MXPA06005563A - Method of partially scrambling a data stream - Google Patents

Abstract

A method of partially scrambling a data stream (6) including transport stream packets (7), each transport stream packet (7) having a header (8) and a payload (9), wherein a sequence of transport stream packets (7) has payloads carrying encoded data elements, arranged in units (15), includes:selecting transport stream packets (7) forming a subsequence of the sequence, and scrambling at least part of the payloads (9) of each transport stream packet (7) in the subsequence. The method further includes monitoring the payloads (9) of at least some of the transport stream packets (7) in the sequence for the presence of data (22) indicating a boundary between two subsequent units (15), and, for selected units (15), including at least one of the transport stream packets (7) carrying data forming part of the selected unit (15) in the sub-sequence.

Description

SI, SK, TR), OAPI (BF, BJ, CF, CG, Cl, CM, GA, GN, - with amended claims GQ, GW, ML, MR, NE, SN, TD, TG). For two-letler codes and other abbrevialions. referto the "Guid-Published: anee Notes on Codes and Abbreviations" appearing at the beginning- with intemalional seareh report no ofeach regular issue of the PCT Gazette.

METHOD TO PARTIALLY MASK A FLOW OF DATA DESCRIPTIVE MEMORY The invention relates to a method for partially masking a data stream, including transport flow packets, each transport flow packet having a header and a payload, wherein a sequence of transport flow packets has payloads transporting coded data elements, arranged in units, including: selecting transport flow packets that form a sub-sequence of the sequence, as well as masking at least part of the payloads of each transport flow packet in the sub -sequence. The invention also relates to a system for partially masking a data stream that includes transport flow packets, each transport flow packet having a header and a payload, wherein a sequence of transport flow packets has payloads that carry coded data elements, arranged in units, including: a port to receive the data flow; and a provision for processing the data in the flow, wherein the system is configured to select transport flow packets that form a sub-sequence of the sequence, as well as to mask at least part of the payloads of each packet of the transportation flow in the sub-sequence. The invention also relates to a computer program. The invention also relates to a signal that carries a data stream, including packets of the transport flow, given a packet of the transport flow having a header and a payload. Examples of said method, system, computer program and signal are known from document WO 03/061289-A1. In the known system, a cable system header selects A V content packets in a packet selector for encryption. The packages selected for encryption are selected so that their lack of reception would severely affect the real-time decoding of a program, as well as any possible post-processing of the recorded content. That is, only critical packets are encrypted. For audio and video, this can be achieved by encrypting packets of "box start" transport containing packetized elementary stream (PES) headers, as well as other headers as part of the payload, because without this information the STB decoder can not decompress the compressed MPEG data; the MPEG-2 streams identify the "frame start" packets with the "Unit Start Indicator" Package "in the transport header In general, packets that carry a payload that contains an image header group or a video stream header can be used to perform the masking technique here. or critical or important packets may also be identified for encryption that could severely inhibit unauthorized viewing, without deviating from the present invention For example, packets of frame I or intra-coding MPEG images could be encrypted to inhibit display of the portion of A problem with the known method is that the encryption of all payloads of TS packets, including headers of PES packets or a selected critical packet, will lead to a flow where a large percentage of packets have an encrypted payload. This makes the method unsuitable for broadcasting to devices with an limited processing capacity for decryption. The mere encryption of only a few of the payloads of the TS packets, including the headers of PES packets or a selected critical packet, will not be sufficient when the payloads of the packets include audio-visual data encoded using advanced compression techniques, a high degree of resilience to the loss of packets. In that case, an unauthorized receiver would be able to decode the content using the data left blank. One purpose of the present invention is to provide a method, system and computer program of the types mentioned above, which allow to mask the payloads of only a few selected transport flow packs, while maintaining effective protection of content in the presence of advanced decoder techniques to deal with missing encoded data. This purpose is achieved by means of the method of compliance with the invention, which is characterized by the monitoring of the payloads of at least some of the transport flow packets in the sequence in relation to the presence of data indicating a boundary between two subsequent units and, for the selected units, including at least one of the transport flow packets that carry data that is part of the unit selected in the subsequence. Since the transport flow packets are selected to form a sub-sequence, i.e., a sequence formed from less than all of the transport flow packets in the sequence, the unmasking load is reduced in a decoder receiving the partially masked flow. Since the encoded data elements are structured in units, they are also processed unit by unit upon decoding. Since the full payload of the transport flow packets is monitored in relation to the presence of data indicating a boundary between two subsequent units, it is possible to ensure that only the selected units become unintelligible. The number and nature can be made dependent on the decoding procedure. This also takes into account the fact that the data that make up the unit can be transported in the payload of a plurality of transport flow packets. Therefore, masking can be customized more closely to the coding technique. In a preferred embodiment, wherein the data stream is a multiplex of elementary streams, the method includes identifying at least one elementary stream that includes the sequence of transport flow packets and monitoring only the payloads of the packets in the stream (s) identified elemental (s). Therefore, a relatively efficient method is provided to partially mask the loads of transport flow packets. In a preferred embodiment, the selected units include units that contain at least part of a coded representation of an image. This mode has a relatively high impact on a decoder. The images contained in the respective units are almost invariably the desired result of the decoder, which can no longer be provided without unmasking the payloads of the transport flow packets containing these selected units. In a preferred embodiment, wherein each of the units contains an indication of the type of data to be tracked and a part containing that data, the type of each unit in the monitored payloads is determined from the indication and the unit it is included among the selected units if the type corresponds to at least one specific type. Therefore, a large reduction in the number of transport flow packets of which the payloads are masked is achieved, while maintaining effective content protection. Preferably, units of types other than the specific type (s) are included randomly among the selected units. Therefore, extra content protection is achieved. In preferred variants of the two above embodiments, the types are defined by the coding technique with which the encoded data elements have been formed. Therefore, those packet payloads containing the most critical units for the decoding process are selectable for masking. This makes content protection more effective, since it is precisely customized to the particular CODEC that is used. In a preferred embodiment, wherein the encoded data elements are decodable using a predictive decoding technique, the specific types include a type of data elements that allows a prediction to be derived only from the decoded data belonging to the data element .

Since these data elements allow random access to a coded data stream, making it inaccessible has a high impact on the decoding process which generally can not start without these coded data elements. In a preferred embodiment, up to a maximum number of transport flow packets following a first transport flow packet that carries data that is part of a selected unit is included in the sub-sequence. This reduces the load on a receiver, while maintaining adequate protection. The maximum number is selected so that error correction becomes impossible. In particular, when all transport flow packets have a standard length, including up to a maximum number of all transport flow packets that carry data that is part of the unit, it avoids masking more payloads than that required to be that the techniques are ineffective in a receiver to deal with the loss of data. Therefore effective protection is achieved, while loading in an authorized receiver that provides access to the partially masked flow is kept as low as possible. According to another aspect, the system according to the invention is characterized in that the system is configured to monitor the payloads of at least some of the transport flow packets in the sequence in relation to the presence of data indicating a boundary between two subsequent units and, for the selected units, to include at least one of the transport flow packets that carry data that is part of the unit selected in the sub-sequence. The system is suitable for generating a partially masked flow where the data elements are well protected against access by unauthorized receivers, but which does not pose great demands on the processing capacity of the authorized receivers. Preferably, the system is configured to include up to a maximum number of transport flow packets that follow a first transport flow packet that carries data that is part of a selected unit in the sub-sequence, as well as provided with a provision to specify the maximum number. This reduces the load on a receiver, while maintaining adequate protection. The maximum number can be specified so that error correction becomes impossible. In particular, when all transport flow packets have a standard length, including up to a maximum number of all transport flow packets that carry data that is part of the unit, it is avoided to mask more payloads than what is required to do that the techniques are ineffective in a receiver to deal with the loss of data. Therefore, effective protection is achieved, while loading on an authorized receiver that provides access to the partially masked flow is kept as low as possible. Since the maximum number can be specified, factors such as the profile of the typical receivers, the way of coding and the number of selected units, can be taken into consideration. In accordance with another aspect, the invention provides a computer program adapted, when operated on a computer, to configure the computer to execute a method according to the invention. In accordance with another aspect, the invention provides a signal that transports a data flow that includes transport flow packets, each transport flow packet having a header and a payload, where a sequence of transport flow packets has payloads that carry coded data elements, arranged in units, each unit being of a certain type, wherein at least part of the payload of each transport flow packet in a sub-sequence of the sequence is masked, where for each unit of a type corresponding to a selected type, the sub-sequence includes at least one of the transport flow packets that carry data that is part of that unit, where at least one of the transportation flow packets includes data indicating a boundary between two subsequent units.

The signal is suitable to be unicast, multicast or broadly disseminated to receivers with a limited processing capacity for unmasking, but is equipped with decoders with a high degree of tolerance to data loss. These include wireless devices in particular. The invention will now be displayed in greater detail with reference to the accompanying drawings, in which: Figure 1 shows schematically a system for partially masking a data flow. Figure 2 schematically shows a data flow comprising transport flow packets in accordance with the MPEG-2 standard. Figure 3 schematically shows the composition of a MPEG-2 transport flow packet header. Figure 4 shows schematically the composition of a Program Elemental Flow packet transported in the data stream of Figure 2. And Figure 5 shows schematically the composition of a Network Adaptation Layer unit as defined in the H standard. .264 / AVC and that is transported in the data stream of Figure 2. The invention will be explained in a way to partially mask a data stream that includes transport flow packets in accordance with the standard of MPEG-2 systems or ISO / IEC 1.3818-1.

In this particular mode, MPEG-2 is used as the delivery mechanism for audio-visual data encoded in accordance with the H.264 / AVC standard. It will be understood that the invention can be applied to other delivery mechanisms than MPEG-2. An example is the Internet Protocol (IP), which defines IP datagrams that have a header and a payload. Similarly, the use of H.264 / AVC as a content coding technique is an advantageous example, thanks to the high degree of compression and acceptable quality that can be achieved. However, an alternative would be MPEG-2 video encoding, where the encoded video is arranged in frames. In the example that will be described in detail, the data flow is in the MPEG-2 transport flow format. The example can be adapted to the program flow format. Therefore, the term transportation flow package does not mean that the use of a particular format is prescribed. It is noted that, when the program flow format is used, the data is transmitted in PES packets, each having a header, and that the PES packets are grouped into packets, each having a packet header. In Figure 1, a video file server 1 provides a stream of MPEG-2 systems that transports H.264 encoded video, among others. A multiplexer 2 divides the Conditional Access (CA) messages into the data flow. CA messages include Ownership Management Messages (EMMs), generated by an EMM 3 generator, as well as Ownership Control Messages (ECMs), generator by a generator. ECM 4. The ECM generator 4 generates Control Words (CWs). These are provided to a masking unit 5 arranged to partially mask the output data stream of the multiplexer 2. The CWs are encrypted under a service key provided by the EMM generator 3, which packs the service keys in the EMMs. Therefore, the CA scheme used corresponds basically to the common masking algorithm of Digital Video Broadcasting (DVB), which is known per se, for example, from the technical report of ETSI 289, and which is not described in more detail in the present. It is noted that the separation into components in Figure 1 has a functional nature. Several with all of them could be combined into a single physical device, just as the functionality of some of these components could be provided separately by different devices. Additionally, the implementation of this functionality could be done in software or hardware. Figure 2 shows a data stream 6 that includes transportation flow pack (TS) 7a-7p. Each TS 7 packet has a TS packet header 8a-8p and a packet payload of TS 9a-9p. The data stream 6 corresponds to a coded bit stream, e.g. a coded data, video or audio stream. For example, the ECMs are transported in a separate elementary stream, while at least one elementary stream received by the masquerader unit 5 is an H.264 encoded video stream. Figure 3 shows the conformation of a TS packet header 8. It includes a sync byte (sync byte, in English) 10, a payload unit start indicator (payload_un¡t_start_indicator, in English) 11, a field of packet identifier (PID) 12, a transport masking control field 13 and, optionally, an adaptation field 14. Other parts are described in ISO / IEC 13818-1. Each elementary stream corresponds to a unique PID. Therefore, the masking unit 5 is able to identify the flow or elementary streams that carry H.264 encoded video. For efficiency purposes, only the TS packets belonging to said flow or flows are processed. As is known from ISO / IEC 13818-1, an MPEG-2 transport stream will contain Program Specific Information (PSI) that identifies the different elementary streams and their assigned PIDs, as well as the type of data transported in the elementary stream. A type, identifiable in a Program Association Table (PAT, for its acronym in English), is H: 264. The masking unit 5 can use the PSI to identify the PIDs of the elementary stream (s) that must be partially masked. In the following, it will be assumed that only an elementary stream should be partially masked. The packets of TS 8 with the PID corresponding to that particular elementary stream are retrieved from the data stream, thereby forming a packet sequence of TS 8 having payloads of TS packets 9 carrying H-coded data elements. 264 Preferably, the encoded data elements contain at least parts of coded representations of images. An image is a frame or field in a video sequence (that is, a complete image or alternating lines). The macroblocks, blocks of sixteen by sixteen luma and chroma samples, are formed from the data of each image. This is done in accordance with a trace sweep, a transfer of the rectangular two-dimensional pattern of the macroblocks in the matrix to a one-dimensional pattern, so that the first introductions in the one-dimensional pattern are from the first upper row of the digitized two-dimensional pattern of left to right, followed by the macroblocks of the following rows, each digitized from left to right. The image is divided into groups of slices, subsets of macroblocks. A slice contains macroblocks that are consecutive in the trace sweep within the group of slices, but not necessarily within the image. Therefore, it is possible to assign any other macroblock in the trace sweep of the image to a separate slice group, for example. A slice is decoded only based on the samples from that slice, that is, independently of the samples in other slices. However, when the data in one slice is lost in the transmission, the data can be used in another slice to "fill the empty space" in the decoded image. The frequency of sampled images provides a division in time, in addition to the topographic division provided by the slices. H.264 is an example of a predictive coding method: a pixel value is coded as the sum of a predicted value and a difference value, where the latter value is reduced as much as possible during coding. For slices containing data from the first image and images defined as random access points, the predicted values are based only on values within that image itself and not on those in images at a different point in the order of transmission. These slices are called intra-coded slices. The remaining slices can be encoded to be uniquely decodable using also the values in other images in the order of a sequence of images. Therefore, it is possible to decode a sequence of coded images by starting in an intra-coded image, but not by starting in another type of coded image. Until now, the description of relevant parts of H.264 has referred only to the so-called Video Coding Layer (VCL). To transmit the encoded slices, they are packaged in the so-called Network Adaptation Layer (NAL) units 15a-15x (Figure 1). Each slice is contained in a single NAL unit 15. The sets of NAL 15 units make up Access Units (AUs). Each AU always contains a primary encoded image, that is, all the macroblocks of the image. It will be noted that the size of the data of a unit of NAL 15 varies according to the number of groups of slices, the type of coding (entropy) used to generate the slices, and so on. Therefore, the packet sequence of TS 7 contains NAL 15 units of varying lengths. Each set of NAL 15 units that forms an AU is transported in a Program Elemental Flow (PES) package 16 which, in turn, is transported by the TS 7 packets. The composition of the packet of PES 16 is shown in Figure 3. This includes a packet start code prefix 17a-d (see also Figure 2). The packet start code prefix 17 is a code with a length of twenty-four bits. In conjunction with a stream identification (stream d, 18), it constitutes a packet start code that identifies the beginning of a PES packet 16. In contrast to the packets of TS 7, which are of a standard fixed length, a pack of PES 16 has a variable length. The length is indicated in a packet length field of PES 19. This field 19 is optionally followed by a header of PES 20. Subsequently comes the payload of the pack of PES 21, ie the units of NAL 15. It is indicated that The stream identification (streamjd, 18) specifies both the type and the number of an elementary stream. Therefore, it can be used as an alternative to identify the elementary stream that should be partially masked by masquerade unit 5. The start of a new PES packet 16 always follows immediately after a packet header of TS 8, in which case the payload unit start indicator (payload_unit_start_indicator, in English) 11 has the value 'TRUE'. There are not two PES 16 packets that start in the same TS 7 packet. Figure 4 shows the composition of a NAL 15 unit in a format for transmission in a byte stream. It is preceded by a start code prefix 22. The start code prefix 22 is a unique three-byte sequence, equal to OxOO 0001 ', embedded in the byte stream as a prefix for each NAL 15 unit. therefore, the start code prefix 22 indicates a boundary between two subsequent NAL 15 units. The start code prefix 22 is followed by a prohibited bit (forbidden_bit, in English) 23, a storage identification component of NAL (nal_storage_idc, in English) 24 and a unit type of NAL (nal_unit_type, in English) 25 Ten different types are defined by the coding technique with which the slices have been formed. One of these is the type of Access Unit Delimiter (AUD, for its acronym in English). The presence of a NAL unit 15 of this type indicates the boundary between subsequent AUs. The first NAL 15 unit of an AU is always of this type. Another possible type is the type of Instant Decoder Recharge (IDR). Said NAL unit 15 contains the Incoded slice. Therefore, the corresponding NAL unit type value (nal_unit_type, in English) 25 indicates that the byte alignment data 26 that follows it, forms part of a data element that represents an uncoded image. Having recovered from the data stream 6 the elementary stream to be partially masked, the masking unit monitors the payloads of TS packets 9 of the TS 7 packets in that stream, in relation to the presence of the code prefix of start 22. Masquerade unit 5 is able to differentiate between the packet start code prefix 17 of a PES pack 16 and the start code prefix 22 indicating a boundary between subsequent NAL 15 units. It is noted that a PES packet start code prefix 17 always follows immediately after a packet header TS 8. Masquerade unit 5 is able to monitor portions of the packet payload of TS 9 separated by the header of the packet. packet of TS 8 by components that indicate the beginning of an elementary flow packet, that is, a PES packet. Once the end of a unit of NAL 15 has been identified, the system determines the unit type of NAL 15 indicated by the unit type of NAL (nal_unit_type, in English) 25. At least if the type corresponds to IDR , one or more of the TS 7 packets carrying data that are part of the NAL unit 15, is selected to be included in a sub-sequence. The payloads of TS packets 9 of the TS 7 packets in the sub-sequence are masked. It is noted that the term "subsequence" is conceptual. In the present it is used to refer to an ordered subset, ie a smaller number of the TS packets 7 in the sequence of the TS packets 7 that make up the elementary stream. Preferably, a series of additional TS 7 packets after the first TS packet 7 that carries data that is part of the selected NAL unit 15, is also selected to be included in the sub-sequence to be masked. These additional TS 7 packets are selected from the same PID-based elementary stream determined from the PID field 12. The system includes means for adjusting the number of additional packets that are selected. Therefore, the operator of a conditional access system can achieve a compromise between the level of content protection and the level of processing capacity that is required to unmask the partially masked stream. The fraction of masked TS 7 packets in the sequence that carries the elementary stream can be relatively low when the selection is based on the presence of NAL 15 units of IDR type. In such cases, additional TS 7 packets may be selected for masking. In an advantageous embodiment, the system is configured or is configurable to select units of types other than IDR randomly. The TS packet payload 7 of at least one of the TS 7 packets carrying the selected unit is also masked. The other types can be all or a subset of the nine defined types other than IDR. In one embodiment, the rate at which additional TS 7 packets are included in the sub-sequence for masking is continuously adjusted. This is done in such a way that the fraction of the TS packets 7 in the sub-sequence with respect to all the TS packets in the sequence representing the elementary stream at or below a certain level is maintained, e.g. 5%. The partially masked data stream is modulated by a transmitter 27. In a preferred embodiment, MPEG-2 transportation flow packets are modulated in Code Division Multiple Access (CDMA) carriers in accordance with system E variant of the Digital Audio Broadcasting standard (ETSI 300 401). Therefore, the transmitter transmits a signal that transports the flow of partially masked data. By using a wireless transmission technique, the partially masked data stream is made available to mobile receiver / decoder devices. These devices are provided with unmasking and decoding capabilities, preferably implemented in software. In this case, the signal, more particularly the way in which the data flow has been masked, allows the unmasking and decoding to be performed with a low processing intensity. It will be clear that the masking is performed at the level of the transport flow, ie masking at least part of the payloads of the TS 9 packets of the selected TS 7 packets. The header of the TS 8 packets, specifically the masked transport control field 13, provides an effective way to signal to the receiver whether a particular packet has a masked payload. Therefore, the receiver does not require to examine the complete content of each packet payload to selectively unmask the encoded data. Instead, you can perform unmasking before processing the contents of the actual payload. The invention is not limited to the embodiment described herein, but may be varied within the scope of the appended claims. For example, the encoded data elements may be transported in the payloads of the transport flow packets defined by other protocols for wireless transmission. In addition, the masking can be performed at the PES level, meaning that the payload of a PES packet 16 determined to carry a NAL 15 unit of a specific type (eg, IDR) is included among the payloads of the packets that They must be masked.

Claims (11)

NOVELTY OF THE INVENTION CLAIMS

1. - A method for partially masking a data stream 6 including transport flow packets 7, each transport flow packet 7 having a header 8 and a payload 9, wherein a sequence of transport flow packets 7 has loads tools carrying encoded data elements, arranged in units 15, including: selecting the transport flow packets 7 that form a sub-sequence of the sequence and masking the payloads 9 of each transport flow packet 7 in the sub-stream. sequencing, monitoring the payloads 9 of at least part of the transport flow packets 7 in sequence, in relation to the presence of data 22 indicating a boundary between two subsequent units 15 and, for the selected units 15, include at least one of the transport flow packets 7 that carry data that is part of the selected unit 15 in the sub-sec uencia.

2. The method according to claim 1, further characterized in that the data stream 6 is a multiplex of elementary streams, the method including identifying at least one elementary stream including the sequence of transport flow packets and monitoring only the payloads of the packets in the identified elementary stream (s).

3. The method according to claim 1 or 2, further characterized in that the selected units include units that contain at least part of a coded representation of an image.

4. The method according to any of the preceding claims, further characterized in that each unit 15 contains an indication 25 of the type of data to be followed and a part 26 containing that data, wherein the type of each unit in the loads Monitored tools is determined from the indication and the unit is included among the selected units if the type corresponds to at least one specific type.

5. The method according to claim 4, further characterized in that the units of types other than the specific type (s) are included randomly among the selected units.

6. The method according to claim 4 or 5, further characterized in that the types are defined by the coding technique with which the coded data elements have been formed.

7. The method according to any of claims 4 to 6, further characterized in that the encoded data elements are decodable using a predictive decoding technique and the specific types include a type of data element that allows a prediction to be derived only from the decoded data belonging to the data element.

8. The method according to any of claims 1 to 7, further characterized in that up to a maximum number of transport flow packets after a transport flow packet that transports data that is part of a selected unit, is included in the sub-sequence.

9. A system for partially masking a data stream 6 including transport flow packets 7, each transport flow packet 7 having a header 8 and a payload 9, wherein a sequence of transport flow packets 7 it has payloads carrying coded data elements, arranged in units 15, including: a port for receiving the data flow; and an arrangement 2-5 for processing the data in the stream, wherein the system is configured to select the transport flow packets 7 that form a sub-sequence of the stream and mask the payloads of each stream packet. transportation 7 in the sub-sequence, where the system is configured to monitor the payloads of at least part of the transportation flow packets in the sequence, in relation to the presence of data 22 indicating a boundary between two subsequent units as well as, for the selected units 15, including at least one of the transport flow packets transporting data that are part of the selected unit in the sub-sequence.

10. The system according to claim 9, further characterized in that it is configured to include up to a maximum number of transport flow packets after a first transport flow packet that transports data that is part of a selected unit in the sub-sequence, as well as provided with a provision to specify the maximum number.

11. A signal that transports a data stream 6 that includes transportation flow packets 7, each transport flow packet 7 having a header 8 and a payload 9, wherein a sequence of transport flow packets has loads tools 8 carrying coded data elements, arranged in units 15, each unit 15 being of a certain type, wherein the payload 9 of each transport flow packet 7 in a sub-sequence of the sequence is masked, wherein for each unit 15 of a type corresponding to a selected type, the sub-sequence includes at least one of the transport flow packets 7 that carry data that is part of that unit 15, wherein at least one of the packets The flow of transportation includes data 22 indicating a boundary between two subsequent units.