Comparison of Radio Propagation Modeling Software

by

Recently discussions regarding the use of various software in the study of coordination modeling have happened.  All claim to implement the same models along with their own variations or empirical derived models.  The model we are most interested in is known as Longley–Rice, or the irregular terrain model (ITM).  This model is optimized to cover 20 to 20,000 MHz for predicting signal strength over real terrain when used with a digital elevation model (DEM).

The ITM model in area propagation mode essentially breaks up the underlying map into tiles of a given resolution, and taking account the DEM heights, computes the signal level in a tile and then onto the adjacent tiles and so on using the now attenuated signals from the source area.  The original code was written in FORTRAN, converted to C++ and has been re-implemented a number of times to speed it up, improve it and add a given vendors “secret sauce”.  The original docs and code may be found on the ITS website.  Keeping in mind the ITM model was used to layout FM and most importantly the VHF/UHF Television spectrum in the 1960’s it’s proven to be a reliable model.  Land Mobile radio has adopted the same basic concepts in needing to predict interference between co and adjacent channel users, albeit with different criteria than TV broadcasters.

As amateurs push to coordinate frequency use in their spectrum, it’s imperative we understand our tools and what software we may use to predict coverage.  This is more important we have repeatable standards which anyone can model and check to keep our coordination bodies honest as well.  FASMA has standardized on SignalServer which is a multithreaded version of SPLAT!, both implement the C++ version of the ITM model.  Also common in amateur circles is the use of RadioMobile, a non-free application for windows operating systems.

Commercially there are many packages used for radio modeling, and two of the most popular are ComStudy and Pathloss.  ComStudy is the de-facto standard used by the LMR coordinators approved by the FCC and implements it’s own proprietary version of the ITM model and ships with an additional land cover database.  Pathloss 5 is mostly known in the Part 101 and microwave Point to Point (PtP) modeling space, but has an ITM mode, and unlike most software, describes the algorithm implemented in great detail.  As Pathloss doesn’t provide source code, we cannot be sure how faithfully they have implemented the ITM model.

The DEM model data used is the 3 arc second resolution data from the Shuttle Radar Topography Mission (SRTM).  This is a digital model of the earth’s elevation in 90×90 meter tiles as a height.  This works well for most areas, but there is higher resolution data 1 arc second and even 1/3rd arc-second available for certain areas of the earth.  Ultra-high resolution DEM data has national defense implications, and is generally not available.  The SRTM data typically includes the tops of trees in forests, which can present a problem if a tower is in a forest.  We can say the height above ground is 100 meters, but if the adjacent trees are 20 meters average height, the model AGL should adjusted to 80m (100-20 tree height), or we will be modeling the system 20m too high.  This same phenomena can present in dense cities as an urban forest area.

 

Modeling Criteria

In comparison we have modeled a test repeater with an omni directional antenna as follows:
<font=monospace>

Latitude:        42.33
Longitude:      -87.91
Frequency:      446 MHz
Antenna Height Above Ground: 97.5 meters
ERP:            707 watts
Signal Criteria: 39 dBuV/m (dBu)
Reliability and confidence criteria - (50, 50)

Signal Server

SignalServer was invoked with the following command:

./signal-server-base.sh -lat 42.328889 -lon -87.912500 -txh 97.54 -f 445 -erp 707 -rt 39 -conf 50 -o blueblue | ./genkmz.sh

This produced the plot below with a radius (based over the water) of 62km.

The reason for basing the plot over the water for radius is it should be the same height and signal strength in all models, as there is nothing there.

This plot took about 150 seconds on a quad core computer.

The Keyhole Markup Language file is here, and a version of the same plot, but red is here.  The red color version is useful for comparing and contrasting with the other results in the various shades of blue.

RadioMobile

RadioMobile is interesting as its non-Free Software, but doesn’t appear to have a commercial use keeping it from being Free Software.  Due to the GUI, it’s popular with a number of amateurs.  The configuration for generating plots is quit complex and poorly documented to get a 1:1 ITM model from it.  RadioMobile is not multi-threaded, in fact the entire GUI locks up while rendering, however a render is completed on a Dell D830 (a laptop circa 2006) in about 60 seconds.  From this we must assume some “speedups” or shortcuts in the code are taken when compared to the ITM reference code.

We followed the FASMA reference guide on modeling in RadioMobile for this plot.

This modeled with an 82km radius over water, and seemed to have greatly overstated coverage compared to the other models.

 

Pathloss 5.1

Pathloss has no easy way to export the configuration for a model.  We had this modeled by a trained professional user of the software.  Pathloss in area mode is exceptional slow, there below took about 12 hours to render.  We found no multi-threading taking place, and like RadioMobile the GUI locked up during this. Based on this, we can assume the code very close to the ITM reference.

The radius here was 65km and compared very closely with the SignalServer model.

The Keyhole Markup file of this is here.

ComStudy 2.2

ComStudy was run by an experienced engineer who has worked LMR across the VHF and UHF business spectrum.  Again there is no easy open way to export the study inputs for ComStudy, but the below files were modeled.  We didn’t have exact measurements of time for the generation of the below plots.

This is a 42km radius and has the smallest area of any model we studied.  The Keyhole Markup file is here.

In ComStudy it also has proprietary land-cover databases which ship with it.  Much like a DEM file, this identifies the type of cover in a given area, which changes the radio propagation characteristics of that area.  We have a model with the LCF enabled and it’s much smaller as well.

Comparisons

SignalServer with PL5

This was the closest comparison, and based on the PathLoss docs, we believe PathLoss has implemented a truthful version ITM algorithm.  This would confirm the SignalServer/SPLAT! model as being correct as well, and we belive this confirms it is the case.

In the below, SignalServer is in red and the PathLoss model is in blue.

The slight variation over water (3km) and the open arc in the PathLoss model would indicate it has two ray turned on by default.  This only comes into effect over large flat reflective areas such as water.  Over land the models align 99%.

Looking at the fringe coverage we see the same tiles, just slightly offset.  This may be due to the datum used internally in PathLoss not being corrected to the WGS84 datum of our coordinates.  This would further confirm the agreement both SignalServer and PathLoss have a true ITM model implementation.

SignalServer with ComStudy

ComStudy claims to implement the same ITM model, however it doesn’t appear to be the same.  We’re not certain why this is, perhaps it’s due to some of the speedup code or just a general desire to mix the ITM with an empirical model as they can call it the ComStudy model.  ComStudy also may take into account the LCF data for land cover and we have presented that below as well.

ComStudy with a 42km vs 62km radius in SignalServer.  ComStudy is the lighter blue color.

We’re not certain why it’s different, but at least in the areas that are covered ComStudy and SignalServer seem to align; looking to the north west above McHenry on the map we can see gaps between the airports and they align in both models.  This may be due to the ComStudy using a different height, correcting for antenna gain, or another issue.  What we do see is that the models do appear to use the same DEM data to determining a valley or non-coverage area.

Zoomed area showing the alignment between areas in ComStudy and SignalServer.

This is the overlay of the ComStudy on Signal Server with the land cover data.  We can see this LCF data has impacted coverage to the north and south over land.  This does give some value to the theory ComStudy may be adjusting the DEM data for forested regions as suspect in the other non-LCF model, and explain some of the descripency between it and SingalServer/Pathloss.

RadioMobile vs. SignalServer

Perhaps in the most interesting study, we found RadioMobile to vastly overstate coverage compared to the other three programs.  This may be due to the primary user base of amateurs who are wowed by the enhanced coverage predictions, thus it encourages many users.  It may also be the “speed ups” done to allow RadioMobile to so fast.  Indeed RadioMobile was the fastest to produce a model of any of the programs tested.

In the above we can see the overstated RadioMobile plot in blue and the SignalServer plot in red.

Here we can see the light blue of the ComStudy in the center, red of SignalServer and then the blue of RadioMobile.

We’re not certian why this is with RadioMobile, but have found ~15 dB of reduction of ERP will bring the model to align more with SignalServer and Pathloss.  This was not tested in other areas and we caution anyone using RadioMobile (or any non-Free Software) to not trust the results.  We cannot recommend using RadioMobile for any real modeling of amateur repeaters due to these issues.

 

Conclusions

The intent of this research was to compare and validate SignalServer’s ITM model with other popular radio propagation software models.  We found a great agreement between SignalServer and Pathloss, confirming a faithful implementation of the ITM model.  ComStudy was shown to be slightly less generous with coverage, but where the models did overlap, we found they agreed.

RadioMobile was the outlier, and was found to be greatly overstating a coverage area, and no alignment with even common areas of coverage in SignalServer.  Based on this the author cannot suggest the use of RadioMobile.

 

References

SignalServer 39 dBu model Keyhole Markup File
Signal Server 39dBu red KMZ 
ComStudy 2.2 with no landcover
Pathloss 5.0 Study
RadioMobile
ComStudy 2.2 with landcover

SRTM mission overview on wikipedia
Shuttle Radar Topography Mission on the JPL homepage

ITM Model homepage on the Institute for Telecommunications Sciences

FASMA Library

SignalServer – Github fork by N9OZB
SPLAT! – Homepage
RadioMobile – Homepage
ComStudy – Homepage
Pathloss – Homepage

Join FASMA at HamCation 2018

by

FASMA will be present at Hamcation 2018 with a booth in the main exhibit hall.  Please come by and speak with the board; feel free to ask any questions you have.

This will be a first for either the FRC/FASMA to have a public outreach presence at the premier hamfest in the southeast.  This is a continuation of the new FRC, now FASMA, boards promise to it’s membership.

 

Please come out and say “hi”.

FASMA will be in both #225 in main hall

Central Florida Fairgrounds and Expo Park
4603 West Colonial Drive
Orlando, Florida 32808

February 9, 2018 | 9AM to 5PM
February 10, 2018 | 9AM to 5PM
February 11, 2018 | 9AM to 2PM

Itinerant Repeater Policy

by

Summary

FASMA receives many requests for low power, limited coverage repeaters for experimental, mobile, temporary “emergency” systems, neighborhood CERT teams and for just plain “fun”.  The FCC refers to such temporary operation as “Itinerant” users. Due to the number of repeaters on the air, it is difficult to find available frequencies to manage this need, due to protection contours of existing machines.  In some areas, there are no available frequencies to assign within the repeater band plan spectrum.

Our technical committee proposes the following plan to accommodate these needs in a simple, easy to understand manner.  We hope this policy will promote more experimentation in repeater systems, add to the body of experienced repeater operators within FASMA’s members and serve the public interest in times of emergency with a planned deployment guide and coordination of, portable, emergency and mobile (such as those used at Hamfests for group communications) repeaters that can add to the general usefulness of Amateur Radio systems.  Plus, we hope it allows everyone to have more fun messing around with Ham Radio gear!

Due to existing band plans and congestion, this proposal only identifies suitable frequencies on the  70 centimeter band.  Research has shown the two meter band is too congested and no standard frequency pairs are available statewide.  220/902/1.2 GHz would likely support such itinerant operation and is best the subject of future study.

There is now a plethora of inexpensive 70cm radios now on the market that can service users of Itinerant repeaters. Itinerant repeater pairs are available on a non-exclusive, first come, first served on line application environment for interested parties.

The application process is designed to help applicants learn and understand the basics of the RF parameters needed to plan a working system.  FASMA staff is available, on a volunteer, as time permits, basis, to consult with applicant’s system details, however, applicants will be required to “do the math” on their application.  Through this process, FASMA hopes to encourage learning and broaden the understanding of the details of repeater planning, service contour definition, interference potential and other facets of repeater systems and add to the available technical talent of Florida amateur radio operators as a whole.

Itinerant Repeater Proposal

These repeaters are the Amateur Radio analog to the new FCC licensed Low Power FM stations now being shoehorned into the FM broadcast band.  They are intended for short range, local, limited coverage.

Itinerant repeaters may use any appropriate modulation formats, either digital or analog. The only limitation to the modulation parameters is that any modulation used must not exceed 20 kHz for wideband or 11.7 kHz for narrowband pairs. Multiple modulation formats may be used on a single machine at any time, even concurrently.  Repeaters may be designated as analog, digital or mixed format by operators.

Itinerant repeaters can be designated by as “closed” systems for their users only. However, in keeping with FCC regulations for the Amateur Service, no form of encryption on either digital streams or analog modulation may be used.

Registration of Itinerant Repeaters is predicated under the parameters illustrated in this policy.

As in any FASMA coordination, any deviation or relocation greater than 8km distance, 5m in height or 10% transmitter power will require re-coordination.

Itinerant repeater sub band frequency pairs:

Wideband Narrowband
Pair Output Pair Output
1 441.8250 1A 441.8250
1B 441.8375
2 441.8500 2A 441.8500
2B 441.8625
3 441.8750 3A 441.8750
3B 441.8875
4 441.9000 4A 441.9000
4B 441.9125
*5 441.9250 *5A 441.9250
*5B 441.9375
6 441.9500
7 441.9625
8 441.9750
9 441.9875

All pairs MUST use a standard offset.  i.e. Receive Frequency = Output Frequency + 5.000 MHz

There are 5 pairs available for wideband (25 kHz) channel use, with 14 pairs for narrowband (12.5 kHz) only use.
Pairs 6-9 are narrow band only, with 1A-4B local option only if 6-9 are in use.

Pair 5 Wideband and 5A/B Narrowband is intended to be a truly itinerant use system.

  • The intent of this channel is for hamfest or other mobile repeaters.
  • No permanent install on this frequency is permitted.
  • Users MAY temporally deploy a system on these frequencies for no longer than 36 hours. During an emergency, this rule is waived for the duration of the emergency.
  • No registration is required for this channel; however, it is encouraged.
  • Narrowband users are encouraged to use pair 5B before 5A

Frequency Agile repeaters

This plan was engineered taking into account the performance of a typical “flat pack” duplexer.  In a UHF duplexer of this type it is common to have 100-150 kHz of usable notch depth bandwidth.  This allows a wideband repeater to use a duplexer tuned to 441.8750/446.8750 on channels 1-5 with acceptable performance.  A narrowband repeater could use a duplexer tuned to 441.9625/446.9625 on channels 5-9 with acceptable performance.  Please note this is simply an explanation for the frequency layout, any duplexer should be individually tested by a competent technician prior to use and FASMA assumes no liability caused by such use.

CTCSS Required

All pairs require CTCSS/DCS/NAC/Color Code/etc. (hereafter CTCSS) transmit and receive by all users at all times.  All users of these pairs MUST use CTCSS transmit and receive in all user radios.  Proposed CTCSS codes must be coordinated via consultation with FASMA coordinators.

System Parameters

Operational system parameters for Itinerant Repeaters must adhere to the specifications described below.  Any deviation must follow normal coordination procedures.

  • Maximum service contour for the purposes of planning co-channel assignments is 20 km (12.5mi) (assuming the maximum 15.3m HAAT (50 ft. Height Above Average Terrain).
  • No protection of service coverage past the above defined service contour description
  • Maximum Transmitter Power Output (TPO): No more than 40 watts into antenna
  • Maximum total allowed Effective Isotropic Radiated Power (ERP) in the system: 110w
  • Maximum allowable antenna gain: 6 dBd.
    This is in the range of the common Diamond X50 and its clones, which is one of the typical antennas for these applications. For better null fill “close up”, such as for CERT applications within a neighborhood, a simple unity gain ¼ ground plane is recommended due to the broader radiation pattern of these antennas.
  • Maximum antenna height to center of radiation: 15.3m (50 ft.)
    A 43-foot tower could wear a Diamond X50 at its apex to qualify. This is technically within the reach of people wanting a local repeater and enabling their deployment in most normal city/suburban areas in Florida.
  • Height above Average terrain must not exceed 60 feet out 25 miles.
    This typical of a suburban location in Florida terrain. This precludes someone living at the base of Bok Tower from applying for an Itinerant repeater system – The HAAT for this location is out of range.

Registration Requirements

Prior registration would be required and these repeaters will be listed in FASMA based repeater coordination lists and reports.  All FASMA requires for this repeater service coordination from the applicant are the following system parameters:

  • Ownership Group/Individual address, contact telephone number and email address
  • Callsign of both the trustee/owner and the repeater itself (if different)
  • GPS coordinates of the location (degrees and minutes to two decimal places)
  • Repeater equipment make, model and maximum power output
  • Feedline loss figure
  • Antenna make and model with gain figure
  • Effective Radiated Power
  • CTCSS tone choice
  • Emission designator (preferable) and/or modulation type (WBFM, NBFM, D-Star, DMR, etc.)

Any proposed change in facilities to increase repeater range would require re-coordination to a “main band” frequency, if such a frequency is available for the proposed site.  In cases where, for a given metropolitan area, a “main band” frequency becomes available, Itinerant repeater operators can request a facilities upgrade to that frequency.  Applications will be processed in accordance with normal coordination procedure.

Interference protection

Itinerant repeaters must accept interference from adjacent channel systems close to their coordinated pairs.  Itinerant repeaters operate on their frequency on a “secondary” basis and must accept any interference from other co-channel or adjacent channel users.  Itinerant repeaters shall not continuously utilize a channel via broadcast or other transmissions. Users shall restrict all transmissions to the minimum practical transmission time and must employ an efficient operating procedure designed to maximize the utilization of the spectrum.  This is analogous to the FCC FB2 station class code in the land mobile spectrum.

The only interference abatement available to Itinerant repeater operators is either antenna terrain shielding or CTCSS tone management.  Only one CTCSS tone shall be used for access to Itinerant repeaters.  Multiple CTCSS tones are not allowed. Changes to CTCSS tone must updated in the repeater’s registration in the FASMA database.

Build Out Notification

Itinerant repeater application and grant will be a fully automated service via the FASMA website.  System owners/trustee will have a maximum 60 days to report operational status (build out).  If operational status is not reported to FASMA after 60 days, the repeater registration will be revoked and the frequency will be available for reassignment.  If any Itinerant system is found to be off the air, the repeater registration will be revoked.

If the operator ceases operation for any reason, the operator shall notify FASMA.

Notes

This is a proposed policy was open for public comment before adoption.   Our policy making is candid process; we continually seek input and new ideas from the community.   We encourage any amateurs wishing to effect change in the operation of coordination in Florida to submit a proposal or comment on existing proposals. We request all comments to rt@fasma.org.

This policy has been enforce since 2018 without issue.  It is adopted for use of all amateurs in Florida.

End of FRC Year Members Meeting

by

This will be the final FRC meeting and all members are invited to attend online.

We’ll be considering the dissolution of the FRC and transfer of FRC assets to FASMA under the following agreement.

 

End Of FRC year meeting
Saturday, December 30, 2017
7:00 pm  |  Eastern Standard Time (New York, GMT-05:00)  |  1 hr
Meeting number (access code): 627 678 048
Meeting password: TvAmQm3w

 
Add to Calendar
When it’s time, join the meeting.
Join from a video system or application
Dial 627678048@frc.my.webex.com
You can also dial 173.243.2.68 and enter your meeting number.
Join by phone
+1-510-338-9438 USA Toll
Global call-in numbers

Minutes of the Emergency FRC Meeting

by

The below are the notes of the emergency members meeting convened on July 6, 2017.

At this meeting the board was censured and removed for cause.  A new interim board of directors was elected and the illegal bylaw changes instituted by the prior board were declared null and void.

An amendment to the Articles of Incorporation was approved to prevent this from happening again.

PDF Version of Minutes

Youtube Video Link

Florida Repeater Council members reject corruption.

by

The FRC has voted to censure and remove incumbent board members during an emergency members meeting.

St Petersburg, July 6, 2017

Summary:

The Florida Repeater Council, Inc. (FRC) membership has voted at an emergency meeting to censure and remove the board of directors and call for a rewriting of bylaws. The ongoing corruption of the board extends back over ten years and recently was exposed by members at the annual meeting in Orlando. During the Orlando meeting the Board voted to expel all members from the FRC and re-elect itself, in violation of the Articles of Incorporation and state law. Bylaw changes giving the president dictatorial like powers were approved and not disclosed until June 28, 2017.

The rebuke of these now former directors by the members and subsequent election of an interim board shows the radio amateurs of Florida have taken the first steps to correct this mismanagement and check the arrogance which permeated the FRC.

Background on today’s actions:

The background of mismanagement by the former board extends over the past decade. Over the past three year members have questioned the board of directors and made formal requests for information from the FRC. Each request for information was met with contempt and even reported to law enforcement as a means to intimidate the membership.

The FRC Reform caucus (http://frcreform.org) was started in 2015 by Bryan Fields W9CR, to shine a light on these matters. In support of these efforts, FRC members assigned proxies to Mr. Fields to be used during the annual members meeting in Orlando February 2017. During this meeting the board refused to recognize Mr. Fields supporters and claimed it was a closed non-members meeting, though it was announced as the annual members meeting. Board member James Deuel N0XIA, further tried to intimidate the assembled members by calling police over a supposed fire code issue a half hour after taking count.

At this Orlando meeting all members were removed from the FRC by the board in violation of the articles of incorporation. The proposed bylaw changes were not presented for review and a request by Mr. Fields to inspect them at the meeting was denied. These new bylaws were finally released on June 28, 2017. These changes gave the president dictatorial like powers and the board became a rubber stamp to presidential fiat.

An emergency members meeting was called July 6, 2017 by Director Mike Gonzalez N2FSU, to address these issues as he was shocked by the actions of the board. At this meeting the only directors who attended either online or in person were President Glenn (Mike) Fletcher NI4M and Secretary Dana Rodakis K4LK.

During the course of this the meeting the membership presented its charges against the directors and board. President Fletcher and Secretary Rodakis were called to explain the mismanagement and their conduct to the members. The only mitigating argument presented was one of “the board has eliminated membership, you are no longer members of the FRC”.

Our members saw through this and voted to censure the board members, remove them from office and call a constitutional convention to rewrite the bylaws. A new board was voted in and tasked with auditing the FRC and correcting the mismanagement. Further it was resolved during this interim time no FRC funds would be spent without unanimous consent of the board and then only for legally required matters.

Censured and removed directors:

Glenn (Mike) Fletcher NI4M
Dana Rodakis K4LK
Ralph Betts W4ORL
Norman Scholer K4GFD
Steve Lowman N4SGL
Douglas Steward KI4NAD
James Deuel N0XIA
Mark Filla N4DES
Joaquin Pidal KZ4TI
Barry Isbelle N2DB
Chuck Lavender W4CLL
Frank Butler W4RH

New FRC board members:

Chairman:
Jon Pearl W4ABC

Directors:
Ryan Owens KJ4SHL
Mike Gonzalez N2FSU
Lu Romero W4LT
Bryan Fields W9CR

About the Florida Repeater Council:

The Florida Repeater Council, Inc. is the recognized coordination body for amateur radio repeaters in Florida. In addition the FRC is responsible for amateur band planing above 50 MHz. The FRC is a non-for-profit members organization and seeks to represent all users of amateur radio spectrum above 50 MHz in the state.

Contact information:

For more information you may contact Chairman Jon Pearl W4ABC or Director Bryan Fields W9CR at 727-409-1194 as the main FRC website (http://florida-repeaters.org) has yet to be migrated.

Members are encouraged to join the FRC Facebook group for more information and updates: https://www.facebook.com/groups/FloridaRepeaters/