TOWARDS QUASI-REALTIME THEFT-EVIDENT MECHANISM FOR PORTABLE ARTIFACTS USING NEAR-FIELD RFID

Kirti Chawla@ , Sunil K. Vuppala$, Puneet Gupta$

UNIVERSITY OF VIRGINIA, CHARLOTTESVILLE@

SET LABS, INFOSYS TECHNOLOGIES$

@ Weblink: http://www.cs.virginia.edu/~kc5dm

$ Weblink: http://www.infosys.com/technology/setlabs-briefings.asp

2 SIGNIFICANT DEVELOPMENTS

2

Quick Biased Inference

Source: Gartner Research (http://www.gartner.com)10 %

PROBLEM STATEMENT

15 %

3

To mitigate/minimize/deter the effect of theft with respect to portable devices using low-cost technology.

Challenges/Requirements

•Space Constraints

•Power Constraints

•Mobility Constraints

•Realtime Evidence of Theft

•Cost/Effort

Candidates

•Wired Technology× Cannot be applied in general due tomobility constraints

•Wireless Technology• Wi-Fi RFID (meets the

requirements)• ZigBee• …

Opportunity/Target Devices

•Portable Audio Player

•Portable Video Player

•Smartphone

•Personal Digital Assistant

•and many more …

Active RFID Tag Passive RFID TagRFID Reader

BRIEF INTRODUCTION TO RFID

4

Definition: RFID or Radio Frequency Identification is a technology for wireless recognition with the help of radio waves. A Tag or Transponder consists of a microchip

that stores data and a coupling element, such as coiled antenna, which is used to communicate via radio frequency. 3 tag-type : Active, Passive & Hybrid A Reader or Transceiver consists of radio frequency module, a control unit and a coupling element to interrogate electronic tags via RF communication 2 operating-fields : Near-Field & Far-Field

A Data Processing System takes data read by RFID reader and stores it in a database. Various operations on the data can be performed (manipulate, view etc.) Applications: Object identification & tracking etc.

Suggestion: Viewers are advised to lookup terms for detailed definitions.

20 % • Has a battery• Long operating range

• No battery• Short operating range • Operates in Near-Field or Far-Field

• Uses load modulation or backscatter for communication with Tag

PROPOSED SOLUTIONDESIGN: BASICS

25 %5

Requirements on underlying RFID carrier

•Operation Field: Near-Field (Frequency ≤ 13.56 MHz), i.e. LF or HF •Operating Distance: Up to 1 meter(s)

Suggestion: Design decision intuition clears up as the presentation goes forward

Signal strength vis-à-vis distance relationship S'=SxF(d)

where. the symbols have following meaning:d = DistanceS = Signal strength at Point of origin of RF SignalS' = Signal strength at point of reference o

6

2

f RF Signal

1; Near-field RF Signal

d F(d) = 1

; Far-field RF Signald

Design Intuition 1: Shorter operating distance low antenna power

How can thisunderlying property of Near-Field RFID signal be used for solving Problem Statement ??

PROPOSED SOLUTIONDESIGN: SYSTEM MODEL

30 %6

Artifact: Fancy name for device/objectICGTA: Increased Chance of Guessing the Thief Around

Design Intuition 2: ICGTA Shorter operating distance

PROPOSED SOLUTIONDESIGN: PACKET FORMAT

35 %

7

Descriptions of Packet Field

•ACN: Artifact Class Number•AN: Artifact Number•ER: Error Resilience•WCN: Wearable-band Class Number•WN: Wearable-band Number•PAWN: Portable Artifact Wearable-band Number•R-PAWN: Reader-PAWN•T-PAWN: Tag-PAWN

Design Intuition 3: Protocol is needed between RFID reader and Tag

Suggestion: PAWN is not same as pawn in CHESS game

How frequently T-PAWN should be fetched from Tag given that we have power constraints ??

RFIDReader

TagT-PAWN

R-PAWN T-PAWN

PROPOSED SOLUTIONDESIGN: ERROR CORRECTION MECHANISM

40 %

8Design Intuition 4: Resilience against intentional/un-intentional corruption required

To minimize the effect of intentional/un-intentional corruption on data-in-transit.

Claim

RFID reader provides error correction mechanisms, like CRC

Suggestion

Error correction mechanism presented here, (uses Hamming Code) adds

multi-bit per byte error-correction with relatively low computation cost on

reader side to augment CRC

PROPOSED SOLUTIONDESIGN: THEFT-EVIDENT ALGORITHM

45 %

9Design Intuition 5: Realtime or Quasi-Realtime evidence of theft is pertinent

T-PAWN Fetch Strategy

•Partition time into 2 slots, viz. Sleep and Sweep Interval•Reader sleeps in Sleep Interval•Reader looks for Tag in Sweep Interval•Sleep Interval ≥ Sweep Interval to meet power constraints•Frequent Entry-Exit of Wearable-band in Reader field, to be monitored using thresholds of time instead of distance.

PA: Portable Artifact with RFID ReaderWB: Wearable-band with TagROI: Region of Interest {bounded by S’}

PA1

PA2

PAN

Your Un-intelligent portable artifact

PROPOSED SOLUTIONDESIGN: OPERATIONS

50 %

10Design Intuition 6: Class numbers facilitates grouping of artifacts, viz. <PA I, WBI>

WB

In above example, pairs are: <PA1, WB>, <PA2, WB> … <PAN, WB>

PROPOSED SOLUTIONIMPLEMENTATION: ERT

55 %

11

ERT - Experimental RFID Tool:

•Written in C++

•~2K source lines

•Lightweight/WIN32 programming model

•Simulates Quasi-Realtime theft scenario

•Supports 1-on-1 operation mode

•Supports selective error-correction mechanism

•User-configurable parameters for Sleep, sweep interval and more

•Support for audible alerts on THEFT_CONDITION

•Support for custom action on THEFT_CONDITION

•Supports multiple error correction modes

•Extensible to Client/Server architecture using WINSOCK

Implementation Intuition 1: Construct a tool to embody design concepts

PROPOSED SOLUTIONIMPLEMENTATION: TEST BED

60 %

12

Sincere Gratitude: Towards Dr. Rajat Moona, Professor, CSE, IIT Kanpur for providing STM NFC Kit

ST Microelectronics CRX14 RFID ReaderST Microelectronics SR176B-A3T/PRY series RFID Tags

Standard USB Cable

Implementation Intuition 2: First iteration, second iteration, ...

RESULTS: ERROR CORRECTION

65 %

13

Multi-bit per byte error-correction to augment CRC on RFID Reader

RESULTS: SNAPSHOT OF ERT

70 %

14

Initialization of ERT

Quasi-Realtime Theft Evidence

MERITS/DEMERITS

75 %

15

1. Low cost prototype embodying design concepts 2. Experimental test-bed for problem statement 3. Version 1.0 contains significant features 4. Minimal effort portable to other platforms 5. Pragmatic solution

1. Extensive field testing required 2. Tag data is not encrypted 3. Tag-cloning possible (But cost of attack increases)

Wisdom: Create, Err, Refine, Create again, …

INTERESTING MERIT

80 %

16

Daily life examples of Faraday Cages(s)

•Elevators•Conference rooms equipped with noise-cancellation technology•…

Immune against operation in Faraday Cage

Lack of proximity of wearable-band to portable artifact in these cages will still result in THEFT_CONDITION

INTERESTING DEMERIT

85 %

17

Frequent theft alarms may be nuisance to people-at-large

Quick solution: User can configure theft alarms (enable/disable) through a user-interface

APPLICATIONS

90 %

18

1. Portable Audio Player 2. Portable Video Player 3. Smartphone 4. Blackberry and more …

1. Documents 2. Wallet 3. Bag 4. ID-Card and more …

IAM: Intelligent Artifact ModelUAM: Un-intelligent Artifact Model

SUBSEQUENT WORK

95 %

19

Further work includes:

•Enabling Sharing operation mode (Intelligent Artifact Model)

•Enabling Multi-tag operation mode (Un-intelligent Artifact Model)

•Enabling Uni-tag operation mode (Un-intelligent Artifact Model)

•Adding support for different types of RFID reader(s)/tag(s)

•Integration with a given portable artifact

QUESTIONS & ANSWERS