| |
PA2AGA > TCPDIG 14.01.98 22:16l 203 Lines 7200 Bytes #-10101 (0) @ EU
BID : TCP_98_7
Read: GUEST
Subj: TCP-Group Digest 98/7
Path: DB0AAB<DB0SL<DB0RGB<DB0MAK<DB0ERF<DB0HSK<PI8DRS<PI8DAZ<PI8GCB<PI8MBQ<
PI8VNW
Sent: 980114/1833Z @:PI8VNW.#ZH2.NLD.EU #:12299 [Hoek v Holland] FBB5.15c
From: PA2AGA@PI8VNW.#ZH2.NLD.EU
To : TCPDIG@EU
Received: from pa2aga by pi1hvh with SMTP
id AA787 ; Wed, 14 Jan 98 18:16:42 UTC
Received: from pa2aga by pa2aga (NET/Mac 2.3.65/7.1) with SMTP
id AA00006087 ; Wed, 14 Jan 98 18:57:22 MET
Date: Wed, 14 Jan 98 18:51:47 MET
Message-Id: <tcp_98_7>
From: pa2aga
To: tcp_broadcast@pa2aga
Subject: TCP-Group Digest 98/7
X-BBS-Msg-Type: B
TCP-Group Digest Wed, 14 Jan 98 Volume 98 : Issue 7
Today's Topics:
TCP/IP - suggestion for new L2 protocol
TCP/IP over radio (2 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.
Loop-Detect: TCP-Group:98/7
----------------------------------------------------------------------
Date: Wed, 14 Jan 1998 00:37:37 +0000 (GMT)
From: Alan Cox <alan@cymru.net>
Subject: TCP/IP - suggestion for new L2 protocol
> - We need a "lighter" error correction, providing a good throughput
> with few retransmissions. How about implementing a
> Reed-Solomon-like EEC algorithmon as an option on the new L2
> protocol? Options could be negotiated on link setup.
If you have a decent FEC - such as Karn's virterbi codec then you dont
need other retransmissions. At the point most frames get through anyway
the overhead of a perfect versus statistically "not bad" protocol swamp
any gain
> packets since ages. A similar specification by PE1CHL exists for the
> AX25 protocol, but to my knowledge nearly no modern AX.25 protocol
> stack supports it. Why? Medium- and high speed NETROM interlinks
The Linux stack does, and that is under the GNU GPL so its easy to use in
other free projects including standalone node code.
> frame in the middle got lost. That's why we need an intelligent frame
> acknowledge technique, some type of ack frame permitting operations
> like "RR 4-12, 14-25, 28-41" (Ack'ing Frames 4 to 12, 14 to 25 and 28
> to 41 because 13, 26 and 27 got lost).
All the TCP evidence with SACK supports this
> a congested router to apply countermeasures like dropping IP
> frames / sending ICMP source quenches (RFC-896!!!) or recomputing
Dont dont dont send ICMP source quench. Its an unfortunate obsolete mistake
that is now viewed in very bad light. Just throw the frame. If you do send
them send them very irregularly and at a controlled rate. Better yet if
you are using a link layer protocol then use the Frame Relay/ DECnet
approach which works far better
Basically you have a "congested" bit routers can set in frames they
forward but maybe have to delay or queue a long time. The other end of
the link replies to frames with the same bit for a while (or a congestion
ack bit) so folks learn there is congestion with no extra frames. You
can also then spoof ICMP source quenches locally at the start/end of the
radio hops if you want to control awkward stacks.
> in use. Also you could play with LZW compression of data payload or
> encryption. By the way, it's time to make the PID (L3 Protocol
LZW has patent stupidities in the US - both LZ66 and predictor dont,
furthermore predictor performs astoundingly well on AX.25 frames
Alan
------------------------------
Date: Wed, 14 Jan 1998 09:16:20 +1100
From: Terry Dawson <terry@perf.no.itg.telstra.com.au>
Subject: TCP/IP over radio
Norito NEMOTO wrote:
> We started release test version
>
> (808kbps 2.4GHz SS tranver used, CSMA, block FEC, fixed routing,
> and needed other's linux or FreeBSD computer.)
>
> Ether-net interface(KIT) : about 18000JPY (140USD?)
> SS RF Trancever. (PRODUCTION) : 60000JPY (450USD)
They both sound like quite good prices.
> If you want these set, we can release now.
How do they interface to the computer? Does the transciever present
an ethernet style interface ? Is it an internal card or some other
way ?
> >> Will you keep us updated on your progress ?
>
> Yes!! Of course.
> And please give us mary suggestion.
Excellent.
I'm interested in what you are doing at layer 3. Do you have any
documentation to describe it ?
Terry
------------------------------
Date: Wed, 14 Jan 1998 21:22:00 +0900 (JST)
From: noriton@netwk.ntt-at.co.jp (Norito NEMOTO)
Subject: TCP/IP over radio
at <34BBE7B4.DBF6E84C@perf.no.itg.telecom.com.au>
terry@perf.no.itg.telstra.com.au wrote
>> > (808kbps 2.4GHz SS tranver used, CSMA, block FEC, fixed routing,
>> > and needed other's linux or FreeBSD computer.)
>> > Ether-net interface(KIT) : about 18000JPY (140USD?)
>> > SS RF Trancever. (PRODUCTION) : 60000JPY (450USD)
>> They both sound like quite good prices.
Thanks, but cation! These need other computer (contorol system).
When become All-in-one production, maybe about 900USD.
>> > If you want these set, we can release now.
Sory, I forgot important point.
These set Japanese amatuer radio band in 2.4GHz.
I dont know where can do SS in AU's 2.4GHz band.
Maybe, it need QSY, but do easy ;-)
>> How do they interface to the computer? Does the transciever present
>> an ethernet style interface ? Is it an internal card or some other
>> way ?
Yes, they have Ethernet interface.
They are fun system. They run at Z80 compatible CPU, but have an ISA slot.
And pruged ET-2000 or other compatible ISA ethernet card for PC.
They use BOOTP and tftp in order to boot up.
Plese see
http://www.drug.prug.or.jp/,
and there is english report at
http://www2s.biglobe.ne.jp/~jh0mrp/ss/
So, I think to need serial interface, but another member said 'not need'.
We have discussion about that now.
DO YOU NEED?
>> > >> Will you keep us updated on your progress ?
>> > Yes!! Of course.
>> > And please give us mary suggestion.
>>
>> Excellent.
>>
>> I'm interested in what you are doing at layer 3. Do you have any
>> documentation to describe it ?
Yes, there is paper(I wrote), but it was writen in Japanese.
If you can read Japanese, I send it to you now.
But cannot, plese wait to few month. I am going to trancerate it.
------------------------------------------------- o o
Norito Nemoto / JH1FBM Nakano-city, Tokyo,Japan \_(((_/
/ \
Packet Radio User's Group / PRUG96 project ___/ n n \__
in charge of FEC chip, MACA, and routing. ( o )
\__ __/_
E-mail noriton@netwk.ntt-at.co.jp \ ^ ^ ^ \_/ |
WWW http://www.nemoto.com/ \ ^ ^ (
Network Applique'tion Laboratory Mascot "TOCKALI kun" \______/\_|
------------------------------
Date: Tue, 13 Jan 1998 19:15:22 +0100
From: loof@vannasbost.ac.se (Vannasbostader, Loof Olof)
subscribe
------------------------------
End of TCP-Group Digest V98 #7
******************************
Read previous mail | Read next mail
| |
