Make the Call with Caller ID Testing

Telephony products with caller ID require testing at the physical and data-link layers.

by Martin Rowe, Senior Technical Editor, T&MW

Caller ID is an analog service by which a telephone central office (CO) switch sends digital information about the incoming call. Individuals and small businesses can choose three types of caller ID CPE products (Table 1). I’ll limit the scope of this article to types 1 and 2; type 3 is an extension of the technology in the other two types.

Table 1. Types of Caller ID.

Type	Description
1	Caller ID--on hook
2	Caller ID with call waiting--off hook
3	Analog Display Services

Testing caller ID functions requires you to simulate the analog environment the CPE will encounter—the physical layer. You also must test your CPE’s software to ensure that it properly decodes the incoming data—the data-link layer. At the physical layer, you have to test for parameters such as signal level, timing, carrier frequency, and noise level. At the data-link layer, you must test for bit errors, bit stuffing, and data validity.

When testing a CPE’s caller ID features, start at the physical layer. Figure 1 shows the transmission format for caller ID, specified in Bellcore standards GR-30-CORE1, TR-NWT-0000312, and TR-NWT-0011883. These standards apply to the US telecom network, while similar standards apply in Canada and other countries.

Figure 1. For type 1 caller ID, a CO switch sends caller ID data between the first and second rings of an incoming call. FSK modulation carries packets of data that include a message byte, data-count byte, and checksum byte.

Type 1 Requirements
In Type 1 caller ID, a CO switch transmits a Bell 202 modem (1200 bps frequency-shift keying, or FSK) signal at least 500 ms after the first ringing pattern ends. To get the CPE’s attention, the signal starts with a preamble consisting of 30 bytes of alternating 0’s (spaces) and 1’s (marks). Next, the carrier typically sends a mark signal for 150 ms (180 bits). Following the mark bits, the CO sends a byte that describes the message type (single-message format or multiple-message format). The CO switch then sends a byte that indicates the number of bytes in the message packet—1 to 255. Finally, the CO switch sends a checksum byte for error detection and ends its transmission with more mark bits.

While simple in concept, Type 1 caller ID requires thorough testing to ensure compliance with Bellcore and Telecommunications Industry Association (TIA) requirements. On the physical level, you must simulate variances in amplitude, frequency, and timing. When testing your UUT’s FSK demodulator, you must vary the frequency of both the mark tones (1200 Hz) and space tones (2200 Hz) by 1% and check for bit errors. You should, however, vary the frequency by more than 1% to give yourself a safety margin.

Because mark tones and space tones differ by 1000 Hz, the telephone network will attenuate those signals differently. The longer the local loop, the more the network will attenuate the tones and the more the tones’ amplitudes will differ. The diagram in the upper-right corner of Figure 2 shows a twist level—the difference between mark and space tone amplitudes that the tester measures. Your CPE receiver should detect mark tones ranging from 10 dB above (+10 dB) to 6 dB below (–6 dB) the space tone’s amplitude.4

Figure 2. Software lets you set the test parameters on a line simulator. For the FSK modulator, you can adjust the mark and space frequencies, twist, level, baud rate, and other parameters. (Courtesy of Advent Instruments)

Telephone local loops often receive noise that can interfere with FSK transmissions and impair data. In a noise test, you must verify that a CPE’s bit-error rate remains within its design parameters in the presence of noise. The amplitude of the interfering signal relative to the FSK transmissions will differ with frequency. Table 2 lists the amplitudes of the interference signals to those of the FSK signal you should use in your tests. You should test the FSK receiver’s sensitivity by varying the signal’s amplitude from –12 dBm to –34 dBm for mark tones and from –12 dBm to –36 dBm for space tones with a 900-V source impedance.5
 

Table 2. Interference Signals and Amplitudes (source: EIA/TIA SP-3674-A)

Because the FSK transmission occurs between the first and second ringing pattern, you must test your CPE with variations in transmission timing. Alter the ring signal duration, the time to send data, and the time to second ring signal. While the standard ring signal duration is 2 s, sometimes the first ring signal is shorter. In fact, the first ring signal may last just 200 ms. You must verify that the receiver will decode the FSK signal following such a short ring signal.

While 500 ms should elapse between the end of the first ring signal and the start of transmissions (Fig. 1), you can’t assume that will always happen. Some phone and computer telephony product manufacturers design their products assuming 0 ms will pass between the first ring signal and the FSK transmission. You can verify your product’s performance by shortening the ring-signal-to-transmit time as required by your design specifications. For testing to meet Bellcore and TIA requirements, vary the ring-to-transmit time from 250 ms to 3.6 s; the CPE under test should still receive data. On the back end, you must allow as little as 200 ms between the end of the mark-off time and the beginning of the second ring signal.

Caller ID with Call Waiting
So far, all the caller ID tests I’ve discussed assume the receiver was on hook (not in use), and therefore could detect a ring. But what happens if your phone is off hook when a call comes in? If you have call waiting, you might like to identify the incoming caller, put the current caller on hold, and take the new call.

If you have caller ID with call waiting (CIDCW) service—type 2 caller ID—you’ll hear two beeps. The beeps alert you and your CPE that the CO switch has FSK data to send. The first beep contains a single tone (440 Hz), called a subscriber alerting signal (SAS), that informs you a new call is waiting. Within 50 ms, the CO sends a beep consisting of two tones—the CPE alerting signal (CAS) tones. CAS tone 1 (2130 Hz) and CAS tone 2 (2750 Hz) get the CPE’s attention.

The CPE then sends an acknowledgement tone (DTMF A or DTMF D) to the CO switch.6 Next, the CPE mutes the voice connection for 500 ms while it receives the caller ID data. If the CO switch fails to send the FSK signal within 500 ms, the CPE should restore the voice connection.7

Because the CPE is off hook, it must correctly identify the CAS tones in the presence of other sounds. If it identifies another sound as a CAS tone (called "talk off"), it will send the acknowledge tone and cut off the talk connection for 500 ms. The tone and loss of voice will annoy callers. On the other hand, equipment that fails to detect a CAS tone ("talk down") causes users to pay for caller ID services they won’t get.

Tests for type 2 caller ID involve noise, timing, and voice. Bellcore has a specification that describes the tests.8 Unfortunately, the type-2 caller ID compliance test can take hundreds of hours to complete. A full compliance test requires tests for talk off at nine amplitudes and tests for talk down at seven amplitudes. Each of the 16 tests takes 96 hrs to complete. You can, however, reduce test time by testing talk off and talk down simultaneously.

To test a CPE’s CAS tone detection, you can simulate the tones that a CO sends to your receiver. Vary the tones’ frequency, amplitude and duration. The CAS tone’s duration (75–85 ms) is key, for you can assume that a similar tone of longer or shorter duration must come from another source. Figure 3 shows a screen from a telephone line simulator that sets the SAS and CAS tone parameters. Table 3 shows some test parameters for CAS testing.

Figure 3. To test caller ID with call waiting, you must vary the frequency and duration of the CAS tones, which inform the CPE of an incoming call. (Courtesy of Rochelle Communications)

Table 3. CAS Parameters (source: Bellcore SR-TSV-002476)

Data-Link Layer Tests
So far, the tests I’ve discussed involve physical parameters such as frequency, timing, and amplitude. Once you’ve established your equipment can detect caller ID signals, you must verify that it can properly process incoming data. At the data-link layer, you’ll test a CPE product’s software.

Caller ID messages have two formats: single-message format and multiple-message format. The data block in Figure 1 shows the standard bit pattern for a single message format caller ID.

With single-message format, one information packet contains the month, day, hour, minute, caller name, and caller number. If the incoming call originates from outside your local calling area, the CO will send the letter "O" in place of the phone number. Callers also have the option of blocking their phone number from caller ID services. A blocked number will produce the letter "P" (private) in place of the phone number. No other data are valid for caller ID.

In single-message format, each caller ID message needs only one message type byte and one data count byte. In multiple-message format, the CO switch breaks the message into several data packets, and each packet requires a message byte and a packet-length byte. The possible data are:

• multiple message header (80h),
• date and time packet (01h),
• calling number packet (02h),
• calling name packet (07h), and
• checksum byte.

With multiple-message format, the CO switch can send the date and time, calling number, and calling name packets in any order. So, your CPE’s software must decode and display the data regardless of the order in which they arrive. The bottom of Figure 4 shows the data in a calling number packet. The field at the bottom of the figure shows the ASCII codes for the number in the calling number field above.

Figure 4. With multiple-message format, you can change the order of the packet transmissions and decide which packets to use in testing. (Courtesy of Advent Instruments)

During times of heavy phone traffic, a CO may send mark bits when it’s serving other callers. The CO sends mark bits only to keep the CPE on-line and so it won’t assume the CO is starting a new message. These mark bits can occur between any of the three boundaries in the data as shown in Figure 1—that is, between the message type byte and data-count byte or anywhere else. Your CPE should tolerate bit stuffing of up to 100 mark bits between data packets with up to 500 mark bits per caller ID session.

After sending a checksum byte, the CO switch may send a series of mark bits before dropping the FSK carrier—called mark out. A CPE product should tolerate up to 4 s of mark bits without losing the FSK carrier. T&MW
 

FOOTNOTES
1. GR-30-CORE, CLASS Feature: Calling Number Delivery FSD 01-02-1051, Bellcore, Piscataway, NJ (www.bellcore.com). December 1994.

2. TR-NWT-000031, CLASS Feature: Calling Number Delivery FSD 01-02-1051, Bellcore, Piscataway, NJ. December 1992.

3. TR-NWT-001188, CLASS Calling Name Delivery Generic Requirements, Bellcore, Piscataway, NJ. April 1995.

4. TIA/EIA SP-3674-A, Telecommunications Telephone Terminal Equipment—Type 1 Caller Identity Equipment Performance Requirements, Section 4.4.3, Telecommunications Industry Association, Arlington, VA. November 4, 1997.

5. Ibid.

6. DTMF A is composed of two tones: 697 Hz and 1633 Hz. DTMF D contains 941 Hz and 1633 Hz. Don’t confuse these two keys with the "2" key (ABC) and the "3" key (DEF) on your phone. The complete set of touch tones consists of 16 pairs of frequencies: those for the 12 keys that appear on all touch-tone phones (1–0 plus * and #) and those for the A, B, C, and D keys. Most phones don’t contain these last four keys.

7. SR-TSV-002476, CPE Compatibility Considerations for the Voiceband Data Transmission Interface, Bellcore, Piscataway, NJ. December 1992.

8. SR-TSV-002578, A Method and Apparatus for Detecting a Dual Tone Signal in the Presence of Speech, Bellcore, Piscataway, NJ. April 1993.
 

FOR FURTHER READING
Armine, Gilbert, Annette Riggio, and Ellis Hill, "Caller ID Goes to Work," Byte (www. byte.com/art/9501/sec11/art9.htm). January 1995.

SR-3004, Testing Guidelines for Analog Type 1, 2, and 3 CPE as Described in SR-INS-002726, Bellcore, Piscataway, NJ. January 1995.