[Vk1tac] RE:: FWD: Ginini 2 metre services

[Neil Pickford]

-------- Original Message --------
Subject: Fw: FWD: Ginini 2 metre services
Date: Fri, 17 Jan 2003 14:46:09 +1100
From: "Mr Gilbert Hughes" <ghughes3@bigpond.com>
To: "Richard Elliott" <richard.elliott@dfat.gov.au>,"Paul Elliot"
<pkbelliot@bigpond.com>, "Paul Bell" <paul.bell@afp.gov.au>,"Neil
Pickford" <neilp@goldweb.com.au>,"Keith Malcolm"
<kmalcolm@tpg.com.au>,"John Woolner" <johnwoolner@auslig.gov.au>,"Mike
Dower" <vk1eng@tpg.com.au>

2nd most recent on last topic

Regards
Gilbert Hughes
mailto:ghughes3@bigpond.com

----- Original Message -----
From: "Owen Duffy" <owen@oda.com.au>
To: "Mr Gilbert Hughes" <ghughes3@bigpond.com>
Cc: "Paul Bell" <paul.bell@afp.gov.au>
Sent: Thursday, 16 January 2003 11:10 AM
Subject: RE: FWD: Ginini 2 metre services


> >===== Original Message From "Mr Gilbert Hughes"
<ghughes3@bigpond.com> =====
> >Hello Owen
> >
> >Can you have a look at Peter W's (he is an active APRS person)
comments
> >and see if you are able to agree.  ???
> >
> >Regards
> >Gilbert Hughes
> >mailto:ghughes3@bigpond.com
> >
> >----- Original Message -----
> >From: "Peter Westerhof" <Peter.Westerhof@aspect.com.au>
> >To: "'Mr Gilbert Hughes'" <ghughes3@bigpond.com>
> >Sent: Tuesday, 14 January 2003 9:41 AM
> >Subject: RE: Ginini 2 metre services
> >
> >
> >> G'day Gil,
> >>
> >>  well what can I say, he may be right in theory, but in paractise
it
> >all
> >> works fine, 10 mobiles and ten stations locally with the addition
of
> >gated
> >> traffic have no adverse effect on our current setup...  As I said
> >before we
> >> have trialed a  WIDE on Ginnini and had alot of success..  Our plan
> >would be
> >> to revert VK1RBM-1 to a RELAY only digi and make Ginnini the full
> >WIDE.  I
> >> don't wnat to start a debate or argument about another WIDE etc but
I
> >think
> >> that since packet activity is just about dead in the ACT let's use
the
> >good
> >> old Aussie attitude and GIVE IT A GO (if only as a trial)....
> >>
> >>
> >> hope it helps.... APRS is becoming more popular, since beginning we
> >have
> >> gone from 1 to 18 plus 8 mobiles(some of wich are our older
members.
> >>
> >>
> >> Enough from me.....
> >>
> >> Cheers
> >>
> >> Peter
> >> VK1NPW
> >>
> >> -----Original Message-----
> >> From: Mr Gilbert Hughes [mailto:ghughes3@bigpond.com]
> >> Sent: Monday, 13 January 2003 6:32 PM
> >> To: Peter Westerhof
> >> Subject: Fw: Ginini 2 metre services
> >>
> >>
> >> Hi Peter
> >>
> >> Got your e mail and message.
> >>
> >> Please see Owen's comments below. In summary I believe you say a
wide
> >at
> >> Ginini would work provided there were fewer other wides within
range,
> >> plus packet messages can also be supported by this system.
> >>
> >> Owen says it wont work.
>
> Gil, you are putting words into my mouth. I did not say that it "won't
work",
> but raised the issue that bandwidth is limited, and a WIDE with very
wide
> coverage might capture more packet demand than the channel can deliver
with
> reasonable probability of success, and might not provide quality of
coverage
> around town as good as the Black Mountain WIDE.
>
> Further, that multiple WIDEs and RELAYs don't create useful source
traffic,
> but add to the utilisation of the radio channel by repeating packets
in the
> hope of improving the probability of delivery, however increasing
channel
> utilisation adversely influences the probability of packet delivery.
>
> RELAYs and WIDEs appear to be deployed in an ad-hoc fashion. For
example, when
> I received this email a couple of nights back, I watched the channel
here in
> Canberra for an hour and observed only one mobile active (although
stationery)
> transmitting position reports every 50 seconds which I received
reliably at a
> distance of 24Km. The position reports were reliably repeated by three
RELAY
> stations that were located within 4Km of the mobile, and the WIDE on
Black
> Mountain. In all, I received 5 copies of the position reports. I don't
> understand the value of three RELAYs (over one or none) in such close
> proximity.
>
> As for the Igate traffic impact on mobile position reports, I make the
> observation that I can hear the Igate station for up to about 4Km from
its
> location, so its service area seems about 12 sq Km. I would estimate
that the
> area of Canberra / Queenbeyan commonly traversed by mobiles is about
400 sq
> Km. So, I estimate that the Igate station being heard over 3% of the
area, is
> likely to interfere with 97% of mobile position reports that occur
while it is
> transmitting, and since it is transmitting for about 16% of the
channel time
> that is apparently available to those mobiles, about 15% of mobile
position
> reports are likely to be clobbered by the 'hidden' Igate station (at
the
> current traffic levels).
>
> I have developed some log analysis tools and analytical traffic models
if
> people are interested in contributing to their refinement and
validation...
> but I think the "suck it and see" approach will be preferred.
>
>
>
> Owen
>
> >>
> >> Feel free to talk to Owen and hopefully get an agreed position that
I
> >> can take to the Committee and members for a decision!
> >>
> >>
> >> Regards
> >> Gilbert Hughes
> >> mailto:ghughes3@bigpond.com
> >>
> >> ----- Original Message -----
> >> From: "Owen Duffy" <owen@oda.com.au>
> >> To: "Mr Gilbert Hughes" <ghughes3@bigpond.com>
> >> Cc: <paul.bell@afp.gov.au>
> >> Sent: Monday, 13 January 2003 11:30 AM
> >> Subject: RE: Ginini 2 metre services
> >>
> >>
> >> > Gil, some thoughts below instream:
> >> >
> >> > >===== Original Message From "Mr Gilbert Hughes"
> >> <ghughes3@bigpond.com> =====
> >> > >(Could one of you please forward this to Karl Makin - I do not
have
> >> his
> >> > >address)
> >> > >
> >> > >Following the installation of the new tower there is a much
greater
> >> > >isolation (about 70 dB compared to 30 odd dB on the old tower)
> >> between
> >> > >the new Tx and Rx antennas.
> >> > >
> >> > >In order to save floor space (and cavity resonators) we can
re-work
> >> the
> >> > >2 metre multi-coupling (currently 14 cavities) to make it easier
to
> >> fit
> >> > >the Ginini 70 cm hub of the linked repeater project, and the 6
> >metre
> >> > >repeater diplexer (modified 70 MHz cavities) and the 2 mtr and
70
> >cm
> >> Rx
> >> > >voting equipment.
> >> > >
> >> > >This has causes us to think about what 2 mtr services we really
> >want
> >> to
> >> > >have on the site.
> >> > >
> >> > >Some thoughts are:
> >> > >
> >> > >1. Only two 2mtr services can easily be engineered eg voice and
> >> packet.
> >> > >
> >> > >2.  Does the current usage of the 4800 baud packet justify
> >continuing
> >> > >the service - (or move it to Bulls head??)
> >> > >
> >> > >3. An APRS wide has been suggested, however we understand that a
> >wide
> >> > >coverage store and forward service could do more harm than good
if
> >> APRS
> >> > >continues to gain in popularity and more 'wides' are installed.
> >> > >
> >> > >4. Is there any advantage to packet users of making the site a
1200
> >> baud
> >> > >hub? (we understand that the answer is no for a number of
reasons.)
> >> > >
> >> > >Does the 2 mtr packet currently  get much use now?
> >> > >
> >> > >Is the usage likely to decline even further?
> >> > >
> >> >
> >> > I ran up a packet station a couple of years ago, and was
interested
> >in
> >> > determining why throughput was so bad on the 2m 1200Baud and 4800
> >Baud
> >> > channels, and exploring ways to improve the situation. The
following
> >
> >> comments
> >> > on packet relate to observations at that time.
> >> >
> >> > The 4800Baud channel was quite congested, which was the prime
reason
> >> for poor
> >> > throughput. The mode of operation of the repeater may have
> >exacerbated
> >> that.
> >> > In a scenario where most stations using the repeater could not
hear
> >> each
> >> > other, there is a high probability of collisions which wastes
> >> bandwidth. This
> >> > problem becomes worse at high utilisation levels where collisions
> >> result in a
> >> > snowball effect.
> >> >
> >> > It is interesting to drill down on what constitutes the traffic.
My
> >> > observations were that it is mainly machines talking to machines
in
> >a
> >> batch
> >> > oriented mode (eg node list transfers, network management, large
> >file
> >> > transfers often of material that is shareware and one might argue
is
> >> has a
> >> > pecuniary aspect and should not be transmitted by amateur
networks).
> >> The
> >> > traffic has grown to a point where the congestion renders the
medium
> >> generally
> >> > unsuited to intereactive traffic, and even email (where in my
> >> experience, it
> >> > would take 1 to 3 days to get a message to a mailbox in the
> >> Netherlands).
> >> >
> >> > Interestingly, the example I quote above was one where I was
> >> collaborating
> >> > with a person in the Netherlands re changes to fix bugs in some
> >packet
> >> > software... but the packet network could not provide a reasonable
> >> > communications service to support its own development - we had to
> >move
> >> to
> >> > Internet email.
> >> >
> >> > My recollection is that Dave Thomson, in one of his fits of
> >> frustration with
> >> > batch transfers spoiling their digi, tore it down.
> >> >
> >> > I formed the view that packet activity was mostly about
establishing
> >> networks
> >> > without any clear objectives of what service the network could
> >> provide, and
> >> > that network overheads and mindless use rendered the networks
> >unusable
> >> for
> >> > interactive traffic and email. Further, there is considerable
> >> resistance to
> >> > change or experimentation.
> >> >
> >> > BTW, the 70cm digi at Ginini was almost unused. The best
performing
> >> (most
> >> > consistent and highest througput) path from my QTH to VK1BBS was
up
> >to
> >> Ginini
> >> > on 70cm, across to the 4800 node, and down to VK1BBS. Even when
the
> >2m
> >> 4800
> >> > channel was very busy and gave hopelessly low thoughput to
stations
> >on
> >> 2m, it
> >> > was possible to get 'priority' access to the 2m node via the 70cm
> >> node.
> >> >
> >> > It seems to me that in 15 years, there has been little progress
in
> >> packet in
> >> > this country.
> >> >
> >> > APRS, although using packet protocols (AX.25), albeit in a
limited
> >> way, is
> >> > more about transmitting packets (in a broadcast mode) without
> >> apparently much
> >> > regard to whether they are received anywhere of use. My
experience
> >is
> >> that
> >> > success rates are low, caused again in part by congestion, and
> >> incorrectly
> >> > aligned transmitters.
> >> >
> >> > My though is that wanting to increase the number of WIDE nodes
that
> >> are heard
> >> > in a community increases congestion. So whilst simple thinking is
> >that
> >> lots of
> >> > WIDES and RELAYS increase the change of a position report being
> >heard
> >> > somewhere usefull, increased traffic causes congestion and
decreases
> >> the
> >> > probability that a position report is not clobbered by another
> >> (hidden)
> >> > station.
> >> >
> >> > My suggestion is that before committing resources to a network of
> >> digis in an
> >> > adhoc fashion, that some service levels be considered, the
traffic
> >> modelled,
> >> > and that the models be validated with some controlled field
trials.
> >> >
> >> > Back to Ginini, the 2m voice repeater seems to get less use as
time
> >> passes. I
> >> > suspect that is part of the changing amateur community that we
don't
> >> > understand very well. Indeed, I see signs that the linked
repeaters
> >> seem to
> >> > get less use the more widespread the linking - perhaps users
> >preferred
> >> the
> >> > 'local' context.
> >> >
> >> > If the moves that I understand are under way to downgrade the
> >> qualification
> >> > requirements for amateur radio (step 1: foundation licence, step
2:
> >> coerce the
> >> > multiple classes of licence to a single class based on the
> >foundation
> >> licence)
> >> > to bolster numbers of licencees (and hopefully active operators),
> >> there may be
> >> > considerable growth in demand on the facilities like 2m repeaters
> >that
> >> offer
> >> > much better service than the CB repeaters.
> >> >
> >> > >Would the regular users agree to either move the service to
Bulls
> >> Head,
> >> > >or close it down?
> >> > >
> >> >
> >> > As noted earlier, my experience of a couple of years ago was that
> >> there is
> >> > considerable resistance to change in the packet network.
> >> >
> >> > Note that a number of modern transceivers come with 1200 and 9600
> >> packet
> >> > modems built in. When I looked around a couple of years ago, the
> >> original HAPN
> >> > (4800) modems had been discontinued by the Vancouver group. I am
not
> >> aware of
> >> > any transceivers with an integrated HAPN (4800) modem. I suspect
> >that
> >> it may
> >> > be easier to get on 9600 packet than 4800, but the shared
> >> infrastructure is
> >> > supports 4800. The ready availability of both TNCs and
transceivers
> >> with
> >> > integral TNC with 9600 support is a factor that perhaps ought be
> >> considered
> >> > when planning future packet resources.
> >> >
> >> > The way ahead might be to continue the 4800 service at Bulls
Head,
> >but
> >> to
> >> > incorporate a 9600 node as well.
> >> >
> >> > It is hard pleasing all of the people all of the time Gil!
> >> >
> >> > Owen
> >> >
> >> > BTW, I have carefully avoided recommending in favour or against
any
> >of
> >> the
> >> > services, but put on paper some thoughts on some of the factors
that
> >> might be
> >> > considered in your deliberations. I hope they are of use.
> >> >
> >> >
> >> > >Lets have an exchange of views as a first step.
> >> > >
> >> > >Regards
> >> > >Gilbert Hughes
> >> > >mailto:ghughes3@bigpond.com
> >> >
> >> > --
> >> > ===================================================
> >> > Owen Duffy  (owen@oda.com.au)
> >> > Phone:  +61 2 6257 3872            (02) 6257 3872
> >> > Mobile: +61 417 408 424            0417 408 424
> >> > ===================================================
> >> >
> >> >
> >>
> >>
> >>
>
>***********************************************************************
*
> >> MIMEsweeper has been used to check this email for security
> >>
>
>***********************************************************************
*
> >>
> >>
>
> --
> ===================================================
> Owen Duffy  (owen@oda.com.au)
> Phone:  +61 2 6257 3872            (02) 6257 3872
> Mobile: +61 417 408 424            0417 408 424
> ===================================================
>
>