| |
PA2AGA > TCPDIG 14.09.96 06:56l 144 Lines 5990 Bytes #-10887 (0) @ EU
BID : TCP_96_190A
Read: GUEST
Subj: TCP-Group Digest 96/190A
Path: DB0AAB<DB0KCP<DB0ULM<DB0LX<DB0RBS<DB0SEL<DB0ZDF<DB0AIS<DB0NDK<DB0ACH<
DB0ACC<PI8DRS<PI8DAZ<PI8APD<PI8GCB<PI8MBQ<PI8VNW
Sent: 960913/2346Z @:PI8VNW.#ZH2.NLD.EU #:1737 [Hoek v Holland] FBB5.15c
From: PA2AGA@PI8VNW.#ZH2.NLD.EU
To : TCPDIG@EU
Received: from pa2aga by pi1hvh with SMTP
id AA17556 ; Fri, 13 Sep 96 23:24:51 UTC
Received: from pa2aga by pa2aga (NET/Mac 2.3.61/7.1) with SMTP
id AA00004629 ; Sat, 14 Sep 96 01:11:08 MET
Received: from pa2aga-10 by pa2aga with SMTP
id AA00004616 ; Sat, 14 Sep 96 01:06:28 MET
Received: from pa2aga-10 by pa2aga-10 (NET/Mac 2.3.61/7.1) with SMTP
id AA00005549 ; Sat, 14 Sep 96 01:06:24 MET
Date: Fri, 13 Sep 96 22:47:58 MET
Message-Id: <tcp_96_190A>
From: pa2aga
To: tcp_broadcast@pa2aga-10
Subject: TCP-Group Digest 96/190A
X-BBS-Msg-Type: B
TCP-Group Digest Fri, 13 Sep 96 Volume 96 : Issue 190
Today's Topics:
high speed coders/decoders (5 msgs)
Send Replies or notes for publication to: <TCP-Group@UCSD.Edu>.
Subscription requests to <TCP-Group-REQUEST@UCSD.Edu>.
Problems you can't solve otherwise to brian@ucsd.edu.
Archives of past issues of the TCP-Group Digest are available
(by FTP only) from ftp.UCSD.Edu in directory "mailarchives".
We trust that readers are intelligent enough to realize that all text
herein consists of personal comments and does not represent the official
policies or positions of any party. Your mileage may vary. So there.
----------------------------------------------------------------------
Date: Thu, 12 Sep 1996 08:15:29 -0800 (PDT)
From: Glenn Elmore <glenne@hpsadr2.sr.hp.com>
Subject: high speed coders/decoders
Phil wrote:
> Raw C/N in a metropolitan area is easy; it's C/multipath that's
> hard. But power-efficient coding can alleviate the problem as
> multipath can be made to look like thermal noise.
The N I was referring to was indeed the combined "noise" from both
external qrm/qrn as well as self induced (multipath) mechanisms.
However getting sufficient signal level in wide bandwidths in an amateur
context is not, in my opinion, easy.
Even delivering the approximately -90 dBm, which is the minimum that
my 230 kbps radios want to see, has not been 'easy' at the user end. It
requires more antenna and path selection than most amateurs are accustomed
to.
The problem of getting adequate signal strength in an amateur context,
even apart from that of mitigating multipath induced by non-optimum
paths, is not trivial. Amateurs don't have the benefit that current
cell systems enjoy of always having one end of the link in an engineered
location. We are dependent on the available density of amateurs, which
affects the mean path length and we are dependent on the quality of
available sites (qths). Neither of these is even nearly as good as
commercial systems enjoy.
Also, these problems only get more severe as frequency is increased
(presumably in the quest for higher data rates). They are severe at L
band but by the time you get to X band or above they are enormous.
> Because the LSI Logic parts are designed for consumer digital
> broadcast receivers, the most integrated (and probably the cheapest)
> parts do decoding only. But their product line does include a
> stand-alone RS encoder, and it's pretty much trivial to encode a K=7
> convolutional code and generate QPSK in your own hardware.
I didn't realize they also had the stand-alone encoder. That could
certainly help. Maybe someone else on here will be able to investigate
the parts in greater detail.
Glenn
------------------------------
Date: Thu, 12 Sep 1996 11:40:48 -0700 (PDT)
From: Phil Karn <karn@qualcomm.com>
Subject: high speed coders/decoders
> Even delivering the approximately -90 dBm, which is the minimum that
>my 230 kbps radios want to see, has not been 'easy' at the user end. It
>requires more antenna and path selection than most amateurs are accustomed
>to.
Hmm, by my calculations -90 dBm at 230kbps corresponds to an Eb/N0
plus thermal noise of 55 dB-K. E.g., if you have a rx temp of 1000K,
you are running at a Eb/N0 of 25 dB. 10,000K corresponds to 15 dB.
One figure or the other seems rather high. (You don't give enough
information to solve separately for thermal noise and Eb/N0 -- do you
have that information?)
With coding you should be able to knock the Eb/N0 down to about 5 dB
even for the high rate code, assuming BPSK or QPSK instead of
noncoherent FSK. So if your rx temp is 1000K, that's a 20 dB savings.
That's the difference between 10W and a kilowatt.
Phil
------------------------------
Date: Thu, 12 Sep 1996 12:09:33 -0800 (PDT)
From: Glenn Elmore <glenne@hpsadr2.sr.hp.com>
Subject: high speed coders/decoders
> Hmm, by my calculations -90 dBm at 230kbps corresponds to an Eb/N0
> plus thermal noise of 55 dB-K. E.g., if you have a rx temp of 1000K,
> you are running at a Eb/N0 of 25 dB. 10,000K corresponds to 15 dB.
> One figure or the other seems rather high. (You don't give enough
> information to solve separately for thermal noise and Eb/N0 -- do you
> have that information?)
It probably is high but these are simple FSK systems and probably don't
work particularly near what is possible. They are, however, not so different
from much of the hardware currently in common use. I believe that the
Metricom Ricochet which runs at only 100 kbps needs something in this
vicinty as well.
But even if there is 10 or 20 dB to be picked up with better selection
of modulation and coding, and I believe we are interested in picking
this up, it needs to be understood that *in an amateur environment*
there is also another 40-60+ dB to be *lost* by poor path/site/antenna
selection, particularly in the bands where we have a lot of spectrum
in which to run high rate systems.
It is difficult to overstate the difference in delivered C/N between a
system which most amateurs might consider reasonable and would call
"line of sight" (probably bearing resemblance to typical vhf/uhf nbfm
arrangemnts) and what a quality installation has. Particularly >1 GHz
and even more so at >10 GHZ I think we are looking at *many* 10's of dB.
To be continued in digest: tcp_96_190B
Read previous mail | Read next mail
| |
