Bellcore's Plans for Caller ID
Bellcore has issued a technical advisory "Voice Band Data Transmission Interface Generic Requirements' (TA-NWT-000030) that details data transmission standards for future Caller ID services. The services directly referred to are: Calling Number Delivery (CND), Calling Identity Delivery on Call Waiting (CIDCW), and Calling Name Delivery (CNAM). While much of the technical data is already known, there are some significant new bits of information we feel people should be aware of.
Vital Statistics
The signaling interface consists of three layers. The first is the physical layer which defines the requirements of analog data transmission.
The transmitted data signal has to meet these parameters (Bell 202):
Modulation Type: Analog, Phase Coherent, Frequency-Shift Keying Mark (Logical 1): 1200 Hz (+/- 12 Hz) Space (Logical 0): 2200 Hz (+/- 22 Hz) Signal Level: -13.5 dBm (+/- 1 dBm) At the point of application to the loop facility into a standard 900 ohm test termination. Signal Purity: Total power of all extraneous signals in the voice-band is at least 30 dB below the power of the signal fundamental frequency. Source Impedance: 900 ohms + 2.16 µF nominal Transmission Rate: 1200 bps (+/- 12 bps) Application of Data: Serial, Binary, AsynchronousThe second layer is the Data Link Layer that deals with error detection through CRC. The third and final layer is called the Presentation Layer. Here, data is converted into ASCII text in a form readable by the customer equipment (Caller ID devices).
Both single and multiple data messages are supported. Single data message format consists of Channel Seizure Signal; Mark Signal; Message Type Word; Message Length Word; Message Word(s); and Checksum Word. Multiple data message format consists of Channel Seizure Signal; Mark Signal; Message Type Word; Message Length Word; the first Parameter Type Word; the first Parameter Length Word; the first Parameter Word(s); any additional Parameter Type Words, Parameter Length Words, or Parameter Words; and Checksum Word.
Each data word consists of an 8-bit data byte. Each data word is preceded by a Start Bit (Space) and followed by a Stop Bit (Mark). Mark can be transmitted between any two words to maintain a continuous signal and cannot exceed 10 bits. The message length word contains the number of words in the message following it, with the exception of the error detection word.
The channel seizure signal is 300 continuous bits of alternating 0's and 1's. This signal is only used for on-hook data transmission and is followed by a Mark signal (logical 1) before the actual data is sent. For future off-hook data transmission, each data message is preceded only by the Mark signal.
The carrier signal consists of 130 ms (+/- 25 ms) of Mark (1200 Hz). The message type word indicates the service and capability associated with the data message.
For instance, the message type word for CND is: 04h (00000100)
In an on-hook state, data transmission takes place between the first and second rings. Transmission doesn't begin until 500 ms of silence has elapsed and has to end at least 200 ms before the next ring begins. This allows for between 2.9 and 3.7 seconds for the entire transmission.
An example of a typical on-hook CND message follows:
04 12 30 39 33 30 31 32 32 34 36 30 39 35 35 35 31 32 31 32 51 04: Message Type Word (4 indicates SDMF) 12-18: Number of Bytes in Date, Time, and Directory Number (Decimal) 30,39: Month - 09 - September (ASCII) 33,30: Day - 30 (ASCII) 31,32: Hour - 12 (ASCII) 32,34: Minutes - 24 (12:24 PM) (ASCII) 36,30,39: Area Code - 609 (ASCII) 35,35,35: Prefix - 555 (ASCII) 31,32,31,32: Suffix - 1212 (ASCII) 51: Checksum WordFuture Features
In an off-hook state, speech transmission will be interrupted for the duration of the data transmission. A tone will be sent for 50-55 ms to alert the customer (CPE Alerting Signal). This tone will probably be a combination of 2130 Hz and 2750 Hz sent at a nominal level of -16.5 dBm/frequency.
Bellcore's explanation for using these particular tones:
"The tone to be generated... must be detectable in the presence of near-end speech and provide for minimal occurrance [sic] of false detections... In addition, the tone must be of tolerable duration and amplitude from a human factors perspective. One of the options that was proposed for such a tone was the DTMF A. This tone did not meet all the performance criteria. As a result, Bellcore researched other options, namely the use of higher frequency dual tones. Based on prior research in Great Britain, Germany, and Japan, it has been established that signal detection performance improves significantly when the alerting signal falls in the upper part of the speech band. For dual tones the frequency pair selected should avoid common harmonic relationships such as 2:1, 3:2, 5:3, etc... Although studies and testing will continue, we expect that this frequency pair will be the final specified alerting tone for the off hook case."
After sending the alert tone, the central office will initiate a 100-120 ms acknowledgment timer and will wait for an acknowledgment from the customer equipment. An acknowledgment signal will consist of either a single DTMF D signal (for the least sophisticated customer equipment) or a DTMF D signal followed by a delay of 45-50 ms, then another DTMF key (0-D) (to identify more sophisticated customer equipment). Each DTMF tone must have a minimum duration of 40 ms.
The actual data will then be transmitted by the central office within 20 ms of the end of the acknowledgment or after the maximum time for an acknowledgment has passed, whichever is greater. If an acknowledgment is not received, the data will not be transmitted and the speech path will be restored.
Right now, one of Bellcore's biggest concerns is the length of the speech path disruption, which can be close to four seconds long. Whether or not customers are willing to put up with that every time another call comes in remains to be seen.