Tài liệu Tín hiệu trong các mạng viễn thông P4 ppt

When exchange B has received the complete called number, it seizes an available trunk T2 to destination exchange C, and sends a seizure signal on the trunk.. 4.1-1, exchange A seizes tru

CHANNEL-ASSOCIATED

INTEREXCHANGE SIGNALING

Channel-associated inter-exchange signaling (CAS)-also known as per-trunk signaling-has been in existence from the beginning of automatic telephony, and was the only form of interexchange signaling until 1976 It is still used in telecommunication networks, but is gradually being replaced by common channel signaling

Early CAS systems were developed independently by individual equipment manufacturers, and exist in many varieties Later CAS systems, notably those developed after the Second World War, show the increasing influence of national and international standards

This section describes three important CAS systems, and their use on frequency-division multiplexed (FDM) analog trunks and time-division multi- plexed (TDM) digital trunks (1.4.5 and 1.5.2)

The acronyms in the figures of this chapter are explained in Section 4.5

4.1 INTRODUCTION

4.1 I Interexchange Signaling Example

Figure 4.1-1 shows a typical interexchange signaling sequence for a call from subscriber S, to subscriber S2 The subscriber signaling for the call is not shown After exchange A has received the called number from S1, it decides to route the call via intermediate exchange B Exchange A seizes an available trunk T,, and sends a seizure signal on the trunk Exchange B responds with a proceed- to-send (or wink) signal, indicating that it is ready to receive the digits of the

68

Copyright  1998 John Wiley & Sons, Inc ISBNs: 0-471-57377-9 (Hardback); 0-471-22415-4 (Electronic)

Local Intermediate Local

Figure 4.1-I Interexchange signaling

called number Exchange A sends the digits, and then cuts through (sets up a path in its switchblock between the subscriber line of S, and T,)

When exchange B has received the complete called number, it seizes an available trunk T2 to destination exchange C, and sends a seizure signal on the trunk Exchange C responds with a wink signal, after which exchange B sends the digits of the called number, and cuts through a path between trunks T, and TZ Exchange C then checks whether called subscriber S2 is idle If this is the case,

C sends a ringing signal to Sz, and ringing-tone on trunk T2 Because exchanges

A and B have cut through, there is a connection between the calling subscriber S1 and exchange C, and subscriber S, hears a ringing-tone

When S2 answers, exchange C cuts through a path between trunk T, and subscriber S, It also sends an answer signal on T2, and exchange B repeats the signal on trunk T, Assuming that originating exchange A is responsible for charging the call, it establishes a billing record that includes the calling and called numbers, the date, and the time of answer

The conversation now begins In this example, called party S2 hangs up first Exchange C sends aclear-back signal to exchange B, which repeats the signal to exchange A Like intraexchange calls, interexchange calls are usually controlled

by the calling party (3.1.1) On receipt of the clear-back, exchange A stops charging, and enters the time when it received the clear-back in the billing record

of the call It also starts a 30-60,second timer It then awaits a clear-forward from calling party S1, or the expiration of the timer, and initiates the release of the connection when one of these events occurs

The release takes place in the following way Exchange A releases its path between S, and trunk T,, and sends a clear-forward signal to exchange B, which releases its path between T, and T2, and repeats the clear-forward to exchange C This exchange then clears its path between T, and called subscriber S,

When exchanges B and C have completed the release of respectively T, and T,, they send release-guard signals (to respectively exchanges A and B) When A and B receive the release-guard, they know that they can again seize respectively T, and T2 for new calls

This scenario is typical for CAS signaling in general, but we shall encounter variations that are specific to individual CAS signaling systems

4.1.2 Groups of Interexchange Signals

Of the four groups of subscriber signals discussed in Section 3.1.2, three also exist in channel-associated interexchange signaling

Supervision Signals (also known as line signals) The signals in this group represent events that occur on the trunk, such as, seizure, proceed-to-send, answer, clear-forward, etc While the majority of supervision signals is used in all CAS systems, there are system-specific differences in the sets of supervision signals

Address Signals (also known as selection signals, digits, or register signals) The digits are used primarily to indicate the called number, but can also have other meanings Like dial telephones, early CAS systems used dial-pulse address signals The systems described in this chapter have multi-frequency (MF) address signaling, similar to the DTMF (dual-tone multi-frequency) signaling

in today’s pushbutton telephones The MF frequencies are system-specific, and different from those used in DTMF

Tones and Announcements (Ringing-tone, busy-tone, etc.) The tones and announcements in interexchange signaling are the same as in subscriber signaling (3.2.5)

4.1.3 CAS Signaling Equipment at the Exchanges

We consider a stored program controlled (SPC) intermediate exchange with a

digital switchblock (Fig 4.1-2) The switchblock provides temporary bi- directional 64 kb/s paths The digital multiplexed ports (DMP) are first-order TDM circuits (b) with frame formats as described in Section 1.5.2

All trunks have CAS signaling The attachment of analog and digital (PCM) trunks to the switchblock is as described in Section 1.7.2 First-order TDM multiplexes carrying PCM trunks (b) are directly connected to DMPs Analog trunks on first-order FDM transmission systems are first demultiplexed into individual four-wire analog circuits (d), which pass through trunk circuits (TC4) and then enter a first-order PCM multiplexer, where they are converted into TDM-multiplexed digital circuits (b)

outgoing supervision signals into the trunk, and detect supervision signals received on the trunk Each TC4 has a control channel (CC) It is used by the processor to order the TC4 to send signals, and by the TC4 to report received signals

The supervision signaling information for the trunks in first-order digital trunk multiplexes is in certain bits of the bit streams (b) Control channel (CC)

to the switchblock has two functions In the first place, it transfers processor commands to set up and release switchblock paths In the second place, it is used for communications between the processor and those DMPs that serve groups of multiplexed digital trunks The processor can order a DMP to send out a supervision signal on a specified trunk in the multiplex, and a DMP reports the supervision signals received from the trunks in its multiplex to the processor

is done with service circuits The exchange has pools of digital multiplexed

service circuits which are attached to DMPs on the switchblock Tone and announcement circuits (TAC) send tones and announcements Digit-senders (SND) and digit-receivers (RCV) send and receive MF address signals to and from the trunks These circuits are signaling-system specific If an exchange has trunks with two CAS signaling systems, it needs separate groups of senders and receivers for each system The circuits are controlled by the control equipment (processor) of the exchange, via control channels (CC)

To send address signals or tones/announcements on a trunk, or to receive address signals from a trunk, the processor seizes an available service circuit of the proper type, and commands the switchblock to set up a path between the circuit and the trunk It then orders a SND or TAC circuit to send specific digits

or tones, or orders a RCV circuit to report the received digits When the sending

or receiving has been completed, the path is released, and the circuit is returned

to its pool

The numbers of TAC, and RCV circuits in an exchange are small compared

to the number of CAS trunks on the

of each type for every lo-20 trunks

exchange, in the order of about one circuit

It is useful to introduce some terms that are frequently found in descriptions of call-control signaling, with the aid of Fig 4.1-1

Outgoing and Incoming These terms can be applied to trunks, and to exchanges An outgoing exchange seizes outgoing trunks, sends forward signals, and receives backward signals, on its outgoing trunks An incoming exchange receives forward signals, and sends backward signals, on its incoming trunks When discussing the signaling between exchanges A and B in Fig 4.1-1, exchange A is the outgoing exchange, and B is the incoming exchange When describing the signaling between exchanges B and C, exchange B is the outgoing exchange, and C is the incoming exchange

Trunk T, (which is seized by exchange A), is regarded at A as an outgoing trunk, and at B as an incoming trunk In the same way, trunk T2 is an out- going trunk at exchange B, and an incoming trunk at exchange C If T, is a one-way trunk, it is always an outgoing trunk at A, and an incoming trunk at B However, if T, is a bothway trunk, its role at exchanges A and B varies per call, depending on whether it has been seized by A or B The same holds true for trunk T,

Originating and Terminating Exchanges The originating exchange in a call is the local exchange serving the calling subscriber, and the terminating (or destination) exchange is the local exchange of the called subscriber In the example of Fig 4.1-1, exchange A is the originating exchange, and C is the terminating exchange

Overlap and En-bloc Address Signaling In the example of Fig 4.1-1, exchange A seizes trunk T1 after it has received the complete called number from subscriber S, Likewise, exchange B seizes trunk T, after receiving the complete number from exchange A This means that, once the exchanges receive a proceed-to-send, they send out the complete called number in one uninterrupted stream This mode of address signaling is called en-bloc register signaling

However, exchanges can generally make route decisions after receipt of just the initial part of the called number For example, if the called number in the example is a subscriber number, consisting of an exchange code followed by a line number (EC-LN), exchange A can seize trunk T1 after receipt of EC from the calling subscriber, and can then send EC to exchange B This exchange again can seize its outgoing trunk T, after receipt of the EC, and send the EC to exchange C Exchanges A and B thus send out the initial digits of the called number while still receiving the later digits of the number, which are sent as soon as they have been received from the calling subscriber, or from the preceding exchange in the connection This mode of register signaling is called overlap register signaling The decision to use en-bloc or overlap address signaling is made by the individual telecoms Overlap address signaling results

in faster call set-ups

Link-by-link and End-to-end Signaling Signaling by two exchanges at the two ends of a trunk is called link-by-link signaling In Fig 4.1-1, supervision and address signaling are link-by-link In general, supervision signaling is always link-by-link Address signaling is link-by-link in most, but not all, CAS signaling systems

In end-to-end address signaling, the digit sender in the originating exchange sends address signals successively to digit receivers in the second, and later exchanges in the connection

4.1.5 CAS Signaling Systems

The CAS signaling systems discussed in this chapter are:

Bell System multi-frequency (MF) signaling,

CCITT No.5 signaling, and

R2 signaling

This section describes the multi-frequency signaling system that was introduced by the Bell System after the Second World War [l-3] It is still in use today, mostly in local U.S networks A nearly identical signaling system,

known as the Rl signaling system [4] and defined by CCITT, is used on international trunk groups in the North American network (for example, groups between the U.S and Canada)

Supervision and address signaling are link-by-link The system can be used on one-way two-wire trunks, one-way and bothway FDM analog trunks, and one- way and bothway TDM digital trunks (see Sections 1.4.5 and 1.52)

The supervision signaling is described in Sections 4.2.1 through 4.2.3 The

MF address signaling is discussed in 4.2.4

4.2.1 Supervision Signaling

In Chapter 3 we have discussed the on-hook and off-hook states of a telephone

In Bell System MF signaling, we speak of the on-hook (idle) and off-hook (in use) states of a trunk These states can be different at the exchanges connected

by the trunk Each exchange continuously sends the trunk state at its end to the other exchange This is known as continuous two-state signaling Changes in trunk state are supervision signals

Supervision Signals The repertoire of supervision signals, and the corres- ponding state changes are:

Seizure Clear-forward

On-hook to off-hook Off-hook to on-hook

Answer Clear-back Proceed-to-send (wink)

On-hook to off-hook Off-hook to on-hook Off-hook pulse, 120-290 ms Consider a trunk between exchanges A and B If the trunk is a one-way trunk that can be seized only by exchange A, this exchange is the outgoing exchange for all calls, and sends forward signals Likewise, exchange B is always the incoming exchange, and sends backward signals If the trunk is a bothway trunk, the exchange that seizes the trunk for a call is the outgoing exchange for that call Figure 4.2-l shows the states, and the forward and backward supervision signals, for a typical call Exchange A is the outgoing exchange

Initially, the trunk is on-hook at both ends Exchange A seizes trunk T, and sends a forward off-hook (seizure signal) Exchange then connects a digit receiver to the trunk, and sends a (backward) wink signal After receipt of the wink, exchange A sends the digits of the called number When the call is answered, exchange B sends an off-hook (answer signal) During the conver- sation, both exchanges are sending off-hook

In this example, the called party clears first, and exchange B starts sending on-hook (clear-back signal) When the calling party clears, exchange A releases the trunk, and sends on-hook (clear-forward) In response, exchange B clears the trunk at its end

The signaling system does not include a release-guard signal Therefore, when outgoing exchange A releases the trunk, it starts a timer that expires after 0.75-1.25 s The exchange does not seize the trunk for a new call until the timer has expired This gives incoming exchange B the time to release the trunk at its end

Double Seizures On bothway trunks, double seizures of the trunk by both exchanges can occur After exchange A seizes a trunk (Fig 4.2-l), it expects to receive a backward change to off-hook that represents the leading edge of the wink signal However, this change may also mean that exchange B (at the distant end of the trunk) is sending a seizure signal After sending a seizure signal, the outgoing exchange thus has to time the duration of the received off-hook The nominal length of the wink is 140-290 ms Therefore, the exchanges are arranged to recognize a return on-hook within say 100-1000 ms as wink If the off-hook duration exceeds 1 s, a double seizure has been detected

There are several ways to deal with a double seizure For example, both exchanges can be programmed to release the trunk, and make a second attempt

to set up their calls, trying to seize a trunk in the same trunk group, or a trunk in

a later-choice group

4.2.2 Supervision Signaling on FDM Analog Trunks

FDM analog trunks can transfer frequencies between 300 and 3400 Hz The exchanges indicate the states of the trunk with a 2600 Hz signaling tone The tone is in band (audible), and should be off when the trunk is carrying a call Therefore, tone-on-idle signaling is used:

Exchange A Trunk Exchange B

A

t

Figure 4.2-2 Four-wire analog trunk circuits (TC4)

Off-hook (in use) Tone off Sending and Receiving the Signaling Tone Figure 4.2-2 is a simplified presentation of the signaling circuitry in the four-wire trunk circuits TC4 of a trunk At each exchange, asignaling tone source (STS) supplies all trunk circuits with the 2600 Hz signaling tone The sending of the tone is controlled by the exchange processor, which controls switch X in the TC4 In the figure, the processor at exchange A has sent an off-hook command to TC4-1, and no tone

is sent on the send channel (S) of the trunk At exchange B, the processor has sent an on-hook command to TC4-2, and switch X connects the tone on the S channel of the trunk In the TC4s, a 2600 Hz tone detector (TD) is bridged across the receive channels (R) In this example, the TD in TC4-2 detects no tone, and reports to its processor that the trunk is off-hook at exchange A The

TD in TC4-1 detects the tone, and reports that the trunk is on-hook at exchange B

There are several aspects of in-band supervision signaling that pose additional requirements on the circuits of the TC4s

Blocking Received Signaling Tone During the conversation, both ex- changes indicate off-hook, and no signaling tone is present in either direction However, in other call states, the signaling tone and other voiceband signals can

be present simultaneously on a trunk Consider a call from subscriber SI, served by local exchange A, to a subscriber served by local exchange C The connection passes through intermediate exchange B Figure 4.2-3(a) shows the

Exchange A Trunk T, Exchange B Trunk T, Exchange C

transmission path in direction C -+ A only The called party has not yet answered, and exchange C has attached tone/announcement circuit (TAC) to trunk T, The circuit is sending ringing-tone At points (p), only ringing-tone is present However, trunk T2 is on-hook at exchange C, and trunk T, is on-hook

at exchange B, and at points (q) the ringing-tone and signaling tone are both present

There are two reasons why TC4-1 and TC4-3 should pass the ringing-tone (or other voiceband signals), but block the signaling tone In the first place, S, should hear ringing-tone only In the second place, supervision signaling is link- by-link This means that the signaling tone from B to A on trunk T, should be controlled by exchange B Therefore, TC4-3 has to block the received signaling tone, which otherwise would “leak” into T,

Figure 4.2-3(b) shows how received signaling tone is blocked Band- efiminationj?lter (BEF) blocks 2600 Hz, but passes other voiceband frequencies The filter is inserted in receive channel R by switch Y, which is controlled by TD When TD is not receiving signaling tone, switch Y is in the position shown, and BEF is not in the receive channel However, when TD receives signaling tone,

it sets Y in the other position, and BEF is inserted into the channel

Exchange A Trunk T, Exchange B Trunk T3 Exchange C

In the first place, a tone detector (TD) does not simply detect the presence of the 2600 Hz tone As shown in Fig 4.2-4(b), the signal received from the trunk

is fed into band-pass filter BPF that passes a narrow band of frequencies around

2600 Hz only, and to band-elimination filter (BEF) that passes all voiceband frequencies, except the frequencies passed by BPF Comparator (COMP) compares the signal strengths at the outputs of the filters

When signaling tone is being received, the output of BPF is stronger than

the output of BEF, even when speech is present at the same time TD thus indicates tone-on (on-hook) to the exchange processor When only speech is received, the output of BEF is stronger than the output of BPF, because speech power is mainly concentrated in the 300-1000 Hz range, and only a small part

of it lies around signaling frequency In this condition, TD indicates off-hook

As a second safety measure, exchange processors do not recognize received changes to on-hook or off-hook as a valid signals until the new state has persisted during a recognition time of a certain length With the exception of winks, the recognition time for most signals is 30 ms However, as an extra protection against talk-off, the recognition time for clear-forward signals is 300 ms Blue-box Fraud In the mid-1960s, the Bell System became aware of “blue box” fraud on FDM analog trunks [3] In principle, all signaling systems with in-band supervision signaling are vulnerable to this type of fraud, in which a subscriber generates in-band supervision and address signals with a device- called “blue box” -attached to the subscriber line The box can generate the

2600 Hz supervision tone, and the MF register signals With the box, a fraudulent subscriber can manipulate the network to avoid charges on long- distance calls

An example is shown in Fig 4.2-5 Suppose that subscriber S, served by local exchange A, has placed a call to an operator at exchange B, and that the connection has been set up on trunk analog trunk (T) Since S has dialed 0 for operator assistance, exchange A does not charge the call, expecting the charging

to be done at exchange B

Once the connection has been set up, exchange A cuts through and there is a voiceband path from S to B This allows S to send blue-box signals to B Local exchange A is unaware of this because, once the connection has been set up, it

Seizure

4 Called Number

Wink

w

Figure 4.2-5 Blue-box signaling Exchange B connects trunk T to digit receiver RCV after receiving the wink signal

monitors only the on-/off hook condition of the subscriber line However, the blue-box signals are accepted in good faith by exchange B

Before the operator has answered, S sends a burst (say l-2 s) of signaling tone This is interpreted by exchange B as a clear-forward signal, and the exchange releases trunk T at its end The end of the burst is interpreted at exchange B as a new seizure signal Subscriber S waits about one second after the end of the signaling tone, and then sends the called party’s number The call

is set up and, since exchange B has not received a dialing pattern that indicates

an operator-assisted call, it assumes that exchange A will charge the call The Bell system has curbed blue-box fraud by vigorous legal prosecution of fraudulent subscribers, and by implementing protective procedures at the exchanges

4.2.3 Supervision Signaling on Digital (PCM) Trunks

In the example of Section 4.1.3, the supervision signaling for trunks in first- order PCM multiplexes is sent and received by the digital multiplex ports (DMP) that attach the PCM multiplexes to the switchblock of the exchange The DMPs have a control channel to the exchange processor, and report the received state (on-hook, off-hook) of the trunks in their multiplexes to the processor The processor commands the state to be sent out on the trunk The DSl frame format of the North American Tl first-order digital transmission systems (1.5.2) is shown in Fig 4.2-6 It consists of consecutive frames Each frame has 24 eight-bit time slots (TS), and one F bit Each time slot

is associated with a particular trunk

Twelve frames form a superframe, in which the frames are numbered from 1 through 12 The DMPs maintain frame and superframe synchronization with the incoming bit stream by locking onto the F bits of the frames, which exhibit a repeating 12.bit pattern:

Superframe

I I(

When locked on to this pattern, a DMP can determine the start of each frame, and of each superframe, in the bit stream In frames 6 and 12, bits 8 (the least important bits of the &bit PCM codes) in the 24 channels are used for supervision signaling This is known as bit robbing The effect of bit robbing on the quality of PCM-coded speech is negligible

The signaling bits in frames 6 and I2 are known as the S, and S,, bits The DMPs update their outgoing signaling bits every 1.5 ms (once per superframe), which is sufficiently fast for supervision signaling The combinations of an S, and

S, bit could indicate four trunk states However, the S, bit in each time slot is set equal to the previous S, bit, resulting in two-state continuous supervision signaling The bit values 0 and 1 represent respectively on-hook and off-hook The signaling bits cannot be heard by the subscribers, and the subscriber’s speech, or a blue box, cannot corrupt the supervision signals This avoids the problems associated with in-band signaling

The MF address signals are combinations of two voiceband frequencies- chosen from a set of six frequencies [1,2,5] The frequency assignments are shown in Table 4.2-l Only 12 of the 15 possible two-out-of-six codes are used in CCITT-Rl signaling Bell MF signaling uses some of the remaining codes in calls that are set up with operator assistance [5]

Address signaling sequences start with a KP (start-of-pulsing) signal, and end with an ST (end-of-pulsing) signal Signaling sequences received without

KP or ST are considered to be mutilated, and discarded by the incoming exchange

The KP signal has a duration of 90-110 ms The duration of the other signals

is 61-75 ms The originating exchange sends the complete called number (en- bloc address signaling), with silent intervals between signals of 61-75 ms The intermediate exchanges can use overlap address signaling

Table 4.2-1 Bell system multi-frequency address signals

Address Signaling Sequences In its simplest form, an address signaling sequence conveys the called number only The called number can be a sub- scriber number, or a national number Let AC(3), EC(3), and LN(4) represent

a three-digit area code, a three-digit exchange code, and a four-digit line number (1.3.1) The address signaling sequences are then: KP-EC(3)-LN(4)-

ST (subscriber number), or KP-AC(3)-EC(3)-LN(4)ST (national number)

A calling subscriber dials subscriber numbers for calls inside his numbering plan area (NPA), and national numbers for calls outside his NPA An originating exchange that receives a national called number from the calling subscriber, or an intermediate exchange that receives a national called number from the preceding exchange in the connection, sends out the called number as either a national number, or a subscriber number This depends on whether the exchange has seized an outgoing trunk to an exchange outside, or inside the NPA

of the called party

Automatic Number Identification In the early years of subscriber-dialed long-distance calling, most local exchanges were not equipped to produce billing records for these calls, and the billing records were generated by the first intermediate (toll) exchange in the connection The local exchange would send both the called and calling numbers, and the toll exchange would handle the charging for the call After the end of the call, the toll exchange would generate

a billing record that included both numbers, the date of the call, and the times

of answer and call clearing The sending of calling numbers is known as automatic number identification (ANI)

After the break-up of the Bell system, calls between subscribers in different local (LATA) networks are billed by either the LATA of the calling party, or by the interexchange carrier (IC) This is a matter of mutual agreement between the LATA and IC carrier

Figure 4.2-7 shows the supervision and address signals for an inter-LATA call originated by subscriber S The call is to be billed by IC exchange B

A subscriber can designate a “default” IC for his inter-LATA calls This information is stored at the local exchange If a subscriber just dials a called number, the exchange routes the call to an exchange of his default IC If subscriber S desires a different carrier for the call, he dials a prefix 10xXx, where= identifies the IC

After the subscriber has dialed:

l-AC(3)-EC(3), (Called national number, default IC), or lOXXX+AC(3)-EC(3) (Called national number, specified IC) or

exchange A knows the desired IC, and the nearest exchange (B) of that IC This example assumes that a direct trunk group connects exchanges A and B Exchange A seizes a trunk T in this group, and sends a seizure signal After

Figure 4.2-7 Transfer of calling and called numbers to exchange B of interexchange carrier

receiving the wink signal, A first sends the national number of the calling subscriber:

KP-I(2)-AC’(3)-EC’(3)-LN’(4)ST

When the subscriber has finished dialing LN(4), the last four digits of the called number, exchange A sends a second digit sequence, which identifies the called national- or subscriber number: KP-AC(3)-EC(3)-LN(4)-ST, or KP- EC(3)-LN(4)ST Exchange B then acknowledges the receipts of both numbers with a wink signal

Sending the calling number first minimizes the elapsed time from the end of address signaling by the calling subscriber to the end of address signaling by the local exchange

The codes in the two information digits I(2) ahead of the calling number characterize the calling line:

I(2) = 00 Identified subscriber line

I(2) = 02 AN1 failure (calling number not included)

I(2) = 06 Call from hotel without room identification

I(2) = 10 Test call

When I(2) = 02 or 06, an operator at exchange B verbally obtains the calling number

4.2.5 Failed Set-ups

Bell MF signaling does not include backward signals to indicate that the set-up

of a connection has failed The exchange where the failure occurs sends a tone

(busy-tone, reorder-tone) or announcement, and the calling party disconnects (sends a clear-forward to the originating exchange) The originating exchange then initiates the release of the connection

4.3 CCITT NO.5 SIGNALING

This signaling system has been developed jointly by the U.K Post Office and Bell Laboratories, and is similar to Bell System MF signaling It was adopted in

1964 by CCITT for use in the international network, and is documented in CCITT Recommendations [6]

CCITT No.5 has been designed especially to operate on TASI-equipped analog trunks (l-6.4) It has been used extensively on long international trunks (underwater transoceanic trunks, satellite trunks) Despite its age, it is still in use

on several international trunk groups

Supervision and address signaling are both link-by-link and in-band

4.3.1 Supervision Signaling

Supervision signals consist of one or two in-band signaling tones TASI compatibility requires a special form of supervision signaling The continuous two-state signaling described in 4.2.1 would defeat the purpose of TASI equipment, because signaling tones would be present in both directions when a trunk is idle, and each idle trunk would therefore occupy a pair of TASI bearer channels

A number of CAS signaling systems use pulsed supervision signals, in which the signals are short bursts (typically 50-150 ms) of signaling tone [ 11 However, this form of signaling is also incompatible with TASI trunks, becausefieezeouts (1.6.4) could shorten the pulses beyond recognition, or suppress them completely

Therefore, compelled (sometimes called continuous-compelled) supervision signaling is used An exchange that sends a supervision signal keeps the signaling tone(s) “on” until it receives an acknowledgment signal (also consisting of one or two signaling tones) from the exchange at the distant end of the trunk

Figure 4.3-l shows a supervision signal on a trunk T between international switching centers (ISC-A) and (ISC-B), as observed on the S (send) and R (receive) channels of the trunk at ISC-A At t,, ISC-A turns on the signaling tone(s) When ISC-B receives the signal it responds with an acknowledgment At t2, the signaling tone(s) of the acknowledgment arrive at ISC-A, and this exchange turns off its signaling tone(s) at t3 ISC-B then notices that the signal from ISC-A has ended, and stops sending its acknowledgment signal Finally, ISC-A notices the end of the acknowledgment at t, Compelled signaling thus successfully transfers a supervision signal, even when the signal or its acknowledgment is subjected to a TASI freeze-out

ISC-A Trunk ISC-B

Signaling Frequencies Two in-band signaling frequencies are used: fi =

2400 Hz, andf, is 2600 Hz There are three signals:f,,f,, and a composite signal that containsf, andf,

Signaling Circuitry in Trunk Circuits The functions and circuitry for super- vision signaling in the TC4 circuits of analog CCITT No.5 trunks are similar to those discussed in 4.2.2 The tone detectors (TD) have to detect the presence of the signaling frequenciesf, and/ or f2 on the receive (R) channel of the trunk Since the signaling frequencies are i&band, a TD that detects their presence has

to block them from entering the switchblock of the exchange In addition, the tone detectors include circuitry that prevents the simulation of signaling tones by the speech (or other voiceband communications) of the subscribers

Because of its in-band supervision signals, CCITI’ No.5 signaling is vulnerable to “blue box” fraud

4.3.2 Supervision Signals and Call Handling Procedures

Table 4.3-l lists the supervision signals, their acknowledgments, and the signaling frequencies A particular physical signal can have several logical meanings, depending on its direction (forward or backward), and the state of the call The supervision signals for a typical call on a trunk seized by ISC-A in Fig 4.3-l are outlined below

Table 4.3-l CCITTNo.5 supervision signals and their acknowledgments

fi = 2400 Hz& = 2600 Hz fwd: forward bkwd: backward

Source: Rec Q.141 Courtesy of ITU-T

Seizure and Proceed-to-send On receipt of the seizure signal, ISC-B attaches a MF digit receiver to the trunk, and then acknowledges with a proceed-to-send signal When ISC-A receives the acknowledgment, it sends the called number in a sequence of address signals

Answer and Clear-back These backward signals have their conventional meanings, and are simply acknowledged at outgoing exchange ISC-A

Clear-forward and Release-guard At the end of the call, outgoing ISC-A releases trunk (T) at its end, and sends a clear-forward signal ISC-B clears the trunk at its end, and then acknowledges with a release-guard signal This indicates that the trunk can be seized again for a new call

Busy-flash is a backward signal sent by ISC-B to indicate that it cannot extend the call set-up (for example, because no trunk to/towards the call destination is available) On receipt of the signal, ISC-A acknowledges It then releases the trunk, and sends a clear-forward signal to ISC-B

If the call arrived at ISC-A on a trunk that has a signaling system which includes signals or messages to indicate the failure of a call set-up, ISC-A informs the preceding exchange, which then releases its trunk to ISC-A

Otherwise, ISC-A connects its incoming trunk to a tone or announcement circuit This alerts the calling subscriber, who then sends a clear-forward to her local exchange, and that exchange initiates the release of the connection The busy-flash reduces the amount of time that international trunks are held in failed set-ups

forward-transfer is used on calls that are set up with operator assistance (Section 4.3.5) The signal is a pulse of nominally 850 ms, and is not acknowledged

Double Seizures Suppose that ISC-A seizes a bothway trunk, and starts sending the seizure signal The seizure may arrive at ISC-B with a delay of up to

600 ms (on satellite circuits) During that time ISC-B may also have seized the trunk

A double seizure is detected when an exchange that has sent a seizure signal receives a fi (seizure signal) instead of the expected fz (proceed-to-send) acknowledgment Both ISC-A and ISC-B detect the double seizure In response, they release the trunk at their respective ends Clear-forward signals are not sent International exchanges are usually arranged to make a repeat attempt to set up a connection that has failed because of a double seizure

Recognition Times An exchange recognizes-and accepts-a received supervision signal only when it has persisted for a certain amount of time The nominal recognition time for the seizure and proceed-to-send signals is 40 ms The recognition times for all other signals are 125 ms

4.3.3 Multi-frequency Address Signaling

CCITT No.5 address signals are combinations of two frequencies, selected from the same set of frequencies (700, 900, 1100, 1300, 1500, and 1700 Hz) that is used in Bell System MF signaling

To avoid clipping of the address signals by TASI equipment, the address signals are sent en bloc The silent interval between the end of the seizure signal and the start of the first address signal is at most 80 ms, and the silent intervals between address signals are 55 ms These intervals are shorter than the 400 ms TASI “overhang” intervals (1.6.4), and the TASI bearer channel that was assigned to the trunk for the seizure signal is therefore not released until all address signals have been sent

Address Signals Table 4.3-2 lists the frequencies of the address signals All address signaling sequences start with a KP signal The KPl signal indicates that the called number is in the form of a national number, and KP2 indicates that the called number is an international number

Code 11 and Code 12 are one-digit addresses for international operators at an ISC

All address signaling sequences include a Z-digit (or: Language digit, Discriminating digit) The value of the digit indicates the calling party category: 0: Subscriber