Dynamic Model-Based Filtering for Mobile Terminal Location - - PowerPoint PPT Presentation

▶
dynamic model based filtering for mobile terminal
SMART_READER_LITE
LIVE PREVIEW

Dynamic Model-Based Filtering for Mobile Terminal Location - - PowerPoint PPT Presentation

Sep 29, 2022 287 likes •813 views

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation Michael McGuire Edward S. Rogers Department of Electrical & Computer Engineering University of Toronto, 10 Kings College Road, Toronto, Ontario, Canada M5S 3G4

slide-1
SLIDE 1

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation

Michael McGuire Edward S. Rogers Department of Electrical & Computer Engineering University of Toronto, 10 King’s College Road, Toronto, Ontario, Canada M5S 3G4 mmcguire@dsp.toronto.edu

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.1/51

slide-2
SLIDE 2

Outline

  • 1. Signal Processing for Future Wireless Communications

Systems.

  • 2. Introduction to Mobile Terminal Location.
  • 3. Zero Memory Estimation
  • 4. Dynamic Estimation
  • 5. Conclusions.
  • 6. Future Work.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.2/51

slide-3
SLIDE 3

Evolution of Wireless Services

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.3/51

slide-4
SLIDE 4

Mobility & Multimedia

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.4/51

slide-5
SLIDE 5

Mobility & Multimedia

3G Systems

  • UMTS

(ETSI) IMT-2000 (ITU) Support user bit rates up to 2 Mbps High mobility environment: 144 kbps Ad Hoc Systems

  • IEEE 802.11

Bluetooth

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.5/51

slide-6
SLIDE 6

Signal Processing for Wireless

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.6/51

slide-7
SLIDE 7

Signal Processing for Wireless

Key problems: Capacity Resource allocation Connection management Channel management

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.7/51

slide-8
SLIDE 8

Signal Processing for Wireless

Present: Reactive control methods Future: Proactive control methods Requires future system state estimation.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.8/51

slide-9
SLIDE 9

State Estimation

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.9/51

slide-10
SLIDE 10

State Estimation

Adaptive estimation Learning model. Adapting to changing model. Estimation techniques Parametric Non-parametric

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.10/51

slide-11
SLIDE 11

Mobility Management

Need to know resources that terminals require in future Prediction of future locations. Channels Handoff algorithm Routing Power/Bandwidth allocation Power control Code selection (CDMA)

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.11/51

slide-12
SLIDE 12

Mobility Management

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.12/51

slide-13
SLIDE 13

Mobile Terminal Location

Locating mobile terminal from radio signal Applications Resource allocation Location sensitive information Emergency communications

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.13/51

slide-14
SLIDE 14

Terminal Location Methods

Handset based Perception of user privacy. Currently greater accuracy. Network based Cheaper terminals. Greater potential accuracy FCC Requirements Configuration Accuracy Requirement

> 67% > 95%

Handset 50 m 150 m Network 100 m 300 m

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.14/51

slide-15
SLIDE 15

Terminal Location Measurements

Received Signal Strength(RSS), Time of Arrival (ToA), Time Difference of Arrival (TDoA). Angle of Arrival (AoA).

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.15/51

slide-16
SLIDE 16

Terminal Location Measurements

Measurement Type Advantages Disadvantages Received Signal Strength (RSS)

  • low cost measurements
  • simple computations
  • low accuracy in large cells

Angle of Arrival (AoA)

  • simple computations
  • specialized antennae
  • low accuracy in large cells

Time of Arrival (ToA)

  • time

measurement re- quired for TDMA/CDMA network operation

  • simple computations
  • synchronized network re-

quired

  • receiver must know time
  • f transmission
  • expensive measurement

Time Difference of Arrival (TDoA)

  • time

measurement re- quired for TDMA/CDMA network operation

  • receiver does not need

time of transmission

  • synchronized network re-

quired

  • expensive measurement
  • complex calculations

TDMA - Time Division Multiple Access, CDMA - Code Division Multiple Access

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.16/51

slide-17
SLIDE 17

Radio Signal Measurements

Non-linear effects make problem more complex

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.17/51

slide-18
SLIDE 18

Radio Signal Measurements

τ(k) is the vector of propagation time measurements for

sample time k

τ(k) = d(k) + ε(k) d(k) is the vector of propagation distances. ε(k) is the vector of measurement noise. z(k) is ToA/TDoA measurement vector: z(k) = Fτ(k) F is the measurement difference matrix.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.18/51

slide-19
SLIDE 19

Geometric Dilution of Precision (GDOP)

High Precision Geometry Low Precision Geometry

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.19/51

slide-20
SLIDE 20

My Contribution

  • 1. Improved Zero Memory Estimation
  • 2. Bounds on Zero Memory Estimation Error
  • 3. Model-based Dynamic Estimation

New Filter Algorithm Developed

  • 4. Bound on Dynamic Estimation Error

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.20/51

slide-21
SLIDE 21

Zero Memory Estimation

Previously proposed techniques are Maximum Likelihood Estimators(MLE). Problems with MLE: Prior knowledge is ignored. Assumed Line of Sight (LOS) propagation model. NLOS is common in urban areas of interest.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.21/51

slide-22
SLIDE 22

Zero Memory Estimation

Observations: Statistical knowledge of terminal position available from hand off algorithm. Propagation survey made during network configuration.

= ⇒ Network has knowledge that can be used for

location.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.22/51

slide-23
SLIDE 23

Zero Memory Estimation

k is sample interval. θ(k) is location of mobile terminal at k. ˆ θ(k) is estimated location of mobile terminal at k.

Survey data: j survey point, j ∈ {1, 2, ..., n}.

θj, location of survey point j. zj, measurement taken at survey point j.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.23/51

slide-24
SLIDE 24

Zero Memory Estimation

Estimated location is weighted average of survey point locations:

ˆ θ(k) = n

j=1 θjh(z(k), zj)

n

j=1 h(z(k), zj)

h(·) is kernel function.

Estimated location is weighted average of survey point locations. Weights determined by kernel functions.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.24/51

slide-25
SLIDE 25

Zero Memory Bounds

NLOS propagation creates discontinuities in propagation equations. Standard bounds (e.g. Cramer-Rao no longer apply). Use other bounds Barankin bounds Weinstein-Weiss bounds.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.25/51

slide-26
SLIDE 26

Simulated Environment

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.26/51

slide-27
SLIDE 27

Zero Memory Results

50 100 150 200 15 20 25 30 35 40 45 50 RMSE (m) Standard Deviation of Range Error Parametric MLE (TDoA) Parametric MLE (ToA) Non-parametric MAP (TDoA) Non-parametric MAP (ToA) Parzen Gaussian (TDoA) Parzen Gaussian (ToA)

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.27/51

slide-28
SLIDE 28

Zero Memory Results

10 15 20 25 30 35 40 45 50 5 10 15 20 25 30 35 40 σd (standard deviation of range error) RMSE (m) WWB ToA Simulated ToA WWB TDoA Simulated TDoA

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.28/51

slide-29
SLIDE 29

Dynamic Estimation

Combine measurements from different sampling periods. Use dynamic model of mobile terminal motion. Dynamic model consists of: Kinematic model. Human Decision model.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.29/51

slide-30
SLIDE 30

Mobile Terminal Motion Model

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.30/51

slide-31
SLIDE 31

Kinematic Model

x(k) is terminal state. u(k) is control input. w(k) is process noise.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.31/51

slide-32
SLIDE 32

Human Decision Model

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.32/51

slide-33
SLIDE 33

Zero Memory Estimator Preprocessor

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.33/51

slide-34
SLIDE 34

Dynamic Estimation

Prediction phase Correction phase

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.34/51

slide-35
SLIDE 35

Dynamic Estimation

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.35/51

slide-36
SLIDE 36

Bounds on Dynamic Estimation

Combine following information sources: Zero Memory Estimator. Dynamic model for mobile terminal motion. Prior distribution for mobile terminal location. Bound calculated on squared error.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.36/51

slide-37
SLIDE 37

Dynamic Estimation Results

Fixed Control Input

20 40 60 80 2 4 6 8 10 12 14 16 Samples RMSE (m) Evaluation Bound Dynamic Filter Zero Memory Estimator

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.37/51

slide-38
SLIDE 38

Dynamic Estimation Results

Changing Control Input

20 40 60 80 4 6 8 10 12 14 16 Samples RMSE (m) Evaluation Bound Dynamic Filter Zero Memory Estimator

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.38/51

slide-39
SLIDE 39

Dynamic Filter Comparison

6 7 8 9 10 11 12 13 14 20 40 60 80 100 RMSE (m) Samples Zero memory estimator Simple Kalman filter (Hellebrandt et al.) Simple Kalman filter Multi-model filter

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.39/51

slide-40
SLIDE 40

Dynamic Filter Comparison

Changing Control Input

2 3 4 5 6 7 8 9 10 20 40 60 80 100 RMSE (m/s) Time (s) Multi-model filter Simple Kalman Filter Simple Kalman Filter (Hellebrandt et al.)

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.40/51

slide-41
SLIDE 41

Dynamic Filter Robustness

6 8 10 12 14 16 0.2 0.4 0.6 0.8 1 RMSE (m) Pr(TURN) multi-model filter, optimized for Pr(TURN)=0 zero memory estimator multi-model filter, optimized for Pr(TURN)=2/3 best multi-model filter for Pr(TURN)

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.41/51

slide-42
SLIDE 42

Dynamic Filter Robustness

7 8 9 10 11 12 13 5 10 15 20 RMSE (m) Maximum Mean Velocity multi-model filter, optimized for C=2.5 m/s2 zero memory estimator

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.42/51

slide-43
SLIDE 43

Dynamic Filter Robustness

7.5 7.6 7.7 7.8 7.9 8 8.1 8.2 8.3 0.1 0.15 0.2 0.1 0.15 0.2 α α (Filter)

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.43/51

slide-44
SLIDE 44

Results

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 5 10 15 20 25 30 35 40 45 Probability Distance Error (m) Dynamic Filter Estimator Zero Memory Estimator

Configuration Accuracy (ToA σd = 15 m) 67% 95% Zero Memory Estimator 12.17 m 21.16 m Dynamic Estimator 7.12 m 14.28 m

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.44/51

slide-45
SLIDE 45

Conclusions

Use all information sources. Model-based estimation gives accurate location estimates. Efficiently combines information from different time periods. Estimation methods are robust. Zero memory estimator robust to changes in noise/propagation model. Dynamic estimator robust to changes in dynamic model.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.45/51

slide-46
SLIDE 46

Future Work

  • 1. Applications of mobile terminal location.
  • 2. Long term motion models.
  • 3. Data fusion.
  • 4. Location of terminals in ad hoc networks.

Location of terminals in hybrid networks.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.46/51

slide-47
SLIDE 47

Applications

Resource allocation Hand off algorithms Many possibilities for collaboration.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.47/51

slide-48
SLIDE 48

Long Term Motion Models

Current dynamic filter based on short term motion models. Long term motion models will improve estimation. Improve motion prediction.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.48/51

slide-49
SLIDE 49

Data Fusion

Use data from multiple information sources. RSS is cheap with wealth of propagation data but has large uncertainty. ToA/TDoA are expensive with low uncertainties. AoA requires special antennae and provides varying accuracy.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.49/51

slide-50
SLIDE 50

Ad Hoc Networks

Examples: Bluetooth, IEEE 802.11 Terminal must be low cost. Limited connectivity between terminals. Hybrid networks

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.50/51

slide-51
SLIDE 51

Final words

Large amount of work to be done. Many applications of results. Potential to develop new estimation and filtering algorithms.

Dynamic Model-Based Filtering for Mobile Terminal Location Estimation – p.51/51