Title of Invention	"POINT-TO-MULTIPOINT COMMUNICATION SYSTEM FOR REPAIRING DATA
Abstract	A method, system, device, and computer code product is disclosed in which a sender (10) transmits data to a plurality of receivers (20) via a point-to-multipoint session. The receiver sends data repair requests to the sender (10) requesting data expected but not received and the sender (10) retransmits the expected but not received data via the point-to-multipoint session. The sender (10) can also schedule point-to-point data repair sessions with individual receivers if the retransmission via the point-to-multipoint session does not correct all errors. The sender (10) can be configured to delay point-to-point repair sessions using a randomization mechanism based on the number of receivers using the point-to-multipoint session.

Title of Invention

"POINT-TO-MULTIPOINT COMMUNICATION SYSTEM FOR REPAIRING DATA

Abstract

A method, system, device, and computer code product is disclosed in which a sender (10) transmits data to a plurality of receivers (20) via a point-to-multipoint session. The receiver sends data repair requests to the sender (10) requesting data expected but not received and the sender (10) retransmits the expected but not received data via the point-to-multipoint session. The sender (10) can also schedule point-to-point data repair sessions with individual receivers if the retransmission via the point-to-multipoint session does not correct all errors. The sender (10) can be configured to delay point-to-point repair sessions using a randomization mechanism based on the number of receivers using the point-to-multipoint session.

Full Text	Data Repair Enhancements for Multicast/Broadcast Data Distribution . . FIELD OF THE INVENTION [0001] The invention generally relates to multicast and broadcast transmission technology and services, that is, services with at least one data source (or sender) and at least one receiver. More particularly, the invention relates to data repair enhancements in a multicast or broadcast transmission. BACKGROUND OF THE INVENTION [0002] For one-to-many (i.e,, point-to-multipoint) services over systems such as IP multicast, IP datacasting (IPDC) and multimedia broadcast/multicast services (MBMS), file delivery (or discrete media delivery or file download) is an important service. Many of the features for delivering files over point-to-point protocols such as file transfer protocol (FTP) and hypertext transfer protocol (HTTP) are problematic for one-to-many scenarios, In particular, the reliable delivery of files - that is the guaranteed delivery of files - using similar .one-to-one (i.e., point-to-point) acknowledgement (ACK) protocols such as transmission control protocol TCP is not feasible, [0003] The Reliable Multicast Transport QRMT) Working Group of the Internet Engineering Task Force (IETF) is in the process of standardizing two categories of error-resilient multicast transport protocols. In the first category, reliability is implemented through the use of (proactive) forward error correction (FEC), that is, by sending a certain amount of redundant data that can help a receiver in reconstructing erroneous data. In the second category, receiver feedback is used in order to implement reliable multicast transport. Asynchronous Layered Coding (ALC, RFC 3450} is a protocol instantiation belonging to the first category, while the NACK-n^or,i-Rri "Reliable Multicast (NOBM) protocol presents an-example of the second category. The details of ALC and NORM protocols are discussed in more detail in publications entitled "Asynchronous Layered Coding (ALC) Protocol Instantiation" (IETFRFC 3450) and "NACK-oriented Reliable Multicast Protocol (Internet Draft) prepared by the Working Group of the IETF. The contents of these publications are fully incorporated herein by reference. [0004] Access networks on which these protocols can be used include, but axe not limited to, wireless multiple-access networks such as radio access networks of the Universal Mobile Teleconmnmications Services (UMTS) system, wireless local area networks (WLAN), Digital Video Broadcasting - Terrestrial (DVB-T) networks Digital Video Broadcasting - Satellite (DVB-S) networks and Digital Video Broadcasting - Handheld (BDVB-H) networks. [0005] Briefly, ALC protocol is a proactive EEC-based scheme that allows receivers to reconstruct mangled packets or packets that have not been received. ALC protocol uses EEC encoding on multiple channels, allowing the sender to send data at multiple rates (channels) to possibly heterogeneous receivers. Additionally. ALC protocol uses a congestion control mechanism to maintain different rates on different channels. [0006] ALC protocol is massively scalable in terms of the number of users because no uplink signalling is required. Therefore, adding more receivers does not put • increased demand on the system. However, ALC protocol is not 100% reliable because reception is not guaranteed, thus it may be generally described as robust, rather than reliable. [0007] NORM, in turn, specifies the use of negative acknowledgement (NACK) messages in order to signal which packets of data (or otherwise defined "data blocks") that were expected to arrive at the receiver were not received at the receiver (or -were received incorrectly). In other words, receivers employ NACK messages to indicate loss or damage of transmitted packets to the sender. Accordingly, a receiver that "missed" some data blocks from a data transmission can send a NACK message to the sender requesting the sender to re-transmit the missed data block or blocks. NORM protocol also optionally allows for the use of packet-level FEC encoding for proactive robust transmissions. [0008] File Delivery over Unidirectional Transport (FLUTE) is a one-to-many transport protocol that builds on FEC (RFC 3452} and ALC building blocks. It is intended for file delivery from sender(s) to receivers) over unidirectional systems. It has specializations which make it suitable to wireless point-to-mulupoint (multicast/broadcast) systems. The details of FLUTE protocol are discussed in more detail in the publication entitled ''FLUTE - File Deliver)/ aver Unidirectional Transport" (Internet Draft) prepared by the above-mentioned Working Group of the IETF. The contents of this publication are fully incorporated herein by reference. [0009] NACK. messages are not generally NORM specific, but they can also be used iu connection with other protocols or systems, such as FLUTE. An ACK is a response message a receiver sends after receiving one onnore data packets to acknowledge they were received correctly. A NACK is a response a receiver sends to the sender about packets that were expected to arrive, but were not received. [0010] When in multicast or broadcast environment the data transmission occurs in a one-to-many fashion. If the transmission is not error free and different receivers are subject to different error rates (for example in MBMS users in different cells.may experience different signal quality and, as a consequence, different error rate);, there is the problem of providing increased data'reliability. This can be achieved through the use of FEC and/or through the use of repair sessions, [0011] FEC provides a certain amount of redundancy to the transmitted data, in order to allow a certain degres of error resilience to enable a receiver to reconstruct the transmitted data. However, one problem of FEC is that it usually does not provide error free error recovery, or it provides full error recovery at the cost of a high use of data redundancy, which increases the channel bandwidth requirements. [0012] A repair session (between receiver and sender) can be employed to complement FEC (to reduce or eliminate the residual channel error rate), or can be used alone as the only method for error recovery. A repair session can occur over a point-to-point channel using a separate session. In this case, all the receivers that have missed some data during the multicast/broadcast transmission, send NACK requests to the sender to request the retransmission of the missing packets. However, if all the receivers miss at least one data packet, all the receivers will establish simultaneously point-to-point connections with the sender causing feedback implosion, Le., congestion in the network (in uplink direction for the large number of NACKs and in downlink direction for the large number of concurrent re-transmission and network connection requests) and overload of the sender. This situation is critical when considering, for example, thousands of users, as maybe the case in MBMS networks, [0013] As such, there is a need for an improved device, system, and method for data repair that is scalable and provides efficient repair of messages in multicast and broadcast environments. SUMMARY OF THE INVENTION [0014] Various embodiments of systems, methods, devices and computer code products are disclosed according to the present invention. The various embodiments are capable of point-to-multipoint communications and can include transmitting data from a sender to a plurality of receivers via a point-to-multipoint session, detennining if any expected data was not received, sending a data repair request to the sender if data is missing, and retransmitting the missing data via the point-to-multipoint session. The sender also can be configured for scheduling and performing point-to-point repair sessions if the point-to-multipornt retransmission does not correct the loss of data problem, [0015] A randomization mechanism can be used to delay point-to-point data repair until after the sender retransmits data indicated as not received via the point-to-multipoint session. The randomization mechanism, can be configured to take into account the number of receivers included in the plurality of receivers. Alternatively (or additionally), the sender can send a point-to-point repair token to the plurality of receivers to announce when point-to-point repair will begin. [0016] In another aspect of the invention, a method for data repair in a system comprising a sender device, at least one receiver capable of receiving data from the sender device, an HTTP server and a FLUTE server is disclosed. The method can include determining that some expected data was not received by the at least one receiver, sending a data request from the at least one receiver to the HTTP server, - an HTTP redirect message from the HTTP server to the at least one receiver, the redirect message including a FLUTE URL sending a session request from the at least one receiver to the FLUTE server, and initiating a FLUTE session between, the FLUTE server and the at least one receiver. BRIEF DESCRIPTION OF THE DRAWINGS [0017] FIG, 1A is a block diagram illustrating a pomt-to-multipoiat transmission scenario in accordance with one embodiment of the invention; [0018] FIG. IB is a block diagram illustrating different missing data repair methods in accordance with embodiments of the invention; [0019] FIG. 2A is a flow chart diagram illustrating one embodiment of a method for data repair according to the present invention; [0020] FIG. 2B is a flow chart diagram illustrating another embodiment of a method for data repair according to the present invention; {0021] FIG. 3 is a block diagram of a system and receiver device in accordance with one embodiment of the invention; [0022] FIG. 4 is a block diagram illustrating a sender device in accordance with one embodiment of the invention, [0023] FIG. 5 is a representation of one embodiment of a repair mechanism according to the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS [0024] There are various methods and systems for repairing data in a multicast or broadcast system. U.S. patent application entitled "Data Repair" (serial no. 10/782,371) filed on February 18,2004, the contents of which are incorporated fully herein, by reference, describes efficient methods for repairing data. This application proposes that after reception of a certain number of HACK requests from receivers., the sender may decide, based on its own decision strategies, to retransmit via point-to-multipoint part of the total number of packets that are NACKed by the receivers, for example, those packets that are most requested from the receivers. The sender may also close the point-to-point connections in order to save network resources. [0025] One drawback with, methods such as these is that retransmitting only the most NACKed packets may not lead to total error recovery in the case where there is little statistical correlation between the NACKL requests of different users. For example, if a particular error situation is such that receiver #1 NACKs for packets 1, 2, and 3, and receiver #2 NACKLs for packets 4, 5, and 6, and so on, the sender may not be able to derive what are the "most requested packets" and, as a consequence, the point-to-multipoint repair may lose its efficiency. The subject invention proposes improved methods, devices and systems for data repair. [0026] Figure 1 A shows a point-to-multipoint data transmission scenario in accordance with an embodiment of the invention. The sender device 10 can be a server, IP-based device, DVB device, GPRS (or UMTS) device or similar device that may use proactive forward error correction, such as an ALC mechanism and/or FEC mechanism, for sending multicast data. blocks (or packets) to receiver devices 20 in a one-to-many fashion. Each receiving device 20 can be configured to send negative acknowledgement NACK messages (or requests) to the sender device 10 concerning missing blocks (blocks not received or received incorrectly), [0027] Data can be transferred from sender 10 to receivers) 20 as objects. For instance, a file, a JPEG image, and a file slice are all objects. The objects can be sent as a series of data blocks. Each data block can have a number -called a source block number (SB'N) or similar identifierj which can be used to identify each block. Blocks can be represented by a set of encoding symbols. An encoding symbol identifier (ESI) or similar identifier, in turn, can indicate how the encoding symbols carried in the payload of a data packet (or block) were generated from the above-mentioned object [0028] In a point-to-multipoint system, a sender 10 can broadcast data blocks or packets representing an obj ect to many receivers 20 simultaneously. If a receiver 20 does not receive all of the packets that it expects, it can send a NACK message back to the server 1 0 indicating which packets were not received. Figure IB illustrates several different data repair methods in accordance with embodiments of the subject invention. In. general, repair of missing data can be performed by using a point-to-point repair session established between one sender 10 and one receiver 20 or by using a point-to-multipoint session between the sender 10 and more than one receiver 20. In a repair session, missing data in total or in part (depending on the case) can be re-transmitted from the sender 10 to the receiver(s) 20 or the whole transmission can be repeated. Repair may be effected from the original sender 10 or from a "third party server" or repair server (or just simply a separate server (not shown)) which has a connection with the original server and is configured to buffer the transmission data/information. This server may, for example, be co-located with the original sender (e.g., an MBMS server, also called BM-SC (Broadcast Multicast— Service Center)), or, for example, be a separte server within an UMTS operator's network. [0029] It has been observed that, in general, reliable multicast systems present the problem of requiring receiver-server control and data messaging which, due to the multiparty nature of multicast, presents scalability problems. There are several areas, in particular, which are of concern. For example: a) limited radio bandwidth and activation resources, where the time it would take to activate many radio channels and the radio bandwidth makes it infeasible to allow many repairs to occur simultaneously, b) limited server capacity, where the server system, which, is providing the "repair content" data, can handle limited numbers of requests (messaging) and associated session context data within a certain time window and a limited amount of simultaneous data transfer sessions: and c) limited md-to-end bandwidth, due to one or more bottlenecks in the overall system. Here the data rate, which could be made available to all the users requiring repair simultaneously, is, in many cases, insufficient to provide this . service. [0030] Thus, one factor which may be used to increase scalability under any or all of these limitations can be to distribute the messaging in time, or avoid it entirely if possible. One embodiment of the subject invention concerns methods, devices, and systems which can enable HACK suppression to provide scalable reliable multicast [0031] One embodiment of the subject invention proposes that all packets that are requested by at least one receiver 20 be retransmitted by the server 10 on the point-to-multipoint bearer. In this embodiment, the receivers 20 can be configured to have both a point-to-point (ptp) bearer and point-to-multipoint (ptm) bearer setup at the same time. The ptp bearer can be used, for example, to service repair requests as described in U.S. Patent Application No. 10/782,371. One embodiment of the subject invention can use randomization rules similar to those described in the aforementioned patent application. However, the embodiment of the subject invention can retransmit the lost data on the downlink ptm bearer instead of using the downlink ptp bearei. [0032] In this embodiment, receivers 20 whose turn to request has not come yet because of the random back-off value they computed, may have the opportunity to repair their own loss by receiving lost packets retransmitted through the ptm channel. If a receiver 20 receives a missing data packet through the ptm channel, it can reconstruct the file using this data and remove the missing data packet from its list of packets to request. It may be possible that a receiver 20 can receive all of its missing data before its computed request .time, in which case it could refrain from making any repair requests at all. 10033] In another embodiment of the subject invention, ptp repair can be offered by a sender 10 in conjunction with the above-described ptm repair mechanism. This may be useful, in particular, for sessions when not all of the receivers 20 are capable of having both a ptp and ptm bearer open, at the same time. In this case, for greater efficiency, the sender 10 may specify a randomization mechanism so as to delay requests for ptp repair. This allows repair on the ptm bearer that may benefit a higher number of receiver 20 to be done first. One way to do so, for example, may be through the use of threshold values (such as X, Y. Z) sent by the sender 10 to the receivers 20. The receiver 20 could then be configured to schedule their repair requests, One sample rule for receivers 20 to schedule repair requests according to one embodiment of the present invention could be; If ptm repair is possible, then uniformly randomize the NACK(s) over a time period X, starting from the end of the initial delivery session; else wait until after a certain rime Y after the initial session ends, and then randomize the TsfACKL(s) over a time period Z. [0034] The sender 10 could also explicitly signal when ptp repair should start. To this end, the sender can aend a ptp repair token to the receivers 20 to announce when ptp repair can start (when ptp repair starts, the ptp repair session can be subject to the normal randomization rules.) Prior to sending the ptp repair token, all repairs are done on the pirn bearer. Receivers 20 that are not capable of having two concurrent bearers (e.g. ptp and ptm) can thus wait for the token before they setup their ptp repair bearer, The repair token can be transmitted using any communication protocol at any of the layers 1-7 of the ISO OSI protocol stack, including, for example, via SDP in a separate "announcement" after the multicast/broadcast transmission. This can also be included in a FLUTE file delivery within a multicast/broadcast transmission, A separate Transport Object Identifier (TOI) value can be used to distinguish between the file content itself and the ptp repair content. In one embodiment of the subject invention, a receiver 20 that has already used ptm repair may also use ptp repair. This can be useful if the pun repair was not successful, i.e. the packet that was resent on the ptm bearer was lost [0035] While randomization can help prevent feedback implosion, it is preferable that back-off times be computed according to the number of receivers 20 in a system in oider to increase efficiency. 'If the backoff times are chosen to small, the risk of feedback implosion may not be minimized, especially if there are a large number of receivers 20 in the session. If, on the other hand, the back-off times are too large, the risk of feedback implosion decreases but the scheme becomes inefficient if there are only a few receivers in the session since each, receiver will be required to wait an unnecessarily large amount of time before being able to make a repair request [0036] If the sender 10 knows the number of receivers 20 in a, session, the sender 10 may be able to scale its randomization values based on the number of receivers 20 to optimize the performance of the system. One such type of session is anMBMS multicast session, where the sender 10 is able to derive the number of receivers 20 as the latter need to signal the session join and leave procedures. In one embodiment, a linear relation between the number of receivers 20 in the session and the randomization values can be used to compute the necessary threshold values. For example, using the randomization method proposed in U.S. Patent Application No. 10/782,371; If below the threshold error rate W then uniformly randomize the NACK(s) over a time period X, starting from the end of the initial delivery session; else •wait until after a certain time Y after the initial session ends, and then randomize the NACK(s) over a time period Z [0037] The values of W, X, Y, and Z cm be fixed and chosen according to the number of participants (number of receivers) in the session. A look-up table, such as the sample one show below, can be stored on the sender device 10 and a look-up into the proper table entry can be used to choose the threshold values. (Table Remove) It should be noted that the above table is merely one sample. Other values and table structures can be used without departaK .a-om the spirit and scope of the invention. [0038] The four values (or in general the values for randomizing the starting time of the repair session) can be conconnnicated from the sender to the receivers via SDP or any other suitable means. The values can be communicated to the receivers anytime between service announcement and the session start time or the litest join time. For example, if a session, is announced now via SDP, and scheduled to start after two hours (or alternatively the latest session joining time after 1.5 hours from the delivery of the service announcement), a second SDP with the randomization'paracQBters can be sent, using a second announcement or token which takes into account the number of receivers 20 that joined the session any time before the start of the session. In tins case, the receivers 20 get an indication of the randomization time, which takes into account the real and updated number of receivers that have joined the session Alternatively, the parameters can be communicated within the FDT of a. FLUTE session or only a subset of these values may vary with the number of receivers. [0039] Turning now to Figure 2A, one embodiment of a method for providing data repair is disclosed. The method disclosed in Figure 2A comprises sending data packets from the sender to a plurality of receivers via a poirrt-to-muitippint session (100). If the any of the receivers determines that it has not received some expected data it sends a NACK massage back to the sender requesting data packets were not properly received and the sender receives these NACK massages (102). Next the sender retransmits the requested data packets to the receivers via the point-to-multipohit session (104). {0040] Another embodiment of the invention is shown in Figure 2B. In this embodiment, the sender indicates the beginning of a point-to-muitipoint session (110) and then collects information about the number of receivers using the session (120). The sender then computes randomization values based on the number of receivers using the session (130) and sends the randomization values to the receivers (140). Next, the sender begins sending data packets to the receivers via the poult-to-multipoint session (150). If any of the receivers does not receive all of the expected data packets, it sends a NACK message back to the sender requesting retransmission of the missed data packets. The sender receives these NACK messages (160) and retransmits the requested data packets on the point-to-multipoint session. Then., the server begins servicing any remaining data repair requests via pomt-to-point session (1 SO). The point-to-point sessions are randomized over a period of time based on the randomization values computed by the sender based on the number of receivers using the point-to-multipornt session. [0041] The data repair methods described herein provide distinct advantages when compared to prior art methods. For example, sending a repair block that a receiver 20 requests via ptp repair via ptm instead of via downlink ptp unloads the ptp channel and helps other receivers 20 that may need the same repair block. Also, scaling the randomization values according to the number of receivers helps avoid the risk of feedback implosion while still minimizing the delay necessary to send requests. [0042] When using a mechanism, such as FLUTE, for transmitting data to many users in an efficient manner, a method for identifying, NACKiag and retransmitting missing data is necessary. In one embodiment of the invention, it is possible to use HTTP to do so. For example, the receiving terminal may use HTTP to request the transmitting server to retransmit a certain file or part of a file. In some cases, a server may be required, to retransmit the files to several terminals or may not have the file and hence may need to redirect the request to another server. One aspect of this invention provides an efficient system and method for ptm repair for FLUTE and other such sessions. [0043] While HTTP requests may be used to obtain a missing set of data from a transmission, it is desirable to have a repair mechanism that can enable a large number of receivers to receive a certain transmission again. In one embodiment, this can be achieved by using the HTTP redirect mechanism with a download URI for FLUTE. The download URI for FLUTE can be analogous to an HTTP URL. That is, it is similar to the HTTP URL, but applicable for FLUTE. The FLUTE URI can refer to a signel file transmitted during a certain FLUTE session. Using the FLUTE URI, a file from a certain FLUTE session can thus be uniquely identified. [0044] In this embodiment, an HTTP server can indicate to the receiver the source from which the file required by the receiver is available. By using the HTTP redirect mechanism, an HTTP server can redirect one or more terminals that have not received a transmission correctly to a FLUTE server that can then start anew ptm repair session for multiple receivers. [0045] In another aspect of the this embodiment, the HTTP redirect mechanism can also be used to indicate to the receiver the location of a FLUTE session description file (such as an SDP file). By using a session description file, the receiver can join a FLUTE session and obtain it file(s) of interest. The client can then cross-reference with already received and stored, session descriptions to check if it has atoeady received this particular session description file. It can then attempt to fetch, the missing file/packets/resource using HTTP/FLUTE, etc. In this example, the FLUTE URI can be the URI of an SDP file that described the FLUTE session to which the file in. question belongs (i.e. the FLUTE session in which it was transmitted and during which the terminal received this file). [0046] In other embodiment, the HTTP redirect may also contain the session description file as payload. This can be done using an absolute URI in the payload and/or by using a redirect code, When a session description file is used to do a redirect, it may be desirable to version the session description file. This can be achieved by using an extension header or by using the metadata envelope. The metadata envelope encapsulates or references the session description, file and provides a versioning mechanism. In this example, the URL of the file can be included in the header of the HTTP redirect message, such as in the form www.example.com/fiie/mp3. [0047] Figure 5 illustrates one embodiment of a method for using the FLUTE URI and HTTP redirect mechanism to launch a ptm file repair session. As shown in Figure 5, on transmission of a FLUTE session, the receiver may realize that is has not received a complete file. In suet a case, the receiver may send a repair request to an HTTP server. The address of the HTTP server may be known to the receiver from repair information transmitted during the FLUTE session. On obtaining the repair request, the HTTP server can decide that it is more efficient to conduct a ptm session using FLUTE rather than several individual ptp repair sessions using HTTP. This may be done for several reasons, such as the number of repair requests received are too high for the HTTP server does not itself possess the requested file. In this case, the HTTP server can send a redirect to the receiver with the FLUTE URI of the file. On obtaining the redirect from the HTTP server, the receiver can send a "session request" to the FLUTE server. The FLUTE server can then start a ptm FLUTE session with potentially many receivers "tuned in." It is possible for the FLUTE and HTTP servers to be separate entities but still be located on the same physical machine. [0048] One example of an SDP file that can be used for session descriptions is: v = 0 c = user123 2890844526 2890842807 IN !P6 2201:Q56D::112E:144A:1E24 s=F!ie delivery session example i— More information t = 28 7339 7496 2873404696 a = source-filter: incl IN IP6 * 2001:2 10: 1 :2:240:96FF:FE25:8EC9 a-flute-tsi:3 m = application 12345 FLUTE/UDPO c = !N 1P6 FF1E:Q3AD::7F2E: 1 72A;1E24J1 m = application 1 2346 FLUTE/VDP 0 c=lN 1P6 FF1E:03AD::7F2E:172A: 1E3Q/1 [0049] This example file describes how a receiver can join aFLUTJi session, It provides information regarding the sender IP address, the number of channels in the session, the destination IP address and port number for each channel in the session, the Transport Session identifier (TSI) of the session, and the start and end times of the session. [0050] Figure 3 illustrates one embodiment of a system 5 and receiver device 20 in accordance with the present invention. The system 5 can include a sender device 10, a transmission network 30, e.g.? an IP network or another fixed network, a •wireless network or a combination of a fixed and wireless (cellular) network, etc., and the receiver device 20, The receiver device 20 can be, for example, a cellular telephone, a satellite telephone, a personal digital assistant, a Bluetooth device, a WLAN device, a DVB device, or other similar wireless device. The receiver' 20 can include an internal memory 21, a processor 22, an operating system 23, application programs 24, a network interface 25, and a NACK and repair mechanism 26. The internal memory application programs 24, The NACK and repair mechanism 26 can enable the NACKing and repair procedures in response to missing or mangled data in a data transmission. The receiver device 20 can be capable of communication "with the sender device 10 and with other devices via the network interface 25 and the network 30. [0051] Figure 4 illustrates one embodiment of a sender device 10 in accordance with the present invention. The sender device 10 can "be, for example, a network server or any suitable device intended for file (or media) delivery. The sender device 10 can include internal memory 11, a processor 12, an operating system 13, application programs 14, a network interface 15, a transmission and repair mechanism 16, and a data storage 17. The internal memory 11 can be configured to accommodate the processor 12, operating system 13, and application programs 14. The transmission and repair mechanism 16 can be configured to enable the transmission of data packets to receiver devices 20. Furthermore, it can be setup to enable re-transmission of data packets in repair sessions. Data to be sent to receiver devices 20 and data to be re-transmitted can. be stored in the data storage 17. Alternatively, date can be stored in a separate device co-located with or outside of the sender device 10. The sender device 10 can be configured to communicate with the receiver device 20 and other devices via the network interface 15 and the network 30. [0052] Procedures relating to repair of missing data can be implemented by software. A computer program product comprising program code stored in the receiver device 20 and run in the processor 22 can be used to implement the procedures at the receiving end of the transmission session, whereas a computer program product comprising program code stored in the sender device 10 and run in the processor 12 can be used to implement the procedures at the transmitting end. [0053] Embodiments of the invention have been illustrated with examples or logical sender/server entitles and receiver units, however, the use of other entities going between for repair requests, and repair responses (if appropriate), are also contemplated and considered within the scope of the subject invention. Such an entity may provide firewall, proxy, and/or authorization services. [0054] While the exemplary embodiments illustrated in the FIGURES and described above are presently preferred, it should be understood that these embodiments are offered by way of example, only. Other embodiments may include, for example, different techniques for performing the same operations. The invention is not limited to a particular embodiment, but extends to various modifications, combinations, and permutations that nevertheless fall within the scope and spirit of the appended claims. WHAT IS CLAIMED IS: 1. A method for data repair in a system capable of point-to-multipoint communications, the method comprising: transmitting data from a sender to a plurality of receivers via a point-to-multipoint session; determining if any expected data was not received; if some expected data was not received, sending a data repair request to the sender requesting that the expected but not received data be resent; and retransmitting from the sender the requested data via the point-to-multipoint session. 2. The method of claim 1, further comprising: after the sender retransmits the requested data, if some data was still not received, scheduling point-to-point repair sessions for specific receivers that expected data that was riot received; sending data still not received to the specific receivers via point-to-point sessions according to the point-to-point repair session schedule. 3. The method of claim 2, wherein scheduling point-to-point repair sessions further comprises specifying a randomization mechanism to randomize point-to-point data repair over a certain time period after the sender has retransmitted the not received data. 4. The method of claim 2, wherein scheduling point-to-point repair sessions further comprises: If pomt-to-multipoint repair is possible for a receiver, then uniformly randomizing data repair requests over a time period X starting from the end of the initial transmitting of data from the sender to the receivers via the point-to-multipoint session; else waiting until a certain time Y after ths initial transmitting of data from the sender to the receivers via the point-to-multipoint session and then randomizing the data repair requests over a time period Z. 5. The method of claim 2, wherein scheduling point-to-point repair sessions comprises sending a point-to-point repair token from the sender to the plurality of receivers to announce when point-to-point repair will begin. 6. The method of claim 3, wherein the randomization mechanism is configured to take into account the number of receivers included in the plurality of receivers. 7. The method of claim 6, further comprising: determining the number of receivers in the plurality of receiver; and computing the randomization values for the randomization mechanism based on the determined number of receivers. 8. The method of claim 7, wherein computing the randomization values further comprises looking up the randomization values up in a look-up table based on me determined number of receivers. 9. The method of claim 1, wherein sending the data repair requests comprises at least one of the plurality of receivers sending the repair request to an HTTP server. 10. The method of claim 9, further comprising receiving a redirect message from the HTTP server to the at least one receiver, sending a session request to a FLUTE server, and initiating a FLUTE session with the FLUTE server. 11. The method of claim 10, wherein the redirect message includes a FLUTE URI for the expected but not received data. 12. The method of claim 10, wherein the redirect message includes a session description file indicating how the at least one receiver can join the FLUTE 13. The method of claim 10, wherein the redirect message includes a reference to the location of a session description file including how the at least one receiver can join the FLUTE session. 14. The method of claim 1, wherein, hi response to a data repair request, a receiver requesting that the expected but not received data be resent is redirected to a session description using a locator of the session description 15. The method of claim 1, wherein in response to a data repair request, a receiver requesting that the expected but not received data be resent is redirected to a' session description by redirect message carrying the session description as the redirect message payload 16. The method of claim 1, wherein in response to a data repair request, a receiver requesting that the expected "but not received data be resent is redirected to a description from where the receiver can subsequently discover E repair session or session description. 17. A point-to-multipoint communication system capable .of data repair, the system comprising: a sender device capable of point-to-multipoint communications; a plurality of receivers capable of receiving data from the sender device; wherein the sender device is configured to transmit, data to the plurality of receivers via a point-to-multipoint session; the plurality of receivers are configured to receive data transmitted by the sender device, determine if any expected data was not received, and, if so, send a data repair request back to the sender device requesting that the expected but not received data be resent; and the sender device is configured to receive data rep air requests from the plurality of receivers and to retransmit the requested data via the point-to-multipoint session. 18. The system of claim 17 wherein the sender device is further configured to schedule point-to-point data repair sessions with the plurality of receivers after retransmission of the requested data and the sender is configured to send expected but not received data to the plurality of receivers via point-to-point sessions. 19. The system of claim 1S 'wherein the sender device is further configured to specify a randomization mechanism to delay point-to-point data repair. 20. The system of claim 19 wherein the sender can determine the numb er of receivers on the point-to-multipoint session and can compute a randomization mechanism that is based on the determined number of receivers. 21. The system of claim 18 wherein the sender is configured to send a point-to-point repair token to the plurality of receives to announce when point-to- point repair -win begin.. 22. The system of claim 18 further comprising a look-up table for determining the point-to-point repair schedule. 23. The system of claim 17,.further comprising an HTTP server and a FLUTE server, wherein the data repair requests comprises sending the repair request to the HTTP server. 24: The system of claim 23, wherein, in response to the data repair request, the HTTP server sends a redirect message to at least one receiver, in response to receiving the redirect message the at least one receiver sends a session request to the FLUTE server, and in response to the session request the FLUTE server initiates a FLUTE session with the at least one receiver. 25. The system of claim 24, wherein the redirect message includes a FLUTE URI for the expected but not received data. 26. The system of claim 24, wherein the redirect message includes a .session description file indicating how the at least one receiver can join the FLUTE session. 27. The system of claim 24, wherein the redirect message includes a reference to the location of a session description file including how the at least one receiver can join the FLUTE session. 28. The system of claim 17, wherein in response to a data repair request, a receiver requesting that the expected but not received data be resent is redirected to a session description using a locator of the session description 29. The system of claim 17, wherein in response to a- data repair request, a receiver requesting that the expected but not received data "be resent is redirected-to a session description by redirect message carrying the session description as the redirect message payload 30. The system of claim 17, wherein in response to a data repair request, a receiver requesting that the expected but not received data be resent is redirected to a description from where the receiver can subsequently discover a repair session or session description, . 31. A computer code product comprising: computer code configured to: transmit data from a sender to a plurality of receivers via a point-to-nmltipoint session; determine if expected data was not received at any of the plurality of receivers; make a data repair request if any data was not received at any of the ulurality of receivers; and retransmit the expected but not received data to the plurality of receivers via the point-to-multipoint session. 32. The computer-code product of claim 31, wherein the computer code is further configured to schedule point-to-point data repair sessions after retransmission of the expected but not received data. 33. The computer code product of claim 31 wherein the computer code is further configured to determine the number of receivers on the point-to-multipoint session and schedule the point-to-point data repair sessions based on the determined number of receivers. 34. A sender device for use in a point-to-multipoint comm.unicati.on system, the sender device comprising: means for transmitting data to a plurality of receivers via a point-to-multipoint session; means for receiving data repair requests from the plurality of receivers requesting expected by not received data; means for retransmitting the expected but not yet received data via a . point-to-multipoint session. 35. The sender device of claim 34 further comprising means for scheduling point-to-point data repair sessions with the plurality of receivers after retransmitting the expected but not yet received data. 36. The sender device of claim 34 wherein 'the sender device further comprises means for determiniag the number of receivers using the point-to- multipomt session wherein the sender is. configured to schedule the point-to-point data repair sessions based on the determined number of receivers. 37. A method for data repair in a system comprisir; sender device, at least one receiver capable of receiving data from the sender de •• ice, an HTTP server and a FLUTE server, the method comprising: determining that some expected data was not received by the at least one receiver; sending a data repair request from the at least one receiver to the HTTP server; sending an HTTP redirect message from the HTTP server to the at least one receiver, sending a session request from the at least one receiver to the FLUTE server; and initiating a FLUTE session between the FLUTE server and the at least one isceiver, 38. -The method of claim 37 wherein the redirect message includes a FLUTE URI for the expected "but not received data. 39. The method of claim 37, wherein the redirect message includes a session description file indicating how the at least one receiver can join the FLUTE session. 40. The method of claim 31, wherein the redirect message includes a reference to the location of a session description file including how the at least one receiver can join the FLUTE session. 41. A method for data repair in a system comprising a sender device and at least one receiver capable of receiving data from the sender device, the method comprising: determining that some expected data was not received by the at least one receiver; sending a data repair request form the at least one receiver; and in response to a data repair request, redirecting the at least one receiver to a session description using a locator of the session description. 42. A method for data repair in a system comprising a sender device and at least one receiver capable of receiving data from the sender device, the method comprising: determioing that some expected data was not received by the at least sending a data repair request form the at least one receiver; and in response to the data repair request, redirecting the at least one receiver to a session description by a. redirect message carrying the session description as the redirect message payload. 43. A method for data repair in a system comprising a sender device and at least one receiver capable of receiving data from the sender device, the method comprising: ddsrrnining that some expected data was not received by the at least one receiver; sending a data repair request form the at least one receiver, and in response to a data repair request, redirecting the at least one receiver to a description from where the receiver can subsequently discover a repair session or session description. 44. A method for data repair in a system capable of point-to-mulnpoint corrmmnication, the method comprising: transmitting data from a sender to a plurality of receivers via apoint-to-nrultipofflt session; determining if any of the plurality of receivers expected data that -was not received; determining the number of receivers using the pomt-to-uiultipomt session; computing randomization values for a randomization mechanism based on the determined number of receivers; scheduling point-to-point repair sessions with any of the plurality of receivers that expected data that was not received; and delaying the point-to-point data repair sessions based on the computed randomization values. 45. A computer code product comprising: computer code configured to: transmit data from a sender to a plurality of receivers via a point-to-multipoint session; determine if expected data was not received at any of the plurality of receivers; make a data repair request if any data was not received at any of the plurality of receivers; determine the number of receivers on the point-to-multipoint session; schedule point-to-point data repair sessions for each receiver that did not receive all expected data,- and delaying the point-to-point data, repair session hased on the number of determined receivers. • '. 46. A sender device for use in a point-to-multipoint communication system, the sender device, comprising: means for transmitting data to a plurality of receivers via a point-to-multipoin.t session; means far receiving data repair requests from the plurality of receivers requesting expected but not received data; means for determining the number of receivers using the point-to- multipoint session; . wherein the sender device is configured to schedule point-to-point data repair sessions with receivers that did not receive all expected data; and delaying the point-to-point data repair session based on the determined number of receivers. 47. A pornt-to-multipoint communication system capable of data repair, the system comprising: a sender device capable of point-to-multipoint communication; a plurality of receivers capable of receiving data from the sender device; wherein the sender is configured to transmit data to the plurality of receivers via a point-to-nuiltipoint session; the plurality of receivers being configured to receive data transmitted by the sender device, determine if any expected data was not received, and if so, send a data repair request .back to the sender device requesting that the expected but not received data be resent; the sender being configured to determine the number of receivers on the point-to-multipoint session and to determine a randomization mechanism based on the determiDed.number of receivers; the sender being configured to schedule point-to-point repair sessions with receivers that expected data that was not received, the point-to-point repair sessions being delayed "based on the randomization mechanism,

Full Text

Data Repair Enhancements for Multicast/Broadcast Data
Distribution
. . FIELD OF THE INVENTION
[0001] The invention generally relates to multicast and broadcast transmission technology and services, that is, services with at least one data source (or sender) and at least one receiver. More particularly, the invention relates to data repair enhancements in a multicast or broadcast transmission.
BACKGROUND OF THE INVENTION
[0002] For one-to-many (i.e,, point-to-multipoint) services over systems such as IP multicast, IP datacasting (IPDC) and multimedia broadcast/multicast services (MBMS), file delivery (or discrete media delivery or file download) is an important service. Many of the features for delivering files over point-to-point protocols such as file transfer protocol (FTP) and hypertext transfer protocol (HTTP) are problematic for one-to-many scenarios, In particular, the reliable delivery of files - that is the guaranteed delivery of files - using similar .one-to-one (i.e., point-to-point) acknowledgement (ACK) protocols such as transmission control protocol TCP is not feasible,
[0003] The Reliable Multicast Transport QRMT) Working Group of the Internet Engineering Task Force (IETF) is in the process of standardizing two categories of error-resilient multicast transport protocols. In the first category, reliability is implemented through the use of (proactive) forward error correction (FEC), that is, by sending a certain amount of redundant data that can help a receiver in reconstructing erroneous data. In the second category, receiver feedback is used in order to implement reliable multicast transport. Asynchronous Layered Coding (ALC, RFC 3450} is a protocol instantiation belonging to the first category, while the NACK-n^or,i-Rri "Reliable Multicast (NOBM) protocol presents an-example of the second

category. The details of ALC and NORM protocols are discussed in more detail in publications entitled "Asynchronous Layered Coding (ALC) Protocol Instantiation" (IETFRFC 3450) and "NACK-oriented Reliable Multicast Protocol (Internet Draft) prepared by the Working Group of the IETF. The contents of these publications are fully incorporated herein by reference.
[0004] Access networks on which these protocols can be used include, but axe not limited to, wireless multiple-access networks such as radio access networks of the Universal Mobile Teleconmnmications Services (UMTS) system, wireless local area networks (WLAN), Digital Video Broadcasting - Terrestrial (DVB-T) networks Digital Video Broadcasting - Satellite (DVB-S) networks and Digital Video Broadcasting - Handheld (BDVB-H) networks.
[0005] Briefly, ALC protocol is a proactive EEC-based scheme that allows receivers to reconstruct mangled packets or packets that have not been received. ALC protocol uses EEC encoding on multiple channels, allowing the sender to send data at multiple rates (channels) to possibly heterogeneous receivers. Additionally. ALC protocol uses a congestion control mechanism to maintain different rates on different channels. [0006] ALC protocol is massively scalable in terms of the number of users because no uplink signalling is required. Therefore, adding more receivers does not put • increased demand on the system. However, ALC protocol is not 100% reliable because reception is not guaranteed, thus it may be generally described as robust, rather than reliable.
[0007] NORM, in turn, specifies the use of negative acknowledgement (NACK) messages in order to signal which packets of data (or otherwise defined "data blocks") that were expected to arrive at the receiver were not received at the receiver (or -were received incorrectly). In other words, receivers employ NACK messages to indicate loss or damage of transmitted packets to the sender. Accordingly, a receiver that "missed" some data blocks from a data transmission can send a NACK message to the sender requesting the sender to re-transmit the missed data block or blocks. NORM protocol also optionally allows for the use of packet-level FEC encoding for proactive robust transmissions.

[0008] File Delivery over Unidirectional Transport (FLUTE) is a one-to-many transport protocol that builds on FEC (RFC 3452} and ALC building blocks. It is intended for file delivery from sender(s) to receivers) over unidirectional systems. It has specializations which make it suitable to wireless point-to-mulupoint (multicast/broadcast) systems. The details of FLUTE protocol are discussed in more detail in the publication entitled ''FLUTE - File Deliver)/ aver Unidirectional Transport" (Internet Draft) prepared by the above-mentioned Working Group of the IETF. The contents of this publication are fully incorporated herein by reference. [0009] NACK. messages are not generally NORM specific, but they can also be used iu connection with other protocols or systems, such as FLUTE. An ACK is a response message a receiver sends after receiving one onnore data packets to acknowledge they were received correctly. A NACK is a response a receiver sends to the sender about packets that were expected to arrive, but were not received. [0010] When in multicast or broadcast environment the data transmission occurs in a one-to-many fashion. If the transmission is not error free and different receivers are subject to different error rates (for example in MBMS users in different cells.may experience different signal quality and, as a consequence, different error rate);, there is the problem of providing increased data'reliability. This can be achieved through the use of FEC and/or through the use of repair sessions,
[0011] FEC provides a certain amount of redundancy to the transmitted data, in order to allow a certain degres of error resilience to enable a receiver to reconstruct the transmitted data. However, one problem of FEC is that it usually does not provide error free error recovery, or it provides full error recovery at the cost of a high use of data redundancy, which increases the channel bandwidth requirements. [0012] A repair session (between receiver and sender) can be employed to complement FEC (to reduce or eliminate the residual channel error rate), or can be used alone as the only method for error recovery. A repair session can occur over a point-to-point channel using a separate session. In this case, all the receivers that have missed some data during the multicast/broadcast transmission, send NACK requests to the sender to request the retransmission of the missing packets. However, if all the receivers miss at least one data packet, all the receivers will establish simultaneously

point-to-point connections with the sender causing feedback implosion, Le., congestion in the network (in uplink direction for the large number of NACKs and in downlink direction for the large number of concurrent re-transmission and network connection requests) and overload of the sender. This situation is critical when considering, for example, thousands of users, as maybe the case in MBMS networks, [0013] As such, there is a need for an improved device, system, and method for data repair that is scalable and provides efficient repair of messages in multicast and broadcast environments.
SUMMARY OF THE INVENTION
[0014] Various embodiments of systems, methods, devices and computer code products are disclosed according to the present invention. The various embodiments are capable of point-to-multipoint communications and can include transmitting data from a sender to a plurality of receivers via a point-to-multipoint session, detennining if any expected data was not received, sending a data repair request to the sender if data is missing, and retransmitting the missing data via the point-to-multipoint session. The sender also can be configured for scheduling and performing point-to-point repair sessions if the point-to-multipornt retransmission does not correct the loss of data problem,
[0015] A randomization mechanism can be used to delay point-to-point data repair until after the sender retransmits data indicated as not received via the point-to-multipoint session. The randomization mechanism, can be configured to take into account the number of receivers included in the plurality of receivers. Alternatively (or additionally), the sender can send a point-to-point repair token to the plurality of receivers to announce when point-to-point repair will begin. [0016] In another aspect of the invention, a method for data repair in a system comprising a sender device, at least one receiver capable of receiving data from the sender device, an HTTP server and a FLUTE server is disclosed. The method can include determining that some expected data was not received by the at least one receiver, sending a data request from the at least one receiver to the HTTP server,
- an HTTP redirect message from the HTTP server to the at least one receiver,

the redirect message including a FLUTE URL sending a session request from the at least one receiver to the FLUTE server, and initiating a FLUTE session between, the FLUTE server and the at least one receiver.
BRIEF DESCRIPTION OF THE DRAWINGS
[0017] FIG, 1A is a block diagram illustrating a pomt-to-multipoiat transmission
scenario in accordance with one embodiment of the invention;
[0018] FIG. IB is a block diagram illustrating different missing data repair methods
in accordance with embodiments of the invention;
[0019] FIG. 2A is a flow chart diagram illustrating one embodiment of a method for
data repair according to the present invention;
[0020] FIG. 2B is a flow chart diagram illustrating another embodiment of a method
for data repair according to the present invention;
{0021] FIG. 3 is a block diagram of a system and receiver device in accordance with
one embodiment of the invention;
[0022] FIG. 4 is a block diagram illustrating a sender device in accordance with one
embodiment of the invention,
[0023] FIG. 5 is a representation of one embodiment of a repair mechanism
according to the present invention.
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0024] There are various methods and systems for repairing data in a multicast or broadcast system. U.S. patent application entitled "Data Repair" (serial no. 10/782,371) filed on February 18,2004, the contents of which are incorporated fully herein, by reference, describes efficient methods for repairing data. This application proposes that after reception of a certain number of HACK requests from receivers., the sender may decide, based on its own decision strategies, to retransmit via point-to-multipoint part of the total number of packets that are NACKed by the receivers, for example, those packets that are most requested from the receivers. The sender may also close the point-to-point connections in order to save network resources.

[0025] One drawback with, methods such as these is that retransmitting only the most NACKed packets may not lead to total error recovery in the case where there is little statistical correlation between the NACKL requests of different users. For example, if a particular error situation is such that receiver #1 NACKs for packets 1, 2, and 3, and receiver #2 NACKLs for packets 4, 5, and 6, and so on, the sender may not be able to derive what are the "most requested packets" and, as a consequence, the point-to-multipoint repair may lose its efficiency. The subject invention proposes improved methods, devices and systems for data repair. [0026] Figure 1 A shows a point-to-multipoint data transmission scenario in accordance with an embodiment of the invention. The sender device 10 can be a server, IP-based device, DVB device, GPRS (or UMTS) device or similar device that may use proactive forward error correction, such as an ALC mechanism and/or FEC mechanism, for sending multicast data. blocks (or packets) to receiver devices 20 in a one-to-many fashion. Each receiving device 20 can be configured to send negative acknowledgement NACK messages (or requests) to the sender device 10 concerning missing blocks (blocks not received or received incorrectly), [0027] Data can be transferred from sender 10 to receivers) 20 as objects. For instance, a file, a JPEG image, and a file slice are all objects. The objects can be sent as a series of data blocks. Each data block can have a number -called a source block number (SB'N) or similar identifierj which can be used to identify each block. Blocks can be represented by a set of encoding symbols. An encoding symbol identifier (ESI) or similar identifier, in turn, can indicate how the encoding symbols carried in the payload of a data packet (or block) were generated from the above-mentioned object
[0028] In a point-to-multipoint system, a sender 10 can broadcast data blocks or packets representing an obj ect to many receivers 20 simultaneously. If a receiver 20 does not receive all of the packets that it expects, it can send a NACK message back to the server 1 0 indicating which packets were not received. Figure IB illustrates several different data repair methods in accordance with embodiments of the subject invention. In. general, repair of missing data can be performed by using a point-to-point repair session established between one sender 10 and one receiver 20 or by

using a point-to-multipoint session between the sender 10 and more than one receiver 20. In a repair session, missing data in total or in part (depending on the case) can be re-transmitted from the sender 10 to the receiver(s) 20 or the whole transmission can be repeated. Repair may be effected from the original sender 10 or from a "third party server" or repair server (or just simply a separate server (not shown)) which has a connection with the original server and is configured to buffer the transmission data/information. This server may, for example, be co-located with the original sender (e.g., an MBMS server, also called BM-SC (Broadcast Multicast— Service Center)), or, for example, be a separte server within an UMTS operator's network. [0029] It has been observed that, in general, reliable multicast systems present the problem of requiring receiver-server control and data messaging which, due to the multiparty nature of multicast, presents scalability problems. There are several areas, in particular, which are of concern. For example:
a) limited radio bandwidth and activation resources, where the time it would take
to activate many radio channels and the radio bandwidth makes it infeasible
to allow many repairs to occur simultaneously,
b) limited server capacity, where the server system, which, is providing the
"repair content" data, can handle limited numbers of requests (messaging) and
associated session context data within a certain time window and a limited
amount of simultaneous data transfer sessions: and
c) limited md-to-end bandwidth, due to one or more bottlenecks in the overall
system. Here the data rate, which could be made available to all the users
requiring repair simultaneously, is, in many cases, insufficient to provide this
. service.
[0030] Thus, one factor which may be used to increase scalability under any or all of these limitations can be to distribute the messaging in time, or avoid it entirely if possible. One embodiment of the subject invention concerns methods, devices, and systems which can enable HACK suppression to provide scalable reliable multicast [0031] One embodiment of the subject invention proposes that all packets that are requested by at least one receiver 20 be retransmitted by the server 10 on the point-to-multipoint bearer. In this embodiment, the receivers 20 can be configured to have

both a point-to-point (ptp) bearer and point-to-multipoint (ptm) bearer setup at the same time. The ptp bearer can be used, for example, to service repair requests as described in U.S. Patent Application No. 10/782,371. One embodiment of the subject invention can use randomization rules similar to those described in the aforementioned patent application. However, the embodiment of the subject invention can retransmit the lost data on the downlink ptm bearer instead of using the downlink ptp bearei.
[0032] In this embodiment, receivers 20 whose turn to request has not come yet because of the random back-off value they computed, may have the opportunity to repair their own loss by receiving lost packets retransmitted through the ptm channel. If a receiver 20 receives a missing data packet through the ptm channel, it can reconstruct the file using this data and remove the missing data packet from its list of packets to request. It may be possible that a receiver 20 can receive all of its missing data before its computed request .time, in which case it could refrain from making any repair requests at all.
10033] In another embodiment of the subject invention, ptp repair can be offered by a sender 10 in conjunction with the above-described ptm repair mechanism. This may be useful, in particular, for sessions when not all of the receivers 20 are capable of having both a ptp and ptm bearer open, at the same time. In this case, for greater efficiency, the sender 10 may specify a randomization mechanism so as to delay requests for ptp repair. This allows repair on the ptm bearer that may benefit a higher number of receiver 20 to be done first. One way to do so, for example, may be through the use of threshold values (such as X, Y. Z) sent by the sender 10 to the receivers 20. The receiver 20 could then be configured to schedule their repair requests, One sample rule for receivers 20 to schedule repair requests according to one embodiment of the present invention could be;
If ptm repair is possible, then
uniformly randomize the NACK(s) over a time period X, starting from
the end of the initial delivery session;
else

wait until after a certain rime Y after the initial session ends, and then randomize the TsfACKL(s) over a time period Z.
[0034] The sender 10 could also explicitly signal when ptp repair should start. To this end, the sender can aend a ptp repair token to the receivers 20 to announce when ptp repair can start (when ptp repair starts, the ptp repair session can be subject to the normal randomization rules.) Prior to sending the ptp repair token, all repairs are done on the pirn bearer. Receivers 20 that are not capable of having two concurrent bearers (e.g. ptp and ptm) can thus wait for the token before they setup their ptp repair bearer, The repair token can be transmitted using any communication protocol at any of the layers 1-7 of the ISO OSI protocol stack, including, for example, via SDP in a separate "announcement" after the multicast/broadcast transmission. This can also be included in a FLUTE file delivery within a multicast/broadcast transmission, A separate Transport Object Identifier (TOI) value can be used to distinguish between the file content itself and the ptp repair content. In one embodiment of the subject invention, a receiver 20 that has already used ptm repair may also use ptp repair. This can be useful if the pun repair was not successful, i.e. the packet that was resent on the ptm bearer was lost
[0035] While randomization can help prevent feedback implosion, it is preferable that back-off times be computed according to the number of receivers 20 in a system in oider to increase efficiency. 'If the backoff times are chosen to small, the risk of feedback implosion may not be minimized, especially if there are a large number of receivers 20 in the session. If, on the other hand, the back-off times are too large, the risk of feedback implosion decreases but the scheme becomes inefficient if there are only a few receivers in the session since each, receiver will be required to wait an unnecessarily large amount of time before being able to make a repair request [0036] If the sender 10 knows the number of receivers 20 in a, session, the sender 10 may be able to scale its randomization values based on the number of receivers 20 to optimize the performance of the system. One such type of session is anMBMS multicast session, where the sender 10 is able to derive the number of receivers 20 as the latter need to signal the session join and leave procedures. In one embodiment, a linear relation between the number of receivers 20 in the session and the

randomization values can be used to compute the necessary threshold values. For example, using the randomization method proposed in U.S. Patent Application No. 10/782,371;
If below the threshold error rate W then
uniformly randomize the NACK(s) over a time period X, starting from
the end of the initial delivery session;
else
•wait until after a certain time Y after the initial session ends, and then
randomize the NACK(s) over a time period Z
[0037] The values of W, X, Y, and Z cm be fixed and chosen according to the number of participants (number of receivers) in the session. A look-up table, such as the sample one show below, can be stored on the sender device 10 and a look-up into the proper table entry can be used to choose the threshold values.

(Table Remove)

It should be noted that the above table is merely one sample. Other values and table structures can be used without departaK .a-om the spirit and scope of the invention. [0038] The four values (or in general the values for randomizing the starting time of the repair session) can be conconnnicated from the sender to the receivers via SDP or any other suitable means. The values can be communicated to the receivers anytime between service announcement and the session start time or the litest join time. For

example, if a session, is announced now via SDP, and scheduled to start after two hours (or alternatively the latest session joining time after 1.5 hours from the delivery of the service announcement), a second SDP with the randomization'paracQBters can be sent, using a second announcement or token which takes into account the number of receivers 20 that joined the session any time before the start of the session. In tins case, the receivers 20 get an indication of the randomization time, which takes into account the real and updated number of receivers that have joined the session Alternatively, the parameters can be communicated within the FDT of a. FLUTE session or only a subset of these values may vary with the number of receivers. [0039] Turning now to Figure 2A, one embodiment of a method for providing data repair is disclosed. The method disclosed in Figure 2A comprises sending data packets from the sender to a plurality of receivers via a poirrt-to-muitippint session (100). If the any of the receivers determines that it has not received some expected data it sends a NACK massage back to the sender requesting data packets were not properly received and the sender receives these NACK massages (102). Next the sender retransmits the requested data packets to the receivers via the point-to-multipohit session (104).
{0040] Another embodiment of the invention is shown in Figure 2B. In this embodiment, the sender indicates the beginning of a point-to-muitipoint session (110) and then collects information about the number of receivers using the session (120). The sender then computes randomization values based on the number of receivers using the session (130) and sends the randomization values to the receivers (140). Next, the sender begins sending data packets to the receivers via the poult-to-multipoint session (150). If any of the receivers does not receive all of the expected data packets, it sends a NACK message back to the sender requesting retransmission of the missed data packets. The sender receives these NACK messages (160) and retransmits the requested data packets on the point-to-multipoint session. Then., the server begins servicing any remaining data repair requests via pomt-to-point session (1 SO). The point-to-point sessions are randomized over a period of time based on the randomization values computed by the sender based on the number of receivers using the point-to-multipornt session.

[0041] The data repair methods described herein provide distinct advantages when compared to prior art methods. For example, sending a repair block that a receiver 20 requests via ptp repair via ptm instead of via downlink ptp unloads the ptp channel and helps other receivers 20 that may need the same repair block. Also, scaling the randomization values according to the number of receivers helps avoid the risk of feedback implosion while still minimizing the delay necessary to send requests. [0042] When using a mechanism, such as FLUTE, for transmitting data to many users in an efficient manner, a method for identifying, NACKiag and retransmitting missing data is necessary. In one embodiment of the invention, it is possible to use HTTP to do so. For example, the receiving terminal may use HTTP to request the transmitting server to retransmit a certain file or part of a file. In some cases, a server may be required, to retransmit the files to several terminals or may not have the file and hence may need to redirect the request to another server. One aspect of this invention provides an efficient system and method for ptm repair for FLUTE and other such sessions.
[0043] While HTTP requests may be used to obtain a missing set of data from a transmission, it is desirable to have a repair mechanism that can enable a large number of receivers to receive a certain transmission again. In one embodiment, this can be achieved by using the HTTP redirect mechanism with a download URI for FLUTE. The download URI for FLUTE can be analogous to an HTTP URL. That is, it is similar to the HTTP URL, but applicable for FLUTE. The FLUTE URI can refer to a signel file transmitted during a certain FLUTE session. Using the FLUTE URI, a file from a certain FLUTE session can thus be uniquely identified. [0044] In this embodiment, an HTTP server can indicate to the receiver the source from which the file required by the receiver is available. By using the HTTP redirect mechanism, an HTTP server can redirect one or more terminals that have not received a transmission correctly to a FLUTE server that can then start anew ptm repair session for multiple receivers.
[0045] In another aspect of the this embodiment, the HTTP redirect mechanism can also be used to indicate to the receiver the location of a FLUTE session description file (such as an SDP file). By using a session description file, the receiver can join a

FLUTE session and obtain it file(s) of interest. The client can then cross-reference with already received and stored, session descriptions to check if it has atoeady received this particular session description file. It can then attempt to fetch, the missing file/packets/resource using HTTP/FLUTE, etc. In this example, the FLUTE URI can be the URI of an SDP file that described the FLUTE session to which the file in. question belongs (i.e. the FLUTE session in which it was transmitted and during which the terminal received this file).
[0046] In other embodiment, the HTTP redirect may also contain the session description file as payload. This can be done using an absolute URI in the payload and/or by using a redirect code, When a session description file is used to do a redirect, it may be desirable to version the session description file. This can be achieved by using an extension header or by using the metadata envelope. The metadata envelope encapsulates or references the session description, file and provides a versioning mechanism. In this example, the URL of the file can be included in the header of the HTTP redirect message, such as in the form www.example.com/fiie/mp3.
[0047] Figure 5 illustrates one embodiment of a method for using the FLUTE URI and HTTP redirect mechanism to launch a ptm file repair session. As shown in Figure 5, on transmission of a FLUTE session, the receiver may realize that is has not received a complete file. In suet a case, the receiver may send a repair request to an HTTP server. The address of the HTTP server may be known to the receiver from repair information transmitted during the FLUTE session. On obtaining the repair request, the HTTP server can decide that it is more efficient to conduct a ptm session using FLUTE rather than several individual ptp repair sessions using HTTP. This may be done for several reasons, such as the number of repair requests received are too high for the HTTP server does not itself possess the requested file. In this case, the HTTP server can send a redirect to the receiver with the FLUTE URI of the file. On obtaining the redirect from the HTTP server, the receiver can send a "session request" to the FLUTE server. The FLUTE server can then start a ptm FLUTE session with potentially many receivers "tuned in." It is possible for the FLUTE and HTTP servers to be separate entities but still be located on the same physical machine.

[0048] One example of an SDP file that can be used for session descriptions is:
v = 0
c = user123 2890844526 2890842807 IN !P6 2201:Q56D::112E:144A:1E24 s=F!ie delivery session example i— More information t = 28 7339 7496 2873404696
a = source-filter: incl IN IP6 * 2001:2 10: 1 :2:240:96FF:FE25:8EC9 a-flute-tsi:3
m = application 12345 FLUTE/UDPO c = !N 1P6 FF1E:Q3AD::7F2E: 1 72A;1E24J1 m = application 1 2346 FLUTE/VDP 0 c=lN 1P6 FF1E:03AD::7F2E:172A: 1E3Q/1
[0049] This example file describes how a receiver can join aFLUTJi session, It provides information regarding the sender IP address, the number of channels in the session, the destination IP address and port number for each channel in the session, the Transport Session identifier (TSI) of the session, and the start and end times of the session.
[0050] Figure 3 illustrates one embodiment of a system 5 and receiver device 20 in accordance with the present invention. The system 5 can include a sender device 10, a transmission network 30, e.g.? an IP network or another fixed network, a •wireless network or a combination of a fixed and wireless (cellular) network, etc., and the receiver device 20, The receiver device 20 can be, for example, a cellular telephone, a satellite telephone, a personal digital assistant, a Bluetooth device, a WLAN device, a DVB device, or other similar wireless device. The receiver' 20 can include an internal memory 21, a processor 22, an operating system 23, application programs 24, a network interface 25, and a NACK and repair mechanism 26. The internal memory

application programs 24, The NACK and repair mechanism 26 can enable the NACKing and repair procedures in response to missing or mangled data in a data transmission. The receiver device 20 can be capable of communication "with the sender device 10 and with other devices via the network interface 25 and the network 30.
[0051] Figure 4 illustrates one embodiment of a sender device 10 in accordance with the present invention. The sender device 10 can "be, for example, a network server or any suitable device intended for file (or media) delivery. The sender device 10 can include internal memory 11, a processor 12, an operating system 13, application programs 14, a network interface 15, a transmission and repair mechanism 16, and a data storage 17. The internal memory 11 can be configured to accommodate the processor 12, operating system 13, and application programs 14. The transmission and repair mechanism 16 can be configured to enable the transmission of data packets to receiver devices 20. Furthermore, it can be setup to enable re-transmission of data packets in repair sessions. Data to be sent to receiver devices 20 and data to be re-transmitted can. be stored in the data storage 17. Alternatively, date can be stored in a separate device co-located with or outside of the sender device 10. The sender device 10 can be configured to communicate with the receiver device 20 and other devices via the network interface 15 and the network 30. [0052] Procedures relating to repair of missing data can be implemented by software. A computer program product comprising program code stored in the receiver device 20 and run in the processor 22 can be used to implement the procedures at the receiving end of the transmission session, whereas a computer program product comprising program code stored in the sender device 10 and run in the processor 12 can be used to implement the procedures at the transmitting end. [0053] Embodiments of the invention have been illustrated with examples or logical sender/server entitles and receiver units, however, the use of other entities going between for repair requests, and repair responses (if appropriate), are also contemplated and considered within the scope of the subject invention. Such an entity may provide firewall, proxy, and/or authorization services.

[0054] While the exemplary embodiments illustrated in the FIGURES and described above are presently preferred, it should be understood that these embodiments are offered by way of example, only. Other embodiments may include, for example, different techniques for performing the same operations. The invention is not limited to a particular embodiment, but extends to various modifications, combinations, and permutations that nevertheless fall within the scope and spirit of the appended claims.

WHAT IS CLAIMED IS:
1. A method for data repair in a system capable of point-to-multipoint
communications, the method comprising:
transmitting data from a sender to a plurality of receivers via a point-to-multipoint session;
determining if any expected data was not received;
if some expected data was not received, sending a data repair request to the sender requesting that the expected but not received data be resent; and
retransmitting from the sender the requested data via the point-to-multipoint session.
2. The method of claim 1, further comprising:
after the sender retransmits the requested data, if some data was still not received, scheduling point-to-point repair sessions for specific receivers that expected data that was riot received;
sending data still not received to the specific receivers via point-to-point sessions according to the point-to-point repair session schedule.
3. The method of claim 2, wherein scheduling point-to-point repair
sessions further comprises specifying a randomization mechanism to randomize
point-to-point data repair over a certain time period after the sender has retransmitted
the not received data.
4. The method of claim 2, wherein scheduling point-to-point repair
sessions further comprises:
If pomt-to-multipoint repair is possible for a receiver, then
uniformly randomizing data repair requests over a time period X starting from the end of the initial transmitting of data from the sender to the receivers via the point-to-multipoint session;
else

waiting until a certain time Y after ths initial transmitting of data from the sender to the receivers via the point-to-multipoint session and then randomizing the data repair requests over a time period Z.
5. The method of claim 2, wherein scheduling point-to-point repair
sessions comprises sending a point-to-point repair token from the sender to the
plurality of receivers to announce when point-to-point repair will begin.
6. The method of claim 3, wherein the randomization mechanism is
configured to take into account the number of receivers included in the plurality of
receivers.
7. The method of claim 6, further comprising:
determining the number of receivers in the plurality of receiver; and computing the randomization values for the randomization mechanism based on the determined number of receivers.
8. The method of claim 7, wherein computing the randomization values
further comprises looking up the randomization values up in a look-up table based on
me determined number of receivers.
9. The method of claim 1, wherein sending the data repair requests
comprises at least one of the plurality of receivers sending the repair request to an
HTTP server.
10. The method of claim 9, further comprising receiving a redirect
message from the HTTP server to the at least one receiver, sending a session request
to a FLUTE server, and initiating a FLUTE session with the FLUTE server.
11. The method of claim 10, wherein the redirect message includes a
FLUTE URI for the expected but not received data.
12. The method of claim 10, wherein the redirect message includes a
session description file indicating how the at least one receiver can join the FLUTE

13. The method of claim 10, wherein the redirect message includes a
reference to the location of a session description file including how the at least one
receiver can join the FLUTE session.
14. The method of claim 1, wherein, hi response to a data repair request, a
receiver requesting that the expected but not received data be resent is redirected to a
session description using a locator of the session description
15. The method of claim 1, wherein in response to a data repair request, a
receiver requesting that the expected but not received data be resent is redirected to a'
session description by redirect message carrying the session description as the redirect
message payload
16. The method of claim 1, wherein in response to a data repair request, a
receiver requesting that the expected "but not received data be resent is redirected to a
description from where the receiver can subsequently discover E repair session or
session description.
17. A point-to-multipoint communication system capable .of data repair,
the system comprising:
a sender device capable of point-to-multipoint communications; a plurality of receivers capable of receiving data from the sender
device;
wherein the sender device is configured to transmit, data to the plurality of receivers via a point-to-multipoint session;
the plurality of receivers are configured to receive data transmitted by the sender device, determine if any expected data was not received, and, if so, send a data repair request back to the sender device requesting that the expected but not received data be resent; and
the sender device is configured to receive data rep air requests from the plurality of receivers and to retransmit the requested data via the point-to-multipoint session.

18. The system of claim 17 wherein the sender device is further
configured to schedule point-to-point data repair sessions with the plurality of
receivers after retransmission of the requested data and the sender is configured to
send expected but not received data to the plurality of receivers via point-to-point
sessions.
19. The system of claim 1S 'wherein the sender device is further configured
to specify a randomization mechanism to delay point-to-point data repair.
20. The system of claim 19 wherein the sender can determine the numb er
of receivers on the point-to-multipoint session and can compute a randomization
mechanism that is based on the determined number of receivers.
21. The system of claim 18 wherein the sender is configured to send a
point-to-point repair token to the plurality of receives to announce when point-to-
point repair -win begin..
22. The system of claim 18 further comprising a look-up table for
determining the point-to-point repair schedule.
23. The system of claim 17,.further comprising an HTTP server and a
FLUTE server, wherein the data repair requests comprises sending the repair request
to the HTTP server.
24: The system of claim 23, wherein, in response to the data repair request, the HTTP server sends a redirect message to at least one receiver, in response to receiving the redirect message the at least one receiver sends a session request to the FLUTE server, and in response to the session request the FLUTE server initiates a FLUTE session with the at least one receiver.
25. The system of claim 24, wherein the redirect message includes a FLUTE URI for the expected but not received data.

26. The system of claim 24, wherein the redirect message includes a
.session description file indicating how the at least one receiver can join the FLUTE
session.
27. The system of claim 24, wherein the redirect message includes a
reference to the location of a session description file including how the at least one
receiver can join the FLUTE session.
28. The system of claim 17, wherein in response to a data repair request, a
receiver requesting that the expected but not received data be resent is redirected to a
session description using a locator of the session description
29. The system of claim 17, wherein in response to a- data repair request, a
receiver requesting that the expected but not received data "be resent is redirected-to a
session description by redirect message carrying the session description as the redirect
message payload
30. The system of claim 17, wherein in response to a data repair request, a
receiver requesting that the expected but not received data be resent is redirected to a
description from where the receiver can subsequently discover a repair session or
session description, .
31. A computer code product comprising:
computer code configured to:
transmit data from a sender to a plurality of receivers via a point-to-nmltipoint session;
determine if expected data was not received at any of the plurality of receivers;
make a data repair request if any data was not received at any of the ulurality of receivers; and
retransmit the expected but not received data to the plurality of receivers via the point-to-multipoint session.

32. The computer-code product of claim 31, wherein the computer code is
further configured to schedule point-to-point data repair sessions after retransmission
of the expected but not received data.
33. The computer code product of claim 31 wherein the computer code is
further configured to determine the number of receivers on the point-to-multipoint
session and schedule the point-to-point data repair sessions based on the determined
number of receivers.
34. A sender device for use in a point-to-multipoint comm.unicati.on
system, the sender device comprising:
means for transmitting data to a plurality of receivers via a point-to-multipoint session;
means for receiving data repair requests from the plurality of receivers requesting expected by not received data;
means for retransmitting the expected but not yet received data via a . point-to-multipoint session.
35. The sender device of claim 34 further comprising means for scheduling
point-to-point data repair sessions with the plurality of receivers after retransmitting
the expected but not yet received data.
36. The sender device of claim 34 wherein 'the sender device further
comprises means for determiniag the number of receivers using the point-to-
multipomt session wherein the sender is. configured to schedule the point-to-point data
repair sessions based on the determined number of receivers.
37. A method for data repair in a system comprisir; sender device, at
least one receiver capable of receiving data from the sender de •• ice, an HTTP server
and a FLUTE server, the method comprising:
determining that some expected data was not received by the at least one receiver;

sending a data repair request from the at least one receiver to the HTTP server;
sending an HTTP redirect message from the HTTP server to the at least one receiver,
sending a session request from the at least one receiver to the FLUTE server; and
initiating a FLUTE session between the FLUTE server and the at least one isceiver,
38. -The method of claim 37 wherein the redirect message includes a
FLUTE URI for the expected "but not received data.
39. The method of claim 37, wherein the redirect message includes a
session description file indicating how the at least one receiver can join the FLUTE
session.
40. The method of claim 31, wherein the redirect message includes a
reference to the location of a session description file including how the at least one
receiver can join the FLUTE session.
41. A method for data repair in a system comprising a sender device and at
least one receiver capable of receiving data from the sender device, the method
comprising:
determining that some expected data was not received by the at least
one receiver;
sending a data repair request form the at least one receiver; and
in response to a data repair request, redirecting the at least one receiver
to a session description using a locator of the session description.
42. A method for data repair in a system comprising a sender device and at
least one receiver capable of receiving data from the sender device, the method
comprising:
determioing that some expected data was not received by the at least

sending a data repair request form the at least one receiver; and in response to the data repair request, redirecting the at least one
receiver to a session description by a. redirect message carrying the session description
as the redirect message payload.
43. A method for data repair in a system comprising a sender device and at
least one receiver capable of receiving data from the sender device, the method
comprising:
ddsrrnining that some expected data was not received by the at least
one receiver;
sending a data repair request form the at least one receiver, and
in response to a data repair request, redirecting the at least one receiver
to a description from where the receiver can subsequently discover a repair session or
session description.
44. A method for data repair in a system capable of point-to-mulnpoint
corrmmnication, the method comprising:
transmitting data from a sender to a plurality of receivers via apoint-to-nrultipofflt session;
determining if any of the plurality of receivers expected data that -was not received;
determining the number of receivers using the pomt-to-uiultipomt session;
computing randomization values for a randomization mechanism based on the determined number of receivers;
scheduling point-to-point repair sessions with any of the plurality of receivers that expected data that was not received; and
delaying the point-to-point data repair sessions based on the computed randomization values.
45. A computer code product comprising:
computer code configured to:

transmit data from a sender to a plurality of receivers via a point-to-multipoint session;
determine if expected data was not received at any of the plurality of receivers;
make a data repair request if any data was not received at any of the plurality of receivers;
determine the number of receivers on the point-to-multipoint session;
schedule point-to-point data repair sessions for each receiver that did not receive all expected data,- and
delaying the point-to-point data, repair session hased on the
number of determined receivers. • '.
46. A sender device for use in a point-to-multipoint communication
system, the sender device, comprising:
means for transmitting data to a plurality of receivers via a point-to-multipoin.t session;
means far receiving data repair requests from the plurality of receivers requesting expected but not received data;
means for determining the number of receivers using the point-to-
multipoint session; .
wherein the sender device is configured to schedule point-to-point data repair sessions with receivers that did not receive all expected data; and
delaying the point-to-point data repair session based on the determined number of receivers.
47. A pornt-to-multipoint communication system capable of data repair,
the system comprising:
a sender device capable of point-to-multipoint communication; a plurality of receivers capable of receiving data from the sender device;

wherein the sender is configured to transmit data to the plurality of receivers via a point-to-nuiltipoint session;
the plurality of receivers being configured to receive data transmitted by the sender device, determine if any expected data was not received, and if so, send a data repair request .back to the sender device requesting that the expected but not received data be resent;
the sender being configured to determine the number of receivers on the point-to-multipoint session and to determine a randomization mechanism based on the determiDed.number of receivers;
the sender being configured to schedule point-to-point repair sessions with receivers that expected data that was not received, the point-to-point repair sessions being delayed "based on the randomization mechanism,

Documents:

http://ipindiaonline.gov.in/patentsearch/GrantedSearch/viewdoc.aspx?id=eqBdWohPPvUVGMF5jxpJlg==&loc=+mN2fYxnTC4l0fUd8W4CAA==

« Previous Patent

Next Patent »

Patent Number

269640

Indian Patent Application Number

5956/DELNP/2006

PG Journal Number

45/2015

Publication Date

06-Nov-2015

Grant Date

29-Oct-2015

Date of Filing

12-Oct-2006

Name of Patentee

NOKIA CORPORATION

Applicant Address

KEILALAHDENTIE 4,FIN-02150 ESPOO, FINLAND

Inventors:

#	Inventor's Name	Inventor's Address
1	LEON,DAVID	43 COURS A. BRIAND 33000 BORDEAUX FRANCE
2	CURCIO,IGOR DANILO DIEGO	HATANPAAN VALTATIE 12 C 57, FIN-331000 TAMPERE,FINLAND,
3	WALSH, ROD	KEIKKAKATU 4B 12, FIN-33580 TAMPERE FINLAND
4	MEHTA HARSH	SUVIKUJA 3D 27 02120 ESPOO FINLAND

PCT International Classification Number

H04L 1/18

PCT International Application Number

PCT/US2004/032414

PCT International Filing date

2004-10-01

PCT Conventions:

#	PCT Application Number	Date of Convention	Priority Country
1	60/614,602	2004-09-30	U.S.A.
2	10/813,343	2004-03-29	U.S.A.