Messages in this thread | | | Subject | Re: ISDN, HDLC, LAPD, tutorial/introduction... | Date | Mon, 8 Jun 1998 23:56:22 +0300 (EEST) | From | Matti Aarnio <> |
| |
> Meir Dukhan <meir@bis.co.il> writes: > > on network protocol in general, Layer 2 and 3 of OSI model, and a > > conceptual/technical description on HDLC, LAPD (is HDLC defined by an > > RFC > LAPD is specified in ITU-T recommendation Q.921, which is (at least was) > on the INFOMAGICS Standards CD (together with lots of other isdn related > stuff). There are also some ftp servers which have mounted this CD (don't > know names by heart).
HDLC is defined by ISO 3309-1979/1991 standards, and the control procedures are described in ISO 4335-1979/1991 standards. (See RFC 1662) The HDLC itself is derived from IBM's SDLC with a couple of bit modifications to generalize control, and address formats "a bit" from its original very restricted lengths. (I recall 1 control octed + 1 address octet per frame at SDLC.)
CCITT/ITU-T chose those technologies as basis for other applications, like for a set of Link Access Protocols (LAP*), of which there exist at least: LAPB (Balanced: X.25) LAPM (Modem: V.42) LAPD (D-channel: Q.921) LAPF (Frame Relay; Q.9221, I think)
Of these LAPD and LAPF are fair bit alike, although certain aspects of the control protocols do differ. LAPB and LAPM are again different animals.
> > Where could I found source code related to these topics ? > For source code, look at drivers/isdn/hisax/isdn_l2.c and others. For other > OSI-related protocol sources, look at net/{x25,802,lapb} from a recent > 2.1.x kernel. The BSD unix sources (netbsd) also contain related stuff.
The OSI protocols are mostly out of vogue these days; Internet came and concured, but OSI has its places in systems where other reasons mandate it -- mostly in telecommunication admin facilities.
If you truly need to dwell into X.25, then you must prepare to configure about all possible parameters that your X.25 interface allows to configure -- in practice no two X.25 networks from different operators are quite alike...
If you want to dwell into ISDN signaling, you are in for a huge task. Within Europe we (mostly) use ETSI's editions of ISDN specifications; often known as "Euro-ISDN", or perhaps as "NET-5". But as late as 1990 there were no harmonized ETSI specifications for pan-european use, and thus Europe has about 30 National ISDN profiles still in existence.
Naturally USA has their own extension sets, and Japanese ISDN is -- well, Japanese decided to use practically all possible extensions bits which CCITT/ITU-T didn't specify, or had left for "national extensions"...
Due to this kind of profileration of national features the ISDN switches sometimes interpret some responses/messages from user equipement in a bit unexpected manners. At one time certain switches had a bug in their analysis paths which ended in main signaling processor doing panic() -- the entire switch rebooted, and as expectable, all connections through it went down :-( Because the telephone is way more "only well-behaving guys play here", one should not put just any arbitary code running in there.
> Henner
/Matti Aarnio <matti.aarnio@sonera.fi>
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu
| |