Multiple multicasting of multimedia streams

[0002] The present application is a continuation-in-part of pending U.S. patent application Ser. No. 08/714,447, entitled “Multimedia Compression with Additive Temporal Layers” currently pending by Navin Chaddha, filed Sep. 16, 1996, assigned to VXtreme Inc., herein incorporated by reference in its entirety.

[0003] 1. Field of the Invention

[0004] The present invention relates to multimedia communications. More particularly, the present invention relates to the efficient delivery of multimedia data to multicast group(s) over a diverse computer network.

[0005] 2. Description of the Related Art

[0006] With the proliferation of connections to the internet by a rapidly growing number of users, the viability of the internet as a widely accepted medium of communication has increased correspondingly. Bandwidth requirements can vary significantly depending on the content of multimedia data being delivered and computational capacity of the client computers receiving the multimedia data. Hence, the ability to efficiently deliver multimedia data to a number of client computers over the internet is limited by how the available bandwidth capacity of the network is utilized to provide video information to a diverse group of client computer.

[0007] In a typical video delivery scheme, for each video stream, a point-to-point connection is provided by the network between the server and each client computer. From the network's perspective, this scheme is inefficient especially when similar content is delivered to a number of client computers. A more efficient method is to multicast “blindly” over the network without any feedback from the client computers, in a manner similar to a wireless television broadcast. One such conventional video encoding and decoding system is described in “An End-to-End Software only Scalable Video Delivery System,” published in Proc. Networks and Operating System Support for Digital Audio and Video, April 1995. Instead of establishing individual point-to-point connections for each client computer, the server multicasts an entire embedded stream for different resolutions and frame rates onto the network as a set of trees. However, “the server has no idea about the decoders at the destinations” (page 136, lines 4-5) (emphasis added). Primary traffic management is performed by not adding branches of the trees carrying the less important bit streams to the lower bandwidth portions of the network. In addition, switches and routers of the network may react to temporary network congestion by dropping packets carrying the less important bits from the embedded stream.

[0008] Unfortunately, with the push multicast model described above, since “the destinations [decoders] are slaved to the flow from the server with no feedback” (page 137, lines 46-47) (emphasis added), the server is incapable of adapting to the actual needs of individual and/or sub-groups of client computers. Packets carrying less important bits are sent to client computers so long as the corresponding portion of the network is capable of carrying the additional information. In other words, the server ignores the actual needs of the client computers. For example, a user at any particular client computer may not be interested in receiving a high resolution and/or a high frame rate video stream, even if the network is capable of supporting the higher bit stream. Alternatively, a particular client computer or its modem may be incapable of processing the higher resolution and/or faster frame rate video stream. As a result, a considerable amount of unused or underutilized information is wastefully multicasted over the network and unnecessarily consumes valuable network resources.

[0009] In view of the foregoing, there are desired improved techniques for adaptively providing scalable multimedia data to a broad range of client computers while efficiently utilizing the valuable network resources.

[0010] A method of interactively providing a number of client computers with a dynamically selectable and scalable range of multimedia data over a diverse computer network including local area networks (LANs) wide area networks (WANs) such as the internet.

[0011] Multimedia data is provided by a server to the client computers includes a base layer and one or more enhancement layers. Enhancement layers can be spatial and/or temporal in nature. Depending on the implementation, the server may also provide information about the multimedia data to the client computers.

[0012] In accordance with one aspect of the invention, the server splits the multimedia data for streaming via multiple multicast group (MMG) addresses. Information about the portion of the multimedia data carried by each MMG is broadcasted to the client computers. Armed with the information about the multimedia data, client computers can intelligently join and leave MMGs as needed.

[0013] In some embodiments, the client computers provide feedback about the usage and/or need for the multimedia data, enabling the server to right-size, e.g., grow and/or prune, the multimedia data for network efficiency. With right sizing, the content of the base layer may be increased or decreased with the corresponding growing and pruning of the enhancement layers. Enhancement layers may also be grown and/or pruned independently of the base layer, i.e., without a corresponding change in the base layer.

[0014] These and other advantages of the present invention will become apparent upon reading the following detailed descriptions and studying the various figures of the drawings.

[0015] FIG. 1 is a block diagram of an exemplary computer system for practicing the various aspects of the present invention.

[0016] FIG. 2 is a block diagram showing an exemplary hardware environment for practicing the invention which includes a web server and client computers, coupled to each other by a computer network.

[0017] FIG. 3 is a block diagram illustrating one embodiment of an encoder in which multimedia data representing an enhanced video image is encoded separately into a base layer and an enhancement layer.

[0018] FIG. 4 illustrates an associated decoder for decoding the base layer and the enhancement layer of FIG. 3 to regenerate a base image and an enhanced image.

[0019] FIG. 5A is a block diagram of an encoder capable of generating additional enhancement layer(s).

[0020] FIG. 5B is a block diagram of yet another encoder capable of generating additional base layer(s).

[0021] FIG. 6A is a block diagram of a decoder capable of decoding additional enhancement layer(s).

[0022] FIGS. 6B and 6C illustrate decoding circuits for regenerating a similar enhanced image from different base layers and enhancement layers.

[0023] FIG. 7A illustrates how multimedia data representing a full temporal sequence can be separated into multiple temporal layers.

[0024] FIG. 7B shows an exemplary spatial arrangement of base layer packets and enhancement layer packets.

[0025] FIG. 8 is a detailed block diagram showing one embodiment of a server and a representative client computer.

[0026] FIGS. 9 and 10 are flowcharts illustrating the adaptive right-sizing of a multimedia stream being transmitted, via a single multicast group, from the server to one or more client computers.

[0027] FIGS. 11 and 12 are flowcharts illustrating the adaptive right-sizing of a multimedia stream being transmitted, via multiple multicast groups (MMGs), from the server to one or more client computers.

[0028] The present invention will now be described in detail with reference to a few preferred embodiments thereof as illustrated in the accompanying drawings. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. It will be apparent, however, to one skilled in the art, that the present invention may be practiced without some or all of these specific details. In other instances, well known process steps have not been described in detail in order to not unnecessarily obscure the present invention.

[0029] FIG. 1 is a block diagram of an exemplary computer system 100 for practicing the various aspects of the present invention. Computer system 100 includes a display screen (or monitor) 104, a printer 106, a floppy disk drive 108, a hard disk drive 110, a network interface 112, and a keyboard 114. Computer system 100 includes a microprocessor 116, a memory bus 118, random access memory (RAM) 120, read only memory (ROM) 122, a peripheral bus 124, and a keyboard controller 126. Computer system 100 can be a personal computer (such as an Apple computer, e.g., an Apple Macintosh, an IBM personal computer, or one of the compatibles thereof), a workstation computer (such as a Sun Microsystems or Hewlett-Packard workstation), or some other type of computer system known to one skilled in the computer art.

[0030] Microprocessor 116 is a general purpose digital processor which controls the operation of computer system 100. Microprocessor 116 can be a single-chip processor or can be implemented with multiple components. Using instructions retrieved from memory, microprocessor 116 controls the reception and manipulation of input data and the output and display of data on output devices.

[0031] Memory bus 118 is used by microprocessor 116 to access RAM 120 and ROM 122. RAM 120 is used by microprocessor 116 as a general storage area and as scratch-pad memory, and can also be used to store input data and processed data. ROM 122 can be used to store instructions or program code followed by microprocessor 116 as well as other data.

[0032] Peripheral bus 124 is used to access the input, output, and storage devices used by computer system 100. In the described embodiment(s), these devices include display screen 104, printer device 106, floppy disk drive 108, hard disk drive 110, and network interface 112. Keyboard controller 126 is used to receive input from keyboard 114 and send decoded symbols for each pressed key to microprocessor 116 over bus 128.

[0033] Display screen 104 is an output device that displays images of data provided by microprocessor 116 via peripheral bus 124 or provided by other components in computer system 100. Printer device 106, when operating as a printer, provides an image on a sheet of paper or a similar surface. Other output devices such as a plotter, typesetter, etc. can be used in place of, or in addition to, printer device 106.

[0034] Floppy disk drive 108 and hard disk drive 110 can be used to store various types of data. Floppy disk drive 108 facilitates transporting such data to other computer systems, and hard disk drive 110 permits fast access to large amounts of stored data.

[0035] Microprocessor 116 together with an operating system operate to execute computer code and produce and use data. The computer code and data may reside on RAM 120, ROM 122, or hard disk drive 120. The computer code and data could also reside on a removable program medium and loaded or installed onto computer system 100 when needed. Removable program mediums include, for example, CD-ROM, PC-CARD, floppy disk and magnetic tape.

[0036] Network interface circuit 112 is used to send and receive data over a network connected to other computer systems. An interface card or similar device and appropriate software implemented by microprocessor 116 can be used to connect computer system 100 to an existing network and transfer data according to standard protocols.

[0037] Keyboard 114 is used by a user to input commands and other instructions to computer system 100. Other types of user input devices can also be used in conjunction with the present invention. For example, pointing devices such as a computer mouse, a track ball, a stylus, or a tablet can be used to manipulate a pointer on a screen of a general-purpose computer.

[0038] The present invention can also be embodied as computer readable code on a computer readable medium. The computer readable medium is any data storage device that can store data which can be thereafter be read by a computer system. Examples of the computer readable medium include read-only memory, random-access memory, magnetic data storage devices such as diskettes, and optical data storage devices such as CD-ROMs. The computer readable medium can also be distributed over a network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

[0039] FIG. 2 is a block diagram showing an exemplary hardware environment for practicing the invention which includes a web server 210 and client computers 231, 232, . . . 239, coupled to each other by computer network 220. Each of server 210 and client computers 231, 232, . . . 239 can be implemented using a computer system such as computer system 100 described above. In this implementation, computer network 220 supports both point-to-point connections and multiple multicast groups.

[0040] In accordance with the present invention, server 210 transmits multimedia data, e.g., video, audio and/or annotation frames, to two or more of client computers 231, 232, . . . 239. Efficiency is accomplished by transmitting the multimedia data in the form of a base layer and one or more enhancement layers to one or more multicast groups, and wherein the multimedia data is shared by two or more of client computers 231, 232, . . . 239. Enhancement layer(s) can be either spatial and/or temporal in nature.

[0041] FIG. 3 is a block diagram illustrating one embodiment of encoder 300 in which multimedia data representing an enhanced video image 312 is encoded separately into a base layer BL(1), representing a base image 316, and an enhancement layer EL(2-1), representing the spatial difference between base image 316 and enhanced image 312, using a suitable encoding technique such as a Laplacian pyramid decomposition algorithm, known to one skilled in the art.

[0042] In this exemplary encoding process, enhanced image 312 is decimated 314, i.e., filtered and sub-sampled, into base image 316 which is compressed 318 to produce base layer BL(1). Next, base layer BL(1) is decompressed 322, upsampled, e.g., by interpolation 326, and then subtracted 342 from enhanced image 316 to produce an error image. The error image is then compressed 344 to produce an enhancement layer EL(2-1) which represents the error data between base image 316 and enhanced image 312. This process can be repeated to produce multiple enhancement layers as described below.

[0043] Conversely, in an exemplary associated decoder 400, as illustrated by FIG. 4, base layer BL(1) and enhancement layer EL(2-1) are used to regenerate a base image I1′ and an enhanced image I2′. Regenerated base image I1′ is produced by decompressing 412 base layer BL(1). Regenerated enhanced image I2′ is produced by combining 426 an interpolated 416 base image I1′ with a decompressed 422 enhancement layer EL(2-1).

[0044] Referring now to FIG. 5A, in encoder 510, additional enhancement layers EL(3-1), EL(4-1) are generated from enhanced images 518, 512, by combining decimation stages 514, 516, compression stages 524, 528, interpolation stages 534, 538, and summation stages 522, 526 to encoder 300, in a manner similar to the generation of enhancement layer EL(2-1) described above.

[0045] Conversely, as shown in FIG. 6A, in decoder 600a, additional enhanced images I3′, I4′ are regenerated from enhancement layers EL(3-1), EL(4-1), by combining decompression stages 622, 632, interpolation stages 612, 616, and summation stages 626, 636 to decoder 400, in a manner similar to the regeneration of enhanced image I2′ described above.

[0046] FIG. 7A illustrates how multimedia data representing a full temporal sequence with frames k, k+1, k+2, . . . k+8, . . . can be separated into a first temporal layer with frames k, k+4, k+8, . . . a second temporal layer with frames k+2, k+6, . . . and a third temporal layer with frames k+1, k+3, k+5, k+7 . . . Accordingly, the first temporal layer is the base layer while the second and third temporal layers are enhancement layers. In this example, in order to optimize network efficiency, the base layer is independent, e.g., includes I frames with complete frame data, while the enhancement layers are additive, e.g., includes P frames with differential data based on other frames. Alternatively, the temporal enhancement layers can also be independent. For example, enhancement frames k+1, k+2, k+3, k+5, k+6, k+7 . . . can also be I frames, but with a corresponding increase in transmission overhead when streamed over network 220.

[0047] FIG. 7B shows an exemplary spatial arrangement of base layer packets BP4, BP5, BP6, BP7 and BP8, and enhancement layer packets EP10, EP12, EP14, EP16 and EP18. In this example, server 210 transmits index-planes for different resolutions in different packets. Hence, base packet BP4 corresponds to the base layer with 4-bits of an index, and base packet BP8 includes the 8^thbit of the index for the base layer. Similarly, enhancement packet EP14 includes the 4^thbit of the index for the first enhancement layer.

[0048] Bandwidth scalability with an embedded bit stream can be accomplished using vector quantization (VQ). In one embodiment, a tree-structured VQ (TSVQ) successive approximation technique is implemented. Accordingly, codewords are arranged in a tree structure, and each input vector is successively mapped (from the root node) to the minimum distortion child node, thereby inducing a hierarchical partition, or a refinement of the input space as the depth of the tree increases. Because of the successive refinement, an input vector mapping to a leaf node can be represented with high precision by the path map from the root to the leaf, or with lower precision by any prefix of the path. Accordingly, TSVQ produces an embedded encoding of the data. Pending U.S. patent application, Ser. No. 08/819,579, entitled “Method and Apparatus for Table-based Compression with Embedded Coding” by Navin Chaddha, filed on Mar. 14, 1997, and assigned to VXtreme Inc., herein incorporated by reference in its entirety, describes several exemplary TSVQ implementations that can be practiced with the present invention.

[0049] FIG. 8 is a detailed block diagram showing one embodiment of server 210 and client computer 231, representative of client computers 231, 232, . . . 239. Accordingly, the following description of the operation of client computer 231 also applies to client computers 232 . . . 239. Server 210 and client computer 231 are coupled to each other via computer network 220 which supports multicast addresses and multiple multicasts.

[0050] Server 210 includes an encoder 812, a source packetizer 814 and a source networking unit 816. Encoder 812 includes a conditional replenishment unit, a Laplacian pyramid encoder with a tree structured hierarchical table lookup vector quantizer (TSHVQ) based on a perceptual distortion measure. Client computer 231 includes a decoder 836, destination packetizer 834 and a destination networking unit 832. Decoder 836 includes a table-lookup vector quantizer with support for color-convertion and dithering. Video decoding is performed on those blocks which change with respect to the previous frame. Encoder 812 can be any one of encoders 300, 510, and decoder 836 can be any one of decoders 400, 600a.

[0051] Server 210 executes the first stage of the encoding process, embedded conditional replenishment, in encoder 812. The blocks which change are encoded using the Laplacian pyramid algorithm with the TSHVQ. Encoder 812 produces the indices for each layer as an embedded stream with different index planes. The first index plane contains the index for the rate 1/k TSHVQ codebook. The second index plane contains the additional index which along with the first index plane gives the index for the rate 2/k TSVQ codebook. Similarly, the remaining index planes have part of the indices for 3/k, 4/k, . . . R/k TSVQ codebooks, respectively. As a result, encoder 812 advantageously produces indices with the embedded prioritized bit-stream. Subsequently, rate or bandwidth scalability can be achieved at client computer 231 by dropping index planes from the embedded bit-stream.

[0052] After encoder 812 has completed generating the embedded bit-stream, source packetizer 814 packages the embedded bit-stream into a number of embedded video stream packets based on RTP protocol and appends the respective packet headers. Depending on the needs of client computers 231, 232, . . . 239, source networking unit 816 can now split the embedded video stream packets into one or more multicast groups in accordance with the present invention.

[0053] In one exemplary multiple multicast (MMG) scenario, as illustrated by FIGS. 7A and 7B, a first multicast group may receive spatial layers BP4 to EP16 and temporal layers T1 to T3, while a second multicast group receives spatial layers BP4 to BP8 and temporal layers T1 to T2.

[0054] In another MMG scenario with joint source-channel coding of two layers, e.g., base layer BL1 and enhancement layer EL(2-1), are split into several embedded layers and sent by source network unit 816 on MMGs after packetization by packetizer 814. Subsets of the embedded bit-stream are sent on different multicast addresses. In this example, spatial layers BP4, (BP5-BP8), (EP12-EP14) and (EP16-EP18) can be sent in temporal layers T1, T2, T3 and T4, resulting in eight different layers which can be sent on different multicast addresses.

[0055] In accordance with one aspect of the invention, client computer 231 intelligently decides which multicast (address) groups to dynamically join or leave. This is possible because server 210 periodically provides updated information to client computer 231 about the different multicast groups, their associated data transfer rates, which portion of the spatial-temporal embedded stream belongs to which MMG, and information about base layer(s) of the embedded stream. A network bandwidth estimation algorithm keeps track of the available network bandwidth of network 220 supporting the respective MMGs. One such network bandwidth estimation algorithm is described in pending U.S. patent application, Ser. No. 08/818,127, entitled “Dynamic Bandwidth Selection for Efficient Transmission of Multimedia Streams in a Computer System” now U.S. Pat. No. 6,292,834 by Hemanth S. Ravi, assigned to VXtreme, Inc., and is herein incorporated by reference in its entirety.

[0056] Armed with the information provided by server 210, client computer 231 is able to reliably predict the both the cost and benefit of joining additional MMGs in addition to the MMG(s) client computer 231 has already joined. Accordingly, if there is network bandwidth available, then client computer 231 joins additional MMG(s) till the bandwidth associated with the MMGs is used up. Conversely, when client computer 231 detects that it is consuming more network bandwidth than is available, client computer 231 leaves some MMGs till the consumption of the network bandwidth is less than or equal to the available network bandwidth. In other words, client computer 231 attempts to efficiently maintain a healthy equilibrium between its need and the availability of the network bandwidth. As a result, the dynamic bandwidth adaptation of the present invention, accomplished by intelligently joining and/or leaving MMGs, advantageously reduces network congestion and packet losses, while optimizing the transfer rates.

[0057] Referring again to FIG. 8, when the data packets arrive at client computer 231, destination networking unit 832 splits the embedded bit stream packets into MMGs. Packets associated with the selected, i.e., joined, are depacketized by destination depacketizer 834 and provided to decoder 836.

[0058] Decoder 836 then use the remaining embedded stream to index a TSVQ codebook of the corresponding rate, e.g., by looking up the reproduction vector in the corresponding rate TSVQ decoder 836 codebook. In this example, since the inverse block transform is performed on the codewords of encoder 812 codebook, there is no need for performing inverse block transforms on decoder 836 codebook.

[0059] As discussed above, computational scalability is provided at decoder 836 by the use of the Laplacian coding scheme and the use of TSVQ. TSVQs achieves computation scalability proportional to the bandwidth of network 220 as the computation performed in lookups is different for different TSVQ codebooks and scales proportionately with the depth of the tree. In sum, the streaming of scalable multimedia data which includes a base layer and at least one enhancement layer by server 210 enables client computer 231 to select different MMGs in search of the best match with the bandwidth of network 220.

[0060] However, because encoder 510 of FIG. 5A provides a single predetermined base layer BL1 and an associated range of enhancement layers EL(2-1), EL(3-1), EL(4-1), the ability of encoder 510 to adapt to the actual needs of client computer 231, 232, . . . 239 is limited to growing and pruning enhancement layers EL(2-1), EL(3-1), EL(4-1).

[0061] In accordance to another aspect of the invention illustrated by FIG. 5B, an expanded encoder 500bgenerates additional base layer(s), e.g., an additional higher content base layer BL(2) and corresponding enhancement layers EL(3-2), EL(4-2) from enhanced images 518, 512 in a manner similar to the generation of base layer BL(1) and enhancement layers EL(2-1), EL(3-1), EL(4-1) described above. Hence, expanded encoder 500bincludes decompression stage 554, compression stages 552, 578, 574, interpolation stages 558, 564, and summation stages 572, 576 in addition to encoder 510.

[0062] Functionally, by including expanded encoder 500bin server 210, in addition to growing and pruning the enhancement layers, server 210 is now able to provide one or more MMGs with a choice from two or more different base layers, upon request by anyone of client computers 231, 232, . . . 239. For example, upon a request for a higher content base layer when the lowest bandwidth actually used by any one of client computers 231, 232, . . . 239, corresponds to image I2′, server 210 begins providing MMGs with base layer BL(2), thereby eliminating the need for combining base layer BL1 with EL(2-1) to generate I2′.

[0063] FIGS. 6B and 6C s how de coding circuits 600band 600c, for generating reconstructed image I3′ from different base layers BL(1) and BL(2), respectively. In this example, decoding circuit 600b, which exists in decoder 600a, reconstructs image I3′ by combining base layer BL(1) with enhancement layer EL(3-1), while decoding circuit 600creconstructs image I3′ by combining base layer BL(2) with enhancement layer EL(3-2). Hence, by expanding decoder 600ato include the additional decoding functionality, such as the functionality provided by decoding circuit 600c, the expanded version of decoder 660ais now capable of receiving a higher content base layer, e.g., BL(2), thereby further improving network efficiency.

[0064] FIGS. 9 and 10 are flowcharts illustrating the adaptive right-sizing of a multimedia stream being transmitted, via a single multicast group, from a server 210 to one or more of client computers 231, 232, . . . 239. As discussed above, server 210 may optionally send information about the content of the multimedia stream to client computers 231, 232, . . . 239.

[0065] In step 910 of FIG. 9, server 210 streams the multimedia data which includes a base layer and at least one enhancement layer to the multicast group. Server 210 also listens for feedback on the use of and/or the need for the multimedia data from one or more of client computers 231, 232, . . . 239 (step 920).

[0066] In accordance with yet another aspect of the invention, server 210 adaptively right-sizes the multimedia data stream in response to the feedback from client computers 231, 232, . . . 239 (step 930). Right-sizing is the process of pruning and/or growing the multimedia data stream to better match the usage of the streams with the needs of client computers 231, 232, . . . 239. There are two related components to right sizing the multimedia stream. First, the higher base layer corresponding to the lowest bandwidth requested by client computers 231, 232, . . . 239 is selected. Next, unused enhancement layers are pruned, i.e., not transmitted to client computers 231, 232, . . . 239.

[0067] Subsequently, depending on the needs of client computers 231, 232, . . . 239, the current base layer may be replaced by a lower content or a higher content base layer. Similarly, depending on the need of client computers 231, 232, . . . 239 and the availability of network 220, the enhancement layers may be grown or pruned. The right-sizing process is repeated until the multicast of the stream is complete (step 940).

[0068] Conversely, as shown in the flowchart of FIG. 10, client computer 231, receives multimedia data via the multicast group (1010). Client computer 231 also provides feedback on the use of and/or need for the multimedia to server 210, thereby causing server 210 to right-size the multimedia data in the manner described above (step 1020).

[0069] FIGS. 11 and 12 are flowcharts illustrating the adaptive right-sizing of a multimedia stream being transmitted, via multiple multicast groups (MMGs), from a server 210 to one or more of client computers 231, 232, . . . 239. As discussed above, server 210 may optionally send information about the content of the multimedia stream to client computers 231, 232, . . . 239.

[0070] In step 1110 of FIG. 11, server 210 streams a first base layer and a first at least one enhancement layer of the multimedia data to a first multicast group. Similarly, server 210 streams a second base layer and a second at least one enhancement layer of the multimedia data to a second multicast group (step 1120). Note that either the first and second base layers, or the first and second at least one enhancement layers, respectively, need to be different.

[0071] In accordance with another aspect of the invention, server 210 receives selection(s) of one or both multicast groups from one or more of client computers 231, 232, . . . 239 (step 1130). Server 210 adaptively right-sizes the multimedia data stream in response to the feedback from client computers 231, 232, . . . 239 (step 1140). As discussed above, right-sizing is the process of pruning and/or growing the multimedia data stream to better match the usage of the streams with the needs of client computers 231, 232, . . . 239.

[0072] Subsequently, depending on the needs of client computers 231, 232, . . . 239, the current first and second base layers may be replaced by a lower content or a higher content base layer. Similarly, depending on the need of client computers 231, 232, . . . 239 and the availability of network 220, the first and second enhancement layers may be grown or pruned. The right-sizing process is repeated until the multiple multicast of the stream is complete (step 1150).

[0073] Conversely, as shown in the flowchart of FIG. 12, client computer 231, receives multimedia data via the MMGs (1210). Client computer 231 then selects one or more multicast group(s) from the MMGs (1220). Client computer 231 is also tasked with providing feedback on the use of and/or need for the multimedia data to server 210, thereby causing server 210 to right-size the multimedia data in the manner described above (step 1230).

[0074] While this invention has been described in terms of several preferred embodiments, there are alterations, permutations, and equivalents which fall within the scope of this invention. For example, instead of right-sizing all MMGs, it is possible to selectively right-size some MMGs while ‘blindly’ multicasting to other MMGs. It should also be noted that there are many alternative ways of implementing the methods and apparatuses of the present invention. It is therefore intended that the following appended claims be interpreted as including all such alterations, permutations, and equivalents as fall within the true spirit and scope of the present invention.