REGISTER  |  SIGN IN
United States Patent5751338
Ludwig, Jr.May 12, 1998
Title
Methods and systems for multimedia communications via public telephone networks
Abstract
Methods and systems for providing multimedia telecommunication services to multimedia workstations communicating with a multimedia central office which includes a digital switch complex coupled to a public digital telephone network, and at least one twisted pair transceiver coupled to at least one twisted pair link in a telephone loop plant. The multimedia central office further includes at least one switch complex operatively associated with the digital switch complex and the at least one twisted pair transceiver. The multimedia central office is capable of transceiving signals with multimedia workstations interfaced to the public digital telephone network, and with multimedia workstations interfaced to the at least one twisted pair link in the telephone loop plant. The signals which are transceived include audio signals, video signals, and digital data signals used in providing the multimedia telecommunication services.
Inventors:Ludwig, Jr.; Lester Frank (Foster City, CA)
Assignee:Visionary Corporate Technologies (Incline Village, NV)
Appl. No.:367976
Filed:December 30, 1994
Current U.S. Class:348/14.12 370/260 379/202.01 379/93.14 379/93.17 709/204 715/733 345/2.2 348/14.11 
Field of Search:348/17,15,7,12,13,14,16,18,19 358/11 364/514 370/84,396,397,398,260 379/93,96,97,98,94,201,220 395/162,200.04
U.S. Patent Documents
3456206July 1969Kwartiroff
3519744July 1970Dorros et al.
3530251September 1970Miyawaki
3568100March 1971Tarbos
3701849October 1972Stapleton
3723653March 1973Tatsuzawa
3873771March 1975Kleinerman et al.
3974337August 1976Tatsuzawa
4005265January 1977Verhoeckx et al.
4054910October 1977Chou et al.
4084181April 1978Mita et al.
4210927July 1980Yumde et al.
4273963June 1981Seidel
4441180April 1984Schussler
4650929March 1987Boerger et al.
4686698August 1987Tompkins et al.
4710917December 1987Tompkins et al.
4716585December 1987Tompkins et al.
4752940June 1988Graham
4760442July 1988O'Connell et al.
4800344January 1989Graham
4847829July 1989Tompkins et al.
5010399April 1991Goodman et al.
5014267May 1991Tompkins et al.
5065396November 1991Castellano et al.
5067018November 1991Sakamoto et al.
5130793July 1992Bordry et al.
5136575August 1992Kuroda
5187571February 1993Braun et al.
5247347September 1993Litteral et al.
5247351September 1993Cho
5283637February 1994Goolcharan
5375068December 1994Palmer et al.
5410343April 1995Coddington et al.
5467212November 1995Huber
5473679December 1995La Porta et al.
5490247February 1996Tung et al.
5528281June 1996Grady et al.
5528286June 1996Goolcharan
5534914July 1996Flohr et al.
5555118September 1996Huber
5555244September 1996Gupta et al.
5557442September 1996Huber
5563882October 1996Bruno et al.
5566179October 1996Kobayashi et al.
5579143November 1996Huber
5617539April 1997Ludwig et al.
Foreign Patent Documents
0354370A1., 0000EP
0497022A1., 0000EP
3507152A1., 0000DE
A 0 244260Nov., 1987EP
A 0 352210Jan., 1990FR
A 0 617560Sep., 1994EP
A 05 268585Jan., 1994JP
A 2282506Apr., 1995GB
Other References
Transactions of the Institute of Electronics, Information and Communication Engineers of Japan, vol. E74, No. 9, Sep. 1991, Tokyo, Japan, pp. 2696-2702. .
Proc. of the 1st International Workshop on Community Networking-Integrated Multimedia Services to the Home -13-14 Jul. 1994, San Francisco (US), pp. 203-209. .
AT & T Visual Solutions Multipoint Control Unit, Product Literature, 1993. .
Chinon America, Inc., Chinon CX-062, Miniature Micro Color CCTV Board Camera, Product Literature, Apr. 1994. .
Communique!, Insoft, Inc. Product Literature. .
Article, Infoworld, "Up-Start Insoft Storms Conferencing", Dated Oct. 10, 1994. .
Insoft Shared Application Resource Environment--Share, Product Literature. .
Insoft Opendve, Your Software Development Bridge for Building Collaborative, Fully-Interoperable Multimedia Applications, Product Literature. .
For-A Company Limited, Operation Manual, Multi Viewer MV-40D (1st Edition-Rev.2). .
For-A Company Limited, MV40D Multi-Viewer, Product Literature. .
For-A Company Limited, MV-90C Multi-Viewer. .
Lightwave Systems, Inc., Communication Systems, Installation Guide Series 4000 Stand-Alone Twisted Pair Video System, STD. Doc.multidot. 840-000-01. .
Maxim Integrated Products, Excerpt From Engineering Journal, "New Amplifiers Simplify Wideband Techniques", Date Unknown. .
Maxim Integrated Products, Wideband Transconductance Amplifiers, MAX435/MAX436, Product Literature, 0042, Rev. 1, Apr. 1993. .
Network Video Technologies, NTV Quality Video Over Ordinary Telephone Wire . . . Rackmount Video Transceiver System, Product Literature. .
Picturetel, M-7000 Multipoint Bridge and Control System, Simultaneous, Interactive Videoconferencing Among Three to Sixteen Locations, Product Literature. .
Picturetel Corporation, Article, Multilocation Teamwork is the Key to Success in the Competitive 90S. .
TUT Systems, Ethernet or 10BASET, Over One Pair Wire, Product Literature. .
TUT Systems, Article TUT Transceivers Target Telcos, ATM, by Reinhardt Krause, Journal Unknown, Data Unknown. .
Videoserver, Inc., Videoserver Series 2000 MCU, Product Literature. .
Videoserver, Inc., Videoserver Series 2000 Conference Scheduling and Reservation Application, Product Literature. .
Data Communications, Videoconferencing, Mix-and-Match Videoconferencing, Startup's MCU Allows Videoconferences Among Sites Using Different Vendors' Codecs Aug. 1992 .
Videoserver, Inc., Videoserver Series 2000 Multipoint Control Unit, Product Backgrounder, The New Standard for Multipoint Conferencing, Product Literature. .
Ameritech Bridged Video Service Press Release, Oct. 11, 1994. .
AT&T Accunet Switched Digital Services, Data Newtorking User Guide, Communications Concert, Jan. 1991. .
AT&T Worldworx Network Services, Product Literature. .
MCI Telecommunications Corporation, Product Literature. .
MFS Communications Company, Inc., Business Plan. .
Pacific Bell, A Pacific Telesis Company, "Affordable Multipoint Videoconferencing is Real, California . . . " Product Literature, Oct. 1994. .
"Fiber Net to Link Video Production Units: 50 Facilities . . . " Photonics Spectra, Nov. 1991. .
"Merrill Tests Video: Brokerage Uses Video-Conferencing System Based on Wtitel DS-3 Fiber-Network Service", Communications Week, Apr. 8, 1991, p. 8. .
Chang, Chen and Messerschmitt, "Video Compositing in the DCT Domain" IEEE Workshop on Visual Signal Processing and Communications, Sep. 1992, pp. 138-143. .
Lei, Chen and Sun, "Video Bridging Based on H.261 Standard", IEEE Transactions on Circuits and Systems for Video Technology, vol. 4, No. 4, Aug. 1994. .
Ludwig, Frontiers in Computer Communications Technology, "A Threaded/Flow Approach to Reconfigurable Distributed Systems & Service Primitives Architectures" ACM Sigcom 1987. .
Ludwig & Dunn, Frontiers in Computer Communications Technology "Laboratory for Emulation and Study of Integrated and Coordinated Media Communication" ACM Sigcomm Aug. 1987. .
Electronic Engineering Times, A CMP Publication, Issue 821, "DSP Kicks ATM Onto Voice Lines", Oct. 31, 1994. .
San Jose Mercury News, "New System Locates Site of Gunshots", Nov. 1994 and Redwood City to Install High-Tech Gunfire Sensor. .
Yun & Messerschmitt "On Architectures for Video Composting", Dept. of Electrical Eng. & Computer Sciences, Multimedia Systems, 1994. .
Boyer & Lukacs, The Personal Presence System--A Wide Area Network Resource for the Real Time Composition of Multipoint Multimedia Communications, ACM Multimedia, Oct. 1994. .
Lukacs, "The Personal Presence System--Hardware Architecture", ACM Multimedia Oct. 1994. .
Press Release, Broadband Technologies, Inc., Dec. 6, 1994. .
PC Week, "You May Be Ready for ATM, But It Isn't Ready For You", Sep. 5, 199 . .
The Wall Street Journal, "Ameritech Sets Charge of 2.5 Billion For Old Equipment; Loss is Seen For Year", p. A4, Nov. 29, 1994. .
Datapoint Corp., News Release, "Datapoint Introduces New Line of Network Video Communications Products", Feb. 15, 1994. .
AT&T, Multimedia Communicator Dataport 2001, Product Literature. .
CLI, Multipoint 2/2M Controllers, Product Literature. .
Bell Laboratories, Bell Laboratories Record, Picturephone, pp. 137-141, May/Jun. 1969. .
Lake & Pate, A Network Environment for Studying Multimedia Network Architecture and Control, 1989. .
Ludwig, L.F., Integration of CAD/CAE With Multimedia Teleconferencing and Messaging via Broadband and Shared Resource Servers, Bellcore, Undated. .
Javelin, Video Transmission System, J411, Product Literature. .
Televideo, Instructions, Product Literature. .
Network World, "AT&T Looks to Extend Mail APIs With Voice, Fax, Video", p. 6, Oct. 10, 1994. .
Im & Werner, Bandwidth-Efficient Digital Transmission Up to 155 MB/S Over Unshielded Twisted Pair Wiring, ICC 1993. .
Klein, P. "Telekommunikation von Angesichtzu Angesicht", Siemens AG. 2323 Telecom Report, No. 5, Sept./Okt. 1986, Erlangen, W. Germany. .
Ramanathan, Rangan, Vin, & Kaeppner, "Optimal Communication Architectures for Multiconferencing in Distributed Systems", Proceedings of 12th International Conference on Distributed Computing Systems, Jun. 9-12, 1992. .
Maeno, Sakata, Ohmori, Watabe & Fukuoka "Distributed Desktop Conferencing System (Mer Maid) Based on Group Communication Architecture" IEICE Transactions, vol. E74 Sep. 1991. .
Weiss, C., "Desk Top Video Conferencingt--An Important Feature of Future Visual Communications", Proceedings of IEEE International Conference on Communications. .
Satellite Source, Mastermind, Product Advertisement. .
Article, "Instant Network Rides on Phone Lines", Electronic Design, Aug. 6, 1987. .
Sheets, W. & Graf, R. "Build This . . . Carrier Current Receiver", Radio Electronics Feb., 1989. .
Sheets, W. & Graf, R., "Build This . . . Carrier Current Audio Transmitter", Radio Electronics, Jan. 1989. .
Teleconcepts, Inc., "Teleconcepts . . . Introduces The Just Plug It In Intercom System", Product Literature. .
Windmaster Manufacturing, Remote Extender, Owner's Manual, Product Literature. .
"A Derring-Do Attitude Propels Wiltel . . . and Creative Expansion", Williams Telecommunications Group Inc. Industry Highlights, Steve Polilli, Apr. 1992..~
Primary Examiner: Kuntz; Curtis
Assistant Examiner: Tieu; Binh K.
Attorney, Agent or Firm:Brooks & Kushman
Claims

What is claimed is:
1. A system for providing a plurality of multimedia telecommunication services to a plurality of multimedia workstations at least two of which are distributed among a first premise and a second premise, the system comprising:
a multimedia central office located at a third premise, the multimedia central office including:
a digital switch complex coupled to at least one of the plurality of workstations via a public digital telephone network;
at least one twisted pair transceiver coupled to a second one of the plurality of workstations via at least one twisted pair link in a telephone loop plant; and
at least one switch complex operatively associated with the digital switch complex and the at least one twisted pair transceiver;
wherein the multimedia central office transceives signals with one of the multimedia workstations interfaced to the public digital telephone network, transceives signals with one of the multimedia workstations interfaced to the twisted pair link in the telephone loop plant, and the signals include audio signals, video signals, and digital data signals; and
wherein an analog video signal is communicated between the multimedia central office and at least one of the multimedia workstations using a plurality of space division video signals, each of the space division video signals being transmitted over a corresponding one of a plurality of twisted pair links.

2. The system of claim 1 wherein the multimedia telecommunication services include application sharing between at least two of the multimedia workstations.

3. The system of claim 1 wherein the multimedia telecommunication services include window sharing between at least two of the multimedia workstations.

4. The system of claim 1 wherein the multimedia telcommunication services include multimedia messaging between at least two of the multimedia workstations.

5. The system of claim 1 wherein the digital switch complex is coupled to a third-party network, wherein the multimedia telecommunication services provided by the system include providing a gateway to the third-party network.

6. The system of claim 1 wherein the multimedia central office is networked to a second multimedia central office via at least one common carrier digital transmission link coupled to the digital switch complex.

7. The system of claim 6 wherein each of at least two of the multimedia workstations is coupled to the multimedia central office by a corresponding one of at least two dedicated digital carriers, and wherein the multimedia central office concentrates data received on the at least two dedicated digital carriers for transmission to the second multimedia central office.

8. The system of claim 1 further comprising an internal premise communication system which couples at least two of the multimedia workstations to the multimedia central office, wherein the at least two of the multimedia workstations share access to the multimedia central office via the internal premise communication system.

9. The system of claim 1 wherein at least two of the multimedia workstations are located within a common user premise.

10. The system of claim 1 wherein the multimedia workstations are located at a plurality of user premises.

11. The system of claim 1 wherein the multimedia central office transceives audio signals having an effective bandwidth of at least 5 kHz, color video signals having an effective bandwidth of at least 3 MHz, and digital data signals having a bit rate of at least 128 kbps via the at least one twisted pair link.

12. The system of claim 1 wherein the twisted pair transceiver comprises:
a plurality of filters responsive to the analog video signal, each of the filters passing a corresponding band of frequencies contained within the analog video signal and producing a corresponding filtered signal based thereupon;
at least one frequency shifter for producing a corresponding frequency-shifted signal based upon the filtered signal from a corresponding one of the filters; and
a plurality of transmitters each transmitting the corresponding frequency-shifted signal over a corresponding one of the plurality of twisted pair links.

13. The system of claim 1 wherein the at least one twisted pair transceiver includes:
a plurality of receivers which receives the space division video signals;
at least one frequency shifter responsive to the receivers, each of the at least one frequency shifter producing a corresponding reconstructed signal; and
a mixer responsive to each corresponding reconstructed signal to form a reconstructed analog video signal.

14. The system of claim 1 wherein the analog video signal contains a chrominance signal and a luminance signal, the chrominance signal being transmitted over a first one of the twisted pair links and the luminance signal being transmitted over a second one of the twisted pair links.

15. The system of claim 1 wherein the at least one twisted pair transceiver includes:
a band-pass filter which produces a band-pass signal in dependence upon the analog video signal, the analog video signal containing a chroma carrier signal and sidebands associated therewith, the band-pass filter passing the chroma carrier signal and the sidebands associated therewith;
a frequency down-shifter, in communication with the band-pass filter, which produces a frequency shifted signal based upon the band-pass signal;
a low-pass filter which produces a low-pass signal in dependence upon the analog video signal, the low-pass filter having a cut-off frequency less than the frequency of the chroma carrier signal;
a first transmitter which transmits a first signal over a first of the twisted pair links, the first signal based upon the frequency shifted signal; and
a second transmitter which transmits a second signal over a second of the twisted pair links, the second signal based upon the low-pass signal.

16. The system of claim 1 further comprising a mixer which mixes an audio signal with one of the space-division video signals to form a mixed signal, wherein the mixed signal is transmitted over one of the twisted pair links corresponding to the space division video signals.

17. The system of claim 16 further comprising a modulator which modulates the audio signal for application to the mixer.

18. The system of claim 1 wherein a digital signal is communicated between the multimedia central office and one of the multimedia workstations over a plurality of twisted pair links using digital inverse multiplexing, wherein a corresponding bit stream carried over each of the twisted pair links has a bit rate higher than that for an ISDN basic rate interface.

19. The system of claim 18 further comprising:
a first digital inverse multiplexer which produces a plurality of digital streams each having a bit rate lower than that of the digital signal; and
a second digital inverse multiplexer in communication with the first digital inverse multiplexer via the plurality of twisted pair links, the second digital inverse multiplexer producing a reconstructed digital signal from the plurality of digital streams.

20. The system of claim 19 further comprising:
a first codec which forms the digital signal in dependence upon an analog signal; and
a second codec coupled to the second digital inverse multiplexer, which forms a reconstructed analog signal from the reconstructed digital signal.

21. A method of providing a plurality of multimedia telecommunication services to a plurality of multimedia workstations at least two of which are distributed among first and second premises, the method comprising:
providing a multimedia central office located at a third premise capable of providing the multimedia telecommunication services;
coupling at least one of the multimedia workstations to the multimedia central office by the public digital telephone network;
coupling at least one of the multimedia workstations to the multimedia central office by at least one twisted pair link within a telephone loop plant;
transceiving signals between the multimedia central office and the multimedia workstations via the public digital telephone network; and
transceiving signals between the multimedia central office and the multimedia workstations via the telephone loop plant;
wherein the signals include audio signals, video signals, and digital data signals; and
wherein the step of transceiving includes a step of communicating an analog video signal between the multimedia central office and one of the second at least one of the multimedia workstations using a plurality of space division video signals, each of the space division video signals transmitted over a corresponding one of a plurality of twisted pair links.

22. The method of claim 21 wherein the multimedia telecommunication services include application sharing between at least two of the multimedia workstations.

23. The method of claim 21 wherein the multimedia telecommunication services include window sharing between at least two of the multimedia workstations.

24. The method of claim 21 wherein the multimedia telcommunication services include multimedia messaging between at least two of the multimedia workstations.

25. The method of claim 21 further comprising the steps of:
coupling the multimedia central office to a third-party network; and
providing, to at least one of the multimedia workstations, a gateway to the third-party network.

26. The method of claim 21 further comprising the steps of:
providing a second multimedia central office; and
networking the multimedia central office to the second multimedia central office via at least one common carrier digital transmission link.

27. The method of claim 26 wherein each of at least two of the multimedia workstations is coupled to the multimedia central office by a corresponding one of at least two dedicated digital carriers, the method further comprising the step of concentrating data received on the at least two dedicated digital carriers for transmission to the second multimedia central office.

28. The method of claim 21 further comprising the step of:
coupling at least two of the multimedia workstations to the multimedia central office by an internal premise communication system, wherein the at least two of the multimedia workstations share access to the multimedia central office via the internal premise communication system.

29. The method of claim 21 wherein the first at least one and the second at least one of the multimedia workstations are located within a common user premise.

30. The method of claim 21 wherein the multimedia workstations are located at a plurality of user premises.

31. The method of claim 21 wherein the multimedia central office transceives audio signals having an effective bandwidth of at least 5 kHz, color video signals having an effective bandwidth of at least 3 MHz, and digital data signals having a bit rate of at least 128 kbps via the at least one twisted pair.

32. The method of claim 21 wherein the step of communicating includes the steps of:
producing a plurality of filtered signals based upon the analog video signal, each of the filtered signals passing a corresponding band of frequencies contained within the analog video signal;
frequency shifting each of at least one of the filtered signals to produce at least one frequency shifted signal; and
transmitting each of the at least one frequency-shifted signal over a corresponding one of the plurality of twisted pair links.

33. The method of claim 21 wherein the step of communicating includes the steps of:
receiving each of at least one space division video signal;
frequency shifting the at least one space division video signal to produce at least one reconstructed signal; and
mixing the at least one reconstructed signal to form a reconstructed analog video signal.

34. The method of claim 21 wherein the analog video signal contains a chrominance signal and a luminance signal, and wherein the step of communicating includes the steps of:
transmitting the chrominance signal over a first of the twisted pair links; and
transmitting the luminance signal over a second of the twisted pair links.

35. The method of claim 21 wherein the analog video signal contains a chroma carrier signal and sidebands associated therewith, and wherein the step of communicating includes the steps of:
band-pass filtering the analog video signal to produce a band-pass signal containing the chroma carrier signal and the sidebands associated therewith;
frequency down-shifting the band-pass signal to produce a frequency shifted signal;
low-pass filtering the analog video signal to produce a low-pass signal containing frequencies less than the frequency of the chroma carrier signal;
transmitting a first signal over a first of the twisted pair links based upon the frequency shifted signal; and
transmitting a second signal over a second of the twisted pair links based upon the low-pass signal.

36. The method of claim 21 wherein the analog video signal contains a chroma carrier signal and sidebands associated therewith, and wherein the step of communicating includes the steps of:
receiving a first signal via a first of the twisted pair links;
receiving a second signal via a second of the twisted pair links;
frequency up-shifting the first signal to produce a third signal; and
mixing the third signal and the second signal to form a reconstructed video signal.

37. The method of claim 21 further comprising a step of mixing an audio signal with one of the space division video signals to form a mixed signal, wherein the mixed signal is transmitted over the one of the twisted pair links corresponding to the one of the space division video signals.

38. The method of claim 21 wherein the step of transceiving the second plurality of signals includes the step of communicating a digital signal over a plurality of twisted pair links using digital inverse multiplexing, wherein a corresponding bit stream carried over each of the twisted pair links has a bit rate higher than that for an ISDN basic rate interface.

39. The method of claim 38 wherein the step of communicating includes the steps of:
inverse multiplexing the digital signal to produce a plurality of digital streams each having a bit rate lower than that of the digital signal;
transmitting each of the digital streams via a corresponding one of the twisted pair links;
receiving the digital streams; and
inverse multiplexing the digital streams to produce a reconstructed digital signal.

40. The method of claim 39 further comprising the steps of:
forming the digital signal in dependence upon an analog signal; and
forming a reconstructed analog signal from the reconstructed digital signal.

41. A method of communicating an analog video signal and an audio signal over a plurality of twisted pair links, the method comprising the steps of:
forming a first space division video signal and a second space division video signal based upon the analog video signal;
mixing the audio signal with the second space division video signal to form a mixed signal; and
transmitting the first space division video signal over a first of the twisted pair links;
transmitting the mixed signal over a second of the twisted pair links.

42. A system for communicating an analog video signal and an audio signal over a plurality of twisted pair links, the system comprising:
means for forming a first space division video signal and a second space division video signal based upon the analog video signal;
a mixer which mixes the audio signal with the second space division video signal to form a mixed signal;
a first transmitter which transmits the first space division video signal over a first of the twisted pair links; and
a second transmitter which transmits the mixed signal over a second of the twisted pair links.
Description
TECHNICAL FIELD

The present invention relates to methods and systems for providing multimedia telecommunications, and more particularly, to methods and systems for providing multimedia telecommunications using existing public telephone system infrastructure and desktop computers.

BACKGROUND OF THE INVENTION

Presently, the cost of both wideband digital transport and interfacing equipment is a deterrent to the widespread usage of multimedia services for collaboration and other business functions. One component of the cost is the tariffs for digital carriers provided by common carriers. Common carriers include public telephone networks registered, regulated, and/or authorized to provide publicly tariffed telecommunication services.

Common carriers discount digital bandwidth according to the amount of bandwidth on a particular digital carrier. Presently, wide-area subscription rates for Fractional T-1 data rates typically break even at about half the rate of a full T-1 and rates for eight T-1 carriers typically match that of a T-3 carrier providing approximately 30 times the bandwidth of a single T-1.

The reason for this 2:1 to 4:1 discounting is that bulk bandwidth delivery and wide-area transport employs less common carrier equipment, management, and control than delivery of the same amount of bandwidth in smaller, unbundled parcels. For a telecommunications concern outside the common carriers, private networks can be made far more cost effective if their design leverages these economies of scale in bandwidth purchases from the common carriers.

The transmission of quality audio and video signals to a multimedia workstation at a user premises is essential in providing multimedia services. In practice, satisfactory quality NTSC/PAL/SECAM video and medium fidelity audio can be transmitted relatively inexpensively through an unshielded twisted pair (UTP) link using analog signals.

Methods and systems of transmitting continuous motion (25-30 frames/second) analog video signals over unshielded twisted pair links range from the AT&T Picturephone of the late 1960's to the disclosure in U.S. Pat. No. 5,283,637 to Goolcharan. However, as the distance of the UTP increases beyond an upper limit, typically beyond approximately 2500 feet, crosstalk and attenuation increase very rapidly. This effect is more pronounced for smaller gauge wires (e.g. 26 gauge) than for larger gauge wires (e.g. 20 or 19 gauge).

Typically, an existing public telephone loop plant, i.e. the access portion of a public telephone network which connects a user premises with a first public telephone building, contains UTP having one or more gauge changes at several junctures. Moreover, the length of the UTP between end locations is often much larger than the physical distance between the locations (a factor of 3.5 is not unusual). At longer UTP loop lengths, the multiple junctures and gauge changes combine to create an uneven high-frequency response.

FIG. 1 illustrates the frequency response of a passive UTP metallic loop without bridge taps between two buildings, each separated from a public telephone central office by approximately 1000 feet. This UTP loop involves a 7100 foot wire length with multiple gauge changes. In this example, the frequency response exhibits significant peaks and notches around 1 MHz. In general, the large notches and peaks which result complicate the compensation for the rapid attenuation at high frequencies.

One approach to this problem is to convert the video to a digital signal for transmission over the UTP. Such an approach requires the steps of converting the video to a digital form, compressing the digital form of the video, and transmitting the compressed digital video over the UTP at bit rates whose analog spectra (dictated by pulse shape) is within the bounds of the UTP. Standard state-machine pulse coding and pulse wave-shaping (such as alternate mark inversion) are typically used to reduce the power spectrum only to multiples of the bit rate of approximately 1.5 (i.e., 1 Mbps can be transmitted over a 1.5 MHz baseband bandwidth). Consequently, even at distances of 2500 feet, digital transmissions attain only about 3-4 Mbps. As a result, significant compression and cost is required to utilize digital signals for transmitting video over the UTP. Although not as costly or confining as 128-384 kbps Fractional T-carrier channels, deployment of codecs compressing two-way, high-quality, full-motion video and audio to 3-4 Mbps at each desktop is costly and problematic.

SUMMARY OF THE INVENTION

It is an object of the present invention to deliver immediate low-cost, wide-area access of multimedia services to multimedia workstations using unshielded twisted pair links within an existing telephone loop plant.

A further object is to reduce the cost of wide-area, wideband digital transport in a system for providing multimedia services to multimedia workstations utilizing desktop computers.

In carrying out the above objects, the present invention provides a system for providing multimedia telecommunication services to a plurality of multimedia workstations. The system comprises a multimedia central office which includes a digital switch complex coupled to a public digital telephone network, and at least one twisted pair transceiver coupled to at least one twisted pair link in a telephone loop plant. The multimedia central office further includes at least one switch complex operatively associated with the digital switch and the at least one twisted pair transceiver. The multimedia central office transceives a first plurality of signals with a first at least one of the multimedia workstations interfaced to the public digital telephone network, and transceives a second plurality of signals with a second at least one of the multimedia workstations interfaced to the at least one twisted pair link in the telephone loop plant. The first plurality and the second plurality of signals each include audio signals, video signals, and digital data signals used for providing the multimedia telecommunication services.

Embodiments of the present invention provide a network approach for immediate, low-cost deployment of advanced multimedia telecommunications services using current technology. In particular, embodiments of the present invention are advantageously capable of utilizing low-cost technologies such as analog audio/video hardware, telephone loop twisted pair, and existing LAN technologies to provide the multimedia telecommunication services. By utilizing the unshielded twisted pair in the telephone loop plant, embodiments of the present invention do not require the deployment of new ATM and fiber optic systems, which may take a number of years before widespread availability and a full, mature implementation is attained. Consequently, global infrastructure investments in telephone loop plants retain much of their value by being capable of providing advanced multimedia telecommunication services. Further, embodiments of the present invention are capable of supporting advanced multimedia services in a flexible manner so that ATM and fiber optic systems may be utilized when economical.

Embodiments of the present invention are capable of delivering low-cost, high-quality video-capable multimedia services including: (i) real-time collaboration services, such as desktop teleconferencing, window sharing, and full-application sharing; (ii) multimedia messaging services, such as multimedia mail and video mail; (iii) access to multimedia information repositories, such as on-line catalogs and video news clippings; (iv) directory services; and (v) gateways to third-party networks, such as a public telephone network and the Internet.

These and other features, aspects, and advantages of the present invention will become better understood with regard to the following description, appended claims, and accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a graph of a frequency response of a UTP loop plant;

FIG. 2 is a block diagram of an embodiment of a multimedia telecommunication system in accordance with the present invention;

FIG. 3 is a block diagram of an embodiment of a multimedia central office in accordance with the present invention;

FIG. 4 is a block diagram of an embodiment of a regional hub office in accordance with the present invention;

FIG. 5 is a block diagram of an embodiment of a national hub office in accordance with the present invention;

FIG. 6 is a block diagram of an example topology wherein fully dedicated digital carriers are employed between multimedia central offices, regional hub offices, and national hub offices;

FIG. 7 is a block diagram of an alternative topology which utilizes switched lines;

FIG. 8 is a block diagram of an alternative topology which utilizes redundant links and offices;

FIG. 9 is a block diagram of an alternative topology which utilizes both redundant links and offices, and switched lines;

FIG. 10 is a block diagram of a three layer network for illustrating topological components of a multipoint conference;

FIGS. 11a-11f are block diagrams illustrating the location of a conference bridge for various non-degenerate 3-party conference cases;

FIGS. 12a-12o are block diagrams illustrating the location of a conference bridge for various non-degenerate 4-party conference cases;

FIG. 13 is a flow chart of an embodiment of a method of resource allocation illustrated for conference bridges;

FIG. 14 is a flow chart of an embodiment of a method of allocation applicable to demand-driven resource allocation illustrated for conference bridges;

FIG. 15 is a block diagram of a compression system to illustrate signal domains;

FIG. 16 is a block diagram of a system for transmitting a digital signal over multiple-lower bandwidth links utilizing inverse multiplexers;

FIG. 17 is a block diagram of a system for transmitting data signals over multiple twisted pair links;

FIG. 18 is a block diagram of a system for transmitting video and audio signals over multiple twisted pair links;

FIG. 19 is a block diagram of a system for transmitting a single analog signal over multiple twisted pair links;

FIG. 20 is a block diagram of a system for transmitting a single analog video signal over two twisted pair links;

FIG. 21 is a block diagram of a space division transmission system of FIG. 20 which includes an audio signal;

FIG. 22 is a block diagram of an automatic compensation system for twisted pair links;

FIG. 23 is a block diagram of a system for carrying a digital stream in near-raw form over twisted pair links;

FIG. 24 is a flow chart of a method of locating distributed conferencing units;

FIG. 25 is a block diagram of an embodiment of a server;

FIG. 26 is a block diagram of an alternative embodiment of a server capable of handling analog sources, switched LAN sources, and DS0-based sources; and

FIG. 27 is a flow chart of a sequence of data communications transactions for invocating a window sharing session or adding a later participant in a conference.

BEST MODE FOR CARRYING OUT THE INVENTION

A block diagram of an embodiment of a multimedia telecommunication system in accordance with the present invention is illustrated in FIG. 2. The multimedia telecommunication system is capable of functioning with two types of multimedia workstations or equivalent room-based systems. A first type of multimedia workstation, indicated by reference numeral 100, is one capable of interfacing with a standardized network interface traditionally used in a public digital telephone network. Examples of standardized network interfacing protocols include, but are not limited to, T-1, T-3, ISDN, SONET, X.21, and X.25. A multimedia workstation interface capable of connection to a standardized network interface is herein referred to as a traditional network interface (TNI).

A second type of multimedia workstation, indicated by reference numeral 102, is one having an interface which is not currently supported by the public digital telephone network. A multimedia workstation interface according to this second type is herein referred to as a nontraditional network interface (NTNI). NTNIs are typically designed based upon the particular workstation rather than the public digital network.

The workstations 100 and 102 are illustrated as being located within one of a plurality of user premises 104a-104e. Five different types of user premises are described for the purpose of illustration. The user premise 104a includes one or more workstations each having a TNI. The workstations in the user premise 104a are directly connected to a local public digital network 106 by means of a conventional digital carrier 108.

The user premise 104b includes one or more workstations, each having a TNI, coupled to an internal premise communication system 110. The internal premise communication system 110 communicates with the workstations via TNI, and provides for sharing of access to the local public digital network 106 via TNI. An example is an ISDN PBX or corporate Fractional T-carrier cross-connect switch.

The user premise 104c includes one or more workstations each having an NTNI. These workstations communicate with the local public digital network 106 via a multimedia central office 112 using conventional twisted pair 114 existing within a telephone loop plant. The multimedia central office 112 employs the same NTNI as the workstations coupled thereto. The local public digital network 106 is coupled to the multimedia central office 112 by means of a digital carrier 113 which employs a TNI.

The user premise 104d includes one or more workstations, each having an NTNI, coupled to an internal premises communications system 116. The internal premises communication system 116 communicates with the workstations via an NTNI. Preferably, the internal premises communication system is capable of providing both interconnection and resource sharing functions, such as the sharing of public network access to the multimedia central office 112, using NTNI used to communicate with the workstations. An embodiment of such an internal premises communication system is described in copending U.S. application Ser. No. 08/131,523.

The user premise 104e includes one or more workstations, each having NTNI, coupled to an internal premises communications system 120. The internal premises communications system 120 includes a system for converting NTNI signals to TNI signals. As a result, the internal premises communications system 120 is coupled to the local public digital network 106 via a TNI such as ISDN or a T-carrier. An embodiment of such an internal premises communication system is described in copending U.S. application Ser. No. 08/131,523.

Other types of premises may be realized based upon a combination of the attributes of the abovedescribed premises 104a-104e. For example, for a premise with all workstations or rooms using NTNI, some parts of the public network access may be done via NTNI and other parts via TNI. Moreover, for a premise with some workstations using NTNI and other workstations using TNI, some parts of the public network access may be provided by a multimedia central office via NTNI and other parts by a public local digital network via TNI. Many other combinations are apparent to those having ordinary skill in the art.

The embodiments described thus far illustrate how workstations or rooms may access both a multimedia central office and a local public digital network regardless of whether NTNI or TNI is used for public network access. This results from the use of the digital carrier 113 which links the multimedia central office 112 and the local public digital network 106. A number of multimedia services can be provided by the multimedia central office 112 (and other network offices to be described), as is discussed later.

Access can be provided to a wide area public digital network 122, representative of one or more arbitrary public networks, using many approaches. In a first approach, the local public digital network 106 is linked to the wide area public digital network 122 by means of a digital carrier 124.

In a second approach, the local public digital network is linked to a regional hub office 126 by a first digital carrier 130. The regional hub office 126 is linked to the wide area public digital network by a second digital carrier 132. The regional hub office 126 is capable of providing networking, topology, and/or services functions not provided by the direct digital carrier link 124. For example, the regional hub office 126 may be capable of switching and transmission concentration for the dedicated lines used as the digital carriers with the multimedia central offices. Further, the regional hub office 126 may be capable of directly providing a number of multimedia services. Also, the regional hub office 126 may act as a strategic location for geographically-sensitive functions such as video messaging storage and conference bridge servers. In addition, regional hub offices may be used as gateways to international wide-area telecommunications services.

The use of a regional hub office in a multimedia telecommunications system in accordance with the present invention is optional. Practically, the use of a regional hub office is dictated by the selected features which are to be provided and economics of a particular implementation.

Selected features and economics of an overall implementation of a multimedia telecommunications system may be served by utilizing one or more national hub offices 134 linked to the wide area public digital network. The national hub offices 134
may provide switching and transmission concentration for any dedicated lines used as digital carriers with the regional hub offices or the multimedia central offices. Further, the national hub offices 134 may directly provide some multimedia services, or act as strategic locations for geographically-sensitive functions.

A political nation, i.e. a country, served by embodiments of the multimedia telecommunications system may have zero, one, or a plurality of national hub offices. A plurality of national hub offices may be redundantly employed in a single political nation for the purpose of providing improved reliability of service. National offices can also be used to provide gateways to international wide-area telecommunication services.

The use of regional hub offices and national hub offices allows for a non-common carrier operator of an embodiment of the present invention to implement a multimedia telecommunication system as a private switched network formed from dedicated digital carrier services offered by a common carrier. More specifically, the non-common carrier operator would interconnect the workstations by means of: (i) switched digital carrier services connecting the premises to a multimedia central office; (ii) dedicated lines among one another in a full or partial mesh topology; and (iii) dedicated lines linking multimedia central offices and TNI premises with regional hub offices and national hub offices. It is noted that a common-carrier operator of an embodiment of the present invention need not implement explicit regional or national offices.

In practice, embodiments of the multimedia telecommunications system include one or more multimedia central offices which provide NTNI access to nearby users as well as network-based multimedia services and reduced-cost transport. Typically, a plurality of multimedia central offices are employed, wherein the multimedia central offices are networked together by means of a conventional common carrier service, or by alternative means such as line-of-site microwave, spread-spectrum radio, satellite, or private fibre optic links. In such a network of multimedia central offices, there may be further value in including additional nodes in the network, namely, one or more of the regional hub offices and the national hub offices as described above.

Alternatively, one or more regional hub offices and national hub offices, optionally supplemented by one or more multimedia central offices, may be directly accessed by users using TNI digital carriers provided by a common carrier. This is beneficial in providing network-based multimedia services with a reduced transport cost but without requiring NTNI for access.

Each of the offices and the networking among them are now described. Architectural variations which may be advantageous for using LAN (local area network) switching hubs as real-time switches for packetized video and audio are presented, followed by network topology aspects.

FIG. 3 is a block diagram of an embodiment of a multimedia central office in accordance with the present invention. The multimedia central office includes a digital switch complex 150 capable of performing multiplexing and inverse multiplexing functions. Users with TNI which employ switched or dedicated digital carriers to access the multimedia central office are connected directly to the digital switch complex 150.

The multimedia central office further includes one or more UTP transceivers 152, to which users with NTNI are connected via the UTP. The UTP transceivers 152 are capable of transceiving audio, video, and digital signals communicated between the UTP and the multimedia central office. Analog video and audio signals transceived by the UTP transceivers 152 are directed to a real-time switch complex 154. The real-time switch complex 154 may contain, for example, one or more analog switches, DSO cross-connects, or switched LAN hubs (used in an effectively collisionless mode) therein. Data communication signals, such as Ethernet, in their original logical form are processed by the UTP transceivers 152 for application to a data hub complex 156.

In such an arrangement, various options for processing UTP-transceived signals are available. For example, a UTP carrying a converted V.35 signal or a converted RS-449 signal may be transceived by the UTP transceivers 152 and connected via the real-time switch complex 154. Here, the UTP transceivers 152 may optionally transform the converted signals back into the original V.35 or RS-449 format. In a second example, a UTP carrying a converted V.35 signal or a converted RS-449 signal may be transceived and transformed back into the original V.35 or RS-449 format by the UTP transceivers 152 for application to the digital switch complex 150.

The combination of the digital switch complex 150, the real-time switch complex 154, and the data hub complex 156 is used to route multimedia communications signals within the multimedia central office. The signal routing is used to provide for the interconnection of users attached to the multimedia central office, or as reachable through the digital carrier 113. Interconnection among users terminated on a common multimedia central office may be achieved as follows.

Interconnection of UTP NTNI users employing unlike NTNIs can be provided by the data hub complex 156 using service modules 160a-160c, coupled to either the real-time switch complex 154 or the digital switch complex 150, capable of performing the necessary conversions. In general, the service modules 160a-160c can provide any number of functions. For example, the service modules may be capable of performing video/audio codec functions, such as transforming between analog video/audio and V.35 or RS-449, or storage functions as may be used for video mail.

Interconnection of UTP NTNI users employing converted V.35 or RS-449 signals with TNI users can be provided by the digital switch complex 150. If the UTP transceivers 152 do not provide the transformations needed to present V.35 or RS-449
signals to the digital switch 150, then the signal is routed through an appropriate service module 160a, as available, via the real-time switch 154. Here, the services modules 160a perform the transformations between converted V.35 or RS-449 and true V.35 or RS-449. If the UTP transceivers 152 do provide the transformation, signals may pass directly from the UTP transceivers 152 to the digital switch complex 150 without the use of service modules.

The service modules 160a include those capable of performing compression or other signal conversion functions required by the service modules 160c. These service modules 160a are accessed by the service modules 160c using the real-time switch
154. This allows for shared use of a first pool of service modules 160a across a second pool of service modules 160c. For example, the service modules 160a may include shared video compression devices while the service modules 160c may include conference bridges and video storage servers. Here, the real-time switch 154 permits sharing of the compression devices for conferencing and storage functions.

In addition, accessing the service modules 160a using the real-time switch 154 supports implementation of complex feature combinations. For example, the service modules 160c may include a pool of K analog multi-point conference bridges while the service modules 160a include a pool of high-performance video codecs. The pool of video codecs includes M of a first type and N of a second type. The real-time switch 154 allows for sharing of these codecs across various ports of the conference bridge. This allows for support of extensive combinations of compression types and conference bridge ports while keeping the values of K, M, and N as small as possible.

The service modules 160b are capable of performing similar functions as the service modules 160c, but do not require signal conversion from the service modules 160a.

Embodiments of the multimedia central office may provide data stream networking wherein data streams are provided to the users to serve various purposes. These purposes include: (i) real-time control signaling for connection and services control; (ii) rapid file transfer of image files used in snapshot and application-sharing elements of real-time collaboration; (iii) real-time drawing and pointing commands used in snapshot and application-sharing elements of real-time collaboration; and (iv) general LAN interconnection.

To support data stream networking, the multimedia central office may be employed as follows. Any UTP NTNI user data streams which carry data communications signals in their original logical form are presented to the data hub complex 156 by the UTP transceivers 152. Any UTP NTNI user data streams which carry data communications signals converted to V.35, RS-449, or a similar interface are presented to data routers 162 with a corresponding V.35 interface, an RS-449 interface, or a similar interface via the digital switch complex 150. Any UTP NTNI user data streams which involve processing or action outside the multimedia central office accesses the digital switch complex 150 via the data routers 162 fitted with an appropriate interface, such as V.35 or RS-449. Any UTP TNI user data streams which involve processing or action within the multimedia central office are linked to the data routers 162 (fitted with appropriate interfaces) by means of connections within the digital switch complex 150. Effectively, these variations act to get all data communications into the data hub complex 156 and/or the data routers 162 as necessary.

The data hub complex 156 is coupled to one or more computers 164. It is usually advantageous for the computers 164 to be coupled to other elements within the multimedia central office, typically using either a serial or a parallel interface. The computers 164 may be employed to perform any of the following functions: (i) providing person-based directory services and routing services; (ii) providing resource-based directory services; (iii) providing resource allocation services; (iv) controlling the connections implemented by the digital switch complex 150 and the real-time switch complex 154; (v) controlling hardware systems within the service modules 160a-160c; (vi) implementing any computer-based service feature (such as call forwarding, person locator, and application sharing host); and (vii) administration, usage, and pre-billing processing functions.

In addition to providing interconnection of users terminated on the common multimedia central office, access may be provided as appropriate to one or more other multimedia central offices, a regional hub office, a national hub office, a public digital network, or a private network. Here, the digital switch complex 150 is coupled to these options via a dedicated or switched digital carrier 166. In some circumstances or embodiments, other means for communication, such as analog radio or satellite, may be used.

FIG. 4 is a block diagram of an embodiment of a regional hub office in accordance with the present invention. This embodiment of the regional hub office may be viewed as a simplified version of the multimedia central office in that no UTP or digital carrier user access is supported.

The regional hub office is connected to one or more multimedia central offices via a dedicated or switched digital carrier 200. The regional hub office may optionally connect with one or more national hub offices, other regional hub offices, or other networks via a dedicated or switched digital carrier 202 connected to a wide area public digital network. In some circumstances, other means for communication, such as analog radio or satellite, may be used.

Data switching is performed by a data hub complex 204, which accesses the digital carriers 200 and 202 by switching within a digital switch complex 206 and data routers 210 fitted with appropriate interfaces such as V.35 or RS-449. The data hub complex 204 accesses one or more computers 212 for controlling the digital switch complex 206, any real time switch 214, and other elements of the regional hub office. The computers 212 control the switches 206 and 214, along with other elements of the regional hub office, by means of serial and parallel interfaces. The computers 212 may further provide the above-described functions for the computers 164 in the multimedia central office.

All through routed connections provided by the regional hub office are realized as point-to-point connections within the digital switch complex 206. For cases where the regional hub office provides other services (such as conferencing or video mail), the digital switch complex 206 dynamically connects the digital carriers 200 and 202 with service modules 216a-216c. The service modules 216a and 216b directly access the digital switch complex 206 while service modules 216c require access to service modules 216a via the real-time switch 214. For example, the service modules 216a may include shared video compression devices while the service modules 216c include conference bridges and video storage servers. Here, the real-time switch 214
permits sharing of the compressing devices. Although also true for real-time switches in multimedia central offices, it may be especially advantageous to implement the real-time switches as a backplane hosting service module cards.

The inclusion of the regional hub office in the multimedia telecommunication system may be based upon engineering and economic matters, which include: (i) improved utilization of bandwidth facilities by concentration of the demand; (ii) improved utilization of services resources by concentration of the demand; and (iii) improved reliability and rapid failure recovery by utilizing redundant backup systems. An improved utilization of many types of service resources and minimization of transport costs can result by locating some of the service resources at the regional hub offices.

FIG. 5 is a block diagram of an embodiment of a national hub office in accordance with the present invention. The national hub office is connected to one or more regional hub offices and/or one or more multimedia central offices via a dedicated or switched digital carrier 230 connected to a wide area public digital network.

Data switching is performed by a data hub complex 234, which accesses the digital carrier 230 by switching within a digital switch complex 236 and data routers 240 fitted with an appropriate interface such as V.35 or RS-449. The data hub complex
234 accesses one or more computers 242 for controlling the digital switch complex 236 and a real time switch 244. The computers 242 control the switches 236 and 244, along with other elements of the national hub office, by means of serial and parallel interfaces. The computers 242 may further provide the above-described functions for the computers 164 in the multimedia central office.

All through routed connections provided by the national hub office are realized as point-to-point connections within the digital switch complex 236. For cases where the national hub office provides other services (such as conferencing or video mail), the digital switch complex 236 dynamically connects the digital carrier 230 with service modules 246a-246c. The service modules 246a and 246b directly access the digital switch complex 236 while service modules 246c require access to service modules 246a via the real-time switch 244. For example, the service modules 246a may include shared video compression devices while the service modules 246c include conference bridges and video storage servers. Here, the real-time switch 244 permits sharing of the compressing devices.

The inclusion of the national hub office in the multimedia telecommunication system may be based upon engineering and economic matters, which include: (i) improved utilization of bandwidth facilities by concentration of the demand; (ii) improved utilization of services resources by concentration of the demand; and (iii) improved reliability and rapid failure recovery by utilizing redundant backup systems. In some cases, an improved utilization of service resources and minimization of transport costs results by locating some of the service resources at the national hub offices.

It is noted that a single, high-quality video stream may be carried by a dedicated 5-10 Mbps switched data-LAN channel, such as a switched-Ethernet connection. For example, a one-way broadcast-quality JPEG video stream requires approximately 7
Mbps of bandwidth, and an uncompressed 16-bit 20 kHz digital audio requires approximately 600 kbps. In the 1-3 Mbps range, MPEG I and II deliver higher compression at good quality, but require significantly more expensive encoders. As a result, it is possible to carry video and audio in packetized digital form over the UTP using conventional LAN protocols. If this approach is taken, some or all of the real-time switches can be implemented as high-bandwidth switching hubs. In some embodiments, one or more of the following architectural variations may be valuable: (i) carrying some or all network and services control channels within the real-time video and audio stream; (ii) carrying some or all shared image files within the real-time video and audio data stream; and (iii) merging of the real-time switches with the data hub complexes.

Embodiments of the multimedia telecommunication system may be realized by a variety of different interconnection topologies for the multimedia central offices, the regional hub offices, and the national hub offices. The principles used in contemporary wide area digital carrier and telephony networking, both public and private, may be employed in embodiments having more than a small number of multimedia central offices, regional hub offices, and/or national hub offices. The incorporation of these principals is illustrated by the example topologies which follow.

In order to simplify the discussion of the topologies, the following abstractions are used. All types of user premises are represented generically as a user site. All types of access between a user site and its serving multimedia central office are represented as a generic user access. In the case of digital carriers, this includes both switched and dedicated services. The combination of a digital carrier between a multimedia central office and a local public digital network, a dedicated connection (not switched) provided by the local public digital network, and a dedicated services digital carrier to a regional hub office is represented as a generic dedicated link. Similarly, the combination of a digital carrier between a regional hub office and a wide area public digital network, a dedicated connection (not switched) provided by the wide area public digital network, and a dedicated services digital carrier to a national hub office is represented as a generic dedicated link. Thus, generic access involves any type of access used to link a generic user site with its serving multimedia central office, while generic dedicated links represent any kind of dedicated (not switched) links between pairs of offices.

FIG. 6 is a block diagram of an example topology wherein fully dedicated digital carriers are employed between multimedia central offices, regional hub offices, and national hub offices. A plurality of user sites 260 are each coupled to a corresponding one of a plurality of multimedia central offices 262 via a corresponding one of a plurality of generic user access links 264. Each of the multimedia central offices 262 is coupled to a corresponding one of a plurality of regional hub offices 266 by a corresponding one of a plurality of generic dedicated links 270. Each of the regional hub offices 266 is coupled to a national office 272 by a corresponding one of a plurality of generic dedicated links 274.

This topology is beneficial in allowing the exploitation of conventional bulk pricing of bandwidth, improving the location of shared service resources, and providing alternative access means for users which use NTNI UTP to communicate with one of the multimedia central offices 262. However, this topology does not provide for redundancy for the purpose of failure immunity. Further, in order to handle heavy traffic, a brute-force approach of increasing the number of trunks in the links 264, 270, and 272 must be employed with this topology.

FIG. 7 is a block diagram of another example topology having provisions for overflow and failure conditions. As with the topology of FIG. 6, a plurality of user sites 280 are each coupled to a corresponding one of a plurality of multimedia central offices 282 via a corresponding one of a plurality of generic user access links 284. Each of the multimedia central offices 282 is coupled to a corresponding one of a plurality of regional hub offices 286 by a corresponding one of a plurality of generic dedicated links 290. Each of the regional hub offices 286 is coupled to a national office 292 by a corresponding one of a plurality of generic dedicated links 294.

Further, a plurality of switched lines are provided by a switched public digital network 300 for handling overflow and failure conditions. Specifically, the multimedia central offices 282 access the switched public digital network via access links 302, the regional hub offices 286 access the switched public digital network via access links 304, and the national office 292 accesses the switched public digital network via access links 306.

Using this topology, a failure or heavy loading of one of the links 290 can be recovered by any of the following steps: (i) setting up a connection through the switched public digital network 300 using access links 302 and 304; (ii) setting up a connection with an alternate one of the regional hub offices 286 through the switched public digital network 300 using the access links 302 and 304; (iii) setting up a connection with one of the multimedia central offices 282 through the switched public digital network 300 using two of the access links 302; or (iv) setting up a connection with the national hub office 292 through the switched public digital network 300 using the access links 302 and 306.

Similarly, a failure or heavy loading of one of the links 294 can be recovered by any of the following steps: (i) setting up a connection through the switched public digital network 300 using access links 304 and 306; (ii) setting up a connection with one of the multimedia central offices 282 through the switched public digital network 300 using the access links 302 and 304; or (iii) setting up a connection with one of the regional hub offices 286 through the switched public digital network 300
using two of the access links 304.

A failure or heavy loading of one of the regional hub offices 286 can be recovered by any of the following steps: (i) setting up a connection with an alternate one of the regional hub offices 286 through the switched public digital network 300
using access links 302 and 304; (ii) setting up a connection with the national hub office 292 through the switched public digital network 300 using the access links 302 and 306; or (iii) setting up a connection with one of the multimedia central offices
282 through the switched public digital network 300 using two of the access links 302.

Similarly, a failure or heavy loading of the national hub office 292 can be recovered by any of the following steps: (i) setting up a connection with one of the regional hub offices 286 through the switched public digital network 300 using access links 302 and 304; or (ii) setting up a connection with one of the multimedia central offices 282 through the switched public digital network 300 using two of the access links 302.

One having ordinary skill in the art will recognize that the trade-off in cost between switched services, dedicated lines, and minimum cost load balancing of the offices 282, 286, and 292 can be exploited in creating a specific implementation of the present invention. In particular, implementations may be created which stochastically minimize operating costs.

FIG. 8 is a block diagram of an alternative topology which utilizes redundant links and offices. The additional links and offices are included in this topology in order to obtain higher reliability and trunk availability. As with the topology of FIG. 6, a plurality of user sites 320 are each coupled to a corresponding one of a plurality of multimedia central offices 322 via a corresponding one of a plurality of generic user access links 324. Each of the multimedia central offices 322 is coupled to at least one of a plurality of regional hub offices 326 by a corresponding at least one of a plurality of generic dedicated links 330. Each of the regional hub offices 326 is coupled to at least one national office 332 by a corresponding at least one of a plurality of generic dedicated links 334.

FIG. 9 is a block diagram of a topology in which redundant links and offices are combined with switched lines. A plurality of user sites 350 are each coupled to a corresponding one of a plurality of multimedia central offices 352 via a corresponding one of a plurality of generic user access links 354. Each of the multimedia central offices 352 is coupled to at least one of a plurality of regional hub offices 356 by a corresponding at least one of a plurality of generic dedicated links
360. Each of the regional hub offices 356 is coupled to at least one national office 362 by a corresponding at least one of a plurality of generic dedicated links 364.

A plurality of switched lines are provided by a switched public digital network 370 for handling overflow and failure conditions. Specifically, the multimedia central offices 352 access the switched public digital network via access links 372, the regional hub offices 356 access the switched public digital network via access links 374, and the national office 362 accesses the switched public digital network via an access links 376.

The topology of FIG. 9 provides improved immunity to both failure and heavy loading in comparison to the topology of FIG. 7 and the topology of FIG. 8.

In practice, multimedia services may be provided by various entities using embodiments of the present invention. A first approach is for a common carrier to provide the multimedia services, either exclusively or in partnership with another entity. A common carrier with a UTP loop plant can use NTNIs for user access to the multimedia central offices. Within the multimedia central offices, each which could be as physically small as a large wiring closet in a building, video compression units ("codecs") can be included in the service modules for conversion of the NTNI signals into streams appropriate for TNI digital carriers.

TNI digital carrier terminations (i.e., CSU/DSUs), multiplexing, and codec equipment can be shared among many users accessing them by means of UTP. There are no known tariffed, publicly-offered services of this type.

UTP transceivers may be used to reconstruct converted LAN signals and present them, via a data hub complex and routers, to the digital switch complex for transmission on TNI digital carriers. Appropriate UTP transceivers, alone or together with conversion systems included in service modules, may be used to reconstruct converted V.35, RS-449, or other interface signals and present them to the digital switch complex for transmission on TNI digital carriers. Dedicated and switched bandwidth can be used to link the multimedia central offices to bulk transmission and switching facilities. TNI digital carrier terminations (i.e., CSU/DSUs), multiplexing, and codec equipment can be shared among many users or leased to specific users.

A common carrier without a UTP loop plant may employ a point-of-presence in key buildings and third party operating loops within and/or emanating from these buildings. In fact, common carriers with existing points-of-presence in key buildings are well suited to utilize embodiments of the invention since the UTP loop plant within large buildings is typically conducive to the simplest and least costly of the UTP techniques discussed earlier.

A second approach is for a third party, which purchases bandwidth from a common carrier, to provide the multimedia services. For users who use UTP with NTNIs for access to a multimedia central office, the bulk-rate purchases of dedicated and switched bandwidth help to further lower the cost for real-time wide-area video connection services. For users who use UTP with NTNIs and are located within a common building or building complex, the multimedia central office can be located within the building and be connected directly to TNI digital carrier access for the building.

For users who use UTP with NTNIs and are located within a common neighborhood, the multimedia central office may be located within any office and may be connected by means of passive UTP loops provided by a common carrier. The passive UTP loops are cross connected at the carrier's central office or wire center. An example of a common carrier passive UTP loop service is Pacific Bell's metallic digital loop tariff 3002.

For users who use a conventional TNI digital carrier to access a multimedia central office, these same bulk-rate purchases of dedicated and switched bandwidth may be accessed to provide lower cost connections out of the area. Typically, this would entail a bulk-bandwidth purchase and an infrastructure investment involved in supporting NTNI UTP users driving the availability and lowering the incremental cost for providing real-time wide-area video connection services to users using conventional TNI digital carriers for access.

In complex cases where there is partial involvement of common carriers with a mix of attributes from each of these cases, one skilled in the art can adapt parts from each of these illustrative approaches.

Abstract pictorials of network connection topologies linking users with centralized multi-point conference bridges do not reveal the important impact of conference bridge location on transmission costs. As an example, if all conference bridges in a nationwide network were located at one site serving the entire nation, almost all multi-point conferences would involve considerable costly long distance telecommunications. Such an approach can be improved upon by migrating some of the conference bridges away from a single national center and into regional and neighborhood locations.

In order to illustrate methods of conference bridge allocation, a block diagram of a three layer network such as those shown in FIGS. 6-8 is considered in FIG. 10. The network comprises a plurality of user sites 400a-400e, a plurality of multimedia central offices 402, a plurality of regional hub offices 404, and a national hub office 406.

A classification system for multi-point call topologies can be created based upon the following designations of components of a multi-point conference. Communication between two users connected to the same multimedia central office (such as a user at the user site 400a communicating with a user at the user site 400b) is designated as a type "IN" communication (i.e., within a neighborhood). Communication between two users connected to the same regional hub office (such as a user at the user site 400a communicating with a user at the user site say 400c) is designated as a type "R" communication (i.e., within a region). Finally, communication between two users connected to the same national hub office (such as a user at the user site 400a communicating with a user at the user site 400d) is designated as a type "M" communication (i.e. the maximum topological span).

One user in a multi-point conference is selected as a reference point for classifying each of the types of connections in the conference. For a K-party conference, there are K-1 such connections which are herein referred to as "components".

Next, a (K-1)-tuple of all connection type designations in the multi-point conference call is created. For a 3-party call, there are nine different 2-tuples: NN, NR, RN, NM, MN, RR, RM, MR, and MM. For a 4-party call, there are twenty-seven possibilities: NNN, NNR, NRN, RNN, etc.

All cases that amount to identical algebraic expressions with respect to commutativity of multiplication can be collapsed into a common representative form or class. For example, for 3-party calls, although there are nine 2-tuples, three are redundant in that they may be written in two different ways (e.g., NR=RN, NM=MN, and RM=MR). Thus, there are six possibilities for the representational classes: N.sup.2, NR, NM, R.sup.2, RM, and MR. For 4-party calls, there are ten possibilities since many of the 3-tuples can be written in 3 or 6 ways. These common representative forms are used to label the class of each multipoint conference. In general, there are K(K-1)/2 possible classes for a K-party conference. The number of ways to write a k-tuple is given by the multinomial coefficient of the expansion of the algebraic expansion (N+R+M).sup.k.

For each multi-point conference class, one or more optimal locations for a multipoint conference bridge can be identified which minimize the cost resulting from long distance transmission. FIGS. 11a-11f illustrate the location of a conference bridge CB for the six possible 3-party cases. FIGS. 12a-12o illustrate the location of a conference bridge CB for 15 different 4-party cases. A summary of optimal locations for a centralized conference bridge by multi-point conference class is provided in columns 3-5 of Tables I and II. A few rarely occurring degenerate cases have been omitted from FIGS. 11 and 12, and Tables I and II, in order to avoid descriptive complexity.

TABLE I ______________________________________ # Codecs Ways to N R M # Codecs @ R Class Write optimal? optimal? optimal? @ N or M ______________________________________ N.sup.2 1 X 3F R.sup.2 1 X 3 M.sup.2 1 X 3 NR 2 X 2F + 1 NM
2 X 2F + 1 RM 2 X 3 ______________________________________

TABLE II ______________________________________ # Codecs Ways to N R M # Codecs @ R Class Write optimal? optimal? optimal? @ N or M ______________________________________ N.sup.3 1 X 4F R.sup.3 1 X 4 M.sup.3 1 X 4 N.sup.2 R 3 X 3F + 1 N.sup.2 M 3 X 3F + 1 R.sup.2 N 3 X X 2F + 2 4 R.sup.2 M 3 X 4 M.sup.2 N 3 X X X 2F + 2 4 M.sup.2 R 3 X X 4 NRM 6 X X 2F + 2 4 ______________________________________

Columns 6-7 of Tables I and II show how many codecs are needed per conference bridge at the indicated location should codecs be required to convert signal formats. For example, if the conference bridges use analog video and audio, any neighborhood connections which are carried as analog video and audio over NTNI UTP can connect directly to a conference bridge while all other connections require a codec. For a given neighborhood with fraction "F" of such analog NTNI UTP users and fraction "1-F" of users requiring a codec to interface to the conference bridge, the number of codecs required at that neighborhood statistically behaves as shown in column 6.

For K>3, some classes have more than one optimal choice because several choices have identical transmission requirements. For example, in 4-party conferences these classes include NRM, NR.sup.2, RM.sup.2, and NM.sup.2 as seen in Table II. In these cases, analytic models can be used to compare the multiple optimal options for ranges of traffic patterns for more efficient use of conference bridge resources for a specified blocking factor. Such an analysis will be explored in more detail below, but typically the best choice for locating conferencing resources in these situations is in the regional hub offices with typical hardware savings around 10% over other choices.

Next, the analytical modeling aspect of the procedure applicable to facility sizing will be described. A flow chart of an embodiment of a method for optimal allocation to facilities is shown in FIG. 13. As indicated by block 420, traffic assumptions are made concerning the overall Erlangian traffic and probability distribution of various classes of conference calls. These assumptions are used to derive probability distributions for each conference class supported by the implementation of the invention, as indicated by block 422. In the absence of other information, the same probability distribution used for point-to-point calls can be used for each communication type to construct the probability for each conference class (using standard independence calculations and multinomial combinatoric coefficients that correspond to the number of ways the class can occur--see column 2 of Tables I and II).

Using these probability distributions and overall assumptions about traffic demand, the pro-rated statistical traffic demand for each conference class is calculated, as indicated by block 424. Using optimal location tables such as Table I and Table II, each class of conference traffic is directed to its optimal location, as indicated by block 426. As indicated by block 428, the method further includes a step of separately aggregating, for each location, the total demand for conference bridges across all conference classes.

As shown in block 430, the method may optionally include overflow procedures when appropriate resources are not available (due to loading or failure). If these optional features are included, probalistic adjustments for overflows across locations may be introduced. As indicated by block 432, the traffic demand for conference bridges and desired blocking performance are applied to the server-variable inverse of the Erlang-B formula to determine the required number of conference bridges at each location.

A flow chart of an embodiment of a method of optimal allocation applicable to demand-driven resource allocation is shown in FIG. 14. Once a request for a conference is received, a step of finding the current location of requested participants is performed, as indicated by block 450. Preferably, this step is performed by querying various registration directory services to obtain the current location of the requested participants actively on the system. If some of the requested participants are unavailable, appropriate call handling elements are employed.

As indicated by block 452, the method further includes a step of determining the conference class for each of the requested participants which are available. A step of finding an optimal conference bridge location for the requested participants is performed, as indicated by block 454. Preferably, this step produces a prioritized list of best conference bridge locations based upon a predetermined look-up table. Optionally, the method may further consider other possible participants which may be added for inclusion in the creation of the prioritized list of best conference bridge locations, as indicated by block 453.

Next, the method performs steps which attempt to allocate conference bridges and communications trunks based upon the prioritized list. Starting with the first priority choice, a step of requesting a conference bridge in the desired location is performed as indicated by block 458. If the request is successful, the necessary trunks are requested in block 460. If either the requested conference bridge is unavailable or the requested trunks are unavailable, the next best choice in the list is attempted.

Embodiments of the above-described allocation methods have several valuable impacts: (i) they save considerable transmission costs by providing a simple classification scheme for selecting the most optimal conference bridge location for each conference call; (ii) they provide a procedure for identifying how many conference bridges are required at each network office; (iii) an additional 10-20% in conference bridge hardware costs can be saved by refining the above procedure to exploit the best allocation choices in thus far "don't care where" cases (namely, R.sup.2 N, M.sup.2 N, M.sup.2 R, and NRM; see Table II); and (iv) they can then be incorporated into the actual control algorithms which allocate resources when users place conferencing calls.

Although this aspect of the present invention is targeted at location of video conferencing resources, the principles can also apply to conventional telephone audio conferencing systems and application sharing server systems.

Next, alternative embodiments of conference bridges for use in the multimedia telecommunications system are presented. Preferred embodiments of the present invention employ continuous presence conference bridges so as to deliver realistic support for multiparty human interaction. The use of continuous presence conferencing bridging provides for the distribution of conferencing resources among offices and demand-driven allocation policies of conferencing resources. This is of value in facilities planning and in minimizing transmission costs. Continuous presence conference bridges provide an improvement over other available wide-area conferencing systems relying on multipoint control units (MCUs), which display only one participant's video image at a time.

The types of continuous presence conference bridging utilized by embodiments of the present invention may be categorized as having either a centralized or a decentralized geographic spread. Further, the types of continuous presence conference bridging may be characterized by the type of video mosaic implementation utilized, namely, analog video domain, pixel domain, discrete cosine transform (DCT) domain, or variable length coding (VLC) domain.

FIG. 15 is a block diagram of a simplified layered architecture of a compression scheme in order to illustrate the aforementioned domains. An analog video source 480 provides an analog video signal to an analog-to-digital converter 482. The analog-to-digital converter 482 digitizes the analog video signal to form a digitized video signal. After conversion, the digitized video signal is applied to a DCT signal processor 484 which provides processing based upon a DCT operation. The output of the DCT signal processor 484 is applied to a compressor 486 which creates a stream of variable length code words. The compressor 486 may employ one of many compression algorithms known in the art.

The variable length code words are then passed to a transmitting communications subsystem 488 which supplies real-time timing and/or file-creation support. A signal transmitted and/or stored by the transmitting communications subsystem 488 is received and/or retrieved by a receiving communications subsystem 490. The receiving communications subsystem 490 acts to recover real-time stream of variable length code words.

The recovered stream of variable length code words is passed to a decompressor 492, which employs a decompression algorithm to invert the action of the compression algorithm in the compressor 486. The decompressor 492 produces a resulting DCT video stream which is applied to an inverse DCT signal processor 494. The inverse DCT signal processor 494 produces a digital video stream which is applied to a digital-to-analog converter 496. An analog video stream produced by the digital-to-analog converter 496 can be presented to an analog video sink 498 such as a display device.

The signals produced by the analog video source 480 and the digital-to-analog converter 496 are regarded as being in an analog video domain, since both contain an analog video signal. The signals produced by the analog-to-digital converter 482
and the inverse DCT signal processor 494 are regarded as being in the pixel domain, since both contain a digital pixel stream. The signals produced by the DCT signal processor 484 and the decompressor 492 are regarded as being in the DCT domain, since both contain a DCT stream. The signals produced by the compressor 486 and the receiver 490 are regarded as being in the VLC domain, since both contain a stream of variable length code words.

Embodiments of the present invention may employ commercially available compression schemes which compute a discrete cosine transform to convert the image from two-dimensional amplitude distribution to two-dimensional frequency distribution. The DCT is the core of many popular deployed and commercially available compression schemes, in particular the ITU-TSS H.261/H.320 standard supported by many video codec manufacturers.

With regard to the use of pixel-domain video compositing in the conference bridges, embodiments of the present invention may use many different approaches. Two specific approaches to pixel-domain video compositing include a centralized implementation and a distributed implementation. The use of pixel-domain video compositing elements (such as the video mosaic units made by Panasonic, For-A, and others) to realize continuous video presence in video conferencing is disclosed in U.S. application Ser. No. 08/131,523 for both centralized and distributed implementations. In addition to these approaches, embodiments of the present invention may use any of the extensions discussed herein to follow. In particular, the distributed implementation of DCT-domain or VLC-domain compositing described herein supplements the art of distributed conference bridges as described in U.S. application Ser. No. 08/131,523.

In the pixel-domain, a 2-by-2 mosaic of four full-motion video regions can be formed by dropping every other pixel in each image dimension from four full-screen pixel streams, and positioning the resulting half-by-half sized regions to form a
2-by-2 mosaic. There are many variations of this, such as forming 3-by-3 mosaics by dropping every 2 out of 3 pixels in each dimension, or forming mixed-region-size mosaics by using different pixel dropping rates for the pixel streams emanating from different sources, as is obvious to one having ordinary skill in the art.

In a wide area situation, at least some of the video images will arrive in the digital communications signal format. Commercially available video mosaic units (such as those made by Panasonic, For-A, and others) provide interfaces for signals in the analog video domain, although they internally perform the video compositing in the pixel domain. The conversions between the analog video domain and the pixel domain are performed by separate internal analog-to-digital and digital-to-analog converters. The reason for this is that low cost interfaces in the analog domain are common and well supported, while low cost interfaces in the pixel domain for these types of interconnections are not available or defined.

More preferably, pixel-domain compositing may interface codec systems in the pixel domain by means of an appropriate interface. In practice, this interface can be engineered to be relatively trivial to implement since both commercially available codecs and video mosaic units typically must convert to and from the analog domain in a similar fashion. Without a coordinated design, the codecs and video mosaic units could well use completely different digital representations of the video frame. There are cost, complexity, and performance advantages to this approach.

Switching among a pool of codecs and a pool of video mosaic units can be done in a conventional digital backplane. Presently, in the analog video domain, this is performed by wideband analog switching matrices, which requires additional hardware for cable transceivers, multiple power supplies, and low-noise wideband analog backplanes. This redundant hardware can be avoided by an appropriate scale server implementation.

The most likely candidate for the analog video interface signals are popular standardized composite signals (NSTC, PAL, and SECAM). Each of these introduces image degradation with each pair of digital-to-analog and analog-to-digital conversions. This can be somewhat problematic in the areas where chroma and luminance frequency bands overlap somewhat and are filtered in different ways at each interface.

As an additional level of improvement, considerable improvements in cost and performance can be made by compositing in the DCT domain. This technique has been explored by others; see "Video Compositing in the DCT Domain", by Chang, Chen, and Messerschmitt, Proceedings of IEEE Workshop on Visual Signal Processing and Communications, Rayleigh, N.C., September 1992. Embodiments of the present invention may employ this approach using the methods of Chang, Chen, and Messerschmitt or related methods.

Effectively, the operations involved in dropping pixels and compositing as a mosaic can be represented mathematically as under-sampling and translation operations. These under-sampling and translation operations have corresponding two-dimensional impulse responses which, when convolved in two dimensions with the video stream frames in the pixel domain, create the desired video mosaic. When this entire process in transformed into the DCT domain, the DCT frames (in the DCT streams within the DCT domain) are multiplied in two dimensions with the DCT of the under-sampling and translation operations, producing the DCT of the desired mosaic. The multiplication is, in general, less complex to implement than the convolution. More importantly, the mosaic can be done without use of the DCT and the inverse DCT signal processors for video signals arriving at the bridge from the wide area. This saves costly hardware, greatly reduces codec delay, and is a natural candidate for incorporation into an appropriate large-scale server implementation.

Even further cost savings and performance improvements can be made by compositing in the VLC domain. This technique has been explored by others; see "Video Bridging Based on H.261 Standard", by Lei, Chen, and Sun, IEEE Transactions on Circuits and Systems for Video Technology, Vol. 4, No. 4, pp. 425-437, August 1994, for example. Distributed implementations of video compositing is also a natural candidate for incorporation into an appropriate large-scale server implementation.

Slight variations allow the aforementioned methods of FIGS. 13 and 14 to be applied to optimal allocation of other related resources within the multimedia telecommunication system. Specifically, the method may be applied to allocate components within service modules such as codecs used to provide digital access to analog video domain conference bridges. Further, the method may be applied to allocate internal components of a conference bridge server, such as the analog-to-digital and the digital-to-analog converters used to provide analog access to pixel, DCT, or VLC domain conference bridges.

More specifically, users of the multimedia system may use either analog NTNI access or digital access to a multimedia central office. For analog video domain conference bridge interfaces, codecs are needed for digital access, but are not needed for analog NTNI access. For conference bridge interfaces in the pixel domain, the DCT domain, or the VLC domain, analog-to-digital and digital-to-analog converters are required to provide analog access to pixel, DCT, or VLC domain conference bridges.

As an example, consider the need for codecs when using analog video domain conference bridges. Since the regional hub offices and the national hub offices do not have analog NTNI access, the number of codecs needed for each conference class (whose optimal location is one of these offices) is simply the number of participants in the conference. This is illustrated in the column 7 of Table I and Table II. At the multimedia central offices and for conference classes where multimedia central offices are the optimal conference bridge location, codecs are needed for each outgoing trunk in conference classes involving users whose access does not terminate on the same multimedia central office. In terms of the communication type designations defined earlier, one codec is needed for every R or M component of the conference class.

Further, if the fraction of users terminating on that multimedia central office with digital access is denoted by a variable F (which ranges between 0 and 1), then the probability a codec is needed is F times the number of users terminating on that multimedia central office which are participating in the conference. In other words, an additional factor of F for the reference user and one each for every N component in the conference class is included in the probability that a codec is needed. This is illustrated in the column 6 of Table I and Table II. The resulting probabilistic (for the multimedia central offices) or deterministic (for the regional hub offices and national hub offices) number of codecs needed per class can then be used to pro-rate the traffic in the steps indicated by blocks 428 and 430 in FIG. 13. Performing the step indicated by block 432 to the pro-rated traffic gives the number of codecs needed for this purpose.

Since commercially available implementations of pixel-domain video compositing elements (such as the video mosaic units made by Panasonic, For-A, and others) are single physical hardware units, centralized implementations of continuous presence conference bridges would appear to be the easiest direct implementation approach to continuous presence video compositing. In U.S. application Ser. No. 08/131,523), however, distributed implementations of video compositing are shown to be advantageous in wide area situations.

The local video of all participants can be included in the mosaic without multiple codec delays (not just those participants who happen to be collocated near a centralized conference bridge in a way that no round-trip codec delay is incurred for them).

In many cases, the trunking requirements which result can be considerably reduced. For example, with two users in location A and two users in location B, the centralized bridge would require 2 trunks if the bridge is collocated at a sufficiently near location A or B (and 4 trunks if the conference bridge is in a third location), while a distributed bridge needs only 1 trunk to give the same results.

Methods and systems for connecting a workstation or a room to a multimedia central office will now be discussed. In general, any means for connecting the workstation or room directly with the multimedia central offices via a public telephone loop plant UTP are sufficient, as are any means for connecting the workstation or room with the multimedia central office through an internal premises communications system via the public telephone loop plant UTP. Clearly, the longer distances of loop plant UTP which can be used, the more valuable reuse of the existing worldwide investment in public telephone UTP loop plants. Thus, it is desirable, but not essential, to carry audio, video, and data signals through as long a length of public telephone loop plant as possible.

A further discussion of methods and systems for interfacing with NTNI workstations and rooms is appropriate. Approaches to computer control distribution of analog video and audio within a premises is taught in U.S. application Ser. No.
08/131,523, and in U.S. Pat. Nos. 4,686,698, 4,847,829, and 5,014,267. These approaches rely on the employment of on-premises codecs for connecting workstations or rooms to the outside world. Embodiments of the present invention avoid the per-premises cost of these codecs by use of a public network service providing NTNIs which link to publicly shared codecs. Additional cost savings and features are realized by designing each resulting element as part of an overall wide area collaboration and multimedia networking system.

In considering how to implement a public network service that avoids the cost of these per-premises codecs through use of NTNIs, audio, video, and data are delivered by either per-premises trunks or per-workstation/room direct lines or per-premises trunks. Preferably, the specifications for the audio, video, and data which are delivered are as follows: (i) data communications supporting rates between 128 kbps and 10 Mbps; (ii) at least 3 MHz, and more preferably, 4-6 MHz analog color video or a perceptual near-equivalent; and (iii) at least 5 kHz, and more preferably, 7-15 kHz analog audio or a perceptual near-equivalent. The alternative of "perceptual near-equivalent" is meant to allow for the inclusion of cases where signal processing is used in such a way to deliver comparable quality with less actual transmission bandwidth.

Various techniques for delivering these requirements over UTP, focusing on the employment of NTNI, but not exclusive thereto, are considered. First, the methods by which workstations or rooms 100 connect with public networks are explained.

Referring back to FIG. 2, user premise 104c includes workstations which connect to the public network UTP loop plant 114 individually, and user premise 104d includes workstations which connect through the internal premises communication network
116 with one another and share UTP loops within the public network UTP loop plant 118. In each of these cases, the workstations do not connect directly with the public network UTP loop plant 114 or 118, but rather are linked by connection means to a loop UTP transceiver. The connection means may include any of the following:

(i) UTP transceivers used for premises networking, as described in U.S. application Ser. No. 08/131,523;

(ii) coaxial cable, as described in U.S. Pat. Nos. 4,686,698, 4,847,829, and 5,014,267;

(iii) fiber optic links; or

(iv) wireless premises radio.

In cases where a user premise does not provide an internal premises network 116, the loop UTP transceiver may be physically located in a premises wiring closet, or may be collocated with each workstation or room system. Alternatively, the connection means may be very short electrical connections within a piece of workstation or room equipment so that the loop UTP transceiver is actually physically incorporated into the workstation or room equipment.

In cases where a user premise does provide an internal premises network 116, the loop UTP transceiver is very likely to be physically located in a premises wiring closet. In this arrangement, the internal premises network 116 provides a concentration function to a pool of shared loop UTP transceivers. Such an arrangement allows a number of workstations and rooms to share a smaller number of UTP trunks and associated transceivers. The connection means may include UTP transceivers, coax cable, fiber optic links, or wireless premises radio.

In either of the two aforementioned cases, the part of the loop UTP transceiver which interfaces with the public network UTP loop plant 114 or 118 typically differs from UTP transceivers used for premises networking in that it typically provides the following functions:

(i) lightning protection and other safety isolation functions;

(ii) additional transmission power and/or receiver sensitivity to compensate for the electrical conditions of the public network UTP loop plant 114 or 118;

(iii) additional frequency compensation to compensate for the electrical conditions of the public network UTP loop plant 114 or 118;

(iv) signal processing and transceiving to compensate for the electrical conditions of the public network UTP loop plant 114 or 118; and

(v) response to remotely controlled installation and maintenance functions (such as loopback, generation/detection of test signals and automated cable compensation adjustments) which in a practical embodiment of the invention may be executed from the multimedia central office.

The loop UTP transceiver employs the public network UTP loop plant 114 or 118 to reach a matching loop UTP transceiver within the multimedia central office. This matching loop UTP transceiver is similar to the loop UTP transceiver used on user premises 104c and 104d but may incorporate differences in packaging for use in a public utility office. The packaging may differ with regard to powering (which may include redundancy provisions), the type of input/output connections and/or cabling employed, and the physical package format. Further, the matching loop UTP transceiver may be capable of: (i) issuing remotely controlled installation and maintenance functions (such as loopback, generation/detection of test signals, automated cable compensation adjustments) which in a practical embodiment of the invention may be executed from the multimedia central office; (ii) performing measurements of response of loop UTP transceiver to remotely executed installation and maintenance functions; or (iii) performing maintenance functions (power checks, input/output connection checks) which are relevant.

A block diagram of a method and system that utilizes inverse multiplexing to carry a single bit stream 510 on a plurality of lower-rate digital carriers 512 is illustrated in FIG. 16. The bit stream 510 is applied to an inverse multiplexer 514
to produce a plurality of lower-rate bit streams. The lower-rate bit streams are communicated through a public digital network 516 over the plurality of digital carriers 512. The lower-rate bit streams are received by an inverse multiplexer 520 which reconstructs the original bit stream. Typically, the inverse multiplexing is bidirectional. This technique is employed, for example, in video teleconferencing where two 56 kbps channels are used to carry a single 112 kbps stream.

Due to the limited bandwidth of a public loop plant UTP, particularly at longer distances, embodiments of the multimedia telecommunication system deliver higher bandwidth signals by "dividing" original signal bandwidth across a plurality of public loop plant UTPs. Moreover, both digital and analog signals may be carried by the UTP using broadened forms of inverse multiplexing.

The division of a signal across multiple physically separated channels is herein referred to as "space division" (in contrast with "time division" used in time-division multiplexers). Space division techniques may be employed to split a signal across multiple UTPs in order to deliver higher bandwidths across longer distances where the transmission properties of the UTP may rapidly degrade (e.g. above 1-2 MHz or even lower analog frequencies).

Digital streams can easily be split and reclocked to create multiple streams operating at lower clock frequencies and, for the range of distances involved, be reassembled into the original digital stream. In analog signals the splitting is relatively straightforward, but the precision frequency transmitter down-shifting, receiver up-shifting, and receiver channel recombining required may be problematic for arbitrary high-fidelity analog signals. In the case of NTSC/PAL/SECAM analog video signals, however, special structural properties of the signal can be exploited.

Methods and systems for splitting a digital stream of a given rate into multiple digital streams of lower rates and reconstructing the digital stream are well-known in the art. Often, the steps in this process are referred to as serial-to-parallel conversion and parallel-to-serial conversion. Typically, a public telecommunications network can not provide a plurality of channels whose relative transmission delays are within a bit-symbol period. Further, in extreme circumstances, differential timing jitter among the plurality of channels can also create problems. Modern inverse multiplexing equipment provide means at the receiving site for measuring the relative transmission delays among the plurality of channels and, based on these measurements, inducing compensating delays on all but the most delayed channel(s) so as to reconstruct the original relative timing used at the transmitting site.

Embodiments of the present invention may directly employ UTP space division (i.e., an adaptation of digital inverse multiplexing) in many ways. For digital control/image/data-communications signals, as illustrated in FIG. 17, inverse multiplexers 530 and 532 are used to transform higher bit rate digital signals 534 into a plurality of lower bit-rate signals 536. Preferably, each of the lower bit-rate signals 536 has a bit rate higher than that for an ISDN basic rate interface. These lower bit-rate digital signals are encoded for transmission over a UTP 540 within the public telephone UTP loop plant 542 by means of UTP transceivers 544. The UTP transceivers 544 include UTP transmitters 544a and UTP receivers 544b. Upon receiving the lower bit-rate signals, one of the inverse multiplexers 530 and 532 reconstructs the original signal.

For video and audio signals, as illustrated in FIG. 18, digital conversions and compression of a signal 560 is performed via a codec 562a. The codec 562a