HiRLoc: Highresolution robust localization for wireless sensor networks

1
HiRLoc High-resolution robust localization for
wireless sensor networks

• Authors L. Lazos and R. Poovendran
• From IEEE Journal on Selected Areas in
  Communications, vol. 24, no. 2, Feb.
  2006.
• Date 2006/3/14

2
Outline

• Introduction
• Related works
• Network model assumptions
• High-resolution range-independent localization
  scheme
• Security threats against HiRLoc
• Performance evaluation
• Conclusion

3
1. Introduction

• WSN need location information
• GPS is not suitable for sensor nodes
• WSN are susceptible to a variety of attacks
• Secure localization issue
• Decentralized and scalable implementation
• Resource efficiency in computation, communication
  and storage
• Range-independence
• Robustness against security threats

4
2. Related works

• For a trusted environment
• Range-dependent
• Based on distance or angle estimates to some
  reference points
• TOA, TDOA, AOA, RSSI
• Range-independent
• Based on the information transmitted from the
  reference points (without time, angle, or power
  measurements)
• GPS-less, range-free, APS

5

• For an untrusted environment
• Range-dependent
• SPINE6 based on bounding the distance of each
  sensor to at least 3 reference points.
• MMSE22 uses minimum mean square estimation to
  filter outliers.
• Range-independent
• SeRLoc18, 19
• ROPE20 need a large number of reference points
• Statistical method21

6
3. Network model assumptions

• Network deployment
• A set of sensors S is randomly deployed with
  density ?s within area A
• Locators are randomly deployed with density ?L

A randomly chosen sensor hears k locators
7

• Antenna model
• Sensors are equipped with omnidirectional
  antennas with maximum power Ps
• Locators
• Equipped with M directional antennas with maximum
  power PL ( PL gt Ps )
• Can simultaneously transmit on each antenna
• Can vary their transmission range from zero to a
  maximum value of R
• Can change their antenna direction

8
4. High-resolution range-independent localization
scheme

• Location determination
• Each locator transmit a beacon at each
  directional antenna
• The locators coordinates
• The angle of the sector boundary lines
• The locators communication range R

9
(No Transcript)
10
(No Transcript)
11

• HiRLoc-I
• Collect all beacons over several transmission
  rounds and compute the intersection of the all
  sector areas.
• HiRLoc-II
• Estimate region of intersection (ROI) after every
  round of transmission and intersect it with the
  previous estimate of the ROI.

12
Range variation
Antenna orientation variation
13
Xmax
Ymax
Xmin
Ymin
14
(No Transcript)
15
Security features

• Encryption of the beacon transmission
• Global shared symmetric key K0 in every locator
  and sensor
• Every sensor shares a symmetric pairwise key KsLi
  with every loactor Li
• Authentication of the beacon transmission
• Collision-resistant hash functions
• Each locator Li has a unique password PWi

16
5. Security threats against HiRLoc

• Wormhole attack
• An adversary deploys a direct link (wormhole
  link) between two points on the network with a
  distance longer than the communication range.
• Record broadcast information and then tunnel it
  to the other

17

• Defending for antenna orientation variation
• Single message/sector per locator property
• Communication range violation property
• Defending for communication variation
• Compute ROI(1) in the first step. (not after
  gather all beacons)
• Attack will be detected by checking ROI(1)
  nROI(j)?

18
(No Transcript)
19
(No Transcript)
20

• Sybil attack
• An adversary impersonates multiple network
  entities.
• impersonate sensors
• Impersonate locators (impact the localization)

21

• Defending for antenna orientation variation
• Lmax Maximum allowable number of locaters heard
  by each sensor.
• If a sensor hears more than Lmax locator, it
  assumes that is under attack (using
  attach-to-closer-locator(ACLA) to determine its
  position)
• Defending for communication variation
• Compute ROI(1) in the first step. (not after
  gather all beacons)
• Attack will be detected by checking ROI(1)
  nROI(j)?

22
(No Transcript)
23
(No Transcript)
24
6. Performance evaluation

• 5000 sensors within a 100100 m2 area.

Total number of locators
The average locators heard by each sensor
Communication range
Set of sensors
The average localization error
25
0.23r
LH15 ? avg. 5 locators/ per sensor
26
200 sensors
27
Reduce 50
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
Conclusion

• Proposed a high-resolution range-independent
  localization scheme
• Higher accuracy
• Fewer hardware resources
• Robust (wormhole and sybil attack)