CA2032617C - Arrangement for reserving and allocating a plurality of competing demands for an ordered bus communications network - Google Patents

Abstract

Abstract The present invention relates to a method and a system for reserving and allocating a plurality of competing demands for an ordered by communications network in response to a plurality of demands from a source. The network includes a plurality of customer sites, each site including one or more endpoints. Each endpoint is connectable to a switching system. The switching system is connectable to one or more other switching systems. A unit is providedfor connecting a first endpoint through the one or more switching systems to a second endpoint. The improvement with respect to the method is comprised of the steps of receiving demands from a customer site for allocating the network for acommunications among a plurality of customer sites; stratifying the received demands in response to a grouping of endpoints to be conferenced at the customersites and allocating the network resources to connect the endpoints to be conferenced in response to the stratified demand. The method includes generatingsetup and terminate times for each endpoint to be connected in a conference;
reserving the endpoints to be connected in the conference; and generating a plurality of bindings for signalling the network and the source as to success ofconnecting the endpoints to be conferenced over the ordered bus network. The improved system includes a unit for receiving demands from a customer site for allocating the network for a communication among a plurality of customer sites; a unit for stratifying the received demands in response to a grouping of endpoints to be conferenced at the customer sites and a unit for allocating the network resources to connect the endpoints to be conferenced in response to the stratified demand. The system includes a unit for generating setup and terminate time for each endpoint to be connected in a conference; a unit for reserving the endpoints to be connected in the conference and a unit for generating a plurality of bindings for signalling the network and the source as to success of connecting the endpoints to be conferenced over the ordered bus network.

Description

. - 2032~7 ARE~ANGEMENT FOR RESE13VING AND ALLOCATING A
PLURALlTY OF COMPETING DEMANDS FOR AN
ORDERED BUS COMMU~CATIONS NElWORK
Back~round oî the Invention Field of the Invention This invendon relates to resource allocadon and, more pardcularly, to allocating a constrained common resource ~mong a plurality of demands for the resource.
Description of the Prior Art The tenn "resource allocation" applies to that class of problems, which has as a common characterisdc the need to physically allocate a restActed or 5~ constrained common resource among a plurality of demands for that resource. For ~ ~
example, in a communicadons system, it is common to allocate network resources ~;
,~ such as trunks for video conferencing on a manual first come, first served basis.
15 Obviously, it would be desirable to automate the allocadon process.
Dynamic storage allocation problems are well known in the art. The ~ -term "dynamic storage allocation" applies to that class of problems, which relates to reserving contiguous space in some storage medium for some period of dme in order - -to satisfy the demands a customer may have for the storage medium. Unfortunately, 20 kno~n dynamic storage allocations algorithms tend to be NP-complete, i.e. they are ~ ~;
not generally solvable in reasonably short periods of time. More specifics can be found in any standard text such as the text by Michael R. Garey and David S.
Johnson endtled "Computers and Intractability - A Guide to the Theory of NP-Completeness", New York: W. H. Freeman and Company (1979). Tn light of the 25 above, it is common to settle on a solution which may represent a halance between the imperfect utilization of the resource and the ~me to find a recommended solution. Notwithstanding, known processes to find a recomrnended solution of the allQcation of the constrained common resource to meet the plurality of demands still -- require excessive arnolmts of time and, therefore, a more ~mely solution remains 30 needed in the ar~ Furtherrnore, flexiibility is needed to better serve the almost unpredictable needs and demands of the customers, who want and need to use the ; common resource. Unfortunately, the common resources often suffers from maladies and cons~aints of its own, not the least of which may be that the common resource is unavailable.
~:

., .
' '~' ;~
` :: 2~32~17 Summar~ of the In~ention These and other problems are solved in accord with the principles of my invention which includes a method and a system for allocating a constrained common resource among a plurality of demands for the resource. In response to a S plurality of demands from a source, the network, which includes a plurali~ of customer sites, each customer site including one or more endpoints, each endpoint being connectable to a switching system, the switching system being connectable to another switching system, comprises means for connecting a first endpoint through one or more switching systems to a second endpoint. The allocation arrangement 10 receives demands from a customer site for allocating the network for a communication among a plurality of customer sites and stratifies the received demands in response to a grouping of endpoints to be conferenced at the customersites. It can then allocate the network resources to connect the endpoints to be~ conferenced in response to the stratified demand including generating setup and ; 15 terminate times for each endpoint to be connected in a conference; reserving the endpoints to be connected in the conference; and generating a plurality of binding for signalling the network and the source as to success of connecting the endpoints to be conferenced over an ordered bus network. The arrangement also stratifies acustomer's demand into sets of strategies for groupings of like endpoints; allocates 20 network resources responsive to characteristics of the different groupings of like endpoints; for each strategr, generates the setup_times and terrninate_times foreach endpoint and reserves the endpoints for a reservation and generates bindings.
The arrangement also, for a first grouping of like endpoints such as codecs, :i alternate accesses, and off netwets, extends a conference signal through a 25 multiplexer or concentrator at a customer site over an access line to a homing DACS; while, for a second grouping of like endpoints such as audio bridges, extends a conference signal from a customer site to a central bridging DACS;
while, for a third grouping of like endpoints such as VCCs, extends a conferencesignal from a customer site to the network; means for generating a DMB sequence 3() number for the reservation.

~;

~' .

~; l `` - 2032~17 ~, ' c -2a-. ., In accordance with one aspect of the invention there is provided a ~
method for reserving and allocating a plurality of competing demands for an : ~ :
ordered bus communications network in response to a plurality of demands from a ~ source, the network including a plurality of customer sites, each customer site .~ 5 including one ore more endpoints, each endpoint being connectable to a switching : system, the switching system being connectable to one or more other switching systems, means for connecting a first endpoint through sa;d one or more switching systems to a second endpoint, where the improvement comprises the steps of: (a) receiving demands from a customer site for allocating the network for a .
~ lû communication among a plurality of customer sites; (b) stratifying the received ~;
;~ demands in response to a grouping of endpoints to be conferenced at the customer sites; (c) allocating the network resources to connect the endpoints to be ~:
conferenced in response to the stratified demand including generating setup and .~
terminate times for each endpoint to be connected in a conference; reserving the . ~:
endpoints to be connected in the conference; and generating a plurality of bindings .`.
for signalling the network and the source as to success of connecting the endpoints :
to be conferenced over said ordered bus network. .
In accordance with another aspect of the invention there is provided .
!'. a system for reserving and allocating a plurality of competing demands for an ordered bus communications network in response to a plurality of demands from a .
source, the network including a plurality of customer sites, each customer site ` ::
including one ore more endpoints, each endpoint being connectable to a switching : .
system, the switching system being connectable to another switching system, means :~:
: for connecting a first endpoint through one or more switching systems to a second : 25 endpoint, wherein the improvement comprises: means for receiving demands from :
, , a customer site for allocating the network for a communication among a plurality ~ of customer sites; means for stratifying the received demands in response to a ; grouping of endpoints to be conferenced at the customer sites; means for ~ ~
allocating the network resources to connect the endpoints to be conferenced in - ~ -response to the stratified demand including (i) means for generating setup and ~`~
;
, .,~
;
.

: ~ ~ ; ~
,. . .

: :

; j 20326~7 -2b-terminate times for each endpoint to be connected in a conference; (ii) means for reserving the endpoints to be connected in the conference; and (iii) rneans for generating a plurality of bindings for signalling the network and the source as to success of connecting the endpoints to be conferenced over said ordered bus 5 network.
These and other problems are solved in accord with the principles of my invention which includes a method and apparatus for allocating a constrained common resource among a plurality of demands for the resource.
Brief Description of the Drawings My invçntion should become more apparent from the following detailed description when taken in conjunction with the accompanying drawing in which: .

:,i, ?;

: :

:: :

: :`

: ~
~ :

~ 20326~7 S~ ;
s,~

FIG. 1 is a block diagram illustrating a network, which is useful in ~: ~
explaining an illustrative embodiment of the prinriples of the present invention, and :::
FIG. 2 illustrates a customer site, which is also useful in explaining an illustrative embodiment of the principles of the present invention, and S FIG. 3 illustrates a first ses of customer demands, which need to be allocated to satis~y she needs of a customer a~d which is helpful in understanding a first example that is also useful in explaining she pAnciples of the present invendon, FIG. 4 illustrates a second set of customer demands, which need to be - -allocated to sasisfy she needs of a customer and which is helpful in understanding a 10 second example that is also useful in explaining the principles of the present inventlon, FIG. 5 illustrates my methodology for allocating a set of custome~ : .
demands, which is helpful in understanding the principles of the present invendon, FIG. 6 illustrates an allocator process for codec-like endpoints for use in~ -15 the allocator of FIG. 5, which is helpful in understanding the principles of the present invendon, FIG. 7 illustrates an allocatQr process for audio bridge-lilce endpoints for . .
use in the allocator of F~G. 5, which is helpful in understanding the principles of the ~:
present invention, :
FIG. 8 illustrates an allocator process for VCC-like endpoints for use in the allocator of FIG. 5, which is helpful in understanding the principles of thepresent invention, and ~ ;~
FIG. 9 illustrates memory and controller for for use in the allocator of ~IG. 5, which is helpful in understanding the principles of the present invention.
25 Detai!edDescri~tion :
Refemng to FIG. 1, each of a plurality of customer sites 110-1 through : 1 l~N is coupled over a respecdve access line 12~1 through 12~N (note that, ~ ~
although N is shown to be eight in FIG. 1, sarne is forpurposes of illustradon and ~ ;
not by way of limitadon since N can be a design parameter, which can be any integer ~ `-30 greater than one) to a rèspecdve digital access and cross-connect sys`tem (DACS) such as DACS 130~1 through DACS 130-M (no~e that, although M ~ shown to be ~;
: four in FIG. 1, same is for pulposes of illustradon and not by way o~ limitation since M can be a design parameter, which can be any integer greater dlan ~ero). DACS
equipment such as the publicly used Digital Access and ~ross-Connect System is . ~ 35 commercially available from American Telephone and Telegraph Company. .
Various features for a DACS including a Digieal Multipoint Bridge (DMB) and ' :, ", - 2~326~7 Opdcal Fiber Linlc Interface are also comme,rcially available and in public use.In my example, customer site 110-1 is coupled to DACS 130-1 by way of access line 120-1; customer sites 110-2 and 110-3 are coupled to DACS 130-2 by way of access lines 120-2 and 120-3, respectively; customer sites 110-4, 110-5, and 5 110-6 are coupled to DACS 130-3 by way of access lines 120-4, 120-5, and 120-6, respectively; and customer sites 110-7 and 110-8 are coupled tO DACS 130-4 by way of access lines 120-7 and 120-8, respectively. Also, the respective DACS arecoupled to each other by way of an ordered bus communications network ar~angement. For exarnple, DACS 130-1 is coupled toDACS 130-2 by way of first 10 internodal link 14~1 and DACS 130-2 is coupled to DACS 130-3 by way of secondinternodal link 140-2 and DACS 130-3 is coupled to DACS 130-4 by way of third internodal link 14~3.
~ ach access liDe 120-i and each internodal link 140 j can include one or more communications paths and each communications path has some predeterrnined 15 bandwiddl and the bandwidth of one communications path may be, but need not be, the same as the bandwidth of another comrnunications path and each , communications path can be a transmission facility, (a) which may directly connect, for example, a customer site 110-i and a DACS 130-k or may directly connect one . DACS with another DACS, for example, DACS 130-1 with DACS 130-2 or (b) which may indirectly connect, for example, a customer site 110-i and a DACS 130-k or may indirectly co:mect one DACS with another DACS, for example, DACS 130-1 with DACS 130-3 through interposed equipment or facilities, which interposed equipment or facilities could include even another DACS lilce DACS 130-2 ineerposed between DACS 13~1 and DACS 13~3.
It is now useful to distinguish between the term "bus" and the term "internodal link." On the one hand, the term internodal link refers to the connection between two DACS, e.g., in FIG. 1, internodal linlcs 140 j, which may directly connect~ for example, one DACS 130-k with another DACS 130-m or which may indirectly connect, for example, the two DACS through interposed equipment or 30 faci}ities like another DACS. On the other hand, the term bus refers to an ordered set of internodal links, which connects a se~ of DACS in a specific ordered configuration like that shown in FIG. 1. As a further aid in distinguishing the terms9 consider the following example. Each DACS 130-k can be connected to one or more other - DACSs like DACS 130-m by a respective internodal link 14~j. Further, when all 35 the DACS in the network are completely interconnected, it is clear that there could be many ordered sets of internodal links. Each ordered se~ of internodal linlcs that .
.
.

.'J
2 ~ 3 2 ~ ~ 7 ,.; 5 3 satisfies a predetennined criteria is called a bus. I have shown only one bus arrangement in FIG. 1, i.e. internodal link 140-1 coupling DACS 130-1 with DACS ~ ~-~/ 130-2 then internodal link 140-2 coupling DACS 130-2 with DACS 130-3 then internodal link 14~3 coupling DACS 130-3 with DACS 130-4, which results in the ordered set of internodal links (140-1, 140-2, 140-3), which is a bus. In the trivial ~ -case where there is only a single DACS, there could be a single bus within the single DACS and there need not ~e any internoda1 links.
;.;l In general, it is desired that the resources of a network be allocated in some timely manner such that equipment at one customer site can comrnunicate with `-`1 10 one or mo~ units of equipment at one or more other customer sites. The term "broadcasting'! is often applied to an arrangement wherein equipment at one ~ ~-customer site communicates in a substantially simultaneous manner with one or ~more units of equipment at one or more other customer sites. Hence, broadcasting is ` -concemed as to the signal source. The term "conferencing" is applied to an 15 arrangement, which, without regard to signal source, involves communication among two or more units of equipment whereby a set of customer sites communicateamong each other.
In accord with one aspect of the principles of my invention, an allocation methodology is set forth by way of which it is possible to allocate the 20 limited or constrained resources of a network such as the bandwidth of a communications path in the network among a plurality of customer demands for that bandwidth and, importantly, to do the allocation in a manner such that the expectations of the customer are sadsfied.
Assume that a customer or source of a demand at customer site 110-3 25 places a demand on a network that a video conferencing connection be set up ~-between customer site 110-3 and all of the other customer sites 110-i shown in FIG.
1. The customer ~pically would have several other demands. For example, the customer may demand that the conference begin at a specific time of a day, called a start dme~ate, and that the conference terrninate at a specific time of the same, or of 30 a different, day, called a stop time-date. E~urther, the custome~ may demand that the communication paths, which are to be allocated, have some minimum bandwidth, which may be subject to a constraint such as the constraint that the network will not -. allocate bandwidth ~eyond some predetermined maximum bandwidth. Still further, i~
the customa may, before the conference begins or during the time interval of the35 conference, demand the flexibility to have certain customer sites dynamically added ` ~`
to, or te~minated from, the conference during Ihe time interval of the conference and, ' 20326:L7 s~ll even further, these added or terminated demands may be made with relatdvelyshort lead-time and may be made either before or during a conference call. The customer may further demand that the time to start or stop the conference be changed or that the minimum bandwidth of the conference be changed. In addidon S to the above, there may be still other customer demands and there may be still other Af, constraints on those demands.
Those are some of the problems, which are solved in accordance with the principles of my invendon. To illustrate those principles, we will shortly describe a procedure, which may be used to generate the bindings for a reselvadon.
10 But before the pr~cedure is described, some addidonal te~ns will need definidon.
; The term "reservadon" refers to a set of demands by a customer such as the set of demands described in the immediately preceding paragraph. The tcrm "binding"
refers to the concept represented by a four-tuple of the form (Xl, X2, X3, X4) where Xl is a first specific posidon within the bandwidth of a commurlicadons path; X2, 15 which for our descripdon is assumed to be larger than X1, is a second specific position within the bandwidth of the communicadons path; X3 is the time-date that the customer wants the conference to be setup; and X4 is the dme-date that the customer wants the conference to terminate. Note that the difference between X2 and X1, i.e. X2 - Xl, corresponds to the minimum bandwidth that the the customer 20 dem~nds, which itself is subject to the constraint that the minirnum bandwidth demanded may not exceed the maximum bandwidth of the communicadons path, which serves the reservation.
To better understand one meaning of first posidon X1 and one meaning of second position X2 consider a well known digital transmission network, which 25 includes channelized bandwidth. For example, it is known that a digital network may include 24 chaMels (called DS0 channels, each DS0 channel operating nominally at about 64 ldlobits per second) in a single muldchannel communications path (called a DS1 channel, which operates nominally at about 1.544 megaWts per second). ~ turn, it is known that 28 DS1 channels may be included in a single 30 mul~ichannel communicadons path ~called a DS3 channel, which operates nominally at about 44.7 megabits per second), etc. Now, first position X1 may be a specific one of the 24 DS0 channels while second position X2, which is larger than X1, may be a specific other one of the 24 DS0 channels.
To better understand the setup-time-date X3 and the terminate-time-date 35 X4, note that X3 and X4 could be respectively some dme-date before and some time-date after the actual start-time and actual stop_time demanded by ~he customer, 2~32~1 7 , ., e.g., the value of X3 in the four-tuple could be less than the actual start-time-date ~,~ demanded by the customer so as to allow the network adequate dme to set up a connection, hence X3 is also called the setup time-date, and may be shortened tosetup time. Accordingly, setup dme X3 is a time-date such that the conference 5 connection, if its connecdon commences be~ore the start time, will satisfy theexpectations of the customer. The difference between the setup time-date and thestart ~rne-date is now described. The time to allow the network to setup a t~ conference connection is called the "res setup time delta ." While the dme to allow the network to add an endpoint to an in-progress conference is called the ~' 10 "room setup time delta." Theres setup dme deltaandroom setup time delta ,~ may~have different values depending on the constraints of the network, but are ` ~ assumed to be zero or greater.
Similarly, the value of X4 in the four-tuple could be a time-date, which `l is greater than the actual stop time~ate demanded by the customer, so as to allow ~ 15 the network adequate dme to disconnect, or tear-down, the connection, hence X4 is '';~t, also called the terminate dme-date, and may be shortened to terminate time.
Accordingly, telminate time X4 is a dme-date such tha~ the conference connection, . if disconnected no later that time X4 but certainly no sooner than the actual ` stop time-date demanded by the customer, will sadsfy the expectations of the t`. 20 customer. The difference between the terminate time-date and the stop time-date is ., now described. The time to allow the network to terminate a conference connecdon ,h~ iS called the "res teardown time delta ." While the time to allow the network to !'~` terminate an endpoint in an in-progress conference is called the "room terminate time delta." Theres teardown time deltaand ;25 room terminate time delta may have different values depending on the constraints of the netwo~k, but are assumed to be zero or greater. ~-Finally, the plural of the term binding, i.e. the term "bindings", means the one or more binding that satisfies the customer's demand.
I''! It is now useful to describe more specifically a customer site 110-i. -' 30 Typically, a custorner site includes one or more units of equipment. Each unit of equipment at a customer site is called an endpoint. Further, there may be various kinds of endpoints~ e.g. endpoints of a first kind, endpoints of a second kind, etc. To exemplify this concept, the ensuing description will consider five kinds of endpoints, - called respecdvely, a codec (for coder/decoder) endpoint, an audio bridge endpoint, a 35 video control console (VCC) endpoint, an alternate access endpoint, and an offnet (for off network) endpoint. A codec like codec 200 in FI&. 2 is a unit of equipment .
:` .
.
~'~

,s~ .s~
.,,,,~,, .. - - . - , . . ~ . . .. . . .

~32~ 7 , on a customer's site, one form of which could encode a video sign~l from a source for transrnission to a sink where the encoded signal could be decoded, for example, to its original form. An audio bridge like audio bridge 220 in FIG. 2 is a unit of equipment on a customer's site, one form of which could be an arrangement, whichS mixes audio signals from a plurality of sources into a single signal that is to be extended to one or more other audio bridges. A VCC like VCC 240 in FIG. 2 is an example of yet another specific unit of equipment, or endpoint, located on a customer's site, one form of which could include capabilides that pennit a customer site to originate a video conference, to communicate with a plurality of other 10 customer sites, e.g. to function as a "broadcaster" of irlformation to rnany sites at ¦ one time, place a reservation request, or to perform other control functions.
Generally, for purposes of this description, on the one handj an alternate access like ~3 alternate access 260 in FIG. 2 is the tenn to describe any unit of equipment, other than a codec, audio bridge or VCC, which functions as an endpoint and is located on 15 a customer's site. On the other hand, an offnet li1ce offnet 280 in FIG. 2, although not necessarily a unit of equipment, can be *ought of as a quasi-unit of equipment, and, as will be made more clear, is considered as a unit of equipment for our purposes. Indeed, an offnet could be a service rather than a product. For example, an offnet ~ould be a member of the Accunet ~R) family of digital services, which is a 20 comrnercially available service. More generally, an offnet, which is a short-hand form of the words off network, refers to a signal processing arrangement, a signal transport arrangement, or any other ar~angement, which ~ or may not be on the customer's site but which extends a signal to the instant i~e~vork for purposes such as interconnecting the o~&et provided signal to one or more endpoints within the25 instant network. A distinction between an offnet and an alternate access is that an alternate access would be a unit of equipment (other than a codec, audio bridge or VCC) on a customer's site while an offnet would be any arrangement other than a unit of equipment on the customer's site.
, It may be noted that a common characterisdc among all the endpoints is 30 that each endpoint will require some minimum bandwidth to communicate with another endpoint and that each endpoint is adapted to extend a signal through a - network to one or more "like" endpoints and that the network has no need to "understand", e.g. no need to decode, the information content of the extended signal.
Using the example of the above five kinds of endpoints; three of those kinds of 35 endpoints, i.e. codecs, audio bridges and VCCs, are examples of "unlilce" endpoints since, for exarnple, a codec would typically not understand the information content , ~

: 2~32~ 7 of a signal extended from an audio bridge.
Like endpoints are endpoints, which would understand signals transmitted therebetween. For example, two endpoints having the same kind of equipment like two codecs, or two audio bridges, or ~wo VCCs would be examples , S of like endpoints. In addition, a customer could demand that a signal be extended ~rom a codec at one customer site to an alte~nate access or to an offnet at another customer site. In this exarnple, the codec and the altemate access or the codec and ~;~ the offnet could be like endpoints. This example assumes that some arrangement would be included in the alternate access or the offnet, which arrangement could10 understand the information content of the extended signal.
In contras~, unlike endpoints are endpoints, which would typically not understand signals transmitted therebetween. For example, a cod~c at one customer ,i site would be an unlike endpoint with respect to an audio bridge or a VCC at another customer site.
Referring to FIG. 2 and keeping ~he description of FIG. 1 in mind, ` ~`
sample customer site 1 lO~i is coupled over access line 120-i to DACS 130-k.
Customer site 110-i can include a plurality of endpoints, each endpoint coupled over a respective communicatdons path through a muldplexer or concentrator 295 thencecoupled over access line 120-i to DACS 130-k. In my example, the endpoints 20 include codec 200, audio bridge 220, video control console (VCC) 240, alternate `
access 260, and offnet 280 coupled through muldplexer or concentrator device 295respeedvely by way of communicadons paths 210, 230, 250, 270 and 290 onto access line 120-i. Although FIG. 2 illustrates one of each of the five kinds of endpoints that were mendoned above, same is not by way of limitation, but only by 25 way of example. A customer site may include od~er combinadons of the shown endpoints, may include other kinds of equipment as endpoints, may include some but not all of the shown kinds of endpoints, et cetera.
Continuing, a customer may request or demand that a connection be established, i.e. that there be "connectivity", between t~o customer sites, which is ' 30 herein referred to as a two point reservation. A twopoint reservadon may reflect a ~-customer's demand for connectivity in order to allow a single signal to be communicated between the two customer sites. Altemadveb, a customer may request or demand connecdvity between more than two customer sites, which is m~
- herein referred to as a muldpoint reservation. A multipoint reservadon may reflect a 35 customer's demand for connectivity in order to allow a single signal to be shared among a plurality of like endpoints such as (a) connectivity to bridge a plurality of ~-;
. . .

. ' ~

2032~7 ., the audio bridges like audio bridge 220 located at different ones of customer sites 110-i so as to allow a first signal to be commonly shared, or broadcast, or conferenced among the connected audio bridges andJor (b) connecdvi~y to connect a plurality of video control consoles like VCC 240 located at different ones of S customer sites 110-i so as to allow a second signal to be commonly shared, or broadcast, or conferenced among the connected VCCs and/or (c) connecdvity to ~r connect a plurality of codec, alternate access, and offnet endpoints like codec 200, alternate access 260 and offnet 280 located at different ones of customer sites 110-i ; so as to allow a third signal to be commonly shared, or broadcast, or conferenced !, 10 among the connected endpoints.
A multipoint reservation may specify the desired connectivity among two or more endpoints or contemplate a predetermined default set of connecdvides.
For example, the customer may demand either (a) that certain codecs be connectedAND that certain aud.io bridges be connected. AND that certain VCCs be connected15 or (b) that certain codecs AND that certain alternate accesses AND ~hat certain offnets be connected AND that a predetermined set of audio bridges and VCCs be connected, which predetermined set is a function of the specific codecs, alternate accesses and offnets that were demanded to be connected by the customer.
One example of alternative (b) above is an embodiment havi.ng a codec 20 200 at a customer site 110-i involved in a conference in which case the ~! predetermined set could include at least one audio bridge 220 AND at least one VCC
240 at the customer site 110-i and further result in the audio bridge 220 and the VCC
` 240 remaining coMected to the conference for so long as the codec 200 remains connected to the conference. In this example, coMectivity would remain among r 25 codec 200 and the endpoints (e.g. audio bridge 220 and VCC 240) in the predetermined set even though a connection among other codecs or other alternateaccesses or other offnets at customer site 110-i has been disconnected. The connecdvity among codec 200 and the endpoints in the predetermined set would remain for so long as a conference remains for codec 200 at the customer site 1 l~i.
30 However, it should be noted that the endpoints in a first predetermined set for first codec 200 could ~Iso be endpoints in a second predetermined set for a second codec or other endpoint at customer site 1 l~i, in which case, even if the first codec 200 is . disconnected ~om the conference, the common endpoints in the first and the second predetermined sets could remain connected by virtue of the second codec remaining 35 or being (e.g. being added to the conference) in the conference.

:

~ 2 ~ ~ 7 . ,~

,, Referencing FIG. 5, my methoclology stratifies (500) customer demand (400) by like endpoints into a plurality of sets of customer demands and then allocates (600, 700, 800) the resources of the network for each set of customer demands with like endpoints and makes the allocation information available (900) to S the network for llse thereby and also feedbacks back (510) the allocation results to the customer. The allocation information is stored and control by way of memory and controller 900 for making the informadon available to both the network and the customer. The controller 910 provides the data requested by allocators 600, 700, and 800 to them and if all allocators are successful in allocating the reservation, the 10 controller removes all tuples in memory associated with this reservation and places the new set of tuples associated with this reservation into the appropriate memories.
If the allocators were successful in allocating the reservation, the controller 910 notifies 510 the customer demand 400 and the network of the success of the methodology and the new tuples. If the allocators were not successful in allocating the reservation, the controller 910 notifies 510 the customer demand 400.
The methodology then iterates the following actions~
1. Stratifies the customer's demands into sets of strategies for ~ -~
groupings of like endpoints;
2. Allocates network resources responsive to characteristics of the 20 diffe~ent groupings of like endpoints;
3. For each strategy, generates the setup times and telminate times for each endpoint;

4. For each strategy, reserves the endpoints for the reservadon;

5. ~or each strategy, generates bindings;

6. For a first grouping of like endpoints such as codecs, alternate accesses, and offnets, extends a conference signal dlrough a muldplexer or ~ -concenlrator at a customer site over an~ access line to a homing DACS;

7. For a second grouping of like endpoints such as audio bridges, extends a conference signal from a customer site to a central bridging DACS;
' 30 8. For a third grouping of like endpoints such as VCCs, extends a conference signal from a customer site to the network; ~
9. Generates a DMB sequence number for the res~vadon; ` ~ ~;
10. Generates cross connects in the DACS to provide connectiYity of -~
like endpoint conference signals;
11. Generates the conference signal internodal connectivity required for the first grouping of codec lilce endpoints; and ::~
:

~l 2032617 1`, . 12. Determines if the network is not available for reconfiguration of the conference during any reconfiguration of the network.
What follows now is a description of each of these actions in furt`her detail.
S As to endpoint demand stratifier 500, endpoints of a first grouping, e.g.
codecs 200, alternate acresses 260, and offnets 280, are allocated acco~ding to a first strategy 600; endpoints of a second grouping, e.g. audio bridges 220, are allocated according to a second strategy 700; endpoints of a third grouping, e.g. VCCs 240, a~e allocated according to a third s~ratégy 800; etc. Each grouping of endpoints includes 10 only like endpoints and no one grouping includes both like endpoints and unlike endpoints.
Cornrnon among each strategy are the processes of generating ~610 in FIG. 6,710 in FIG. 7, and 810 in FIG. 8), setup ~nes X3 and ternunate times X4 for the endpoints reserving (620 in FIG. 6, 720 in FIG. 7, and 820 in F~G. 8), the 15 endpoints and generating (630 in FIG. 6, 730 in F~G. 7, and 830 in FIG. 8). bindings (Xl,X2,X3 andX4) Before going into the details of three such s~rategies, an example of a process of allocating bandwidth to a conference on an access line or on an internodal linl~ is provided through the example shown in FM. 3. The example is chosen so as 20 to also focus on a process of generating bindings.
Referring to ~IG. 3, assume that a plurality of customer demands gives rise to a corresponding plurality of bindings, which are labelled in summary form in FIG. 3 as 31~1, 31~2 and 310-3. The customer demands may be allocated to two communica~on paths 32~1 and 32~2 of an access line 12~i or an internodal link 25 14~k. Here assume the paths are part of internodal link 140 k. Customer demand 31~1 gives rise to a four-tuple binding on communicadons path 320-1 of internodal link 14~k. The four-tuple includes a first specific position Xl within the bandwidth of communications~path 320-1 of Cl; a second specific position~X2 within the bandwidth of communications path 32Q-1 of C6; a setup time X3 of tl; and a 30 terminate time X4 of tS. Customer demand 310-2 gives Ase to a four-tuple binding on communicadons path 320-1 of internodal link 140-k. The four-tuple includes a first specific posidon Xl within the bandwidth of communications path 320-1 of (~
- a second specific position X2 within the bandwidth of cornmunicalions parh 320-1 of C12 where for this example p equals 12; a setup dme X3 of t2; and a telminate time -35 X4 of t4. Customer demand 310-3 gives rise to a four-tuple on communicadons path 32~2 of intemodal link 140-k. The ~our-tuple includes a first speci.fic position Xl .
: ~:
. ~

. , -, c ~32~17 '~ - 13-wi~n the bandwidth of communications path 320-2 of Cl; a second specific position X2 within the bandwidth of comrmmications path 320-2 of C5; setup time X3 of t2; and a te~ninate time X4 of t4. For purposes of the example, the bindings are stored in a binding memory 930 shown in FIG. 9. Further, for the pu~poses of5 this example, assume that there are no two-tuples (X3, X4) stored in unavailability memory 970 also shown in FIG. 9. Unavailability memory 970 contains sets of tw~tuples (X3, X4) where X3 and X4 have the same time definitions as for a binding. The two-tuples are stored in unavailability memo~y 970 and serve a function of indicadng that no cornmunica~ion path was located as being available10 from the time-date beginning at setup time X3 to and including terminate time X4.
Assume a new customer demand 330-1 is received. A binding is needed on internodal link 140-k with a four-tuple of (Xl=l, X2=S, setup time X3=t3, tern~inate time X4=t6), where (X2 - Xl) corresponds to the minimum bandwidth necessary to sadsfy the customer demand and the setup time and terminate time 15 corresponds to the time period the connectivity is required to be invol~red in the conference.
New demand 330-1 can be sadsfied by a binding 310-4. Hence, new demand 33~1 can be sadsfied as demand 310-4, which is served on communicadons path 32~2, which gives rise to a four-tuple binding on communicadons path 320-2 `
20 of internodal link 140-k. The four-tuple includes a first specific position Xl within the bandwidth of communicadons path 320-2 of C6; a second specific position X2 within the bandwidth of communicadons path 320-2 of Cl l; setup time X3 of t3;
and a terminate dme X4 of t6. :: ;
So much for an example of generating a binding As mentioned, each respecdve grouping of like èndpoints has its ;
respective strategy for allocating the network resources for conferencing purposes.
To illustrate a first strategy, my methodology includes an arrangement ~-600 for allocating network resources for conferencing a first gr~uping of like endpoints such as codecs, alternate accesses, and offnets. Codecs, alternate accesses, 30 and offnets can be coMected in a conference by extending a conference signal from a unit of equipment ~ike 220, 260 or 280) through a multiplexer or concentrator 295 ~-over an access line 12~i to a homing DACS 130-~c. Thereby, ~e endpoints are connected to their respective homing DACS and, fi~her, the respective homing DACSs are connectable among themselves. Thereupon and at any dme after being so 35 connected, each unit of equipment, which has its conference signal extended over its access line to its homing DACS and whose horning DACS is so connected, is able to -:-:
. , . '~
. . ~:

. ~' r~
`: ~
~0~26~ 7 ,, ~'¦ participate in the conference.
For purposes of identificadon and by way of example only, the signal to be conferenced will be referred to as the "conference signal". To establish connectivity of the conference signal, the methodology also includes the process of . 5 extending the conference signal from each endpoint in customer site 110-i, which ; endpoint is an endpoint for the conference being demanded by the customer, through multiplexer or concentrator 295 to an access line 12~i. That access line 120-i is called a conference signal access line and the connectivity, which is being established, is called the "conference signal access line connectivity". DACS 130-k ~0 to which the conference signal is transmitted over the conference signal access line 120-i is called the "homing DACS" for each specific endpoint in customer site 11~i, which is a pardciparlt in the conference.
To illustrate a Iirst strategy, my methodology includes a process of generadng 610 setup dmes X3 and terminate times X4 in response to the customer 15 demanded start time and stop_time for each endpoint in the first grouping of like endpoints.
The process of generadng the setup times and terminate times for the endpoints ~610 in FIG. 6, 710 in PIG. 7, and 810 in FIG. 8) is a two step process.
The setup time, X3, is generated on an endpoint by endpoint basis. Each endpoint; 20 that has a start time equal to the start dme of the reservadon, has a setup dme set to the start time of the reservatdon less the dme res setup time delta, i.e., X3 =
start time - res se~up dme delta. Each endpoint that does not start at the start dme of the reservadon, has the setup dme set to the start time of the endpoint less the timeroom setup ~ne delta, i.e., X3 = start time -room setup time delta. Note 25 that the res setup time delta may be larger than the room setup time delta to allow for additional time for such acdons as the dme necessary for the network to verify the connecdvity is coIrect and correctly setup for this reservadon~
The terminate time, X4, is generated on an endpoint by endpoint basis.
Each endpoint Ihat has a terminate time equal to the terminate time of the 30 reservation, has a terminate time set to the terminate dme of the reservation plus the dme res terminate dme delta, i.e., X4 = terminate time +
res terminate dme delta. Each endpoint that does not te~minate at the terminate dme of the reservation, has the terminate time set to the terminate time of the endpoint plus the the dme room terminate time delta, i.e., X4 =
35 terminate time+room terminate dme delta. Notethatthe res terminate dme deltamaybe-largerthantheroom terminate time deltato ' .

~i3 ~2617 allow for addidonal time for such a~tions, for example, as the time necessary toverify the network has disconnected the conference and preventative maintenance has occurred. ` -;~ The endpoints are then reserved 620 for the conference and the bindings S may be concu~ently generated 630 in a manner similar to that described for theexample in PIG. 3. We now describe the process of reserving endpoints (620 in FIG.
6, 720 in FIG. 7, and 820 in FIG. 8). The process of reserving endpoints uses the endpoint availability memory 980. The endpoint availability memory 980 stores a tw~tuple (X3, X4) where X3 and X4 have tHë same definitions as in bindings, for an 10 endpoint which indicates that the endpoint is not available for the time period beginning at X3 and terminadng at X4. Given a setup time and terminate time of the endpoint which would satisfy the customers demand, my methodology searches the endpoint availability memory for any two-tuple (X3-i,X4-i) where the following conditions (a) X3-i is less than X4 (b) X3 is less than X4-i are true. If any two-tuple 15 is found that matches the above conditions, the endpoint is considered unavailable (e.g, because it is involved in another conference) and the customer demand can not be satisfied. If no two-tuple is found, my methodology generates, but does not store, a two-tuple (X3,X4) for the endpoint into the endpoint availability memory 980.
The bindings are generated 630 in a manner similar to that described for 20 the e~cample in ~IG. 3. The binding arrangemen~ assumes an ordered bus `
configuration that allows one to find a signal to be conferenced from among the customer's demands.
My methodology then iterates the above to find the connecdvity required to connect all of the endpoints in the first grouping of like endpoints, which 25 are at each of the other customer sites 1 lOj, which are to be pardcipants in the conference, to their respecdve homing DACS 130-m so as to meet the customer's demand.
My methodology thereafter generates the identity of all of the homing DACSs and, responsive to a predefined set of buses, generates a set of buses, which 30 interconnect all of the homing DACSs. Responsive to the set of generated homing DACS buses, my methodology generates a set of bindings to interconnect the homing DACSs during the time interval (a) that each respective hon~ng DACS has at least one endpoint participating in the conference or (b) that each respective homing DACS, independent of having an endpoint participadng in the conference 35 for the entirety of the conference (i.e., note that a DACS can be a homing DACS
until all endpoints at all customer sites for which it is the homing DACS are ;~

.

--~ ` 2~32B~ 7 , disconnecte~ from the conference. Thereupon, it looses its status as a homing DACS. However9 it may continue its status as a DACS, which is an element of an internodal link between other DACSs, which others DACS condnue their status as homing DACs3, is a DACS, which is an element of an internodal link that is a busS for the confer~nce. This connectivity is calle~d the "conference signal internodal connectivity". Based on the location of the endpoints needing connectivity in a reservation, conference signal internodal connectivity may or may not ~e ~equired.
For example, if codec 200 at customer site 110-2 and codec 200 at customer site 110-3 are involved in a conference, no intemodal links are required to connect the 10 homing DACS, since both codecs have the same homing DACS, namely DACS
130-2, involved in the conference.
My methodology then operates in response to whether the reservadon type 640 is a two point reservation or a multipoint reservadon. For a two point reservation, it is ne essary to generate cross connects (6~0) in response to which the I5 conference signal access line connecdvity of one endpoint may be cross connected to both the conference signal internodal connectivity (if any) and the conference signal access line connecdvity of the other endpoint.
To generate the DACS cross connects 650 for a two point reservadon, my methodology searches the set of generated bindings and matches those bindings20 that have communications paths that are coupled to the same unit of equipment. B~r way of example and referring to FIG. 1, a communicadons path on access line 120-1 and a communications padl on intemodal link 14~2 terrninate on the same unit of equipment and are involved in the conference for the same penod of dme. For those bindings that haw communications paths that terminate on the same unit of 25 equipment, my methodology generates a six-tuple (Xl-l, X2-1, Xl-2, X2-2, X3, X4) where (Xl-l) and ~X2-1) are respectively the Xl and X2 of one binding, Xl-2 and X2-2 are respectdvely, the Xl and X2 of another binding, X3 and X4 are the X3 and X4 of one of the two bindings (since they have the same X3 and X4), which will be stores in cross connect memory 940. It should be noted that each binding on an 30 access line will be involved in generating one DACS cross connect and a binding on an intemodal link will be involved in generadng two DACS cross connects, one foreach DACS tbat it is connected to it. It should also be noted that if the set ofgenerated bindings does not use an internodal link, that the bindings on the twocommunicadons paths on the access lines will be cross connected.

..~ , :~ ~
, ..

~J 2~g~

On the other hand, ~or a multipoint reservation, because of the binding a~rangement for conferencing endpoints of the first grouping of like endpoints, which in this example are codecs, alternate accesses, and offnets, the methodology does not need to determine l:)ACS cross connects. Rather, the cross-connect would 5 be accomplished as a roudne function of a standard DACS.
To illustrate a second strategy, rny methodology includes an ?,` arrangement 700 for allocadng network resources for conferencing a second grouping of lilce endpoints such as for conferencing audio bridges. Audio bridges can be connected in a conference by extending the conference signal from an audio 10 bridge 220 to a central bridging DACS. The conference signals from all the audio bridges and like endpoints participating in the conference can be connected through a DMB located at the central bridging DACS. The DMB provides a mixing function of all audio signals connected to a DMB sequence number so that each audio signal extended to the DMB seiquence number is able to receive the mixed audio signal 15 from all the other extended audio signals that are connected to the DMB sequence number and, hence to the eonference.
At this juncture, some additional definidons are needed.
The term "channel group" refers to the concept represented by a two-tuple of the form ~Xl, X2) where Xl and X2 have the same bandwidth definitions as y 20 for a binding. Therefore, a channel group represents a subset of the bandwidth of a communicadons path. A "circuit" is then a set of channel groups that define one connectivity ar~angement between two units of equipment where the units of equipment could be endpoints or other than endpoints, for ex~mple, the units of equipment could be DACS switches or, for that matter, any units of equipment and25 where the units of equipment could have other units of equipment interposed therebetween. ~ ~
~ The term "schedulable circuit" refers to a circuit, wh*h has the -characterisdc that it (i.e., the circuit) is available to be reserved for use to meet a ; specific customer demand for a specific conference. A schedulable circuit could be 30 marked with a flag to indicate that it is reserved for a specific conference and it is unavailable for a differen~ conference during the time inten~al of the specific conference and, of some significance, it is included in a set of bindings, which would ~ ;i meet the customer's demand. When a schedulabie circuit is included in a set of bindings, the schedulable circuit is said to be "reserved" for the time interval defined 35 in the bindings, i.e., from setup time X3 to terminate time X4, USillg the set of channel groups (Xl,X2) defined in the circuit. One example of a schedulaWe circuit . ~ ~

.

f~ 0 3 2 6 ~L 7 '~ ~ is an audio p~ivate line circuit between an audio bridge 220 at a customer site 110-i and a central bridging DACS.
The ~erm "central bridging DACS" refers to any of the DACS 130-k in the network, which DACS includes a standard digital multipoint bridge lDMB) for S merging or mixing a plurality of audio signals whereby each audio bridge 220 at each customer site 110-i, which is participating in the conference, has coupled to it the plurality of audio signals. It should also be noted that there may be more than one cen~ral bridging DACS in the networic.
Continuing with my description and by way of example, the audio 10 bridge signal, which is also the conference signal, from each audio bridge like endpoint can be extended to the central bridging DACS via one of two mechanisms.The first mechanism allows allocator methodology 700 to generate 730 a set of bindings over the access line and internodal links to pro~ride a connecdon to the central bridging DACS. The second mechanism uses a "schedulable" audio ~` 15 private line. My allocator methodology provides the flexibility to allow the customer demand to be satisfied using either mechanism.
To illustrate the first mechanism where the methodology determines a set of bindings, my methodology includes a binding arrangement for generating bindings, which assumes an ordered bus configuration that allows one to find a 20 signal to be conferenced from among the customer's demands. To establish connectivity of the conference signal, the methodology includes the process of extending the conference signal from each endpoint at customer site 110-i, which ~i endpoint is an endpoint like an audio bridge endpoint for the conference being demanded by the customer, through multiplexer or concentrator 295 over an access25 line 120-i to its homing DACS.
To illustrate the second mechanism where the audio bridge signal to be !: extended to the central bridging DACS utilizes an audio private line circuit, my ` methodology includes a binding arrangement for reserving the audio bridge's audio private line circuit. To establish connectivity of the conference signal, the 30 methodology includes the process of extending the conference signal from eachendpoint at customer site 110-i which endpoint is an audio bridge like endpoint for the conference being demanded by the customer to the cen~ idging DACS. This process reserves the audio private line circuit, which is retrieved from the audio private line circuit memory 950 and which provides connectivity from the endpoint 35 at customer site 1 lO~i to the central bridging DACS for the pe~iod of time the endpoint is involved in ~e conference.

~ i ~
` ~ 2~26~7 iy 1 9 .
My methodology iterates the a~ve for each endpoint, which endpoint is an audio bridge like endpoint for the conference being demanded by the customer,based on the mechanism each endpoint utilixes to connect to the central bAdging DACS, to find the connectivity required to connect all of the audio bridge like 5 endpoints in each of the other customer sites 110j, which are to ~e participants in the conference, to their respective central bndging DACS so as to meet the customer's demand.
This process includes scheduling the schedulable circuits, scheduling or reserving 760 a DMB sequence number, and generating 750 the required DACS
10 cross connects. The circuit may or may not be a communications path on the access line or through the bus configuration.
My me~hodology also reserves a ~MB sequence number 760 for the conference. The DMB sequence number to be used is determined by cycling through all possible DMB sequence numbers available on each DMB. My 15 methodology generates a four-tuple (X3, X4, X5, X6), where X3 and X4 are the same as defined for a binding, where X5 is the identity of the specific l:)MB, which contains the DMB sequence number to be used, and where and X6 is the DMB
sequence number to be used for reserving the DMB sequence number used by the audio bridges in a reservadon. A DMB sequence number X5 can be used to serve 20 one reservadon but cannot be used to serve two reservadons at the same time.
The methodology attempts to find the DMB sequence number using two sets of dmes. If the first set of times to be used is not able to reserve a DMB
sequence number, the second set of times is used. The first set of times places an incremental time called buffer time around the demanded setup time X3 and 25 terminate time X4 to allow possible customer demands such as a demand for an extension of the reservadon once it is up. The second set of times uses the setup time X3 and terminate time X4 without the buffer time included. The first set of times used are the setup time and terminate time with a buffer of res buffer time subtracted or added to each time. The second set of times used are ' 30 the actual setup time and terminate time.
The methodology determines a DMB sequence number for a ~ -reservation. Determining the DMB sequence number to be reserved can be done using any of several alternatives. For example and not by way of limitation, themethodology could use the first DMB sequence number which has not previously 35 been used OR the methodology could use the DMB sequence number which has the greatest time interval between other conferences that will be served by the Dh IB

. ~, : ~
.

2~3~ 7 ~J
'i, '1 `I , sequence number OR the methodology could cycle through the set of DMB sequence ; numbers fo~ each reservation ignoring any DMB sequence numbers which are serving other conferences during the time of the reservation OR the methodology could cycle thrwgh the set of DMB sequence numbers on a daily basis for each S reservation in that day ignoring any DMB sequence numbers which are serving other conferences during the time of the reseIvation.
As one example and to show one alternadve by way of which a DMB
'~ sequence number can be reserved, we will now describe the procedure where the `s~ methodology cycles through the set of DMB sequence numbers on a daily basis for 10 each reservation in that day ignoring any DMB sequence numbers which are serving other confierences during the time of the reseNadon. The methodology of reserving a DMB sequence number will use the last DMB sequence number memory 990 in ~` which is stored three-tuple (X8, X9, X10) where X8 is the DMB that contains DMB
sequence number X9 is the last DMB sequence number cycled through for the day . 15 spesified by X10, and X10 is the date imbedded in the setup time X3 of the reservations which will use this three-tuple. The methodology retrieves the set of three-tuples from the last DMB sequence number memory where the X10 is the day of the setup time of the reservation. Chosing the first tuple, the methodology increments ~he X9 by ono. The process of incrementing X9 by one may cause the A'. 20 value of X9 to bo reset to the first DMB sequence number available in DMB X8 if X9 is greater than the number of DMB sequence numbers maintained by the DMB
~; X8. The methodology then searches the DMB sequence number memory 960 for any four-tuples (X3-i, X4-i, XS-i, X6-i) that the following conditions (a) XS-i = X8, (b) X6-i = X9, (c) X3-i is less than X4 + buffer time, (d) ~3 - buffer_time is less 25 than X4-i are true. If any tuples are retrieved from memory, the methodology increments X9 and iterates the above search. If no tuples are retrieved from memory, the me~hodology generates a four-tuple (X3, X4, X8, X9) and replaces thetuple retrieved from the last DMB sequence number memory 990 with the three-tuple (X8, X9, X10) where the value of X9 contains the newly incremented value of -30 X9. It should be noted that the DMB dlat this DMB sequence numberis associated ~
with may have a maximum number of audio bridges that may be connected to the ; `
DMB. If this is the case, the methodology would also need to sea~ch the endpoint- memory to detennine all the audio bridges which are using the DMB at each given -moment of time and whether or not this maximum number of audio bridges limit has35 been reached. If the limit has been reached, the methodology would not generate a tuple and would con~nue. Further, if the maximum number of audio blidges being ` ~
. ` -'`,''~''''.

`; ~

~ ~0326~7 ,~

served by the network are are less than this limit, the methodology does not need to perfonn this additional check. The methodology continues to iterate the above until either a tuple is generated or the value of X9 is equal to the old value of X9 which was re~ieved from the last DMB sequence number memory 990. If a tuple was ~;~ S generated the methodology is finished reserving a DMB sequence number for the g reservation. If the tuple was not generated, the methodology continues. It should be noted, that the methodology must search the D~fB sequence number for tuples even~; if the value of X9 has not be cycled (reset to the first DMB sequence number available on the DMB) because the DMB sequence number may be serving another 10 conference that has a setup time earlier than the day of the setup time of the reservation currently being reserved.
The metilodology iterates the above by selecting the next three-tuple (X8, X9, X10) from the set of tuples retrieved from last DMB sequence number memory. ~ no tuple was generated, the methodology repeats the above process with15 the change that the buffer_time is set to zero for purposes of the search.
If the conference is in-progress, the methodology searches the DMB
sequence number mernory for any tuples that the following conditions (a) X3-i isless than X4, (b~ X3 is less than X4-i, (c) XS-i = X5, (d) X~i -- X6 are true. If tuples are retrieved from memory, the customer demand is denied. If no tuples are 20 retrieved, the methodology modifies the X4 of the old reservation of the DMB
sequence number for this reservation with the new value of the X4. The methodology may simply update the value of X4 if the new value of X4 is less than the old value of X4 reserved in the DMB sequence number memory. It should be noted, that bæause the conference is in-progress, the methodology uses the same 25 DMB sequence number ~hat is cuIsently being served by the conference. By way of example, the methodology couid detelmine a new DMB sequence number for the conference.
To illustrate a third strategy, my methodology includes an ~rangement ~or conferencing like endpoints such as a third grouping of like endpoints including 30 VCCs. VCCs can be conferenced by a firs~ mechanism of extending a conference signal to the network by direct distance dialing (DDD) or by a second mechanism of extending the conference signal through the multiplexer or concentrator over theaccess line to the homing DACS. My binding methodology provides the flexibility to aDow file customet dom md to bo sahsfied using dther mechimism.

: ::

; 2~2617 , - 22 -.".
I'o illustrate the first mechanism where the conference signal is ~JI extended to the netwo~k by DDD, my methodology includes an arrangement which assumes a public switched network (PSN) to find a signal to be conferenced from among the customer's demands. The connectivity for a VCC 240 in a customer site S 110-i with another VCC requires that each VCC be available during the dme that it is needed for the conferencing. The direct d;istance dialing will use a public switched i network which is assumed to always allow the VCCs to connect to the network > unless the direct distance dialing is not available, for example, due to a failure of its , network.
To illus~ate the second mechanism where the conference signal is extended to the network over the access line, my methodology includes a binding ? ' arrangement 830 for conferencing the VCCs. To establish connectivity of the r conference signal, the binding methodology includes the process of extending the conference signal from each endpoint in customer site 110-i, which endpoint is an lS endpoint for the conference being demanded by the customer, through multiplexer or concentrator 295 to an access line 120-i. The connectivity from the endpoint over the access line to the homing DACS satisfies ~e customer's demand. Further, the , methodology could be constrained to place the VCC signal on a specific bandwidth , position within the communicadons path.
My binding methodology iterates the above for each endpoint, which endpoint is a VCC like endpoint for the conference being demanded by the customer, based on the mechanism each endpoint utilizes to connect to the network, to find the connectivity required to connect all of the VCC like endpoints in each of the other customer sites 110-j, which are to be participants in the conference to the network so 25 as to meet the customer's demands. ~
On the one hand, bindings ~or a future conference call may result in a -new set of bindings in which all the elements Xi may change~ This is tme becauseall the old bindings have a setup time X3 greater than (i~e. in a temporal sense later than) the current time. On the other hand, bindings for an in-progress conference -'30 call are generated using bindings, which are retrieved from the bindings memory. -This is because all acdve bindings of an in-progress conference must maintain the ~
same Xl and X2 and X3 elements during the conference, i.e. only the terminate_time ` ;
`X4 may change.
No~e that, for an in-progress conference, setup dme X3 is less ~an (i.e.
35 in a temporal sense earlier than) the current time whereas tenninate time X4 of both the acdve bindings and the new bindings is greater than the current time. Note also ,.~ ,.

;3 "'':~':~' :: ' ~s~
2~32~ 7 , ~i ; . - 23 -that some changes to an in-progress conference may result in a new set of bindings in which all elements Xi may change. This is true for those old bindings having a setup dme X3 ~eater than the current time. Also, this is because the new set of bindings will have a setup time X3, which is greater than the current time, meaning S that the changes to an exisdng conference will occur in the future.
If a conference is currently in-progress when a new demand is received for any bindings that have the old terminate time earlier than the current time or any Wndings that are no longer serving the conference, because they have been terminated from the conference, the binding methodology does not allow the new 10 demand to be accepted. This is because the new demand attempts to alter a binding which no longer serves the conference. Since the binding no longer serves the conference, the binding can not be used to generate a set of new bindings, which will ~3 occur in ihe future for this conference.
Funher, for an in-progress conference or a conference that has been 15 terminated~ once a set of new bindings is generated for the demand (some based on the old bindings), whether it is a new demand or a change to a demand, the previously stored set of bindings can be deleted from the bindings memory and the new set of bindings is substituted sherefor in the bindings memory.
We now describe a procedure9 which uses my methodology, to generate 20 the needed binding whether the needed binding would be for an access line 12û-i or for an internodal link 140-k. For purposes of illustradon only, our example contemplates the needed binding to be for an internodal link but the methodology is broader.
The conference signal, which is in need of a binding, may be of a type, 25 which can be classified as a "service type." A service type is a classification for use by the network resources when satisfying customer demands. The network resourcesmay be dedicated resources for that type of service, shared resources for that type of semce and one or more other types of services, or may be of a special quality tomeet the needs of ~he demands of that service type. ~
The methodology for finding a binding allows for a set of predetennined ~ -parameters to be used as a function of the type of link being used (access line or internodal link), as a funcdon of the type of like endpoints being demanded by the customer, and as a function of the network constraints. The parameters can bç
represented as elements of a six-tuple of the form (starting cha~inel, 35 stopping channel, direction, service type, buffer time, minimurn bandwidth) where the starting channel is the lowest channel within the bandwidth of the , ~ 2 ~

~,i communications path that may be used in a binding, stopping channel is the highest chMnel within the bandwidth of the communications path that may be used in a binding, direcdon is the manner the methodology should look within the starting channel and stopping channel to find available bandwidth, service type S indicates that only those comrnunications paths that support the specified service type should be considered, buffer time is the minimum dme before X3 and after X4 tQ use when searching for an available binding, and minimum bandwidth is the minimum bandwidth demanded by the customer.
For purposes of exarnple and not by way of limitation, if the proredure 10 were provided a set of predetennined parameters where the starting channel = C0, stopping channel = Cp, direcdon = upward, service type = "a codec like endpoint signal", buffer time = zero, and minimum bandwidth = 6; tlie procedure could perform the following actions to generate a binding. The procedure searches eachcommunicadons path on internodal link 140-k in the communicadons path memory 15 920 which allows a service to be placed on it of the type "a codec like endpoint signal." The set of communications paths ~etrieved from the communicadons path memory may be ordered by the binding methodology to increase the dme that a communications path is unused between the setup time and terminate time of bindings and can therefore increase the ILkelihood of sadsfying future changes to 20 previously sadsfied customer demands. By way of example and not by way of lirnitation, the ordering of communicadon paths might be such that the first communicadons path has the greatest dme interval between when the new reservation demand is requested and when the next and previous demands on the ;
communicadons path occur, the second communicadons path has the second greatest 25 time interval, etc. for all the communication paths retrieved from the communications path memory.
My methodology iterates through each communications path in the above ordered list un~l a binding is generated.
For purposes of this exarnple, the order generated by the binding -- 30 methodology of ite rating through the communicatdons paths is first path 320-1 and second path 32~2. Other communicadon paths for internodal link 140-k, such as a ~ -communicadon path that only supports audio bridge signals are not included in the iteradve process because they do not allow a binding for a signal of signal type = "a `~
- codec like endpoint signal" to be placed on the communications path. : --, . .. ~.
.. -,."".'.`.'".
. ~"~
. . .
. ' ~

6 ~ 7 ~3 For each communication path in the ordered list, the procedure searches for an available minimum bandwidth between the sta~ng channel and the stopping channel. The procedure searches this range of stardng channel and stopping channel in the manner indicated by the direction element of the six-tuple.
S If the direction is upwæd, the procedure searches the range by starting at thestarting channel and stopping at the stopping channel to find ~he minimum ;~ bandwidth requested. If the direction is downward, the procedure searches the range -by starting at the stopping channel and stopping at d~e starting channel to find the minimum bandwidth requested. In the example and based on the direction element 10 in the six-tuple, the procedure will search each communications path in the ordered list starting at channel C0 and stopping at channel Cp. For each channel within the range to be searched, the procedure defines a channel range (Xl, X2), where X2 -Xl corresponds to the minimum bandwidth element of dle sLx-tuple such that X2 is ~ -not ~reater than the stopping channel element in the six-tuple. In the example, the 15 first channel range to be defined is (Xl = (:0, X2 = CO + 6 - C6) on communications path 320-1. The last channel group to be defined on communications path 32~1 would be (Xl = Cp - 6 = 6, X2 = Cp = 12).
~ or each channel group defined, the procedure searches the binding memory 930 for collisions of the defined channel range with other customer 20 demands. A "collision" is defined to occur when an exisdng binding in bindingmemory defined by the four-tuple (Xl-l, X2-1, X3-1, X4-1) and a channel group (Xl, X2) and the desired setup time X3 and desired terminate dme X4 of that channel group have the following condidons (a) Xl is less d~an X2-1, (b) Xl-I isless than X2, (c) X3 - buffer dme is less than X4-1, and (d) X3-1 is less than X4 +
25 buffer time are all tsue. If a collision occurs, the procedure continues to iterate through the range of channels in the directdon indicated by the direction elment ln the ` six-tuple. If all the channels within the specified range for a communi~ations path have been searched, t~he procedure iterates through the remaining communicationspaths in the ordered list.
In the example of FIG. 3, demand 31~1 causes a collision for channel ranges searched (1,6), (2,7), (3,8), (4,9), (5,10), and (6,11). Demand 310-2 causes a collision for channel ranges searched (2,7), (3,8), (4,9~, (5,10), (6,11), and . (7,12). At this point in the processing of the demand, thc procedure has exhausted -all possible channel ranges on communications path 32~1 and begins to attempt togenerate a binding of communications path 320-2. In this example, demand 320-3 causes a collision for channel ranges searched (1,6), (2,7), (3,8), (4,9), and (5,10).

. ' ,~
:

2~32~ 7 , .

The first cha}mel range tO not cause a collision is channel range (6,11).
Whereupon, the procedure generates a binding with a four-tuple (Xla6, X2=11, , X3=t3, X4=t6).
My binding methodology iterates the above actions first with a S buffer time equal to the buffer time element specified in the six-tuple and if unable to generate a binding, performs the same iterations again, but with the buffer time set equal to zero, which attempts to check for collisions with the exact times desired ~ in the binding.
,~C!~ Note that a first signal, which is being transmitted on a communicadon 10 path for a specified dme, uses a set of contiguous channels. Other signals may use other channels on that communicadon path at the same time provided they do not overlap channels of the first signal.
The communicadon path must also be available during the dme ~ -re~luested. ~he binding methodology searches a unavailability memoTy 970 (which 15 can be suitably generated and stored in software), to determine if the cornmunicadon path is not available duling the desired dme beginning at dme-date X3 and terminating at dme-date X4.
Binding signals on the access line communicadon paths can be limited `~; by the binding methodology to a particular range of channels to be searched and a 20 direction of the search which is different from the limitadons on the range of channels to be searched and the direcdon of the search for internodal linkj or for that manner, access lines which connect a DACS 140-k to a set of offnet endpoints, tofind the necessary bindings. For example, the communicadon paths on an access ~ -line 120-i may be limited by ~e binding methodology to searching channels from 25 Cm to Cp-n in the downward direction, where Cm is some predetermined channel on that acccss line and n is some predetermined integer value equal to or greater than zero. Whereas, the cornmunication paths on an internodal link 140j may be limited by the binding medlodology to searching channel from C0 eo Cp in ehe upwa~d direction.
We now describe a procedure, which uses my methodology, to generate the ne~ded information for passing a signal from an endpoint at a customer site 110-i :
~3 through the multiplexer or concentrator 295 at customer site 110-i. This process is called generating a cross connect for the muleiplexer or concentrator 295 at customer - -~
- site 110-i. For purposes of illustration only, a signal is needed from endpoint 200 at customer site 110~i to pass ~hrough multiplexer or concentrator 295 a~ customer site - -110-i. The cross connect that my methodology establishes is a five-tuple (X7, X1, ~,.

:
~

~3~

~i X2, X3, X4) vvhere X7 is the endpoint providing the signal, Xl, X2, X3, X4 are the same defini~ons for a binding. The Xl, X2, X3 and X4 elements of the five-tuple are the same elements as the binding for this conference signal on the access line, which is called the conference signal access line connectivity. My binding 5 methodology generates the five-tuple after the access line binding has been generated and the resu}ting five-tuple is stored in the cross connect memory 940.
We now describe the procedure, which uses my methodology, to extend the signal of an endpoint to the homing DACS. The procedure finds a binding for the endpoint on the access line using the predetermined pa~ameters for that access 10 line. The procedure then generates the cross connect for the multiplexer or concentrator 295 at customer site 110-i.
A circuit may be reserved, similar to the manner a schedulable circuit may be reserved using network resources. The distinction between a reserved circuit and a reserved schedulable circuit is that a reserved schedulable circuit reserves the 15 circuit for a specific conference, whereas the reserved circuit does not. The reserved eircuit is typically used for reserving additional circuits in the net vork not associated with a reservadon, however, may satisfy some other customer demand not associated with any customer demand. An example of a Nserved circuit that utilizes the network resources on the instant network is called a stadc circuit signal. When static 20 circuit signals are allocated, these will be allocated starting at channel C0 and move up to Cm- 1 in the upward direction, where Cm is the same predetermined channel on an access line as described above. The value Cm can be predetermined for each access line, or comrnunications path on the access line, and can be suitable stored in the communications path memory 920.
Reservations are scheduled using a delta scheme. The methodology uses the bindings currendy in the binding memory for the customer demand (if they exist) and determines the necessary changes to those bindings to generate new bindings. If - a required binding did not change from the one in binding memory, where the two endpoints of the communicadons path is the same, the desired rninimum bandwidth 30 is the same, and the setup time and terminate time are the same, the binding methodology uses the existing binding in binding memory. If an old binding is nolonger required, it is not maintained in the new set of bindings for the customer demand. If the required binding does not exist, the methodology finds a set of bindings to sadsfy the customer demand and adds the new se$ of bindings into the set 35 of bindings required to satisfy the reservation. If the required binding has changed from the one in ~e binding memory, the methodology handles these re~uests . . . .

~2~17 :~.

differently. ~f the change occurs for an in-progress binding (i.e., when setup time p X3 is less than current time), the sarne channel groups are used with the new times X3 and X4 as specified in the new customer demand. If the change occurs for a future bincling (i.e., when setup time X3 is j~eater than current time), the oldS binding may be removed from the exisdng set of bindings to satisfy the reservation and a new binding can be determined. If, however, the old binding is not changed, the old binding would be added to the set of bindings which satisfy the reservation.
If the old binding's X4 is greater than the new demand's X4 and the setup times are the same, the binding methodology will use the old bindings except that the new 10 demand's X4 will replace the old binding's X4 and the binding methodology will not check for collisions.
When a reservation is provided, a reservation setup time, start dme, and stop time will be provided with the reservation specificadon. If a se~up drne is ~
provided, thç generated setup times and terminate times for end points (610, 710, 1 -15 810) will use this time instead of the generated reservadon setup time.
All times that are checked within the methodology for scheduling ~` -endpoints or their associated audio or control portions will be based on the setup_time and terminate time. The methodology attempts to find network bindingsand the DMB sequence number using two sets of dmes. The first set of dmes places20 a buffer of dme around the request to allow possible extensions of the reservation ~ ~
once it is up. The second set of times, uses the network actual times The first set of ` ~ ~;
dmes used are the setup time and terminate time with a buffer of res buffer timeminutes subtracted/added to each dme. The second set of dmes used are the actualsetup time and telminate dme.
2S Endpoints are scheduled to ensure that they are available. This is done `
by checking that the endpoint has not been allocated for the time period that the endpoint is in the reservation. No time buffering occurs when scheduling the -endpoints. Each endpoint in the reservation is also checked to see if it has been placed in una~railability memory 970. If the endpoint is a codec, the codec, the ~ -30 customer site 1 l~i whère the codec is located, and the multiplexer or concentrator 295 at customer site 1 l~i are checked if they have been placed in unavailability -memory 97Q ~
i Given a specific time period, a valid binding is determined by searching ;~ `
all the communication paths between the two desired locations. For each 35 communication path, methodology detennines what services is traveling on that ~ -communication path during the period necessary. The records found indicate . ``'.`

2 ~ 3 2 ~ 7 ;' I

channel ranges that are not usable for the desired time period. The available channel ranges are generated. The available channel ranges are then checked to satisfy the request. The channel ranges are checked based on ~he boundanes deffned (lower and upper channel ranges) and the direction. A request is satisfied if the desired S bandwidth falls within the boundary and the entire bandwidth is available.
Refernng to FIG. 4, there is shown an example to illustrate a customer demand, also called a reservation, which may consist of several demands. In FIG. 4, a customer reservadon may demand (i) a minimum bandwidth through the network of 12 contiguous channels of bandwidth; (ii) a conference start time-date of 1:00 pm 10 June 15, 1990; (iii) a conference terminate_~dme-date of 5:00 pm June 15, 1990; and (iv) a conferewe where the audio bridges and VCCs at the customer sites for the codecs, alternate access, or offnet endpoints exist. Further in the example, assume the following further customer demands:

(1) offnet endpoint 280 at customer site 110-1 is to pardcipate in the 15 conference with a setup dme-date of 2:00 pm June 15, 199Q and a tenninate time-date of 4:00 pm June 15, 1990;

(2) codec endpoint 200 at customer site 110-2 is to participate in the conference with a setup dme-date of 1:00 pm June 15, 1990 and a telminate time-date of 5:00 pm June 15, 1990;
;`

~3) codec endpoint 200 at customer site 110-3 is to participate in the conference with a setup time-date of 1:00 pm June 15, 1990 and a terminate time-date of 5:00 pm June 15, 1990;

(4) codec endpoint 20Q a~ customer site 11~4 is to participate in the conference with a start time-date of 1:00 pm June 15, 1990 and a telminate time-25 date of 5:00 pm June 15, 1990;
,, , ~
(S) codec endpoint 200 at customer site 110-5 is to participate in the conference with a start time-date of 1:00 pm June 15, 1990 and a telminate time-date of 5:00 pm June 15, 1990;

~6) codec endpoint 200 at customer site 110-6 is to par~cipate in the 30 conference with a start time-date of 1:00 pm June 15, 1990 and a terminate time-2~32~1~

date of S:OO pm June 15, 1990;

(7) codec endpoint 200 at customer site 11~7 is to participate in the conference with a start time-date of 1:00 pm June lS, 1990 and a terrninate time~
.. date of S:OO pm June lS, 1990; and ~: ..

(8) codec endpoint 200 at cus~omer site 110-8 is ~o participate in the conference with a start dme-date of 2:00 pm June 15, 1990 and a tenninate time- ~
date of 4:00 pm June 15, 1990. - . .
s Hence, FIG. 4 recites to a set of customer demands. This set of ~ -.
customer demands can be stratified according to a first strategy of a set of like .~
;. 10 endpoints such as: ~ `

(1) offilet endpoint 280 at customer site 110-1 is to participate in the conference widl a setup dme-date of 2:00 pm June lS, 1990 and a tenninate dme-date of 4:00 pm June 15, 1990; -(2) codec endpoint 200 al customer site 110-2 is to par~icipate in the 15 conference with a setup dme-date of 1:00 pm June lS, 1990 and a terminate time-date of S:OO pm June 15, 1990; . . ` .

(3) codec endpoint 200 at customer site 11~3 is to pardcipate in the conference with a setup time-date of 1:00 pm June lS, 1990 and a terminate dme~
date of S:OO pm June lS, l990; :~
.,~
(4) codec endpoint 200 at customer site 110-4 is to participate in the conference with ~ start time-date of 1:00 pm June lS, 1990 and a terminate time-date of ~:00 pm June lS, 1990;

~5) codec endpoint 200 at customer site l lO-S is to par icipate in the conference with a start time-date of 1:00 pm June 15, 1990 and a terminate_time-25 date of S:OO pm June lS, 1990;

(6) codec endpoint 200 at customer site 11~6 is to participate in the conference with a start time-date of 1:00 pm June lS, 1990 and a tennina~e time-date of S:OO pm June lS, 1990; :
..,.,.
.
.
~:

i~; ~ 2~32B17 ,~, (7) codec endpoint 200 at customer site 110-7 is to participate in the conference with a start time-date of 1:00 pm June lS, 1990 and a terminate time-date of S:00 pm June lS, 1990; and (8) codec endpoint 200 at customer site 110-8 is to participate in the S conference with a stan time-date of 2:00 pm June lS, 1990 and a terminate time-date of 4:00 pm June lS, 1990.
This set of customer demands can be stratified according to a second strategy of a set of like endpoints such as (1) audio bridge 220 at customer site 11~1 is to participate in the 10 conference with a setup time-date of 2:00 pm June lS, 1990 and a terminate Sime-date of 4:00 pm June lS, 1990;

~ 2) audio bridge 220 at customer site 11~2 is to participate in the conference with a setup time-date of 1:00 pm June 15, 1990 and a telmina~e time-date of 5:00 pm June lS, 1990;

lS (3) audio bridge 220 at customer site 11~3 is to participate in the conference with a setup time-date of 1:00 pm June lS, 1990 and a terminate time-~: date of S:00 pm June lS, 1990;
i ~
. (4) audio bridge 220 at customer site 1104 is to participate in the . conference with a start time-date of 1:00 pm June 15, 1990 and a te~ninate time-~: 20 date of S:00 pm June lS, 1990;.``
(S) audio bridge 220 at customer site 110-S is to par~cipate in the conference with a start time-date of 1:00 pm June 15, 1990 and a terminate time-date of S:00 pm June lS, 1990;

(6) audio bridge 220 at customer site 110-6 is to participate in the : ;
conference with a start time-date of 1:00 pm June lS, 1990 and a terminate time-date of S:00 pm June 15~ 1990;
":
(7) audio bridge 220 at customer site 110-7 is to participate in the conference with a start time-date of 1:00 pm June lS, 1990 and a tern~inate time-. .

,Y ~

i :
s 2~32~7 ~
- 32 - :
date of 5:00 pm June 15, 1990; and ~: ~

~8) audio bridge 220 at customer site 110-8 is to participate in the ~: .
conference with a start time-date of 2:00 pm June 15, 1990 and a terminate time-date of 4:00 pm June 15, 1990. This set of customer demands can be stra~ified S according to a third strategy of a set of like endpoints such as :. ~, '' ','~
(1) VCC 240 a~ customer site 110-1 is to participate in the conference ~`
with a setup time-date of 2:00 pm June 15, 1990 and a terminate time-date of 4:00 pm June 15, 1990;
; . ~
(2) VCC 240 at customer site 1 10-2 is to participate in the conference .~:.
10 with a setup time-date of 1:00 pm June 15, 1990 and a terminate time-date of 5:00 pm June 15, 1990; . .

¦ (3~ VCC 240 at customer site 110-3 is to participate in the conference -with a setup_time-date of 1:00 pm June 15, 1990 and a terminate_time-date of 5:00 pm June 15, 1990;
.

(4~ VCC 240 at customer site 110-4 is to participate in the conference ~
with a start_time~ate of 1:00 pm June 15, 1990 and a terminate time-date of 5:00 `~:
pmJune15,1990; :~ ~

(S) VCC 240 at customer site 110-5 is to pardcipate in the conference ~ `
with a start_time-date of 1:00 pm June 15, 1990 and a terminate time-date of 5:00 pm June 15, 1990;
. ~ . :
(6) VCC 240 at customer site 110-6 is to participate in the conference with a start time-date of 1:00 pm June 15, 1990 and a terrninate_time-date of 5:00 : .
' pm June 15, 1990; ~
..
(7) VCC 240 at customer site: l 10-7 is to participate in the conference - 25 with a start_time~ate of 1:00 pm June 15, 1990 and a tenninate_time-date of 5:00 pm June 15, 1990; and (8) VCC 240 at customer si~e 110-8 1s to parhcipate in rhe conference ~ . ` '~
"'`
~,...

~ ` 2~3~7 with a start_time-date of 2:00 pm June 15, 1990 and a terminate time-date of 4:00 pm June lS, 1990.
While the example of customer demand shown in FIG. 4 does not show two codecs or like endpoints participadng in the same conference at one customerS site, my methodology allows such a customer demand. Further and by way of example, if codec 200 at customer site 110-2 is to participate in a conference with a setup time-date of l:OOpm June lS, 1990 and terminate time-date of S:OOpm June lS, 1990 and alternate access 260 at customer site 110-2 is to participate in the same conference with a setup time-date of 5:00 pm June lS, 1990 and terminate time-10 date of lO:OOpm June lS, 1990 and audio bridge 220 at customer site 110-2 is a common unit of equipment in the predetermined set, the audio bridge 220 at customer site 110-2 will participate in the conference for the setup time-date of 1:00 pm June lS, l9gO and teIminate time-date of lO:OOpm June lS,l99Q The setup time date X3 and terminate time-date X4 are generated for audio bridge 220lS at customer site 11~2 by using the earliest setup time-date and the latest ~ terminate time-date of the set of codec like endpoints at customer site 110-2 .` involved in the conference such that at least one of the set of codec like endpoints is r involved in the conference at any time from the earliest setup time-date to the latest termiinate time-date. In the example above, at least one of the set of codec, alternate 20 access, and offnet at customer site 110-2 is involved in the conference from l:OOpm . June lS, 1990 until lO:OOpm June 10,1990. It may be the case that an endpoint can be involved in a conference muldple times and that the endpoin~s of the predetermined set for that endpoint may also be involved in a conference multiple times.
For puIposes of the example, all VCC connecdvity will be done using DDD as the manner of extending the VCC signal to the network.
. Although my invention has been described and illustrated in detail using certain examples, it is to be understood that the same is not by way of limitation.
Hence, the spiri~ and scope of my invendon is limited only by the terrns of the 30 appendedclai~

. .

Claims (28)

1. A method for reserving and allocating a plurality of competing demands for an ordered bus communications network in response to a plurality of demands from a source, the network including a plurality of customer sites, eachcustomer site including one or more endpoints, each endpoint being connectable to a switching system, the switching system being connectable to one or more other switching systems, means for connecting a first endpoint through said one or more switching systems to a second endpoint, wherein the improvement comprises the steps of:
(a) receiving demands from a customer site for allocating the network for a communication among a plurality of customer sites;
(b) stratifying the received demands in response to a grouping of endpoints to be conferenced at the customer sites;
(c) allocating the network resources to connect the endpoints to be conferenced in response to the stratified demand including generating setup and terminate times for each endpoint to be connected in a conference;
reserving the endpoints to be connected in the conference; and generating a plurality of bindings for signaling the network and the source as to success of connecting the endpoints to be conferenced over said ordered bus network.

2. The method defined in claim 1 further comprising the steps of:
(d) stratifying a customer's demand into sets of strategies for groupings of like endpoints;
(e) allocating network resources responsive to characteristics of the different groupings of like endpoints;
(f) for each strategy, generating the setup times and terminate times for each endpoint;
(g) for each strategy, reserving the endpoints for a reservation; and (h) for each strategy, generating bindings.

3. The method defined in claim 2 further comprising the steps of:
(i) for a first grouping of like endpoints selected from the group consisting of codecs, alternate accesses, and offnets, extending a conference signal through a multiplexer or concentrator at a customer site over an access line to a homing DACS to which said endpoints are connected.

4. The method defined in claim 2 further comprising the step of:
(i) for a second grouping of like endpoints including one or more audio bridges, extending a conference signal from a customer site to a central bridging DACS on which said audio bridge resides.

5. The method defined in claim 3 or 4 further comprising the step of:
(j) generating instructions for controlling said homing DACS to provide connectivity of like endpoint conference signals.

6. The method defined in claim 3 further comprising the step of:
(j) generating bindings for connections between said one or more switching systems required for the first grouping of codec like endpoints.

7. The method defined in claim 2 further comprising the step of:
(i) for a third grouping of like endpoints including one or more video conference controllers, extending a conference signal from a customer site to the network.

8. The method defined in claim 2 wherein said bindings generation step (h) further comprises the steps of:
responsive to a service type and to a bandwidth and to a start_time and to a stop_time from a source of demand, searching for one or more communication paths, which satisfy the demand;
responsive to the demand satisfying communication paths, searching the communication paths in a predetermined direction and range of bandwidth to identify the bandwidth in the paths, which satisfy the source demand.

9. The method defined in claim 8 wherein said bindings generation step (h) further comprises the steps of:
generating a buffer time around the start_time and the stop_time;
responsive to the buffer time, searching for a bandwidth which does not collide with other source demands.

10. The method defined in claim 2 wherein said bindings generation step (h) further comprises the steps of:
responsive to a delta scheme and reponsive to the current_time of the demand, determining whether a new source demand can be satisified using existingbindings, or whether new bindings need to be generated;

responsive to the determining step, providing a set of bindings to satisfy the source demand.

11. The method defined in claim 1 further comprising the step of:
(i) determining if the network is not available for reconfiguration of a conference during any reconfiguration of the network.

12. The method defined in claim 1 wherein said bindings generation step further comprises the steps of:
responsive to a service_type and to a bandwidth and to a start_time and to a stop_time from a source of demand, searching for one or more communication paths, which satisfy the demand;
responsive to the demand satisfying communication paths, searching the communication paths in a predetermined direction and range of bandwidth to identify the bandwidth in the paths, which satisfy the source demand.

13. The method defined in claim 12 wherein said bindings generation step further comprises the steps of:
generating a buffer time around the start_time and the stop_time;
responsive to the buffer time, searching for a bandwidth which does not collide with other source demands.

14. The method defined in claim 1 wherein said binding generation step further comprises the steps of:
responsive to a change to a previously processed demand and responsive to the current_time of the demand, determining whether a new source demand can be satisfied using existing bindings, or whether new bindings need to be generated;responsive to the determining step, providing a set of bindings to satisfy the source demand.

15. A system for reserving and allocating a plurality of competing demands for an ordered bus communications network in response to a plurality of demands from a source, the network including a plurality of customer sites, eachcustomer site including one or more endpoints, each endpoint being connectable to to a switching system, the switching system being connectable to another switching system, means for connecting a first endpoint through one or more switching systems to a second endpoint, wherein the improvement comprises:

means for receiving demands from a customer site for allocating the network for a communication among a plurality of customer sites;
means for stratifying the received demands in response to a grouping of endpoints to be conferenced at the customer sites;
means for allocating the network resources to connect the endpoints to be conferenced in response to the stratified demand including (i) means for generating setup and terminate times for each endpoint to be connected in a conference;
(ii) means for reserving the endpoints to be connected in the conference; and (iii) means for generating a plurality of bindings for signalling the network and the source as to success of connecting the endpoints to be conferenced over said ordered bus network.

16. The system defined in claim 15 further comprising:
means for stratifying a customer's demand into sets of strategies for groupings of like endpoints;
means for allocating network resources responsive to characteristics of the different groupings of like endpoints;
for each strategy, means for generating the setup_times and terminate_times for each endpoint;
for each strategy, means for reserving the endpoints for a reservation;
and for each strategy, means for generating bindings.

17. The system defined in claim 16 further comprising:
for a first grouping of like endpoints selected from the group consisting of codecs, alternate accesses, and offnets, means for extending a conference signal through a multiplexer or concentrator at a customer site over an access line to a homing DACS to which said endpoints are connected.

18. The system defined in claim 16 further comprising:
for a second grouping of like endpoints including one or more audio bridges, means for extending a conference signal from a customer site to a central bridging DACS on which said audi bridge resides.

19. The system defined in claim 17 or 18 further comprising:

means for generating instructions for controlling said homing DACS to provide connectivity of like endpoint conference signals.

20. The system defined in claim 17 further comprising:
means for generating bindings for connections between said one or more switching systems required for the first grouping of codec like endpoints.

21. The system defined in claim 16 further comprising:
for a third grouping of like endpoints including one or more video conference controllers means for extending a conference signal from a customer site to the network.

22. The system defined in claim 16 wherein said last mentioned bindings generation means further comprises:
responsive to a service_type and to a bandwidth and to a start_time and to a stop_time from a source of demand, means for searching for one or more communication paths, which satisfy the demand;
responsive to the demand satisfying communication paths, means for searching the communication paths in a predetermined direction and range of bandwidth to identify the bandwidth in the paths, which satisfy the source demand.

23. The system defined in claim 22 wherein said last mentioned bindings generation means further comprises:
means for generating a buffer time around the start_time and the stop_time;
responsive to the buffer time, means for searching for a bandwidth which does not collide with other source demands.

24. The system defined in claim 16 wherein said last mentioned bindings generation means further comprises:
responsive to a change to a previously processed demand and reponsive to the current time of the demand, means for determining whether a new source demand can be satisified using existing bindings, or whether new bindings need to be generated;
responsive to the determining step, means for providing a set of bindings to satisfy the source demand.

25. The system defined in claim 15 further comprising:

means for determining if the network is not available for reconfiguration of a conference during any reconfiguration of the network.

26. The system defined in claim 15 wherein said bindings generation means further:
responsive to a service_type and to a bandwidth and to a start_time and to a stop_time from a source of demand, means for searching for one or more communication paths, which satisfy the demand;
responsive to the demand satisfying communication paths, means for searching the communication paths in a predetermined direction and range of bandwidth to identify the bandwidth in the paths, which satisfy the source demand.

27. The system defined in claim 26 wherein said bindings generation means further comprises:
means for generating a buffer time around the start_time and the stop_time;
responsive to the buffer time, means for searching for a bandwidth which does not collide with other source demands.

28. The system defined in claim 15 wherein said binding generation means further comprises:
responsive to a delta scheme and reponsive to the current_time of the demand, means for determining whether a new source demand can be satisified using existing bindings, or whether new bindings need to be generated;
responsive to the determining step, means for providing a set of bindings to satisfy the source demand.