A Prototype Positioning System based on Digital Audio Broadcast - - PowerPoint PPT Presentation

a prototype positioning system based on digital audio
SMART_READER_LITE
LIVE PREVIEW

A Prototype Positioning System based on Digital Audio Broadcast - - PowerPoint PPT Presentation

A Prototype Positioning System based on Digital Audio Broadcast Signals Duncan Palmer, Terry Moore, Chris Hill Institute of Engineering Surveying and Space Geodesy The University of Nottingham United Kingdom Overview Why use the DAB


slide-1
SLIDE 1

A Prototype Positioning System based on Digital Audio Broadcast Signals

Duncan Palmer, Terry Moore, Chris Hill

Institute of Engineering Surveying and Space Geodesy The University of Nottingham United Kingdom

slide-2
SLIDE 2

Overview

– Why use the DAB signal? – Technical Characteristics – Positioning Potential – Hardware/Software – Network Geometry Simulation (HDOP) – Early results – Conclusions

slide-3
SLIDE 3

Digital Audio Broadcast Signal

Why use the DAB signal?

– Designed for dynamic receivers (car radios) – Uses Single Frequency Networks (SFNs) synchronised by GPS – Two National and many Local/Regional SFNs – > 85% UK coverage – Terrestrial signal power up to 1000× higher than GNSS signal power

slide-4
SLIDE 4

DAB Signal Characteristics

Signal Basics

  • Operates in VHF Band III (170 - 240 MHz)
  • Channel bandwidth = 1.512 MHz
  • Uses Coded Orthogonal Frequency Division

Multiplexing (COFDM) modulation technique

  • Spread spectrum technique delivering slow-rate

data using > 1500 sub-carriers

slide-5
SLIDE 5

DAB Signal Characteristics

System Clock

  • DAB system clock frequency = 2.048MHz
  • Fundamental DAB Unit T obtained by:
  • All units in system can be derived from this value
  • Describes time periods in the temporal domain
  • Speed of light travels ≈ 146m in one unit of T
slide-6
SLIDE 6

DAB Signal Characteristics

Transmission Frame

  • Consists of three channels broadcast sequentially:

– Synchronisation Channel – used for signal acquisition

  • Composed of 1 Null and 1 OFDM symbol (length ≈ 2.5ms)

– Fast Information Channel – used for multiplex data

  • Composed of 3 OFDM symbols (length ≈ 3.7ms)

– Main Service Channel – used for “music” data

  • Composed of 72 OFDM symbols (length ≈ 89.7ms)

Total Frame Length 96ms

slide-7
SLIDE 7

DAB Signal Characteristics

Synchronisation Channel Structure

  • First two symbols of the transmission frame - Null &

Time Frequency Phase Reference (TFPR) symbols

  • Guard Interval (GI) inserted before each OFDM

symbol of length 504T

  • GI is a replica of the last 504T of each OFDM symbol

& inserted before that symbol starts

= 504T 504T Null TFPR

slide-8
SLIDE 8

Tx1 = Waltham

DAB Signal Characteristics

Transmitter Identification (TII)

1 1 1 1 0 0 0 0

11100001 Pattern ID defines Region as: East Midlands

Tx1

Tx2 = Mapperley

Tx2

1 2 3 4 8-bit pattern where 4 of 8 sub- carriers switched on

slide-9
SLIDE 9

Positioning Potential of DAB

  • To use the Synchronisation Channel for Time

Difference of Arrival (TDOA) measurements

– Subsequent transmissions start in the Guard Intervals

  • f first received transmission

– Transmitter locations are known from TII

slide-10
SLIDE 10

Hardware

The Universal Software Radio Peripheral (USRP)

  • USRP Comprises:

– FPGA Motherboard – USB 2.0 Interface – 4× 64 MS/s 12 bit ADC – 4× 128 MS/s 14 bit DAC – Digitises up to 16MHz spectral bandwidth * – External clock input option * Dependent on daughter- board front-end.

Ettus

– TVRX receiver daughter-board front end used for this project – Frequency range 50 – 870 MHz – Maximum receiving bandwidth 6MHz

slide-11
SLIDE 11

Hardware

DAB Antenna

  • 360° Beamwidth
  • Frequency Range:
  • 200 – 240 MHz
  • Gain 2.2 dBd
slide-12
SLIDE 12

Software

GNU Radio – USRP designed for use with GNU Radio – Runs on most Linux platforms (Ubuntu in this case) – Software Defined Radio processing blocks constructed using Python/C++ code (all open source) – Signal demodulation either done on-board or raw data recorded for post-processing – Matlab or similar to post-process data

BBC Nottingham Local Digital One

slide-13
SLIDE 13

Two DAB Signals

Captured in Frequency Domain

slide-14
SLIDE 14

Initial Test Region

Nottinghamshire/Leicestershire

Nottingham:

  • Mapperley
  • Waltham

Leicester:

  • Copt Oak
  • Houghton-
  • n-the-Hill
slide-15
SLIDE 15

Matlab Simulation Results

TDOA HDOP Map – Approach 1

  • TDOA HDOP Simulation based on 4 synchronised DAB Tx

locations in the Notts/Leics area using a single network

  • Gives three independent TDOA measurements
slide-16
SLIDE 16
  • Problem 1: Many areas will not receive national

signals from more than two transmitters

  • Solution: Multi-network solution (i.e. Both

National networks simultaneously)

  • Problem 2: Most transmitter sites broadcast more

than one network (e.g. Both National networks and a local network)

  • Solution: Combination of two local networks

which are unlikely to share transmitter sites

  • Involves a different strategy using two pairs of

synchronised transmitters

Matlab Simulation Results

Difficulties to overcome

slide-17
SLIDE 17

Conceptual Network

Time Difference of Arrival

A1 & A2 Synchronised B1 & B2 Synchronised Differences r1-r2 & r3-r4

slide-18
SLIDE 18

Matlab Simulation Results

TDOA HDOP Map – Approach 2

  • TDOA HDOP Simulation based on the same transmitters but

using 2 pairs of synchronised transmitters – two local networks

  • Gives two independent TDOA measurements
slide-19
SLIDE 19

TDOA Measurement Process

Find First Null Symbol

Find end of first Null Symbol in the temporal domain by testing against pre-defined value

Search Direction

slide-20
SLIDE 20

TDOA Measurement Process

Define Symbols

First 504T (GI) Last 504T Both extracted in time domain

slide-21
SLIDE 21

TDOA Measurement Process

Compare Extracted Data

From right to left, data practically overlaps perfectly until this cut-off point

slide-22
SLIDE 22

TDOA Measurement Process

Calculate Time Delay

Signal arrival delay of 68T 68T = 68 × 0.48828µs 68T = 33.203µs delay 0.48828µs ≈ 146m

slide-23
SLIDE 23

Initial Positioning Results

East Midlands Test Region

A1 - MAPPERLEY B1 – HOUGHTON A2 - WALTHAM B2 – COPT OAK

880m 90m

slide-24
SLIDE 24

TDOA Measurement Process

New TFPR CIR Approach

  • Described measurement system used for rough

signal acquisition

  • New algorithm to use Channel Impulse Response

(CIR) Method using the TFPR symbol

  • TFPR values known to receiver, so cross-correlation

technique used

  • Will provide multiple TDOA measurements per

network

slide-25
SLIDE 25

Error Sources

  • TII information indicates that only 2 transmitters are

received per network... ‒ A 3rd much weaker transmitter in some areas could make timing cut-off measurement “diffuse”

  • Although synchronised by UTC, deliberate timing biases

can be inserted as part of the SFN design to avoid ISI

  • Cross-correlation of TFPR symbol should give better

TDOA than time delay measurement

  • Geometry of each network affects HDOP values
  • No terrain correction currently

‒ Possible Multipath interference

slide-26
SLIDE 26

Summary

  • DAB signal contains components usable for positioning

purposes

  • Low-frequency, terrestrial signal provides good power

and horizontal geometry of transmitters

  • Early HDOP simulations indicated good coverage in UK,

particularly in urban areas where GPS difficulties could

  • ccur
  • Geometry of networks affects HDOP values

‒ Network designed for comms NOT navigation

  • Use of 2 pairs of transmitters from different local

networks most likely solution

  • Improvements expected using second algorithm

‒ CIR approach over the TFPR symbol

slide-27
SLIDE 27

Duncan Palmer PhD Candidate The University of Nottingham University Park Nottingham NG7 2RD UK

  • Telephone:

+44 (0) 115 951 3880

  • Fax:

+44 (0) 115 951 3881

  • Email:

isxdp2@nottingham.ac.uk

  • WWW:

www.nottingham.ac.uk/iessg

Contact Details Questions?