Final Report

7/21/2019 Final Report

http://slidepdf.com/reader/full/final-report-56da682647949 1/35

FM/HD Radio Mapping (ECE4007 L01) 1

Final Report

FM/HD Radio Mapping

ECE4007 Senior Design Project

Sections L01, FM/HD mapping

Thomas ShanksLeandro Franca

Brian Casey

Submitted

April 30, 2009




TABLE OF CONTENTS

EXECUTIVE SUMMARY ........................................................................................................ III

1. INTRODUCTION............................................................................................................. 4

1.1 Objective ................................................................................................................ 4

1.2 Motivation.............................................................................................................. 5

1.3 Background ........................................................................................................... 5

2. PROJECT DESCRIPTION AND GOALS .................................................................... 6

3. TECHNICAL SPECIFICATIONS ................................................................................. 8

4. DESIGN APPROACH AND DETAILS ....................................................................... 10

4.1 Design Approach................................................................................................... 7

4.2 Codes and Standards .......................................................................................... 10

4.3 Constraints, Alternatives, and Tradeoffs.......................................................... 18

5. SCHEDULE, TASKS, AND MILESTONES ............................................................... 19

6. PROJECT DEMONSTRATION................................................................................... 20

7. MARKETING AND COST ANALYSIS ...................................................................... 24

7.1 Marketing Analysis............................................................................................. 14

7.2 Cost Analysis ....................................................................................................... 26

8. SUMMARY ..................................................................................................................... 16

APPENDIX A............................................................................................................................. A1




EXECUTIVE SUMMARY

The aim of this project is to actively map FM signal quality and strength quality by use of an

automated mobile system. This system will provide radio stations, transmitter/receiver design

companies, and end users with a service that will show information for any station. Current

mapping involves mathematical modeling using FCC data. NPR performed a vehicular

mapping. However, their scope was different in that they were interested in adjacent channel

interference. The design group’s system provides actual coverage with measurements of the

analog RF power and signal to noise ratio of the audio. It will show the signal strength, whether

HD is present, and will give information about the audio quality. The measurements will be

compared with similar results from NPR studies.

The system consists of a vehicle with an FM receiving antenna and a GPS unit controlled by an

automated computer system. Use of said system requires the station to send a test tone for

calibration and to further increase the accuracy of reading. Several radios will be used to test the

audio from various types of radios. Tasks have been separated between the group to ensure

design milestones will be met on time. Upon completion, a demonstration will be performed,

and data will be analyzed and compared with NPR Labs studies. The total project cost to re-

create the project has been estimated at $28,892.




FM/HD Radio Mapping

1. INTRODUCTION

Radio stations and government regulators generally rely on mathematical modeling to determine

the effective coverage area of their broadcasts. These mathematical models are based on rough

estimates of signal quality in the presence of approximated terrain and interference levels. They

do not accurately account for many factors that affect the signal, including multipath fading,

terrain type, attenuation from man-made obstructions, and interference caused by on-channel or

ineffectively filtered adjacent IBOC (in-band on-channel) HD Radio® digital signals within

analog FM receivers.

1.1 Objective

The aim of this project is to create an active mapping system for FM radio reception

quality and signal strength using an automated vehicle-mounted system. The design

places an FM antenna on a vehicle and feeds the antenna’s output to multiple consumer

receivers within the vehicle to measure the actual received signal quality with the devices

used in that test. The output from these receivers is processed to determine the quality of

the audio. A GPS unit is used to record the location of each measurement so that they

can be subsequently mapped. Audio is sampled and processed at each location to

determine the signal-to-noise ratio received. The audio is statistically examined to

determine the statistical characteristics of the noise and to measure or visualize the effects

of such parameters as field strength, distance from the transmitter, bearing from the

transmitter to the location of reception, type of radio used, adjacent station interference,

and IBOC carrier presence on the distribution.




1.2 Motivation

The system provides radio stations, transmission and reception hardware companies,

advertisers, and end users with a service that will measure current coverage and validate

the results of existing models of coverage prediction. This information can be used to

show the station’s current coverage and to help determine if antenna patterns are met or

transmission system upgrades are called for. The system also measures the local

behavior of interference and the statistical distribution of noise versus under real-world

conditions and locations. Additionally, stations can see the actual measured IBOC

interference to their analog signal at various actual locations within their market.

Publication of these results will also show end users where they can receive a station’s

FM signal and whether stronger antenna/amplifiers are useful or required to receive a

station’s signal in their respective areas.

1.3 Background

Current coverage mapping involves use of simple terrain elevation and transmit location

data to predict the radio terrain coverage [1]. Variables such at transmitter power,

antenna pattern, and antenna height which are published for each station by the FCC are

used to make these predictions.

NPR performed a study in 2006-2008 in which they investigated the affect of adjacent

radio stations on HD Radio reception [2]. They designed an all-inclusive system that

measured field strength of the desired channel and of two adjacent channels in a mobile

environment. It also determined whether HD Radio was successfully decoded at each

location. Locations were marked by GPS. Raw data was recorded to a memory card,




then was later analyzed using MATLAB. From this data, a model of HD Radio coverage

prediction was developed.

Amateur radio product engineer Brian Beezley has investigated the noise that a station’s

own IBOC signal induces in the reception of that station’s analog signal in average

consumer receivers. He has posted detailed explanations and measurements of this

adjacent-signal interference phenomenon in depth [3]. He finds that the primary cause of

IBOC-induced noise is imperfect stereo decoding methods in consumer receivers. He

states that such noise can be eliminated either though careful filtering at the intermediate-

frequency (IF) level or through post-detection filtering before stereo is decoded. Since

most of the IBOC-induced noise comes in through the stereo difference signal decoding

process, Beezley finds that switching a radio to mono, when possible, also removes most

of the noise. He indicated that most home analog FM tuners are susceptible to this noise

issue, but that most car FM tuners are not because of their narrower IF filters.

2. PROJECT DESCRIPTION AND GOALS

The system provides FM radio stations with measurements of the area where their analog FM

radio signal can be received.

The system is also capable of measuring where the IBOC digital radio signal can be received.

However, this feature was not tested because the digital IO hardware that the project team was to

be loaned was not made available as arranged. Additionally, the system can measure the effects

that the transmitted digital radio signal has on the reception of the station’s analog signal. This is




done by comparing reception over the coverage area with and without IBOC transmission

enabled., but sufficient equipment (DIO, susceptible analog radio) were not available within the

budget to test or use this feature. and, therefore on the analog signal’s effective coverage area.

Since these detrimental effects occur because of imperfect filtering of adjacent signals within the

radio, susceptible consumer analog FM radios, residential and portable units in particular, would

need to be selected and tested in the field. This would allow a more general sense of the average

consumer radio’s IBOC signal rejection performance. NPR Labs in the area and conditions

measured. The project team was unable to purchase any analog radios beyond the non-

susceptible vehicular unit within the budget allotted. NPR Labs

The System has the following features:

• Maps signal RF power and analog FM SNR received in light of actual city RF noise

• Visualizes the results in a manner that can easily be compared with FCC-providedcoverage maps or with predictions of coverage and interference provided by NPR Labs toall CPB-funded stations

Field measurements can be compared existing computer model predictions to validate the

model’s applicability to that particular station and environment [4] or to determine whether a

station’s transmission system or the RF environment are the source of interference. The project

team has provided a method to perform a visual comparison with the FCC-provided coverage

prediction map using Google Maps. The NPR Labs project team in Washington, DC committed

to producing predictions of the HD Radio and FM analog coverage with and without IBOC

enabled for Georgia Tech’s WREK [5], the target station for design, testing, and validation.

Unfortunately, due to changes in funding and staffing, NPR Labs was unable to provide




numerical or visual predictions of WREK’s current coverage to numerically or graphically

compare to the group’s measurements as originally planned.

3. TECHNICAL SPECIFICATIONS

The field measurement system consists of a vehicle with an FM receiving antenna and a GPS

receiver connected to equipment within the vehicle. For each GPS location, the system will

record the following measurements:

• Receive power (from antenna) of the target station's analog FM radio signal (ReceivedPower (dBm))

• Calculate Signal to noise ratio of audio from radios Power of the audio signal over the

rms power of the noise(dBm)

• Measured from one channel of the output audio that is decoded from the stereo analogFM signal by several representative consumer radios of different types:

o audio signal level when level calibration tone present ("FM-Stereo Audio Calib.(dBFS)")

o average audio noise level when no modulating audio present ("FM-Stereo AudioNoise (dB below calibration)")




Use of the system requires the station under test to transmit a pulsing audio signal, to enable FM

stereo, and to disable all other FM subcarriers.

Unit Min Max Acc. ± Meas./sec

Received Power dBm -90 -10 2 5

Field Strength dBuV/m 30 100 8 1

RF Noise Power dBc -70 -20 10 5

IBOC Carriers Enabled T/F F T - 5

HD Radio Receivable T/F F T N/A 1

FM-Stereo Audio Calib. dBFS - - 2 0.01

FM-Stereo Audio Noise dB cal -70 -10 3 20

Table 1. Minimum Design Specifications.




4. DESIGN APPROACH AND DETAILS

4.1 Design Details

Description of Design

The field measurement system as currently implemented receives the FM radio signal

from a monopole magnet-mount antenna placed on top of a vehicle. This is connected

via a cable TV amplifier (gain = 25 dB) to a spectrum analyzer, an HD Radio receiver,

and several different types of consumer analog FM radios through an RF splitter, as

shown in Figure 1. The cable from the antenna was matched from 50 Ohms to 75 Ohms,

using a 12th wave matching network to improve the VSWR between the antenna and the

following components. Using an antenna meter the antenna was properly shortened to be

¼ wavelength of the target station’s frequency. The antenna was shortened so that its

VSWR at the target frequency was 1.0. This measurement was performed while the

S p l i t

t e r

Figure 1. Schematic of proposed design.




antenna was attached to the roof of the vehicle. The length of said antenna after being

shortened is 36 inches long. Preceding the amplifier is an RF bandpass filter. This filter

offered for use to the group by Dr. Aaron Lanterman of the Georgia Tech Electrical and

Computer Engineering Department. The purpose of this filter is to block all RF signals

outside of the FM band. This prevents saturation of the amplifier when driving near

stations such as in the television bands. The antenna, amplifier, and matching network are

selected in order to provide the components with an RF signal that best match that of the

signal straight off the antenna. From the antenna, the signal is split four ways, by means

of a satellite TV splitter, to the spectrum analyzer, two radios, and a 75-ohm termination.

The original design called for three radios; however, due to budget restraint only two

were used in system. Software is set up to support three radios however, and hence the

termination is a placeholder for future updates. The spectrum analyzer used has an

optical to serial cable for data acquisition and hardware control. The GPS also has a

serial interface. The RS-232 serial output of both the GPS and the spectrum analyzer are

connected through serial-to-USB converters to the central USB hub. USB audio cards

then attach to the radios using RCA to mini headphone jack (1/8”). These connect to the

microphone/line in of the audio cards. The audio cards connect to the USB hub as well.

This USB hub connects to the laptop where the user can configure the system and control

automation.

Controlling the above hardware is LabView 8.5 with the additional Sound and Vibration

Toolkit (SVT) installed. The design group used the evaluation versions of both LabView

and SVT on the system laptop. The software was developed as a state machine, for

modularity and ease of debugging. There are several states in the software for testing




various hardware, and setting up said hardware to be ready for an automation mode. The

LabView automated data acquisition is shown in the flow chart in Figure 2.

Figure 2. Automation software flow chart.






Figure 3. Screenshot of test tone.




Data Acquisition

A GPS receiver will be placed on the vehicle so that the current location, direction, and

speed using NMEA standard RS-232 connection. The data output from the GPS receiver,

the channel power from the spectrum analyzer, and the audio output from the radios will

be connected to a computer located within the vehicle. The computer will analyze the

signals using LabView that is capable of processing the data in real time to determine the

transmission mode of the target station and configure the measuring equipment

accordingly. LabView sets up the spectrum analyzer for channel power collection. The

spectrum analyzer is centered at the radio station’s frequency, with a channel power

bandwidth set to 500 kHz. LabView takes a 500 ms sample from each audio card. It first

S p l i t t e r

Figure 4. Signal loss/gain through system.




calibrates the threshold by scanning the values, and determining the max 1 kHz tone

value. It then determines the power of the signal in dB. Once the power level of the 1

kHz tone drops, and is below the threshold of -10 dB, the noise is calculated. An ITU-

468R weighting filter is applied to the noise channel. This filter is more representative of

what the human ear is actually hearing as far as noise. The filter does not need to be

applied to the signal because the filter would apply no change to the 1 kHz frequency od

the periodic level-setting tone. The RMS power of this weighted noise is determined

from the signal of silence following each 1 kHz pulse. The signal to noise ratio is then

calculated as the measured power of the signal minus the measured power of the noise.

Processing and Visualization

The data collected by the LabView system is post-processed using a Python script and

visualized using GNUplot 4.0. The files to do this in a POSIX environment are found on

the project’s website. The user calls the included bash shell script “plot.sh”, which calls

the other tools itself. The shell script runs the Python post-processing functions located

in “parse.py” located in the same directory, then calls GNUplot, which is assumed to be

located in “/tmp/gnuplot”, to produce the images. “plot.sh” requires that the input

measurement file from LabView and the output directory for the plots be specified on the

command line, in that order. These scripts were tested to run on both Cygwin and Linux

systems.

The Python post-processing code calculates the distance to the measured points from the

GPS data. The common Haversine formula is used to accomplish this. The bearing from

the transmitter to that point is also calculated XX CITE http://www.movable-




type.co.uk/scripts/latlong.html XX. The set of data points is then saved into another

file, “points.txt”, in the output directory so that the post-processing output can be viewed

and so that GNUplot can access the data.

GNUplot is then called several times to produce a set of plots showing the relationship

between various variables and to produce interpolated color maps of the RF signal

strength and the signal to noise ratio at each point. An HTML index file is also

automatically created in the output directory to facilitate convenient viewing of the

several plots. Some information, such as the location of the transmitter and the axis

limits for the plots, is included in the scripts for reasons of visual comparability.

Descriptions of these values is included in comments within the code.

The Google Earth visualization of the data requires simply loading the .kml files from the

project website. These files specify the location that the image is loaded from and tell

Google Earth where to place the image on the screen. To load image data from other

sources, the user would edit the file name text in these simple, small XML files.

GNUplot is unable to provide plots that are completely free of white margins, so

automation of this process was not possible.

4.2 Codes and Standards

HD Radio has been standardized by the National Radio Systems Committee as NRSC-5-

B [10]. This standard describes the air protocol and the data stream in great detail. It

also describes the emitted spectrum and provides maximum bounds for the desired signal

and for spurious emissions transmitted by stations.




The common practice for testing FM radio receivers is described in IEEE standard 185-

1975, Standard Methods of Testing Frequency Modulation Broadcast Receivers [13],

which is still in use despite being withdrawn [11] ,[12]. The standard was not available

to access or use within the budget permitted. However, the project was designed with

methods similar to those specified in the standard.

The ITU-468 weighting filter

4.3 Constraints, Alternatives, and Tradeoffs

The design group considered the approach of using a vehicle’s native antenna. Upon

further thought and discussion with John Kean of NPR Labs [2], the design team decided

that due to the somewhat directional nature of this antenna proper results could not be

obtained. Due to high level of wear and tear on the tube of the transmitter’s exciter, the

group will not able to turn on and off the IBOC signal rapidly but more along the order of

tens of seconds. The spectrum analyzer being a handheld and slightly less capable could

cause some issues with various measurements. The use of the spectrum analyzer for

obtaining field strength measurements of non-preselected channels could too cause

issues.

The fact that the design team plan to use a single antenna causes the issue of noise due to

a splitter, and amplifier being introduced into the signal path. The major tradeoff is

commonality vs. a more real signal.




Some researchers in this field have access to testing zones that offer RF free zones and

large turntables to rotate antennas for calibration and testing. Due to a minimum budget

use of such locations will not be possible. Antenna system design will be tested in a very

high RF zone, Georgia Tech. To compensate numerous tests will be made to generate

calibration data for the system.

In order to be able to send test tones of various sorts from the FM transmitter, testing will

only be permitted at late hours of the night. The design team will be performing data

acquisition testing at hours such at 3-5am. Time is limited, and therefore a month has

been set aside for data acquisition.

5. SCHEDULE, TASKS, AND MILESTONES

The group designed and built this system as a team. However task responsibilities have been

divided among group members based on expertise and resource availability. Table 2 shows the

scheduled task, start dates, end dates, person responsible for the task, and the degree of difficulty

of each task.

The project was completed in the following phases-

• Phase I - Preliminary Design - This step included the design and , such as hardware

drivers, working on the software interface, and choosing the antenna and the radios to be

tested.

• Phase II - System Integration – software and hardware will be put together, calibration

measurements will be determined, and the vehicle will be prepared.




• Phase III – Testing – the vehicle will be driven around to acquire data that will be

analyzed and presented.

Appendix A shows the projected project Gantt chart, while appendix B shows the actual completion

dates.

6. PROJECT DEMONSTRATION

To start, all the system components must be attached as describe in previous sections

and since there is a large amount of power and current being used by the system, a

power inverter with output of at least 400W connected directly to the car battery is

required.

Project Definition 14 days 1/5/2009 1/22/2009 1/22/2009

meet with advisor 7 days 1/5/2009 1/13/2009 2/4/2009 teamdefine project scope 7 days 1/14/2009 1/22/2009 2/11/2009 team

Preliminary Design 14 days 1/23/2009 2/10/2009 2/13/2009

communication with DAQ hardware 11 days 1/23/2009 2/6/2009 2/6/2009 Brian

antenna and radio selections 14 days 1/23/2009 2/10/2009 2/16/2009 Leandro

software interface 11 days 2/2/2009 2/13/2009 2/5/2009 Brian

hardware and software working 1 day? 2/16/2009 2/16/2009 3/15/2009 milestone

System Integration 15 days 2/16/2009 3/6/2009 3/17/2009

automated DAQ control 7 days 2/17/2009 2/25/2009 3/12/2009 Brian

data control and analysis 7 days 2/26/2009 3/6/2009 3/2/2009 Thomas

calibration 4 days 2/26/2009 3/3/2009 RF- 4/12/2009 Brian

vehicle preperation 3 days 3/4/2009 3/6/2009 3/8/2009 team

functional system 1 day 3/7/2009 3/7/2009 3/17/2009 milestone

Data Acquisition 30 days 3/8/2009 4/16/2009 4/16/2009

driving/data collection/analysis 30 days 3/8/2009 4/16/2009 4/16/2009 Leandro

presentable maps 1 day 4/17/2009 4/17/2009 4/17/2009 milestone

Demonstration 5 days 4/20/2009 4/24/2009 4/22/2009

demo for advisor 5 days 4/20/2009 4/24/2009 4/22/2009 team

Table 2. Project Schedule, Tasks, and Milestones.




• Connect a laptop with LabView and Sound Vibration Toolkit to the system

and have the file base_loop.vi ready to run ( GUI shown in Figure XX)

• Start the system and tune all radios to the target frequency

• Verify if each individual radio output signal is calibrated to the line level

output voltage, in case it is not, tune the volume on to radio to achieve the

calibrated signal

• Set the device IDs of the spectrum analyzer, GPS by selecting the specific

COM port

• Toggle the audio card switch on from the front panel GUI application

• Determine and enable the device ID of the audio card will from the front

panel. Press the button labeled “set signal power” in the spectrum analyzer

section of the front panel

Once this is performed, the spectrum analyzer is set to the correct configuration to

perform continuous signal power measurements. A check of the GPS and audio cards is

important to insure automation will be gathering valid data. The test tones may now be

started and the system can be set in automation mode by clicking “run auto”. The chart

on the VI’s front panel updates continuously to show the SNR from the radios as

measured at the moment as a confidence indicator of continued functionality of the

measurement system. After system is visibly running, the vehicle can then begin to drive

the test path. Data is recorded only while the GPS unit indicates that the speed of the

vehicle is greater than 20 miles per hour (32.2 km/h) and that position data is current and

correct.




Figure 6 shows a map of the path driven in Figure 7. This map is of the SNR values. Gnuplot is

used to generate this map and interpolate the non driven areas. This map is then overlaid in

Google Earth and a file such as that seen in Figure 8 is generated. This information along with

the location map give a usable map for stations.

Figure 5. Labview base_loop.vi front panel.




Figure 7. Path driven to prove system.

Figure 6. SNR interpolated data.




7. MARKETING AND COST ANALYSIS

7.1 Marketing Analysis

NPR Labs were the pioneer and essentially the only organization to analyze and map FM

analog and HD broadcasting data. This was performed as a research project funded and

sponsored by Kenwood USA, Harris Corp. [11] and CPB, but no commercial systems are

yet available. The designed system has similarities to the NPR Labs such as power and

coverage mapping, but differs in that audio SNR processing that was not available in the

system designed by NPR Labs . The FM/HD mapping system will enable radio stations

to analyze and predict their FM analog and HD area of coverage. Additional data that

will include audio quality ranges at various data points will be available. For example,

after testing a station information about the SNR of the audio will be available for a given

Figure 8. Mapped RF data.




distance range. Information obtained from final capture results and analysis will be

incomparably useful and beneficial to radio stations.

7.2 Cost Analysis

The development of FM/HD mapping system costs $28,892 as shown in Table 3, and considering a 15%

daily profit over the cost minus labor cost including two on-field engineers paid $75/hr, vehicle, gas and

maintenance, the net profit is of approximately $2064,2 per day of consulting service.

Labor charges include only cost charged towards project, and this amounts to 100 hours per person.

Actual time spent (including class lectures, reports, meetings, etc.) would be near 300 hours per group

member.

Item Planned Final Our Source

Rohde & Schwarz FSH3 Spectrum

Analyzer

8,200 8,200 -

www.tequipment.net

Consumer HD Radio 150 99 - www.smartwareetc.com

Laptop 800 800 - www.newegg.com

NI LabVIEW Professional 4,299 4,299 - www.ni.com

BR-355 GPS 43 43 - www.semsons.com

Labor ($35/hr per person) 8,400 10,500 -

USB to PS/2 adapter 10 30 10 www.walmart.com

Amplifier 4-way Splitter 20 20 20 www.radioshack.comMisc. components 320 385 185

Audio Card - 90 90 www.frys.com

Car maintenance 60 60 60

Gas/200 miles 40 40 40

NI Sound and Vibration - 3,999 - www.ni.com

RF BPF - 250 -

Magnet Mount and Antenna 77 77 77 www.cheapham.com

Total (in USD) 22,429 28,892 482

Table 3. System design estimated development cost.

Item Planned Final Our Source

Rohde & Schwarz FSH3 Spectrum

Analyzer

8,200 8,200 -

www.tequipment.net

Consumer HD Radio 150 99 - www.smartwareetc.com

Laptop 800 800 - www.newegg.com

NI LabVIEW Professional 4,299 4,299 - www.ni.com

BR-355 GPS 43 43 - www.semsons.com

Labor ($35/hr per person) 8,400 10,500 -

USB to PS/2 adapter 10 30 10 www.walmart.com

Amplifier 4-way Splitter 20 20 20 www.radioshack.comMisc. components 320 385 185

Audio Card - 90 90 www.frys.com

Car maintenance 60 60 60

Gas/200 miles 40 40 40

NI Sound and Vibration - 3,999 - www.ni.com

RF BPF - 250 -

Magnet Mount and Antenna 77 77 77 www.cheapham.com

Total (in USD) 22,429 28,892 482

Table 3. System design estimated development cost.




An estimate in finalizing future work would cost an additional 100 hours per group member. This would

include time spent fully calibrating and implementing SVT, and drive times.

8. SUMMARY AND CONCLUSIONS

Currently the system can perform the following function:

• Fully automated data capture of GPS, RF power, and SNR

• Tag data as “good” or “bad”

• Tag files as “HD on” or “HD off“

• Data can be plotted using gnuplot

• Gnuplots can be overlaid on Google Earth

Future work and improvements:

• Fully integrated SVT for SNR determination

• Full coverage drive to map stations full RF and SNR coverage

• Use NPR study to select representative radios

After building this system and running the test, the team concluded that some of the initial goals did not

work out quite as expected. The RF power decays nearly exactly as expected as can be seen from Figure

9. The SNR, however, is not quite as clean and following the SNR as expected. As can be seen from

Figure 9, the SNR values have a trend that does decay but as it does the values fan out and the deviation

significantly increases. There are many reasons that this could be the case. Some of the reasons include

multipath issues of the audio, indirect signal acquiring, low levels of signal being more susceptible to

attenuation, and the AGCs of the radio kicking in for all said RF level fluctuations. Even with these

fluctuations, SNR can still be mapped by this system. More research should be performed in the field of

SNR capture (signal processing) of this sort of audio to further increase the accuracy of this aspect of the

system. Another initial goal was to observe the interference caused by HD on the analog RF signal by

observing the difference in SNR coverage. Results to this question cannot be determined from the data.

As stated before the SNR has a trend of spanning out as the vehicle moves further and further from the

transmitter. This spanning out, or increase in standard deviation of the decaying SNR values, masks any

trend of the HD affecting the analog signal quality. Tables 4 and 5 show the mean and standard deviations

of RF and SNR one with HD on and one with HD off. Figures 9 and 10 show the audio SNR of verse

distance from the transmitter kilometer. It can be seen that the SNR does decay as distance increases as




expected. HD interference is entirely radio dependent however and could in fact be shown with future.

Radios that represent degrees of HD interference should be used to accurately determine if the

interference is relevant and in deed destructive.

SNR with HD off

0

10

20

30

40

50

60

70

80

0-5 km 5-10 km 10-15 km 15-20 km 20-25 km 25-30 km 30-35 km

Distance from tx (km)

S N

R (

d B )

Home Radio

Car Radio

Figure 9. Chart of SNR means with HD off.

HD OFF

distance RF mean RF stddev SNR1 mean SNR1 stddev SNR2 mean SNR2 stddev

0-5 km -43.34 3.07 72.58 4.25 58.98 3.20

5-10 km -48.09 4.65 70.26 5.09 57.40 4.89

10-15 km -57.10 3.47 65.36 8.16 55.83 4.97

15-20 km -61.44 2.72 64.66 6.03 54.20 4.61

20-25 km -64.17 3.04 58.49 9.40 52.05 5.66

25-30 km -69.55 2.45 49.10 9.25 47.68 6.18

30-35 km -70.28 2.49 49.34 10.33 48.69 5.44

Table 4. Mean and standard deviation values with HD off.




The RF power levels, shown in Figures 11 and 12, show that the power decays in an expected logarithmic

fashion. Also looking at Tables 4 and 5 the standard deviations of the power can be viewed. These

standard deviations show that measurements of the RF power are very precise. Through lab testing

performed on the system as previously described, a calibration value was determined. Use of this

calibration value ensures accuracy. The accuracy as well as the precision of the RF values ensure valid

and good data to be used in mapping analog coverage.

SNR with HD on

0

10

20

30

40

50

60

70

80

0-5 km 5-10 km 10-15 km 15-20 km 20-25 km 25-30 km 30-35 km

Distance from Tx (km)

S N R ( d B )

Home Radio

Car Radio

Figure 10. Chart of SNR means with HD on.

HD ON

distance RF mean RF stddev SNR1 mean SNR1 stddev SNR2 mean SNR2 stddev

0-5 km -38.98 6.32 73.05 3.04 58.93 3.84

5-10 km -48.39 4.85 69.77 6.16 57.68 4.18

10-15 km -57.08 3.55 65.95 7.04 55.51 4.70

15-20 km -61.53 2.75 63.63 8.36 54.77 4.35

20-25 km -64.30 3.09 60.06 9.31 52.05 5.15

25-30 km -69.68 1.93 50.83 9.70 47.30 5.16

30-35 km -70.71 1.44 49.11 9.02 45.51 5.74

Table 5. Mean and standard deviation values with HD on.




Some items have been changed throughout the course of this design. The design team decided that use of

software that was already available (gnuplot) for interpolating and mapping this sort of data was more

RF Power vs. Distance from Transmitter (HD on)

y = -10.462Ln(x) + 2.8163

R2 = 0.843

-80

-70

-60

-50

-40

-30

-20

-10

0

10

0.2 4.1 7.7 12.0 15.5 19.1 22.7 26.2 32.5Distance (km)

R F P o w e r ( d B m )

Figure 11. Power vs Distance from Transmitter with HD on.

RF Power vs Distance (HD off)

y = -8.1059Ln(x) - 12.741

R2 = 0.7562

-80

-70

-60

-50

-40

-30

-20

-10

0

2.323913059 5.53888115 9.538955664 13.33316447 17.62751722 21.09460785 24.41676021 29.19922287

Distance (km)

R F P o w e r ( d B )

Figure 12. Power vs Distance from transmitter with HD off.




practical than reinventing the wheel and making this same software with Matlab. It was initially the

design team’s goal to turn on and off HD signal rapidly throughout the driving path. However, due to

foresight in issues that may cause at the transmitter, separate driving runs were made with HD and with it

off. The group had intended in initially continuously sampling the audio and calculating the SNR in the

back ground. This would allow the SNR to just be sampled from this background software. However,

the driver combinations of Labview, directX, and the audio cards would not allow this. In order to read

from a card a connection to it must be opened. This connection must first be closed before another sound

card can be connected to. Due to this issue the software needed to be modified to sample half second

captures of audio, and alternate between cards.

Additional work that would be highly beneficial to the design is to use a more accurate interpolation of

the data which resides outside the driven path. This would have to accurately model the loss of this

particular RF signal frequency. The interpolation being used is more of a generic interpolation.




9. REFERENCES

[1] Theodric Technologies LLC, “Radio-Locator,” 2009, http://www.radio-locator.com/.

[2] J. Kean, “An Improved Coverage Prediction Method for HD Radio,” NAB Broadcast Engineering

Conference Proceedings, 2008, pp.137-145.

[3] “Self-Noise,” Ham-Radio.com, 2008, http://www.ham-radio.com/k6sti/hdrsn.htm.

[4] NPR Labs , Final Report to the Corporation for Public Broadcasting Digital Radio Coverage &

Interference Analysis (DRCIA) Research Project, National Public Radio, Washington, DC,

2008.

[5] J. Kean, NPR Labs (private communications), January 27, 2009.

[6] Sony XDR-F1HD, ham-radio.com, February 2009, http://ham-radio.com/k6sti/xdr-f1hd.htm.

[7] Rohde and Schwarz FSH3 Spectrum Analyzer Manual, http://www2.rohde-

schwarz.com/R&S_FSH3_manual.

[8] J. Kean ([email protected]), “RE: Analog FM Coverage/Interference Measurement and

Prediction under HD Radio for Georgia Tech academic project,” Email, February 1, 2009.

[9] J. Kean, “HD Radio Coverage Measurement and Prediction,” NAB Broadcast Engineering

Conference Proceedings, 2006, pp.353-359.

[10] “In-band/on-channel Digital Radio Broadcasting Standard,” NAB, April, 2008,

http://www.nrscstandards.org/SG/NRSC-5-B.asp.

[11] “Updating IEEE 185-1975,” ham-radio.com, February 1, 2009, http://www.ham-

radio.com/k6sti/ieee.htm.

[12] “A Study of Co-Channel and Adjacent-Channel Interference Immunities of Subsidiary

Communications Authorization (SCA) FM Broadcast Receivers, ” FCC, December 1 2000,

http://www.fcc.gov/oet/info/documents/reports/trb_99-3_sca.doc.




[13] IEEE Standards, http://standards.ieee.org/db/status/status.txt.

[14] J. Lawhorn and J. Broo, “Tomorrow RadioSM

Project Announces Stellar Test Results, Declares

Victory in Multi-Channel HD RadioSM

Research,” Nation Public Radio, January 9, 2004,

http://www.npr.org/about/press/040109.tomorrowradio.html.

[15] “Calculate distance, bearing and more between two Latitude/Longitude points”http://www.movable-type.co.uk/scripts/latlong.html



FM/HD Radio Mapping (ECE4007)

APPENDIX A: Planned Project Gantt Chart



FM/FM Radio Mapping (ECE4007 L01)

APPENDIX B: Actual Project Gantt Chart



Final Report

Documents