1
ADDITIVE PCM SPEAKER CIRCUIT FOR A TIME
SHARED ISDN CONFERENCE ARRANGEMENT
CROSS-REFERENCE TO RELATED 5
APPLICATIONS
This Application is related to co-pending U.S. patent applications Ser. Nos. 87-7-018 and 87-7-033, which have the same assignee as the present Application. This Application is also related to co-pending U.S. patent application Ser. No. 87-7-040, having the same inventive entity and being assigned to the same Assignee as the present Application.
BACKGROUND OF THE INVENTION 15
The present invention generally pertains to a conference calling arrangement for an integrated services digital network and more particularly to improvements for speaker selection and broadcasting for multi-port 20 conference circuitry.
U.S. Pat. No. 4,466,093 teaches a time shared conference arrangement. U.S. Pat. Nos. 4,466,092 and 4,466,094 also teach similar techniques. The assignee of the present application is the successor in all rights, title 2j and interest to the assignee of the U.S. Pat. No. 4,466,093 .
This patent teaches the broadcasting of the "loudest speaker" to the other conferees. A series of on-line comparisons is made between the PCM magnitude lev- 3Q els of up to three conferees. Such on-line comparisons require an amount of switching system real time. Next, this patent teaches the comparison of the loudest detected speaker to a threshold level. If the PCM magnitude of the speakers' values exceeds this threshold, then 35 this loudest speaker's PCM sample is broadcast to the other conferees during the appropriate time slots. If the PCM magnitude of the loudest conferee did not exceed the threshold, then the identity of the loudest conferee stored during the previous sampling time period was 40 selected and the corresponding PCM sample broadcast to the other conferees during the appropriate time slots.
This technique provided for rapid switching between the speakers introducing noise. As a result, the overall audible quality broadcast to the conferees is substan- 45 tially degraded when two or more parties speak simultaneously.
When multiple sound sources are present, the human ear hears a combination of all the various sources. Some of the sources are louder in amplitude than others. The 50 PCM samples, which represent speech sound, of modern digital networks are not directly combinable. Therefore, the human ear listening via a digital switching system does not hear a true combination of each conferee's speech. 55
PCM data samples are nonlinear logarithmic representations of human speech. Since these PCM samples are nonlinear, they are difficult to combine.
It is an object of the present invention to provide a PCM speaker control circuit for a time shared confer- 60 ence arrangement of an integrated services digital network (ISDN) which provides for low noise output and true representation of each conferee's speech.
SUMMARY OF THE INVENTION 65
In accomplishing the object of the present invention a novel additive PCM speaker circuit for a time shared ISDN conference arrangement is provided.
2
An ISDN system includes a number of system subscribers connected to the system network. The network connects at least three of the system subscribers in a conference call arrangement by transferring PCM voice samples in consecutive time slots. The time shared conference arrangement has a timer for generating a plurality of periodic pulses. Also included is a first converter which is connected to the ISDN network. The first converter produces linear values corresponding to each of the PCM voice samples.
A buffering arrangement is connected to the timer and to the first converter. The buffering arrangement sequentially stores the corresponding linear values of each conference subscriber. The buffering arrangement also stores, during sequential time slot,, the linear values of the subscribers in a second conference call.
A multiplexer is connected to the buffering arrangement and to the timer. The multiplexer transmits pairs of the stored linear values during successive time slots. An adder is connected to the multiplexer. The adder arithmetically combines the pairs of linear values to produce a resultant value.
Lastly, a second converter is connected between the adder and the network. The second converter produces a PCM value corresponding to each resultant value of the adder. Then, the second converter transmits these PCM values to the network. The transmitted PCM values are the equivalent PCM values corresponding to each resultant value of the adder.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of the conference circuit embodying the principles of operation of the present invention.
FIG. 2 is a layout of the channel input to and output from the conference circuit via the network.
FIGS. 3A and 3B are a flow chart of a method for converting linear values to PCM logarithmic values.
DESCRIPTION OF THE PREFERRED
EMBODIMENT
In FIG. 1, a conference circuit block diagram of an integrated services digital network (ISDN) is shown. ISDN systems typically provide for switching both data and voice communications. The conference circuit shown herein is applied only to PCM voice samples. That is, the circuit is used only for conferences between voice subscribers. In this application, only PCM voice samples are received from the ISDN system's time switched network. PCM voice data is input to the conference circuit via the received data lead RXD.
The ISDN systems transmits multiplexed voice and data at a rate of 3.088 megabytes per second. The RXD lead is connected to other logic (not shown) of the network that segregates the voice channels from the data channels. This system logic also orients the PCM voice samples of each group of three conferees, which comprise the conference call, in three adjacent channels or time slots. There are a maximum of 24 channels of information in a frame and hence a maximum of eight three party conference groups.
FIG. 2 shows that callers Al, Bl and CI, with their PCM voice samples in channels 1, 2 and 3 respectively, comprise conference call 1. Similarly, speakers A2, B2 and C2 with their PCM voice samples in channels 4, 5 and 6 respectively, comprise conference call 2. Up to eight time shared conference calls may be handled by this conference circuit arrangement.
