```From: floyd@tanana.polarnet.com (Floyd Davidson) Newsgroups: comp.dcom.telecom.tech Subject: Re: T1 signalling Date: 5 Oct 1998 01:10:38 GMT Grimm wrote: >On Oct 2, James Carlson wrote: > >>When you're doing robbed bit signaling on a T1, the eighth bit out of >>each channel is stolen once every six frames. You can tell which >>frame is which by looking at the pattern of the 193rd bit (the frame >>sync bit). In a 24 frame sequence, these stolen bits form a 4 bit >>word for each channel every 3 milliseconds. That word is commonly >>called the "ABCD bits." > > >Quick question for you James or anyone who can clarify. Is robbed bit >signaling done for all variations of line coding/framing used for T1s: ami >or b8zs, sf or esf. Or are there different techniques used? > >Thanks >Steve Here is a brief sketch of what each different variation is and how it relates to other T1 variations. B8ZS/AMI B8ZS and AMI are line coding variations. B8ZS coding is always to be preferred over AMI because it guarantees 1's density regardless of the data being sent, thus allowing a 64Kbps clean DS0 channel. AMI encodes each 1 bit as a pulse, and alternates the polarity of each pulse that is sent. A string of 0 bits results in no pulses. B8ZS works by recoding any string of 8 sequential zero bits into a pattern which includes a bipolar violation at the 4th and 7th bit locations, and begins with a pulse polarity that depends upon the last sent 1 bit polarity. (If the last sent 1 bit was a positive pulse, then an 8-zero string is replaced with 0 0 +1 -1 0 -1 +1. If the last 1 bit was a negative pulse the polarity of the 1's is opposite that sent for a positive 1.) The pattern is replaced at the receive end with the original 8 zero bits. Maintaining 1's density is required for the receiving end to acquire clock synchronization. B8ZS has no effect on either choice of framing or on Robbed Bit Signaling. SF/ESF and RBS SF (SuperFrame) and ESF (Extended SuperFrame) are two variations in framing. With either type RBS (Robbed Bit Signaling) uses the 8th bit in every 6th frame, and with either type of framing RBS is not necessarily used (it can be optioned off). SF With SF, (also commonly referred to as D4 Framing or D3/D4 Framing, which is the same as D1D or D2 with the exception of DS0 channel sequence numbering) the sync bit (that one bit inserted in the 193rd position of every frame) is used only as a pattern to detect frame sync. A superframe consists of 12 frames, and the sync bit pattern for those 12 frames is 1000 1101 1100. The pattern is then repeated every 12 frames and by monitoring it the receiving end can determine which bit is the sync bit, thus allowing it to also know where a frame begins and ends. (There is also a variation of the above used for SLC-96 applications where only sync bits in the odd frames are used for sync, and the even frame sync bits are made available for a data link channel.) ESF With ESF (less commonly referred to as D5 Framing) the sync bit sequence takes 24 frames to make a superframe, and the pattern of sync bits is 001011, which are applied only to frames 4, 8, 12, 16, 20, and 24. The 193rd bit in frames 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21 and 23 are used as a data link channel. The 193rd bit in frames 2, 6, 10, 14, 18, and 22 are used to pass a CRC-6 code for each superframe block. ESF is always to be preferred over SF, if for no other reason than the CRC-6 data with each frame allows in service monitoring for bit errors. ROBBED BIT SIGNALING Robbed Bit Signaling (RBS) is a system that uses the 8th bit from each DS0 channel to pass that DS0's supervision status on frames 6 and 12 with SF, and frames 6, 12, 18, and 24 with ESF. Frames 6 and 12 are the A and B bits, while frames 18 and 24 are the C and D bits, respectively for each DS0 channel (C and bits are available only with ESF). Typically the C and D bits are never actually used, though there exists a list of 16 options defining such features as Call Forwarding. ESF can be optioned to provide none of the ABCD bits, to duplicate the A bit in all four time slots, or to duplicate the A and B bits in the C and D time slots, but none of those are commonly seen. The following is a comprehensive chart for status of AB signaling bits as used for E&M signaling, DX signaling, FX Loop Start, and FX Ground Start circuits. *** ************* E and M Signaling ************* *** E & M Signaling: IDLE CONDITION office A loop digital carrier loop office Z ==> M-gnd A = 0 B = 0 E-open ==> CO -------------------------------------------- CO <== E-open A = 0 B = 0 M-gnd <== E & M Signaling: OFFICE A OFFHOOK, OFFICE Z ONHOOK office A loop digital carrier loop office Z ==> M-bat A = 1 B = 1 E-gnd ==> CO -------------------------------------------- CO <== E-open A = 0 B = 0 M-gnd <== E & M Signaling: OFFICE Z OFFHOOK, OFFICE A ONHOOK office A loop digital carrier loop office Z ==> M-gnd A = 0 B = 0 E-open ==> CO -------------------------------------------- CO <== E-gnd A = 1 B = 1 M-bat <== E & M Signaling: BUSY (BOTH OFFICES OFFHOOK) office A loop digital carrier loop office Z ==> M-bat A = 1 B = 1 E-gnd ==> CO -------------------------------------------- CO <== E-gnd A = 1 B = 1 M-bat <== *** **************** DX Signaling *************** *** DX Signaling: IDLE CONDITION office A loop digital carrier loop office Z ==> A-gnd A = 0 B = 0 A-gnd ==> CO -------------------------------------------- CO <== A-gnd A = 0 B = 0 A-gnd <== DX Signaling: PBX A OFFHOOK, PBX Z ONHOOK office A loop digital carrier loop office Z ==> A-bat A = 1 B = 1 A-bat ==> CO -------------------------------------------- CO <== A-gnd A = 0 B = 0 A-gnd <== DX Signaling: PBX Z OFFHOOK, PBX A ONHOOK office A loop digital carrier loop office Z ==> A-gnd A = 0 B = 0 A-gnd ==> CO -------------------------------------------- CO <== A-bat A = 1 B = 1 A-bat <== DX Signaling: BUSY (BOTH PBXs OFFHOOK) office A loop digital carrier loop office Z ==> A-gnd A = 1 B = 1 A-bat ==> CO -------------------------------------------- CO <== A-bat A = 1 B = 1 A-gnd <== NOTES: The B lead is always -20 VDC. *** *************** FX Loop Start *************** *** FX Loop Start: IDLE CONDITION office A loop digital carrier loop office Z ==> TR-bat A = 0 B = 1 TR-bat ==> CO -------------------------------------------- STA <== TR-Hi Res A = 0 B = 1 TR-Open <== FX Loop Start: Station Off Hook (Initiating Call) office A loop digital carrier loop office Z ==> TR-bat+dt A = 0 B = 1 TR-bat+dt ==> CO -------------------------------------------- STA <== T-Lo Res A = 1 B = 1 TR-short <== FX Loop Start: Central Office Ringing (Initiating Call) office A loop digital carrier loop office Z ==> TR-bat+ring A = 0 B = 0 TR-bat+ring ==> CO -------------------------------------------- STA <== TR-Hi Res A = 0 B = 1 TR-Open <== FX Loop Start: Busy office A loop digital carrier loop office Z ==> TR-bat A = 0 B = 1 TR-bat ==> CO -------------------------------------------- STA <== T-Lo Res A = 1 B = 1 TR-short <== NOTES: The B bit from the CO follows the ring cycle, 0 state equals ringing present, 1 state equals ringing not present. Tip is +BAT and Ring is -BAT. *** *************** FX Ground Start *************** *** FX Ground Start: IDLE CONDITION office A loop digital carrier loop office Z ==> T-open,R-bat A = 1 B = 1 T-open,R-bat ==> CO -------------------------------------------- STA <== TR-open A = 0 B = 1 T-bat,R-open <== T=-48v R=-30v FX Ground Start: Station Off Hook Sequence (Initiating Call) office A loop digital carrier loop office Z (Station end applies Ring GND) CO -------------------------------------------- STA <== T = -48v A = 0 B = 0 T = -48v <== R = 0v R = gnd ==> TR-dialtone A = 0 B = 1 TR-dialtone ==> T = 0v T = gnd R = -48v R = -48v CO -------------------------------------------- STA (CO returns Tip GND and Dialtone) (Station end applies loop closure and removes Ring GND) CO -------------------------------------------- STA <== T-Lo Res A = 1 B = 1 TR-short <== FX Ground Start: Central Office Ringing (Initiating Call) office A loop digital carrier loop office Z ==> TR-bat+ring A = 0 B = 0 TR-bat+ring ==> CO -------------------------------------------- STA <== TR-Hi Res A = 0 B = 1 T-bat,R-open <== FX Ground Start: Busy office A loop digital carrier loop office Z ==> TR-bat A = 0 B = 1 TR-bat ==> CO -------------------------------------------- STA <== T-Lo Res A = 1 B = 1 TR-short <== NOTES: The B bit from the CO follows the ring cycle, 0 state equals ringing present, 1 state equals ringing not present. One article in this thread mentioned that manuals for test equipment are a superb source of information about digital systems, and the above is a prime example! It doesn't come from a test set manual, but almost every bit of it is in an old manual, "The T1 Fact Finder Book", published by a test equipment manufacturer. The particular copy I have is a bit dated now, but in 1989 it sold for \$5.00, from: Phoenix Microsystems, Inc. 991 Discovery Drive Huntsville, AL 35806 A little poking on the net came up with the fact that Communications Technology Corp purchased them in 1994, and their web site is http://www.phoenixmicro.com/ Of course, everything that is not in "The T1 Fact Finder Book" is probably of little or no use in the field, but might be essential knowledge for design engineering staff. Bellcore Technical Reference documents, such as TR-INS-00342 "Hi-Capacity Digital Special Access Service -- Transmission Parameter Limits and Interface Combinations", and various ANSI documents should be obtained for definitive descriptions. Floyd -- Floyd L. Davidson floyd@ptialaska.net Ukpeagvik (Barrow, Alaska) floyd@barrow.com ``` ```From: floyd@tanana.polarnet.com (Floyd Davidson) Newsgroups: comp.dcom.telecom.tech Subject: Re: T1 signalling Date: 13 Oct 1998 14:47:23 GMT James Carlson wrote: >vandry@Mlink.NET (Phillip Vandry) writes: >> Of course it's highly unlikely that this would occur, but is there any >> provision for preventing this same bit pattern from appearing in some >> other bit position than the sync bit, such that the receiver might >> mis-sync? > >Not exactly. Framers are required to watch for this pattern to repeat >several times before declaring synchronization, and they transmit only >sync data while they're doing it. So, if sync is lost, no user data >is transmitted and the only pattern that exists on the wire should be >the real sync pattern. Not quite correct. Lets back up just one step farther and consider all of the possible conditions. If a DS1 receiver (lets call it the near end) detects no incoming signal, its associated transmit side will send an AIS (Alarm Indication Signal) pattern toward the distant end. That is all ones with no framing, and hence the distant end receiver cannot acquire frame lock while receiving an AIS signal. Hence if both the transmit and receive sides are disconnected, both ends will be sending all ones. If the near end receive side is then reconnected, the receiver end will detect an incoming signal (AIS, which it cannot acquire framing on) from the distant transmitter. The near end will then begin to send a "yellow alarm" signal on its transmit towards the distant end. Since we still have that side disconnected, the distant end will not see the yellow alarm data and will continue sending all 1s. A yellow alarm signal is generated by forcing the 2nd bit of every channel to 0 and is required to last for at least 1 second. Note that a yellow alarm signal will be framed data that the distant receiver can acquire frame lock on. In the above scenario, if the other pair is then reconnected, the distant receiver will also detect an incoming signal (the yellow alarm pattern), and will then begin sending a yellow alarm pattern too. At that point both receivers are seeing valid framed data and each, totally independent of the other, will begin searching for a sync bit pattern in every 193rd bit. Totally independent of the distant end, the receiver will be skipping a bit each time the right bit is not detected and begin the sequence again. When multiple sequences of the correct framing pattern have been received the receiver declares itself to have acquired frame sync, and allows its transmit side to send normal data for the 2nd bit of each channel. >> Same question wrt ESF, though the CRC-6 would make it even more unlikely... > >Same technique. You're always doing one of two things. Either you >and the peer are out of sync and you're searching for a sync pattern >and no user data is present, or you're both in sync and you're just >comparing the sync patterns in the known locations. Once again, there are several possible states, not just two. The only difference from SF is where the ESF sync bits are located and what the pattern expected is. User data is in fact present. >The "extra" sync bits that allow data transfer form a 4Kbps >synchronous data channel called FDL (facilities data link) that's used >for all sorts of low-level performance and status checks. It also >allows non-HDLC BOPs (bit-oriented patterns) for gross error >conditions, like yellow alarm. A yellow alarm is not related to the DL (Data Link) and indeed is available independent of which framing is used. The DL can be made available to a customer as a 4kbps data channel, but normally it carries the performance data referred to above. An idle bit pattern of 0111 1110 is sent, and one of two formats are used for data. One format is a 15 byte message based on CRC receive violations and is sent every 1 second. A second format consists of repeated 16 bit codes which send "alarms, commands, and responses". I have no idea what the "alarms, commands, and responses" are, however! Floyd -- Floyd L. Davidson floyd@ptialaska.net Ukpeagvik (Barrow, Alaska) floyd@barrow.com ``` ``` ``` ``` ```