Note: Descriptions are shown in the official language in which they were submitted.
11140~Z
This invention relates generally to telephone
conferencing circuits, and more particularly to telephone conferencing
circuits employing digital techniques and wherein the digital signals
involved are neither summed nor decoded for the purposes of conferencing.
Conferencing circuits are well known in the field of
telephony. In general terms, a conference circuit is a circuit for
allowing three or more parties (or conferees) to talk to one another
at the same time. For the purposes of this description, a conference
circuit will also include a conversation between two parties, since
one embodiment of the invention to be described allows for conversations
between two to five conferees.
Early conference circuits, employed in analogue telephone ~ -
systems, provided conferencing by summing all the signals of all the
participants and transmitting this resultant signal to all the
conferees, with the exception of the talker who received the resultant
signal minus his own signal. As the telephone technology advanced
into the world of digital techniques, simple summing and subtracting
no longer provided an easy solution to the problem of conferencing.
Some prior art approaches to conferencing with digital
techn~ques were simply to convert the digital signals to analogue
signals, perform an analogue conferencing, and re-convert the resultant
analogue conference signal into a digital signal. One example of
such an approach is shown in U.S. patent 3,970,797 dated July 20, 1976
to D.A. Johnson and Wm. C. Towle. It is, however, cumbersome to
conference in this manner if it is possible to conference directly in
digital format. Additionally, the converting to analogue and
reconverting to digital adds distortion to the signals involved.
An improvement of the analogue summing of signals
for conferencing is to do the summing directly with digital signals.
Since the digital signals are commonly not linear, but rather are
Pulse Code Modulated (PCM, e.g. the ~-255 code is used in North America),
. ~ -- 1 -- ... .
' ' ,'',
- ~ . - . - ...... . . . . . . ..................... . . .
. ,., . . : ~ ., . - ., . .~ ~ . . ' . . ' . '- . '
O~:
it is necessary to first linearize the digital signals, add them,
and then re-code (all the while remaining in the digital domain).
U.S. patent 3,924,082 dated December 2, 1975 to S.E. Oliver and
N.R. Winch describes one such system.
A further modification in conferencing circuits is to
provide a digital conferencing circuit which performs the conferencing
function directly using the coded digital signals. U.S. patent
4,007,338 dated February 8, 1977 to D.W. McLaughlin, U.~. patent
4,022,~91 dated May 10, 1977 to M.J. Kelly et al, and U.S. patent
4,031,328 dated June 21, 1~77 to S.G. Pitroda, all depict such a
conferencing circuit.
The aforementioned conference circuit of the wholly
digital type (exemplified by U.S. patent 4,007,338) ;s well described
at column 2, lines 34 to 52 of U.S. patent 4,007,338, wherein it is
stated: -
"In other words, assuming, for example that channels 3,
5 and ~ are engaged in a 3-way conference, during the channel 3 time
slot, the samples from channel 5 and channel 9 are read and compared,
and the larger of the two samples is transmitted to channel 3.
Subsequently, during the channel 5 time slot, the samples from
channel 3 and channel 9 are read and compared, and the larger of the
two samples is transmitted. During the channel 9 time slot, the
operation is repeated, with the largest of the channel 3 and 5 samples
being transmitted. It can be seen that during each channel time slot,
the incoming information is written into the fixed channel number in
the information memory and the outgoing information is read from some
other channel specified in the control memory. This operation of
writing and reading the information memory goes on continuously during
each channel. The next time frame, new PCM [pulse code modulation]
samples are written into the information memory."
The conference circuit disclosed and claimed herein,
. . .
, . . ~. . ..
U~z
produces the sa~e result as does the ju5t described conference circuit
of U.S. patent 4,007,338, i.e. assuming a three-way conference between
channels 3, 5 and 9, during the channel 3 time slot the sa~ples from
channels 5 and 9 (previously stored) are compared and the larger of
the two samples is transmitted to channel 3, etc.
However, the conference circuit of the present invention
performs this operation in a different manner, which is made more
apparent later in this description, and performs it with circuitry that
is much less complicated and involved. In simplistic terms, the
conference circuit of U.S. patent 4,007,338 requires memory for each
conferee and it operates on the stored data in a parallel fashion.
The conference circuit of the present invention requires memory for
only N-l of N conferees, and it operates on the stored data in a
serial fashion. In short, the present invention provides a relatively
simple apparatus and method for conferencing in the digital domain.
In simplistic terms, one embodiment of the present
învention provides for a three party conference in the following
manner. A digital time slot interchanging network (i.e. a time
switch~ intermediate the digital switching network and the conference
circuit functions to present the TDM (time division multiplex) PCM
signals for the three conferees on a common PCM bus, each PCM signal
in a distinct time slot on the bus. The conference circuit itself
comprises two eight-bit shift registers connected in a series circuit
relationship. The first eight-bit shift register is selectively
responsive to the signals (i.e. PCM words) appearing on the PCM bus
only during the time slots of the three conferees. The second eight-
bit shift register is selectively responsive to the signals stored in
the first shift register and receives signals from the first shift
register whenever the first shift register is receiving data. The
contents of the two shift registers are examined and the contents of
the register containing the largest magnitude binary number ~ignoring
.,, . - , . ~ : ...................... ~ ................. .. .
.,
~ 4(~F~2
the sign) is output to a transmit PCM bus. This transmit PCM bus
is interfaced with a digital time slot interchanging network
(i.e. a time switch~ to match the time slots of the conference circuit
with the time slots of the respective conferees in the digital switching
network.
The following simple example will help to illustrate
more clearly the operation of the invention. Assume that there are
three callers (or conferees) AA, BB and CC who desire a conference or
three part~ cal~. When the proper interconnections are made in a
digital switching network the circuit functions as follows: the
speech ~aveforms of conferees AA, BB and CC are sampled periodically
(e.g. 8000 times per second) and encoded to produce digital (i.e. PCM)
speech samples. The resultant PCM speech samples are interleaved
~i.e. time multiplexed) so that several encoded waveforms can be carried
on one PCM bus. PCM samples from conferee AA appear on the bus only
during time slot A; PCM samples from conferee BB appear on the bus
only during time slot B; and PCM samples from conferee CC appear on
the bus only during time slot C. In the general case, these time
slots can appear on any PCM buses in the switching network.
In order to provide the conferencing function, the first
step is to gather the time slots from the three conferees onto a
common PCM bus and assign each one a distinct time slot. This function
is performed by the digital time slot interchanging network (i.e. time
switch~ such that party M is assigned time slot X on the common
conference bus, party BB is assigned time slot Y on the common conference
bus, and partyCC is assigned time slot Z on the common conference bus. ~:
Assuming the repetitive cyclic order of X, Y and Z for the time slots,
the information in time slot X (i.e. party AA) is stored in the first
shîft register. When time slot Y appears, the contents of the first
register are shifted into the second register and the contents of time
slot Y ~i.e. party BB) are stored in the first register. When time slot Z
- 4 -
- - - . . .. .. ~ . . . . .
- ... .. : .. : , .: . . .. .: ~ . .
- . - . . . . . . . .
... . .. .. . ... . . .
4~2
appears, the contents of the first register are shifted into the second
register and the contents of time slot Z (i.e. party CC) are stored in
the first register, etc., for subsequent time slots X, Y and Z until
the call is terminated. Just before the time that the first time
slot Z appears, a decision is made as to the absolute magnitudes of
the digital numbers stored in the two registers. The contents of the
register having the largest digital number (ignoring sign) are applied
to the transmit PCM bus of the conference circuit, during time slot Z.
A digital time slot interchanging network interfaces the transmit PCM
bus with the digital switching network so that time slot Z is
connected to the time slot of party CC in the switching network, etc.
Stated in other terms, the present invention is a
conference circuit for connection in a digital telephone system
intermediate a first TDM bus for carrying, in N distinct time slots,
PCM words to the conference circuit, and a second TDM bus for
carrying, in N distinct time slots, PCM words from the conference
circuit, wherein N is a positive integer, 2 ' N ' 128, the conference
circuit comprising: serial storage means for storing the N-l most
recent PCM words received from the first TDM bus during N-l of the
N time slots; circuit means for determining, according to a
predetermined criterion, which one of the PCM words stored in the
storage means meets the criterion; and routing means for routing
the PCM word meeting the criterion to the second TDM bus in the
next one of the N time slots.
Stated in yet other terms, the present invention is a
conference circuit for connection in a digital telephone system
intermediate a first TDM bus for carrying, in N distinct time
slots of each frame, PCM words to the conference circuit, and
a second TDM bus for carrying, in N distinct time slots of each
frame, PCM words from the conference circuit, wherein N is
a positive integer, 2 ' N ' 5, the conference circuit comprising:
-,
~., . j . .. ; . : . . :
11~40~Z
N-l shift registers for storing the N-l most recent PCM words received
from the first TDM bus, during N-l of the N time slots; magnitude
comparator means for determining which one of the shift registers
contains the PCM word having the largest absolute binary value; and
routing means for routing the contents of the shift register containing
the PCM word having the largest absolute binary value to the second
TDM bus in the next one of the N time slots.
Expressed in still other terms, the present invention
is a method of providing a conference circuit interconnection in a
digital telephone system for an N party conference, wherein N is a
positive integer, 2 < N < 32, and wherein the telephone system has a
first TDM bus for carrying, in N distinct time slots of each frame,
PCM words to the conference circuit, and a second TDM bus for
carrying, in N distinct time slots of each frame, PCM words from
the conference circuit, the method comprising: storing the N-l
most recent PCM words received on the first TDM bus during the
N-l of the N time slots; determining, according to a predetermined
criterion, which one of the N-l PCM words that have been stored
meets the criterion; and outputting on the second TVM bus, during
the next one of the N time slots, the PCM word meeting the criterion.
The invention will now be described in more detail ~ -with reference to the accompanying drawings, wherein like parts in
each of the several figures are identified by the same reference
character, and wherein:
Figure 1 is a greatly simplified block diagram of a
typical telephone system showing the interconnection therewith of
the present invention,
Figure 2 is a simplified block diagram of the preferred
embodiment of the present invention for a three-party conference circuit;
Figure 3, comprising parts a, b, c and d, i5 a timing
diagram, useful for understanding the operation of the circuit depicted
in Figure 2;
- 6 -
.-
~ .. .
-- 11140~Z ::
Figure 4 is a simplified block diagram of the preferred
embodiment of the present invention for a conference circuit having
between two and five conferees.
The conference circuit of the present invention will
first be described in general terms in the context of the environment in
which it operates, as exemplified by Figure 1. Figure 1 depicts three
telephone sets 10, 11 and 12 which are employed by three conferees, one
conferee per telephone set. Each telephone set 10, 11 and 12 has
associated with it a line circuit 13, 14 and 15 respectively, as is well
known. Line circuit 13 converts the analogue signal that it receives
from telephone set 10 to a PCM (pulse code modulation) signal on a TDM
(time division multiplex) line 16. Additionally, line circuit 13 converts
the digital PCM signals received on line 17 into analogue signals and
relays these analogue signals to telephone set 10 via analogue line 18.
Line circuit 14 functions in a similar manner with lines 19, 20 and 21
corresponding respectively to lines 16, 17 and 18; similarly line
circuit 15 functions in a similar manner with lines 22, 23 and 24
corresponding respectively to lines 16, 17 and 18. Digital switching
network 25 is any standard digital switching network that would serve to
interconnect any of the telephone sets 10, 11 and 12 together, in a one-
to-one relationship to permit a two-party interconnection. As a detailed
description of digital switching network 25 is not required for an
understanding of the present invention, its operation will not be
described further.
Conference circuit 30 is connected intermediate a
conference receive TDM bus 31 and a conference transmit TDM bus 32 as
shown in Figure 1. TDM bus 31 and TDM bus 32 are each a time division
multiplex bus providing thirty-two time slots, each time slot containing
one 8-bit PCM word. Time switch 33 interfaces the receive TDM bus 31
and the lines 16a, l9a and 22a from switching network 25. Line 16a
carries, in a particular time slot, the information from line 16, after
it has been switched by switching network 25. Similarly, line l9a
7 -
~il4~
carries, in a particular time slot, the information from line 19, and
line 22a carries, in a particular time slot, the information from line 22.
The function of time switch 33 is to interconnect the
information (i.e. PCM word) contained in a particular time slot on line 16a
with one particular and distinct time slot on TDM bus 31. Similarly, time
switch 33 functions to interconnect the information contained in a
particular time slot on line l9a with one particular time slot on TDM
bus 31, different from the time slot assigned to the PCM word from
line 16a. And finally, time switch 33 functions in a similar manner to
assign to the time slot of interest on line 22a a distinct time slot on
TDM bus 31 distinct and different from the time slots assigned to the PCM
words ~rom lines 16a and 22a. This can be seen in Figure 3a, wherein the
time slots of interest on TDM bus 31 are indicated by the letters X, Y
and Z.
The operation of conference circuit 30 will be discussed
later in greater detail in conjunction with Figure 2. In simplistic
terms, conference circuit 30, in the three-party conference depicted in
Figure 1, receives PCM words to be conferenced during three different
time slots (i.e. X, Y and Z~ out of the thirty-two time slots available
on TDM bus 31. Conference circuit 30 selects (according to a criterion ~ -
to be discussed later) one PCM word which has been received during one
of the time slots of interest and outputs that PCM word on transmit
TDM bus 32. Time switch 34 serves to assign (or allocate) one time
slot of interest on bus 32 to the appropriate time slot on line 17a.
Similarly, time switch 34 allocates the remaining two time slots of
interest on bus 32 appropriate time slots on lines 20a and 23a,
respectively. Line 17a is connected to line 17 via switching network 25,
and similarly, lines 20a and 23a are connected to lines 20 and 23
respectivel~ via switching network 25.
3Q Figure 2 depicting, in simplified form, the actual
circuitry of conference circuit 30 will now be described. The circuitry
is interconnected as shown in Figure 2 and attention is directed to
- 8 -
- , . . . , . ~ ~. , .
~ ~L1~L4 0 ~Z
it. Figure 3 contains waveforms helpful in understanding the operation
of Figure 2 and it may be useful to refer to Figures 2 and 3
simultaneously.
The 8-bît shift registers 37 and 38 are enabled by a
logic signal from OR gate 39. OR gate 39 is responsive to three
signals from time slot clock 40. Clock 40 is designed to output a
high logic signal on each of its output lines (only three of which
are shown) for the duration of the time slots of interest. In other
words, line 41 carries a high logic signal only for the duration of
time slot X, line 42 carries a high logic signal only for the duration
of time slot Y, and line 43 carries a high logic signal only for the
duration of time slot Z. Clock pulses (from a bit clock, not shown)
are applied continuously to clock pulse inputs 44 and 45 of shift
registers 37 and 38 resepectively. The clock pulses have a frequency
o~ eight times every time slot and there are thirty-two ti~e slots for
each 125 microsecond frame.
Let the time slot of interest on line 16a (Fig. 1,
i.e. the time slot on line 16a carrying information from telephone
set 10) be referred to as time slot A, containing a PCM word "a".
Similarly, let the time slot of interest on line l9a be referred to
as time slot B containing a PCM word "b", and the time slot of interest
on line 22a be referred to as time slot C containing a PCM word "c".
Time switch 33 ~Fig. 1) has switched or rearranged
time slots A, B and C so that they now appear during time slots X, Y
and Z respectively, of TDM bus 31 (see Fig. 3a). It should be noted
that the time slots X, Y and Z can be any three different time slots
of the thirty-two time slots (or channels) available on TDM bus 31.
As can be seen from Figure 3a the PCM word "a",
originally in time slot A, appears in the second time slot of the frame,
i.e. time slot X of TDM bus 31. PCM ~ord ''a" is stored in 8-bit shift
register 37. Figure 3b depicts the contents of shift register 37,
_ g _
, . . .
- :
40N;2
and from Figure 3b it can be seen that word "a" has been completely
entered into register 37 at the end of time slot X and it is stored
in register 37 until the beginning of time slot Y. At the beginning
of time slot Y, word b starts to be entered into register 37 and is
completely entered by the end of time slot Y. At the same time that
word b is being entered into register 37, word a is being shifted out
of register 37 and into register 38. At the end of time slot Y, when
word b has been complete1y entered into register 37, word a has been
completely entered into register 38 (see Figure 3c). Words a and b
are stored in registers 38 and 37 respectively until the beginning
of time slot Z.
During the interval just before words a and b are
completely entered into registers 38 and 37 respectively, magnitude
comparator 48 compares the magnitude of the four most significant bits,
exclusive of the sign bit, of each word a and b in the registers 38
and 37 respectively. It should be noted that this magnitude comparison
occurs before the PCM words "a" and "b" are completely entered into
their respective shift registers. This can be seen from Figure 2
sfnce the T input of comparator 48 is connected to the third through
sixth bit locations (counting from the left of the shift register) of
register 37; the S input of comparator 48 is similarly connected to
the third through sixth bit locations of register 38. Conse~uently,
when comparator 48 makes the comparison of interest, the four most
significant bits of the PCM word are contained in the third through sixth
bit locations of registers 37 and 38, the seventh bit location of the
register contains the sign bit and the eighth bit location contains the
least significant bit of the preceding PCM word stored in the register.
Magnitude comparator 48 provides a digital indication
(logic high or low~ on its output terminal 49 indicative of which
register 37 or 38 contains the PCM word having the larger magnitude
Cignoring 5ign~. The output on terminal 49 is high if S > T, otherwise
.'. .
- 10 -
-- ~il4G~2
the output is low. In the event that S = T, the output on terminal 49
is low, thereby giving priority to the contents of register 37.
The D terminal of D-type flip-flop 50 is responsive
to the logic signal appearing on the output terminal 49 of magnitude -
comparator 48. Flip-flop 50 has a clock pulse input 55 from a channel
clock ~not shown). The channel clock, applied to input 55, supplies
one pulse every time slot, coinciding with the last bit pulse of the
time slot and of the same duration as the bit pulse (recall that each
time slot, or channel, has eight bit pulses as produced by the bit
clock, and there are thirty-two time slots to one frame). Consequently, :
flfp-flop 5Q changes its output state only once per time slot, during
the time period of the last bit of the time slot so as to have the
correct output for application to AND gates 51 and 52 at the beginning
of a new time slot.
Flip-flop SO, AND gate 51, AND gate 52, and OR gate 53
form a routing circuit 54. Routing circuit 54 is interconnected as
shown in Figure 2 and functions to shift either the PCM word stored
in the register 37 to TDM bus 32 or the PCM word stored in register 38
to TDM bus 32, during the time period that the output of OR gate 39 is
high. Recall that the output of OR gate 39 is high- (i.e. the circuitry
is enabled~ only during time slots X, Y and Z. Consequently, a word
is output on TDM bus 32 only during time slots X, Y and Z. This can be
seen in Figure 3d.
Time switch 34 (Figure 1) allocates or switches the
PCM words appearing on TDM bus 32 so that during time slot Z, the
PCM word (a or b) appearing on TDM bus 32 at that time period is
switched to the appropriate time slot on line 23a; during time slot X,
the word (b or c) appearing on TDM bus 32 at that time period is
switched to the appropriate time slot on line 17a; and during time
slot Y, the word ~a or c) appearing on TDM bus 32 at that time period
is switched to the appropriate time slot on line 20a.
.'
.
- .. . .. , -:
~1140~Z
The operation of conference circuit 30 can be
summarized in simplistic terms as follows. PCM words are continually
appearing at the input of shift register 37 via TDM bus 31. Only
during the time slots X, Y and Z of bus 31 are the shift registers 37
and 38 enabled so as to receive selectively the appropriate PCM words
applied to the shift registers. In the simplistic example described
previously, a PCM word "a", appearing during time slot X is entered
into register 37. During the next time slot of interest on TDM bus 31,
i.e. time slot Y, PCM word "b" is entered into register 37, and the
PCM word "a" previously contained in register 37 is shifted into
register 38. The magnitude comparator 48 compares the four most
significant bits of the PCM words stored in registers 37 and 38
(i.e. PCM words a and b) to determine which PCM word is the largest
(excluding sign~ and produces a logic output signal at its output
terminal 4~ indicative of this. Routing circuit 54 is responsive to
the output of magnitude comparator 48 and routes the largest PCM word
(a or b~ to the transmit TDM bus 32 during the next time slot of
interest, i.e. during time slot Z.
During time slot Z, PCM word c appears on TDM bus 31
and is entered into register 37. PCM word b, previously contained in
register 37,is shifted into register 38. PCM word "a" previously
contained in register 38 is either routed via routing circuit 54
(in particular, via AND gate 51 and OR gate 53) to TDM bus 32 if it
were the largest of PCM words a and b or else it is simply "lost".
If PCM word b were the largest, as well as getting shifted into
register 38, PCM word b would also get routed via routing circuit 54
(in particular, via AND gate 52 and OR gate 53) to TDM bus 32.
During the next time slot X, another PCM word a is
introduced into register 37, PCM word c, previously contained in
register 37 is shifted into register 38, and PCM word b, previously
contained in register 38 is shifted out. The same selection of the
- 12 -
.
.. . . .. . . . . . . . . . .
1 1 1 4 0E~
largest PCM word (ignoring sign~ contained in registers 37 and 38
occurs, and the large~t of the PCM words contained in the registers is
routed to TDM bus 32, etc.
While the conference circuit 3Q of Figure 2 is the
preferred one envisaged by the inventors for their particular
application, variations can be made therein. One modification would
be to use more than just the four most significant bits (excluding the
sign bit~ to determine which PCM word contained in a register has the
largest numerical value. The five most significant bits, or the six
most significant bits of the PCM word could be used in the determination
of the largest value. It would also be possible to use less than four
bits to determine the PCM word having the largest absolute value.
Figure 4 will now be discussed. The conference circuit 60
of Figure 4 operates on the same principles as does the conference
circuit 3~ of Figure 2. Conference circuit 60 is capable of handling
a conference call of between two and five parties, as a consequence
its circuitry is more complex than that of conference circuit 30, but
since the basic principle is the same, conference circuit 60 will not
be described in great detail.
As with conference circuit 30 (Fig. 2), conference
circuit 6Q is connected between TDM bus 3l and TDM bus 32 as shown
in Figure 4. Shift registers 6l, 62, 63 and 64 operate in a similar
fashion to shift registers 37 and 38 of circuit 30 (Fig. 2). Each
shift register 61, 62, 63 and 64 has a clock pulse input 65, 66, 67
and 68 respectively from a bit clock (not shown); the bit clock produces
clock pulses having a frequency of eight times every time slot and there
are thirty-two time slots in each l25 microsecond frame. Since
conference circuit 60 can handle a two-party, a three-party, a four-
party~ or a five-party conference, AND gates 69, 70 and 71 along with
switches 72, 73 and 74 are required. The circuit of Figure 4 is
depicted with the switches 72, 73 and 74 all closed so as to provide
- 13 -
. . .
,, ....... . , , - ~ .
. . .
.140~Z
a five-party conference. When closed~ as shown in Fig. 4, switches 72,
73 and 74 apply a high logic signal, from terminal 75, to the AND gates
69, 70 and 71 respectively. If switch 74 were open, a four-party
conference would be provided by conference circuit 60, if switches 73
and 74 were open a three-party conference would be provided by circuit 60,
and if all the switches 72, 73 and 74 were open, a two-party conference
would be provided by circuit 60. In short, when a high logic signal
is not applied to one of the ~ND gates 69, 70 and 71, that AND gate
blocks signals from reaching the subsequent shift register and thereby
limits the number of parties (or conferees) that circuit 60 can
accommodate. The number of conferees (or parties) handled by circuit 60
is the number of consecutive shift registers functioning in the circuit
plus one. It should be noted that switches 72, 73 and 74 are shown
in Figure 4 in simplified form and this function would normally be
provided by solid-state switches. Additionally, the switches 72, 73
and 74 would normally be controlled so that they are closed or open
in the following pattern: all switches 72, 73 and 74 are open;
switch 72 is closed, switches 73 and 74 are open; switches 72 and 73
are closed, switch 74 is open; and all switches 72, 73 and 74 are
closed. This provides of course two-party, three-party, four-party
and five-party conferences, respectively. The shift registers 61, 62,
63 and 64 are enabled via high logic on line 76 which carries logic
signals analagous to the ones produced by OR gate 39 of Figure 2;
i.e. the registers 61, 62, 63 and 64 are enabled only during the time
slots of interest.
The output of each shift register 61, 62, 63 and 64
is applied to the eight-to-one data selector 77. As its name implies,
eight-to-one data selector 77 has eight switched inputs (only four of
which are shown) and one output (i.e. TDM bus 32). The logic signals
applied to control terminals 78, 79 and 80 determine which one of the
eight input terminals is connected to the output. In circuit 60 of
- 14 -
.
-
11~4~Z
:
Figure 4, this means that selector 77 functions to route a PCM word .. .
contained in one of the shift registers 61, 62, 63 and 64 to TDM
bus 32. Enable input 88 receives logic signals from line 76 so that
an output appears on TDM bus 32 only during the appropriate time slots.
The control signals to be applied to terminals 78, 79
and 80 are determined as follows. Comparator 81 receives on its
S input th.e four most significant logic bits from the PCM word
contained in register 61 and, on its T input, comparator 81 receives
th.e four most significant logic bits from the PCM word contained in
register 62. If the binary value of the bits on the S input is larger :
th.an the binary value of the bits on the T input the output of
comparator 81 is a logic high, otherwise it is a logic low. Selector 82
functions to route the four most significant binary bits that had the
largest value as determined by comparator 81 to another comparator 83.
This process is accomplished by selector 82 performing the Boolean
algebra operation A R + B R to the signals appearing at its
inputs A, B and R.
Comparator 84 and selector 85 make a similar selection
based on the PCM words contained in registers 63 and 64. Comparator 83
then makes the final comparison between its S and T inputs with its
S input receiving the four most significant bits (excluding sign) of
the larger PCM word from registers 61 and 62, and its T input
receiving the four most significant bits (excluding sign) of the larger
PCM word from registers 63 and 64.
Three bit latch 86 functions to keep track of which
sh.ift register contains the largest PCM word and to transmit this
data, in the form of a three-bit binary number to eight-to-one data .
selector 77 to control the operation of selector 77. One input of
three-bit latch 86 is responsive to the output of comparator 83,
.
3Q another input of latch 86 is responsive to the output of comparator 81, .
and the final input of latch. 86 is responsive to the output of
- 15 -
1~40~Z
comparator 84. Latch 86 has a clock pulse input 87 from a channel
clock (not shown~. The channel clock, applied to input 87, supplies
one pulse every time slot, coinciding with the last bit pulse of the
time slot and of the same duration as the bit pulse (recall that
each time slot, or channel, has eight bit pulses as produced by the
bit clock, and there are thirty-two time slots to one frame).
Consequently, latch 86 changes its output state only once per time
slot, during the time period of the last bit of the time slot so as
to have the correct output for applying to selector 77 at the beginning
of a new time slot.
The equipment employed in the preferred embodiments
of Figures 2 and 4 is as follows:
Comparators 48, 81, 83, 84 Texas Instruments No. 7485
Shift Registers 37, 38, 61, 62, 63, 64 Texas Instruments No. 74164
Selectors 82, 85 Texas Instruments No. 74157
Three bit latch 86 Texas Instruments No. 74175
Eight-to-one Data Selector 77 Texas Instruments No. SN74151
While the conference circuit 60 of Figure 4 is the
preferred one envisaged by the inventors for their particular
application, variations can be made therein. One modification would be
to have up to thirty-two conferees; or even more depending upon the
equipment used. The number of conferees is limited (on technical grounds)
solely by the number of time slots available in one frame; the maximum -
number of conferees that a conference circuit constructed according
to the present invention can handle is equal to the number of time
slots available in one frame. One main telephone system in use
today employs thirty-two time slots per frame. Consequently, the
maximum number of conferees is limited to thirty-two. It should be
noted that a conference call with thirty-two participants (or conferees),
while technically possible-using the present invention, would likely be
too unwieldly due to the human factors involved, and a conference call
- 16 -
~4(;~Z `~
would most likely have fewer participants than thirty-two. It should
also be noted that internally some telephone systems employ more
than thirty-two time slots per frame. For example, Western Electric's
number 4 ESS employs 128 time slots per frame, and in that case the
maximum number of conferees would be 128 (although it is very unlikely
that a conference call of this magnitude would be made).
It should also be noted that instead of employing
the shift registers of Figures 2 and 4 a tapped delay line could also
be used (being tapped after every 8 bits~.
.