx Proofs in the paper! Computations offline Assume knows mobility profiles

Smallest Demand First (SDF) LDF highly sub-optimal  chosen capacity depends on larger edge demand value SDF similar to LDF, except In step 1, replace larger edge demand value by smaller value Additional step: For each vertex, remember the largest edge demand 𝑑 𝑣 𝑚𝑎𝑥 incident on it In 𝑆 𝑀𝐶 , choose capacity 𝑐 𝑣 =max{max 2 𝑘 ∀𝑘 s.t. 𝑣∈𝑆 𝐺 𝑘 , 𝑑 𝑣 𝑚𝑎𝑥 } Result 5: SDF is a 𝑂(𝑚𝑛) time 4𝛽-approximation algorithm for mix cover where 𝛽 is approximation ratio of WVC−2Approx Computations offline Assume knows mobility profiles

Outline Mix Cover (MC) Problem Algorithms Evaluation and Results Let’s see if it makes any difference

Experimental Setup Input graph constructed using real vehicular traffic data 2 US states, Florida and Virginia 3 sizes of road network, 25%, 65% and 100% of total state municipalities 3 different distributions of vertex weight, constant (1), uniform (between 1 and 100) and Gaussian (mean=50, sd=10) Edge demands chosen from real traffic intensities Algorithms implemented in MATLAB, executed on multi-core computer Results average over 100 runs

Solution Quality Naïve solution: Select all vertices in final solution Ratio of LDF/SDF solution cost to naïve strategy cost Naïve solution: Select all vertices in final solution SDF outperforms LDF in both cases for all graph sizes SDF achieves as low as 34% of the cost of the naïve solution Performance best for uniform vertex weight distribution and worst for constant distribution v/e v/e LDF Florida SDF LDF Virginia SDF

Execution Efficiency Duration (in seconds) of algorithm execution SDF runs slower compared to LDF in both cases for all graph sizes Algorithms fastest when vertex weight constant and worst when selected from a Gaussian distribution LDF Florida SDF LDF Virginia SDF

Results for LP-based Algorithm Too slow for large graphs Executed on reduced Florida graph of 515 and 1024 vertices For 515 vertices, ratio of solution cost compared to naïve strategy improves to 0.24 (better than LDF and SDF) Execution time is twice compared to LDF and four times that of SDF For 1024 vertices, execution time increased by a factor of 20

Conclusion Mix Cover: cost-efficient mix zone placement that guarantees mixing coverage Modeled as a generalization of weighted vertex cover problem Never been studied Model general enough and applicable to other scenarios Approximation algorithms using Linear programming LDF and SDF based on “Divide and Conquer” approach Results Proposed algorithms provide solution quality and execution time guarantees Experimentation using real data and standard computation resources show feasibility Mix cover has never been studied, both by the privacy and the combinatorics community Combinatorial hardness result not surprising; dynamic road conditions reinforces need for fast approximation algorithms murtuza.jadliwala@epfl.ch

Backup Slides

How to obtain mix zones? Silent mix zones Passive mix zones Turn off transceiver Passive mix zones Where adversary is absent Before connecting to Wireless Access Points Encrypt communications With help of infrastructure Distributed

bluetoothtracking.org People do that for fun Marketing value: Most popular Nokia phone, brand of GPS module in car Social value: How do people move in a city

Pleaserobme.com Or people do that for malicious motives

Mix networks vs Mix zones Alice home Mix Zones Mix node Mix node Alice work Bob Alice Mix node - Nodes move on road network - Road network = restricted network - Mix zones must be placed

Assumption Central authority periodically computes optimal mix cover offline Knows the (dynamic) node or traffic intensity on roads Knows mixing cost at each intersection Nodes or vehicles access the latest mix cover computation from the central authority Computations offline Assume knows mobility profiles

Solution Size SDF performs better than LDF in Florida Number of vertices in the final solution SDF performs better than LDF in Florida LDF performs better than SDF in Virginia Algorithms do not optimize solution size; depends on road network topology Solution size between 46% and 58% of the total number of vertices v/e v/e LDF Florida SDF LDF Virginia SDF