Note: Descriptions are shown in the official language in which they were submitted.
'~ ~' CA 02405902, 2002-10-O1
i ,
1
2
3
4
7
8
9
11
12 ' TIME DIVISION PARTIAL ENCRYPTION
13
14
CROSS REFERENCE TO RELATED DOCUMENTS
16 This application is related to U.S. provisional patent application serial
17 number 601296,673 filed June 6, 2001 to Candelore, et al. entitled "Method
for
18 Allowing Multiple CA Providers to Interoperate iri a Content Delivery
System by
19 Sending Video in the Clear for Some Content, and Dual Carriage of Audio and
Dual
Carriage of Video and Audio for Other Content"; and provisional patent
application
21 serial number 60/304,241 filed July 10, 2001 to Unger et al., entitled
"Independent
22 Selective fncryptions of Program Content for Dual Carriage", and
provisional
23 patent application serial number 60/304,131 filed July 10, 2001 to
Candelore et al.,
24 entitled "Method for Allowing Multiple GA Providers to Interoperate in a
Content
Delivery System by Partial Scrambling Content on a Time Slice Basis" and to
U.S.
26 provisional patent application serial no. 601 ; filed on October 26, 2001
27 to Candelore et al., entitled "Television Encryption Systems", docket
number SNY-
28 R4646P, which are hereby incorporated herein by reference.
Docket No.: SNY-84646.02 -1- PATENT
CA 02405902 2002-10-O1
o ,
1 ~ This application is being filed simultaneously with patent applications
2 docket number SNY-84646.01 entitled "Critical Packet Partial Encryption" to
Unger
3 et al., serial number ; docket number SNY-84646.03 entitled
4 "Elementary Stream Partial Encryption" to Candelore, serial number
' ; docket number SNY-84646.04 entitled "Partial Encryption and
and docket
6 PID Mapping to Unger et al., serial number
7 number SNY-84646.05 entitled "Decoding and Decrypting of Partially Encrypted
8 Information" to Unger et al., serial number These simultaneously
9 filed patent applications are hereby incorporated by reference herein.
11 COPYRIGHT NOTICE
12 A portion of the disclosure of this patent document contains material which
13 is subject to copyright protection. The copyright owner has no objection to
the
14 facsimile reproduction of the patent document or the patent disclosure, as
it
appears in the Patent and Trademark Office patent file or records, but
otherwise
16 reserves all copyright rights whatsoever.
17
18 FIELD OF THE INVENTION
19 This invention relates generally to the field of encryption systems. More
particularly, this invention relates to systems, methods and apparatus for
providing
21 partial encryption and decryption of digital of television signals.
22
23 BACKGROUND OF THE INVENTION
24 Television is used to deliver entertainment and education to viewers. The
source material (audio, video, etc.) is multiplexed into a combined signal
which is
26 then used to modulate a carrier. This carrier is commonly known as a
channel. (A
27 typical channel can carry one analog program, one or two high definition
(HD)
28 digital program(s), or several (e.g. nine) standard definition digital
programs.) In a
29 terrestrial system, these channels correspond to government assigned
frequencies
Docket No.: SNY-84646.02 -2- PATENT
a ~' CA 02405902 2002-10-O1
1 frequencies and are distributed over the air. The program is delivered to a
receiver
2 that has a tuner that pulls the signal from the air and delivers it to a
demodulator,
3 which in turn provides video to a display and audio to speakers. In a cable
system
4 the modulated channels are carried over a cable. There may also be an in-
band
or out-of-band feed of a program guide indicating what programs are available
and
6 the associated tuning information. The number of cable channels is finite
and
7 limited by equipment/cable bandwidth. Cable distribution systems require a
8 significant capital investment and are expensive to upgrade.
9 Much of television content is valuable to its producers, therefore copyright
holders want to control access and restrict copies. Examples of typically
protected
11 material include feature films, sporting events, and adult programming.
Conditional
12 access (CA) systems are used to control availability of programming in
content
13 delivery systems such as cable systems. CA systems come as matched sets-one
14 part is integrated into the cable system headend and encrypts premium
content, the
other part provides decryption and is built into the set-top boxes (STB)
installed in
16 user's homes. Several CA systems are used in the cable industry including
those
17 provided by NDS (Newport Beach, CA), Motorola (Schaumberg, IL) and
Scientific
18 Atlanta (Atlanta, GA). This matched set aspect of CA systems has the,
effect that
19 the "legacy" vendor is locked in as the supplier of additional STBs. Since
the
various technologies for conditional access are not mutually compatible (and
are
21 often proprietary), any new potential supplier is forced to license the
legacy CA.
22 Thus, the cable operator finds itself unable to acquire newer technology or
23 competing technology from other set-top box manufacturers since the
technology
24 owners are often unwilling to cooperate, or charge reasonable license fees.
This
inflexibility can be especially troublesome when cable companies with
disparate CA
26 systems are merged. Service providers would tike more than one source for
STBs
27 for any number of reasons.
28 Once a cable operator picks an encryption scheme, it is difficult to change
29 or upgrade the content encryption scheme without introducing a backward
Docket No.: SNY-84646.02 -3- PATENT
CA 02405902 2002-10-O1 f
1 compatible decoding device (e.g. set-top box). Providing multiple mode
capability
2 in new set-top boxes to handle multiple encryption systems can add
substantial cost
3 to any.new set-top box, providing that the technology can be made available
to the
4 STB vendor to provide the multiple.decryption capability.
The only known current option to avoiding domination by the legacy vendor
6 (short of wholesale replacement) is using "full dual carriages. Full dual
carriage
7 means that transmission is duplicated for each encrypted program - once for
each
8 type of CA encryption to be used. To provide full dual carriage, the headend
is
9 enhanced to provide each form of CA simultaneously. Legacy STBs should not
be
impacted and should continue to perform their function despite any change.
11 However, full dual carriage often comes at an unpalatable price because of
the
12 bandwidth impact, thus reducing the number of unique programs available.
13 Generally, the number of premium channels suffers so that the number of
options
14 available to the viewer are limited .and the value that can be provided by
the cable
operator is restricted.
16 A conventional cable system arrangement is depicted in FIGURE 1. In such
..
17 a system, the cable operator processes audiolvideo (AN) content 14 with CA
18 technology from manufacturer A (system A) using CA encryption equipment 18
19 compliant with system A at the cable system -headend 22. The encrypted A/V
content along with system information (S1) 26 and program specific information
21 (PSI) 27 is multiplexed together and transmitted over the cable system 32
to a
22 user's STB 36. STB 36 incorporates decrypting CA equipment from system A
23 (manufacturer A} 40 that decrypts the AN content. The decrypted AN content
can
24 then be supplied to a television set 44 for viewing by the user.
In a cable system such as that of FIGURE 1, digital program streams are
26 broken into packets for transmission. Packets for each component of a
program
27 (video, audio, auxiliary data, etc.) are tagged with a packet identifier or
PID. These
28 packet streams for each component of all programs carried within a channel
are
29 aggregated into one composite stream. Additional packets are also included
to
Docket No.: SNY-84646.02 -4- PATENT
CA 02405902 2002-10-O1
1 provide decryption keys and other overhead information. Otherwise unused
2 bandwidth is filled with null packets. Bandwidth budgets are usually
adjusted to
3 utilize about 95% of the available channel bandwidth.
4 Overhead information usually includes guide data describing what programs
are available and how to locate the associated channels and components. This
fi guide data is also known as system information or SI. S1 may be delivered
to the
7 STB in-band (part of the data encoded within a channel) or out-of-band
(using a
8 special channel dedicated to the purpose). Electronically delivered SI may
be
9 partially duplicated in more traditional forms - grids published in
newspapers and
magazines.
11 In order for a viewer to have a satisfying television experience, it is
generally
12 desirable that the viewer have clear access to both audio and video
content. Some
13 analog cable systems have used various filtering techniques to obscure the
video
14 to prevent an unauthorized viewer from receiving programming that has not
been
paid for. In such a system, the analog audio is sometimes sent in the clear.
In the
1fi Motorola VideoCipher2 Plus system used in C-band satellite transmissions,
strong
17 digital audio encryption is used in conjunction with a relatively weak
protection of
18 , the analog video (using sync inversion). In airline in-flight movie
systems, the
19 availability of audio only through rental of headphones has been used to
provide
the full audio and video only to paying customers.
21
22 ' BRIEF DESCRIPTION OF THE DRAWINGS
23 The features of the invention believed to be novel are set forth with
24 particularity in the appended claims. The invention itself however, both as
to
organization and method of operation, together with objects and advantages
26 thereof, may be best understood by reference to the following detailed
description
27 of the invention, which describes certain exemplary embodiments of the
invention,
28 taken in conjunction with the accompanying drawings in which:
29 FIGURE 1 is a block diagram of a conventional conditional access cable
Docket No.: SNY-84646.02 -5- PATENT
CA 02405902 2002-10-O1 ''
1 system.
2 FIGURE 2 is a block diagram of a system consistent with one embodiment
3 of the present invention in which dual encrypted audio is transmfitted along
with
4 clear video.
~ FIGURE 3 is a block diagram of a system consistent with an embodiment of
6 the present invention in which portions of programming are dual encrypted
7 according to a time slice mechanism.
8 FIGURE 4 is a flow chart of a dual encryption process consistent with
certain
9 embodiments of the present invention.
FIGURE 5 is a flow chart of a decryption process consistent with certain
11 embodiments of the present invention.
12 FIGURE 6 is a block diagram of a system consistent with an embodiment of
13 the present invention in which portions of programming are dual encrypted
on a
14 packet basis.
FIGURE 7 is a.flow chart of a dual encryption process consistent with certain
16 embodiments of the present invention.
17 FIGURE 8 is a flow chart of a decryption process consistent with certain
18 embodiments of the present invention.
19 FIGURE 9 is a block diagram of a system consistent with an embodiment of
' the present invention in which system information is encrypted and
programming
21 is sent in the clear.
22 FIGURE 10 is a block diagram of a generic system consistent with various
23 embodiments of the present invention. .
24 FIGURE 11 is a block diagram of a first embodiment of implementation of an
encryption system consistent with embodiments of the present invention in a
cable
26 system headend.
27 FIGURE 12 is a block diagram of a second embodiment of implementation
28 of an encryption system consistent with embodiments of the present
invention in a
29 cable system headend.
Docket No.: SNY-84646.02 -6- ' PATENT
CA 02405902 2002-10-O1 ~~
r
1 FIGURE 13 is a flow chart of an overall encryption process used to
2 implement certain embodiments of the present invention in a cable system
3 headend.
4 FIGURE 14 is a block diagram of a first embodiment of a set-top box
implementation of a decoding system consistent with embodiments of the
6 present invention.
7 FIGURE 15 is a block diagram of a second embodiment of implementation
8 of a decoding system consistent with embodiments of the present invention in
a
9 cable system STB.
FIGURE 16 is a block diagram of a third embodiment of implementation of
11 a decoding system consistent with embodiments of the present invention in a
12 cable system STB.
13 FIGURE 17 illustrates the PlD remapping process carried out in one
14 embodiment of a set-top box PfD re-mapper.
FIGURE 18 is a block diagram of an exemplary decoder chip that can be
16 utilized in a television set-top box consistent with the present invention.
17
18 DETAILED DESCRIPTION OF THE INVENTION
19 While this invention is susceptible of embodiment in many different forms,
there is shown in the drawings and will herein be described in detail specific
21 embodiments, with the understanding that the present disclosure is to be
22 considered as an example of the principles ofthe invention and not intended
to limit
23 the invention to the specific embodiments shown and described. In the
description
24 below, like reference numerals are used to describe the same, similar or
corresponding parts in the several views of the drawings. The terms "scramble"
26 and "encrypt" and variations thereof are used synonymously herein. Also,
the term
27 "television program" and similar terms can be interpreted in the normal
28 conversational sense, as well as a meaning wherein the term means any
segment
29 of A/V content that can be displayed on a television set or similar monitor
device.
Docket No.: SNY-84646.02 -7- PATENT
'. CA 02405902 2002-10-O1 I~
r ,
1 OVERVIEW ..
2 Modern digital cable networks generally use CA systems that fully encrypt
3 digital audio and video to make programming inaccessible except to those who
.
4 have properly subscribed. Such encryption is designed to thwart hackers and
non-
subscribers from receiving programming that has not been paid for. However,
as'
6 cable operators wish to provide their subscribers with set-top boxes from
any of
7 several manufacturers, they are frustrated by the need to transmit multiple
copies
8 ~ of a single program encrypted with multiple encryption technologies
compliant with
9 the CA systems of each STB manufacturer.
This need to carry multiple copies of the programming (called "full dual
11 caririage") uses up valuable bandwidth that could be used to provide the
viewer with,
12 additional programming content. Certain embodiments of the present
invention
13 address this problem in which the bandwidth requirements to provide an
equivalent
14 to multiple carriage are minimized. The result could be described as
"Virtual Dual
S
Carriage" since the benefits of full dual carriage are .provided without the
full
16 bandwidth cost. Several embodiments of the present invention are presented
17 herein to accomplish effective partial scrambling. These embodiments vary
by the
18 criteria used to select the portion to encrypt. The portion selected in tum
affects the
19 additional bandwidth requirements and the effectiveness of the encryption.
It may
be desirable to use one encryption process or several processes in combination
in
21 a manner consistent with embodiments of the present invention.
22 Certain of the implementations of partial dual encryption described herein
23 utilize an additional (secondary) PID for each duplicated component. These
24 secondary PIDs are used to tag packets that carry duplicated content with
an
additional encryption method. The PSI is enhanced to convey information about
the
26 ~ ~ existence these new PIDs in such a way that inserted PIDs are ignored
by legacy
27 STBs but can be easily extracted by new STBs.
28 Some implementations of partial dual encryption involve duplicating only
29 certain packets tagged with a given PID. Methods for selecting which
packets to
Docket No.: SNY-84648.02 -8- PATENT
CA 02405902 2002-10-O1 I.
1 encrypt are detailed hereinafter. The original (i.e. legacy) PID continues
to tag the
2 packets encrypted with legacy encryption .as wel l as other packets sent in
the clear.
3 The new PID is used to tag packets encrypted by the second encryption
method.
4 . Packets with the secondary PID shadow the encrypted packets tagged with
the
primary PID. The packets making up the encrypted pairs can occur in either
order
6 but, in the preferred implementation, maintain sequence with the clear
portion of the
7 PID stream. By use of the primary and secondary PIDs, the decoder located in
the
8 set-top box can readily determine which packets are to be decrypted using
the
9 decryption method associated with that set-top box, as will be clear upon
consideration of the following description. The processes used to manipulate
PIDs
11 will be described later in greater detail.
12 The encryption techniques described herein can be broadly categorized
13 (according to one categorization) into three basic variations - encrypting
just a
14 major portion (i.e. audio), encrypting just the St, and encrypting just
selected
packets. In general, each of the encryption techniques used iwthe embodiments
16 disclosed herein seek to encrypt portions of the an AN signal or associated
17 information while leaving other portions of the AN signal in the clear to
conserve
18 bandwidth. Bandwidth can be conserved because the same clear portion can be
19 sent to all varieties of set-top boxes. Various methods are used to select
the
portions of infbrmation to be encrypted. By so doing, the various embodiments
of
21 this invention eliminate the traditional "brute-force" technique of
encrypting the
22 entire content in one specific scrambling scheme, which predicates the
redundant
23 use of bandwidth if alternate scrambling schemes are desired. In addition,
each of
24 the partial dual encryption schemes described herein can be used as a
single
partial encryption scheme without departing from embodiments of the present
26 invention.
27 The various embodiments of the invention use several processes, alone or
28 in combination, to send substantial portions of content in the clear while
encrypting
29 only a small amount of information required to correctly reproduce the
content.
Docket No.: SNY-84848.02 -9- PATENT
CA 02405902 2002-10-O1
1 ~ ~ Therefore the amount. of. information transmitted that is uniquely
encrypted in a
2 particular scrambling scheme is a small percentage of the content, as
opposed to
3 the entire replication of each desired program stream. For purposes of the
4 exemplary systems in this document, encryption system A wial be considered
the
legacy system throughout. Each of the several encryption techniques described
6 above will now be described in detail.
7 The various embodiments of the invention allow each participating CA
8 system to be operated independently. Each is orthogonal to the other. Key
sharing
9 in the headend is not required since each system encrypts its own patents.
Different
key epochs may be used by each CA system. For example, packets encrypted with
11 Motorola's proprietary encryption can use fast changing encryption keys
using the
12 embedded security ASIC, while packets encrypted with NDS' smart card based
13 system use slightly slower changing keys. This embodiment works equally
well for
14 Scientific Atlanta and Motorola legacy encryption.
16 ENCRYPTED ELEMENTARY STREAM
17 Turning now to FIGURE 2, one embodiment of a system that reduces the
18 need for additional bandwidth to provide multiple carriage is illustrated
as system
19 100. In this embodiment, the system takes advantage of the fact that
viewing
television programming withaut audio is usually undesirable. While there are
21 exceptions (e.g., adult programming, some sporting events, etc.), the
typical viewer
22 is unlikely to accept routine viewing of television programming without
being able
23 to hear the audio. Thus, at headend 122, the video signal 104 is provided
in the
24 clear (unencrypted) while the clear audio 106 is provided to multiple CA
systems
for broadcast overthe cable network. In the exemplary system 100, clear audio
106
26 is provided to an encryption system 118 that encrypts audio data using
encryption
27 system A (encryption system A wil I be considered the legacy system
throughout this
28 document). Simultaneously, clear audio 1 ~ is provided to encryption system
124
29 that encrypts the audio data using encryption system B. Clear video is than
Docket No.: SNY-84646.02 -1 O- PATENT
CA 02405902 2002-10-O1 ~ -
1 multiplexed along with encrypted audio from 118 (Audio A) and encrypted
audio
2 from 124 (Audio B), system information 128 and program specific information
129.
3
4 After distribution through the cable system 32, the video, system
information,
program specific information, Audio A and Audio B are al l delivered to set-
top boxes
6 36 and 136. At legacy STB 36, the video is displayed and the encrypted audio
is
7 decrypted at CA system A 40 for play on television set 44.' Similarly, at
new STB
8 136, the video is displayed and the encrypted audio is decrypted at CA
system B
9 140 for play on television set 144.
Audio has a relatively low bandwidth requirement compared with a complete
11 AN program (or even just the video portion). The current maximum bit rate
for
12 stereophonic audio at 384 Kblsecond is approximately 10°l° of
a 3.8Mb/second
13 television program. Thus, for dual carriage of only encrypted audio (with
video
14 transmitted in the clear) in a system with ten channels carried with 256
QAM
(quadrature amplitude modulation); a loss of only about one channel worth of
16 bandwidth would occur. Therefore, approximately nine channels could be
carried.
17 This is a dramatic improvement over the need to dual encrypt all
channels,.which
18 would result in a decrease in available channels from ten to five. Where
deemed
19 necessary, e.g., sporting events, pay per view, adult programming, etc.,
dual
encryption of both audio and video can still be carried out, if desired.
21 Both legacy and new set-top boxes can function in a normal manner
22 receiving video in the clear and decrypting the audio in the same manner
used for
23 fully decrypting encrypted AN content. if the user has not subscribed to
the
24 programming encrypted according to the above scheme, at best the user can
only
view the video without an ability to hear the audio. For enhanced security
over the
26 video, it possible to employ other embodiments of the invention (as will be
27 described later) here as well. (For example, the SI may be scrambled to
make it
28 more difficult for a non-authorized set-top box to tune to the video
portion of the
29 program.) Unauthorized set-top boxes that have not been modified by a
hacker, will
Docket No.: SNY-84646.02 -11- PATENT
CA 02405902 2002-10-O1
1 blank the video as a result of receipt of the encrypted audio.
2 Authorized set-top boxes receive Entitlement Control Messages (ECM) that
3 are used to get access criteria and descrambling keys. The set-top box
attempts
4 to apply the keys to video as well as the audio. Since the video is not
scrambled,
it simply passes through the set-top boxes' descrambler unaffected. The set-
top
6 boxes do not care that the video is in-the-clear. The un-modified and un-
subscribed
7 set-top boxes behave as being un-authorized for the scrambled audio as well
as the
8 clear video. The video, as well as the audio which was actually scrambled,
will be
9 blanked. An on-screen display may appear on the TV stating that the viewer
needs
'to subscribe to programming. This desirably totally inhibits the casual
viewer from
11 both hearing and viewing the content.
12 fn one embodiment of the present invention, the encrypted audio is
13 transmitted as digitized packets overthe AIV channel. Two (or more) audio
streams .
14 are transmitted encrypted according to the two (or more) encryption systems
in use
by the system's set-top boxes. In order for the two (or more) STBs to properly
16 decrypt and decode their respective audio streams, SI (system information)
data are
17 transmitted from the cable system's headend 122 that identifies the
particular
18 channel where the audio can be found using a transmitted Service Identifier
to
19 locate the audio. This is accomplished by assigning the audio for system A
is a first
packet identifier .(PID) and assigning the audio for system B a second packet
21 identifier (PID). By way of example, and not limitation, the following
program
22 specific information (PSI) can be sent to identify the location of the
audio for two
23 systems, one using NDS conditional access and one using Motorola
conditional
24 access. Those skilled in the art will understand how to adapt this
information to the
other embodiments of partial encryption described later herein.
26 The SI can be separately delivered to both legacy and non-legacy set-top
27 - boxes. It is possible to send SI information so that the legacy and non-
legacy set-
28 top boxes operate essentiallywithout interference. 1n the SI delivered to
legacy set-
29 top boxes, the VCT (virtual channel table) would state that the desired
program, e.g.
Docket No.: SNY-84646.02 -12- PATENT
CA 02405902 2002-10-O1
s v
1 HBO referenced as program number 1, is on Service ID "1" and that the VCT
2 access control bit is set. The network information table (NIT) delivered to
that first
3 STB would indicate that Service ID "1" is at frequency = 1234. In the Sl
delivered
4 to non-legacy set-top boxes, the VCT would state that the desired program,
e.g.
HBO referenced as program number 1001, is on Service ID "1001" and that the
6 VCT access control bit is set. The network information table delivered to
the non-
? legacy STB would indicate that the Service ID "1001" is at frequency 1234.
The
8 following exemplary program association Table PS1 data are sent to both
legacy
9 and non-legacy set-top boxes (in MPEG data structure format):
11
Docket No.: SNY-84648.02 -13- PATENT
CA 02405902 2002-10-O1
1
2 PAT sent on PID=0x0000
3 PAT 0x0000
4 - Transport Stream ID
- PAT version
6 - Program Number 1
7 - PMT 0x0010
8 ' - Program Number 2
9 - PMT 0x0020
- Program Number 3
11 - PMT 0x0030
12 - Program Number 4
13 - PMT 0x0040
14 - Program Number 5
- PMT 0x0050
16 - Program Number 6
1 T - PMT 0x0060
18 - Program Number 7
19 - PMT 0x0070
- Program Number'8
21 - PMT 0x0080
22 - Program Number 9
23 - PMT 0x0090
24 - Program Number 1001
- PMT 0x1010
26 - Program Number 1002
27 - PMT 0x1020 '
28 - Program Number 1003
29 - PMT 0x1030
- Program Number 1004 , .
31 - PMT 0x1040
32 - Program Number 1005
33 - PMT 0x1050
34 - Program Number 1006
- PMT 0x1060
36 - Program Number 1007
37 - PMT 0x1070
38 - Program Number 1008
39 - PMT 0x1080
.
- Program Number 1009
41 - PMT 0x1090
42
43 The following exemplary program map table PSI data are selectively
44 received by legacy and non-legacy set-top boxes (in MPEG data structure
format):
Docket No.: SNY-84646.02 -14- PATENT
CA 02405902 2002-10-O1
i
1
2 . PMT sent on PID=0x0010
3 PMT 0x0010
4 PMT Program number 1
PMT Section Version 10
6 PCR PID 0x0011
7 Elementary Stream
8 - Stream Type (Video 0x02 or 0x80) .
9 - Elementary PID (0x0011)
- Descriptor
11 - CA Descriptor (ECM) for CA provider #1
12 Elementary Stream
13 - Stream Type (Audio 0x81)
14 - Elementary PID (0x0012)
- Descriptor
16 - CA Descriptor (ECM) for CA provider #1
17
18 ~ PMT sent on P!D=0x1010
19 PMT 0x1010
PMT Program number 1010
21 PMT Section Version 10
22 PCR PID 0x0011
23 Elementary Stream
24 - Stream Type (Video 0x02 or 0x80)
- Elementary PID (0x0011)
26 - Descriptor
27 - CA Descriptor (ECM) for CA provider #2
28 Elementary Stream
29 - Stream Type (Audio 0x81)
- Elementary PID (0x0013)
31 - Descriptor
32 . - CA Descriptor (ECM) for CA provider #2
33
34
36 Considering an example wherein it is desired to deliver programming in a
37 system using either Motorola or Scientific Atlanta as well as NDS CA, the
above
38 communications are consistent with the PSI delivered by both Motorola and
39 Scientific Atlanta in their CA systems, with only minor changes. The
program
association table (PAT) is changed to reference an additional program map
table
41 (PMT) for each program. Each program in this embodiment has two program
42 numbers in the PAT. In the table above, program number 1 and program number
43 1001 are the same program except that they wil I reference different audio
PIDs and
Docket No.: SNY-84646.02 -15- PATENT
CA 02405902 2002-10-O1
1 CA descriptors. Changes in the system to create multiple PMTs and to
multiplex
2 new PAT and PMT information with the data stream can be made to
appropriately
3 modify the cable system headend equipment. Again, those skilled in the art
will
4 understand how to adapt these messages to other partial encryption schemes
described herein. An advantage of this approach is that no special hardware or
6 software is required for headend or for legacy and non-legacy set-top boxes
to
7 deliver audio that is both legacy and non-legacy encrypted using this
scheme.
8 This technique deters the user from use of premium programming which has
9 not been paid for by rendering it inaudible, but a hacker may attempt to
tune the
video. To combat this, the mechanisms employed in other encryption techniques
11 consistent with the present invention (as will be described later) can be
employed
12 simultaneously, if desired. Since closed captioning is generally
transmitted as a
13 part of the video data, the user can still obtain readable audio
information in
14 conjunction with clear video. Thus, although adequate for some
applications, the
. present technique alone may not provide adequate protection in all
scenarios. In
16 another embodiment, video packets containing closed captioning information
as a
17 part of the payload can additionally be scrambled.
18 In an alternative embodiment, only the video may be dual encrypted with
19 , separate PIDs assigned to each set of encrypted video. While this may
provide a
more secure encryption for general programming (since video may be more
21 important than audio), the amount of bandwidth. savings compared with full
dual
22 carriage is only approximately ten percent, since only the audio is shared
amongst
23 all the set-top boxes. However, this approach might be used for certain
content,
24 ~ e.g. adult and sports, and help reduce the bandwidth overhead for that
content
while the audio encryption approach may be used for other content types. In
the
26 Digital Satellite Service (DSS) transport standard used for the DirecTVT""
service,
27 the ardio packets can be identified for encryption by use of the service
channel
28 identifier (SCID) which is considered equivalent.
29
Docket No.: SNY-84646.02 -16- PATENT
CA 02405902 2002-10-O1
r
1 TIME SLICING
2 ~ Another embodiment consistent with the present invention is referred to
3 herein as time slicing and is illustrated in FIGURE 3 as system 200. In this
4 embodiment, a.portion of each program is encrypted on a time dependent basis
in
a manner that disrupts viewing of the program unless the user has paid for the
6 programming. This embodiment of the invention can be implemented as
partially
7 encrypted video and clear audio, clear video and partially encrypted audio
or
8 partially encrypted video and audio. The duration of the time slice that is
encrypted,
9 taken as a percentage of the total time, can be selected to meet any
suitable
desired balance of bandwidth usage, security against hackers. In general,
under
11 any of the embodiments described herein, less than 100 percent of the
content is
12 encrypted to produce a desired partial encryption. . The following example
details
13 partially encrypted video and audio.
14 By way of example, and not limitation, consider a system which has nine
programs that are to be dual partially encrypted according to the present
exemplary
16 embodiment. These nine channels are fed to the cable headend as a
multiplexed
17 stream of packets and are digitally encoded using packet identifiers (P-ID)
to identify
18 packets associated with a particular one of the nine programs. In this
example,
19 assume that those nine programs have video PIDs numbered 101-109 and audio
PIDs numbered 201-209. The partial encryption, according to this embodiment is
21 time multiplexed among the programs so that only packets from a single
program
22 are encrypted at any given time. The method does not need to be content
aware.
23 With reference to TABLE 1 below, an exemplary embodiment of a time slice
24 dual encryption scheme consistent with an embodiment of the invention is
illustrated. For program 1 having primary video PID 101 and primary audio PID
201,
26 during the first time period, packets having PID 101 and PID201 are
encrypted
27 using encryption system A, while the others representing the other programs
are
28 sent in the clear. In this embodiment, secondary PIDs are also assigned to
both the
29 video and the audio. The secondary PlDs .are PID 111 for video and PID 211
for
Docket No.: SNY-84646.02 -17- PATENT
CA 02405902 2002-10-O1 y
1 audio respectively for program 1. The packets with the secondary PIDs are
2 encrypted using encryption system B during the first time period. The next
eight
3 time periods are sent in the clear. Then for time period 10, packets having
any of
4 the above four PIDs are again encrypted followed by the next eight time
periods
being sent in the clear. In a similar manner, during the second period of
program
6 2 having primary video PID 102 and primary audio PID 201 are encrypted using
7 encryption system A and packets with their associated secondary PIDs are
8 encrypted using encryption system B, and during the next eight time periods
are
9 sent in the clear, and so on. This pattern can be seen clearly in TABLE 1 by
10- examination of the first nine rows. Both audio and video packets, or audio
11 alone or video alone can be encrypted according to this technique, without
12 departing from the invention. Also, the audio and video can have their own
13 individual encryption sequence. In TABLE 1, P1 indicates time period number
1,
14 P2 indicated time period number 2 and so on. EA indicates that the
information is
encrypted using CA system A and EB indicates that the information is encrypted
16 using CA encryption system B.
Docket No.: SNY-84848.02 -1$- PATENT
CA 02405902 2002-10-O1
1
~ '
2 PROti.IDEO AUCNO P9 P4 P5 P6 P7 P8 P10P11P12
PID PID s ~ o
v ...
3 1 PID PID EA clearclearclearclearclearclearclearclearEA
clearclear...
101 201
4 2 PiD PID cl~rEA clearclearolearclearclearclearclearclearEA
clear...
102 202
3 PID PID clearclearEA clearclearclearclearclearclearclearclearEA .
103 203
4 PID PID clearclearclearEA
clearclearclearclearclearclearclearclear...
104 204
7 6 PID PID clearclearclearclearEA
clearclearclearclearclearclearclear...
105 205
$ 6 PID PID clearclearclearclearclearEA
clearclearclearclearclearclear...
106 206
9 7 PID PID clearclearclearclearcl~rclearEA
clearclearclearclearclear...
107 207
8 PID PID clearclearclearclearclearclearclearEA
clearclearclearclear...
108 208
11 8 PID PID clearclearclearclearclearclearclearclearEA
clearclearclear...
109 209
12 1 PID PID EB EB
111 211 ~
-
13 2 PID PID EB EB
112 212
14 S PID PID EB EB
113 213
4 PID PID EB
114 214
5 PID PID EB
115 215
17 ~ P1D PID EB
116 216
7 PID PID EB
117 217
19 8 PID PID EB
118 218
8 PID PID EB
119 219
21 TABLE 1
22 In orderto retain compatibility with an established legacy encryption
system
23 (encryption system A), the encrypted periods for each of programs one
through nine
24 are encrypted using encryption system A. Legacy STB equipment will accept
such ,
partially encrypted A/V data streams passing unencrypted packets and
decrypting
26 encrypted packets transparently. However, it is desired to obtain dual
encryption
27 using both encryption system A and encryption system B. fn order to achieve
this,
28 a specified program is assigned both primary PIDs {e.g., for program 1,
video PID
29 101 and audio PID 201 ) and a secondary PID (e.g., for program 1, video PID
111
and audio PID 211 ) to cant' the elementary data streams for a given premium
31 channel.
32 - With reference to L=IGURE 3, system 200 generally depicts the
functionality
33 of the cable system headend 222 wherein N channels of clear video 204 at
the
34 headend 222 are provided to an intelligent switch 216 (operating under
control of
a programmed processor) which routes packets that are to be transmitted in the
36 clear to be assigned a primary PID at 220. Packets that are to be encrypted
are
Docket No.: SNY-84846.02 -19- PATENT
CA 02405902 2002-10-O1
1 routed to both conditional access system A encrypter 218 and to conditional
access
2 system B encrypter 224. Once encrypted, these encrypted packets from 218 and
3 224 are assigned primary or secondary PIDs respectively at 220. System
4 information from 228 is multiplexed or combined with the clear packets, the
system
A encrypted packets and the system B encrypted packets and broadcast over the
6 cable system 32.
7 For discussion purposes, if the period of the time slice is 100 milli-
seconds,
8 then as shown in TABLE 1, there are on average one and a fraction encrypted
9 periods totaling 111 mini-seconds each second for all nine-programs. If the
period
is 50 milii-seconds, then there are on average two and a fraction encrypted
periods
11 totaling 111 milli-seconds. A non-subscribing box attempting to tune video
would
12 obtain a very poor image if it could maintain any sort of image lock and
the audio
13 would be garbled.
14 The PSI for a partially scrambled stream is handled slightly differently
from
the dual audio encryption example above. Essentially, the same Sf and PAT PSI
16 information can be sent to both legacy and non-legacy set-top boxes. The
17 difference lies with the PMT PSI information. The legacy set-top box parses
the
18 PMT PSI and obtains the primary video and audio PIDs as before. The non-
legacy
19 set-top box obtains the primary PIDs like the legacy set-top box but must
look at the
CA descriptors in the PMT PSI to see if the stream is partially scrambled. The
21 secondary PID is scrambled specifically for a particular CA provider,
consequently
22 it makes sense to use the CA descriptor specific to a particular CA
provider to
23 signal that PID. The invention can allow more than two CA providers to co-
exist by
24 allowing more than one secondary PID. The secondary PID shall be unique to
a
particular CA provider. The set-top box know the CA ID for the CA it has, and
can
26 check all CA descriptors for the relevant one for it.
27 v1/hile it is possible to send the secondary PID data as private data in
the
28 same CA descriptor used for the ECM, the preferred embodiment uses separate
CA
29 descriptors. The secondary PID is placed in the CA PID field. This allows
headend processing equipment to "see" the PID without having to parse the
private
Docket No.: SNY-84646.02 -20- PATENT
CA 02405902 2002-10-O1
1 data field of the CA descriptor. To tell the difference between the ECM and
2 secondary PiD CA descriptor, a dummy private data value can be sent.
3
4
PMT sent on PID=0x0010
6 PMT 0x0010
7 PMT Program number 1
8 PMT Section Version 10
9 PCR PID 0x0011
Elementary Stream
11 - Stream Type (Video 0x02 or 0x80)
12 - Elementary PID (0x0011)
13 - Descriptor
14 - CA Descriptor (ECM) for CA provider #1
- CA Descriptor (ECM) for CA provider #2
1 - CA Descriptor (Secondary PID) for CA provider
fi #2
17 Elementary Stream
18 - Stream Type (Audio 0x81)
19 - Elementary PID (0x0012)
- Descriptor
21 - CA Descriptor (ECM) for CA provider #1
22 - CA Descriptor (ECM) for CA provider #2
23 - CA Descriptor (Secondary PID) for CA provider
. #2
24
26 CA
Descriptor
for
CA
Provider
#2
(ECM)
29 Descriptor
Tag: Conditional Access (0x09)
31 Length: 4 Bytes
32 Data
33 - CA System ID:.Ox0942 (2"d CA provider)
34 - CA PID (0x0015)
36
37
Docket No.: SNY-84646.02 -21- PATENT
CA 02405902 2002-10-O1
1
CA Descriptor for CA Provider #2 (Secondary PID)
4 Descriptor
Tag: Conditional Access (0x09)
6 Length: 5 Bytes
7 Data
8 - CA System ID: 0x1234 (2"d CA provider)
9 - CA PID (0x0016)
- Private Data
11
12
13 Legacy STB 36 operating under CA system A receives the data, ignores
14 the secondary PIDs, decrypts the packets encrypted under CA system A and
presents the program to the television set 44. New or non-legacy STB 236
16 receives the SI 228. It receives PSI 229 and uses the PMT to identify the
17 primary and secondary PID, called out in the second CA descriptor,
associated
18 with the program being viewed. The packets encrypted under CA system A are
19 discarded and the packets encrypted under CA system B with the secondary
PID
are decrypted.by CA system B 240 and inserted into the clear data stream for
21 decoding and display on television set 244.
22 FIGURE 4 illustrates one process for encoding at the cable system headend
23 that can be used to implement an embodiment of the present invention
wherein CA
24 system A is the legacy system and CA system B is the new system to be
introduced.
As a clear packet is received, at 250 for a given program, if the packet (or
frame)
26 is not to be encrypted (i.e., it is not the current time slice for
encryption for this
27 program), the clear packet (C) is passed on to be inserted into the output
stream
28 at 254. .If the current packet is to be encrypted by virtue of the current
packet being
29 a part of the encryption time slice, the packet is passed for encryption to
both
packet encryption process A 258 and packet encryption process B 262. The
31 encrypted packets from encryption process A at 258 (EA) are passed on to
254 for
32 insertion into the output stream. The encrypted packets from encryption
process
33 . B at 262 (EB) are assigned a secondary PID at 264 for insertion into the
output
34 stream at 254. This is repeated for all packets in the program.
Docket No.: SNY-84646.02 -22- PATENT
CA 02405902 2002-10-O1
~, ~,
1 FIGURE 5 illustrates a process used in the STB 236 having the riewly
2 introduced CA system B for decrypting and decoding the received data stream
3 containing C, EAand EB packets having primary and secondary PIDs as
described.
4 When a packet is received at 272, it is inspected to see if it has a the
primary PID
of interest. if not, the packet is examined to see if it has the secondary PID
of
6 interest at 274. If the packet has neither the primary or secondary PID, it
is ignored
7 or dropped at 278. Any intervening packets between the EA and EB packets
that
8 are not the primary or secondary P1D are discarded. It is an implementation
and
9 mainly a buffering issue whether a decoder can receive multiple EA or EB in
a row
before receiving the replacement matched EA or EB packet. Also, just as easy
to
11 detect for secondary packets that come before and not after the primary
packet.
12 It is also possible to design a circuit where either case can happen - the
secondary
13 packet can before or after the primary packet. If the packet has the
primary PID of
14 interest, the packet is examined at 284 to determine if it is encrypted. If
not, the
packet (C) is passed directly to the decoder at 288 for decoding. If the
packet is
16 encrypted at 284, it is deemed to be an EA packet and is dropped or ignored
at 278.
17 In some implementations, the primary packet's encryption does not get
checked at
18 284. Rather, its simple position relative to the secondary packet can be
checked
19 ~ at 284 to identify it for replacement.
If the packet has the secondary PID at 274, the PID is remapped to the
21 primary PID at 292 (or equivalently, the primary PID is remapped to the
secondary
22 PID value). The packet is then decrypted at 296 and sent to the packet
decoder at
23 288 for decoding. Of course, those skilled in the art wilt recognize. that
many
24 variations are possible without departing from the invention, for example,
the order
of 292 and 296 or the order of 272 and 274 can be reversed. As mentioned
earlier,
26 284 can be replaced with a check of primary packet position with respect to
the
27 secondary packet. Other variations will occur to those skilled in the art.
28 Legacy STB 36 operating under the encryption system A totally ignores the
29 secondary PID packets. Packets with the primary PID are decrypted, if
necessary,
and passed to the decoder without decryption if they are clear packets. Thus,
a so
Docket No.: SNY-84646.02 -23- PATENT
CA 02405902 2002-10-O1
1 called "legacy" STB operating under encryption system Awill properly decrypt
and
2 decode the partially encrypted data stream' associated with the primary PID
and
3 ignore the secondary PID without modification. STBs operatingunder the
4 encryption system B ace programmed to ignore all encrypted packets
associated
with the primary PID and to use the encrypted packets transmitted with the
6 secondary PID associated with a particular channel.
7 Thus, each dual partially encrypted program-has two sets of PIDs associated
8 therewith. If, as described, the encryption is carried out on a period-by-
period
9 basis, for the system shown with an appropriate time slice interval, the
picture will
be essentially unviewable on a STB with neither decryption.
11 In order to implement this system in the headend 322 of FIGURE 6, the SI
12 and PSI can be modified for inclusion of a second set of CA descriptor
information.
13 ~ Legacy set-top boxes may not be able to tolerate 'unknown CA descriptors.
14 Consequently, alternatively, in the set-top box, it may be possible to
"hard code"
offsets from the legacy CA PIDs for both the content PIDs andlor the Si/PSl
and
16 ECM PIDs. Alternatively, parallel PSI may be sent. For example; an
auxiliary PAT
17 can be delivered on PID 1000 instead of PID 0 for the non-legacy set-top
boxes. It
18 can reference auxiliary PMTs not found in the legacy PAT. The auxiliary
PMTs can
19 contain the non-legacy CA descriptors. Since auxiliary PMTs would not be
known
to the legacy set-top boxes, there would not be any interoperation issue.
21 In systems where system A corresponds to legacy set-top boxes
22 manufactured by Motorola or Scientific Atlanta, no modifications to the
STBs are
23 required. For the system B compliant STBs, for dual carriage of partially
encrypted
24 programs as described herein, the video and audio decoder are adapted to
listen
to two PIDs each (a primary and a secondary PID) instead of just one. There
may
26 be one or more secondary shadow PIDs, depending on the number of non-Legacy
27 CA systems in use, however a specific set-top box only listens to one of
the
28 secondary PIDs as appropriate for the CA method being used by that specific
STB.
29 In addition, ideally the encrypted packets from the PID carrying the mostly
clear
video or audio are ignored. Since ignoring "bad packets" (those that cannot be
Docket No.: SNY-R464fi.02 -24- PATENT
CA 02405902 2002-10-O1
1 readily decoded as is) may already be a function that many decoders perform,
thus
2 requiring no modification. For systems with decoders that do not ignore bad
3 packets, afiltering function can be used. It should be understood that the
time slice
4 encryption technique could be applied to just the video or the audio. Also,
the video
may be time slice .encrypted while the audio is dual encrypted as in the
earlier
6 embodiment. The time slice technique may be applied to multiple programs
7 concurrently. The number of programs that encrypted during a period of time
is
8 mainly an issue of bandwidth allocation, and although the example discusses
9 scrambling a single program at a time, the invention is not limited by that.
Other
combinations of encryption techniques described in this document will also
occur
11 to those skiNed in the art.
12
13
14 MT" AND N PACKET ENCRYPTION
Another embodiment consistent with the preserit invention is referred to
16 herein as M~' & N packet encryption. This is a variation of the embodiment
17 illustrated in FIGURE 3 as system 200. In this embodiment, packets of each
PID
18 representing a program are encrypted in a manner that disrupts viewing of
the
19 program unless the user has paid for the programming. In this embodiment, M
represents the number of packets between the start of an encryption event. N
21 represents the number of packets that are encrypted in a row, once
encryption
22 takes place. N is less than M. If M=9 and N=1, then every nine packets
there is an
23 encryption event lasting 1 packet. If M=16 and N=2, then every sixteen
packets
24 there is an encryption event lasting two packets. Each packet to be dual
partially
encrypted is duplicated and processed using CA system A 218, and CA system B
26 224 as in the previous embodiment. The difference in operation between this
27 embodiment and the time slicing technique previously is in the operation of
switch
28 216 to effect the selection of packets to encrypt under control of a
programmed
29 processor.
Docket No.: SNY-84848.02 -25- PATENT
CA 02405902 2002-10-O1
1 By way of example, and not limitation, consider a system which has nine
2 channels of programming that are to be dual encrypted according to the
present
3 exemplary embodiment. These nine channels are digitally encoded using packet
4 identifiers (PID) to identify packets associated with a particular one of
nine
programs. In this example, assume that those nine programs have video PIDs
6 numbered 101-109 and audio PIDs numbered 201-209. The encryption, according
7 to this embodiment is random program-to-program so that packets from other
8' programs may be encrypted at the same time. This is illustrated in TABLE 2
below
9 in which M=6 and N=2 and in which only video is encrypted, but this should
not be
considered limiting. The method does not need to be content aware. In TABLE 2,
11 PK1 indicated packet number 1, PK2 indicates packet number 2, and so on.
12
Docket No.: SNY-84646.02 -26- PATENT
CA 02405902 2002-10-O1
1
2 PRO(3.VIDEOPK1PK2 PK3 PK4 PKb PlCBPKT PK8 PK8 PK10PK11PK12
3 1 PID EA EA chr clearclearGearEA EA clearclearclear' ...
101 cl~r
4 2 PID chrclearclearEA EA dearclearclearclearEA EA clear...
102
3 PID clearclearEA EA clearclearchr clearEA EA cl~rchr ...
103
4 PID clearclearclearEA EA clearclearclearchr EA EA chr ...
104
7 5 RID clearclearEA EA clearclearclearcl~rEA EA clearclear..
105
6 PID EA chr clearclearclearEA EA clearcV~rcl~rchr EA ...
106
7 PID EA EA clearclearclearclearEA EA chr clearclearcl~r ...
107
8 PlD clearEA EA clearclearGearclearEA EA clearchr clear...
108
1 1 9 PID EA clearclearclearclearEA EA clear~ GearclearEA ...
109 clear
12 1 PID EB EB EB EB ...
111
'~ 3 2 PID EB EB EB EB ...
112
14 3 PID EB EB EB EB ...
113
4 PID EB EB EB EB ...
114
16 5 PID EB EB EB EB ...
115
17 6 PID EB EB EB EB ...
116
18 7 PID EB EB EB EB ...
117 .
19 8 P1D EB ~EB EB EB ...
118
9 PID EB EB EB EB ...
19
21 TABLE 2
22
23 In the example of TABLE 2, each program is encrypted fully independently
24 of the others using the M=6 and N=2 encryption scheme. Again, the
illustrated
example encrypts only the video, but audio could also be encrypted according
to this
26 or another arrangement. If applied to just the video, audio may be dual
scrambled
27 or time slice encrypted as in earl ier embodiments. Alternatively, if
applied to just the
28 audio, the video may be time sliced as in the earlier embodiment.
29 Those skilled in the art will recognize that many variations of the
technique
can be devised consistent with the partial scrambling concepts disclosed
herein.
31 For example, a pattern of five clear followed by two encrypted followed by
two clear
32 followed by one encrypted (CCCCCEECCECCCCCEECCE...) is consistent.with
Docket No.: SNY-84648.02 -27- PATENT
CA 02405902 2002-10-O1
1 variations of the present partial encryption concept, as are random, pseudo-
random
2 and semi-random values for M and N may be used for selection of packets to
3 encrypt. Random, pseudo-random or semi-random (herein collectively referred
to
4 as "random" herein) selection of packets can make it difficult for a hacker
to
algorithmically reconstruct packets in a post processing attempt to recover
recorded
6 scrambled content. Those skilled in the art will understand how to adapt
this
7 information to the other embodiments of partial encryption described later
herein.
8 Some of the embodiments can be used in combination to more effectively
secure the
9 content.
11 DATA STRUCTURE ENCRYPTION
12 Another partial encryption method consistent with embodiments of the
present
13 invention uses a data structure as a basis for encryption. By way of
example and
14 not limitation, one convenient data structure to use for encryption is an
MPEG video
frame. This is illustrated (again with video only) in TABLE 3 below in which
every
16 tenth video frame is encrypted. In this embodiment, each program's ten
frame
17 encryption cycle is distinct from each other channel, but this should not
be
18 considered limiting. This concept can be viewed as a variation of the time
slice or
19 M~' and N partial encryption arrangement (or other pattern) based upon
video or
audio frames (or some other data structure) with the exemplary embodiment
having
21 M=10 and N=1. Of course, other values of M and N can be used in a similar
22 embodiment. fn TABLE 3, F1 represents frame number 1, F2 represents frame
23 number 2 and so on.
24
Docket No.: SNY-84646.02 -28- PATENT
CA 02405902 2002-10-O1
1
2 PROG.VIDEOF1 F2 F3 F4 F5 F6 F7 F8 F8 F10 F11 F12
3 1 PID EA clearGearclearclearclearclearclearclearclearEA clear...
101
4 2 PID clearclearclearEA chr clearclearclearclearclearclearclear...
102
3 PlD clearclearEA clearclearclearclearclearclearclearclearclear...
103
6 4 PID clearclearclearclearEA
clearclearclearclearclearclearclear...
104
7 5 PID clearclearclearEA
clearclearclearclearclearclearclearclear...
105
8 6 P1D EA clearclearclearclearclear- clearclearclearEA clear...
106 clear
9 7 PID chrEA clearclearclearclearclearclearclearclearclearEA ...
107
8 PID clearEA clearclearGar clearclearclearchr clearclearFJ1 ...
108
11 9 PID EA clearclearclearGearclearchr clearclearclearEA clear...
109
12 1 PID EB EB ...
111
13 2 PID EB ...
112
14 3 PID EB ...
113
4 PID EB ...
114
16 5 PID EB ...
115
17 6 PID EB EB ...
116
18 7 PID EB EB ...
117
19 8 PID EB EB ...
118
9 PID EB EB ...
119
21 TABLE 3
22
23 Thus, again each encrypted program has two sets of PIDs associated
24 therewith. If, as described, the encryption is carried out on a period-by-
period basis,
for the system shown, the picture will be essentially unviewable. For a nine
program
26 system at 30 frames per second as depicted, approximately three frames per
second
27 will be encrypted. For viewers who are not entitled to view the program,
their STB
28 will be unable to capture much more than an occasional frozen frame as the
STB
29 constantly attempts to synchronize and recover. Viewers who have subscribed
to
the programming will be able to readily view the programming. The bandwidth
cost
31 for such an encryption arrangement depends upon the frequency with which
the
32 encryption is applied. In the above example, an extra factor of 1/9 of data
are
Docket No.: SNY-84646.02 -29- PATENT
CA 02405902 2002-10-O1
1 transmitted for each program. .In this example, approximately one program's
worth
2 of bandwidth is used. With a greater number of programs, fewer packets per
3 program are encrypted and the security of the encryption system may degrade
4 somewhat. As in the randomized M and N method, random.frames may be
selected.
Choosing random frames, in the video case, would help guarantee that all frame
6 types would be affected - intra-coded frames (I frames), predictive-coded (P
7 frames), Bi-directional-coded (B frames) and DC frames.
8 In a variation of the invention, it may be possible to encrypt fewer packets
to
9 achieve an acceptable level of security: That is, perhaps in a system of
nine
programs, only one frame per second may need to be encrypted to achieve
11 acceptable levels of security. In such a system, the overhead becomes one
12 encrypted period per second per program or approximately 1 /30 of data
transmitted
13 in overhead. This level of overhead is a dramatic improvement over the 50%
loss
14 of bandwidth associated with full dual carriage of encryption under two
encryption
systems. In another variation of the invention, it may be possible to encrypt
only
16 certain video frames to achieve an acceptable level of security. For
example, for
17 MPEG content, only intra-coded frames (I frames) may be scrambled to
further
18 reduce the bandwidth overhead and still maintain an acceptable level of
security.
19 These offer significant improvement over the bandwidth required for full
dual
carriage.
21
22 CRITICAL PACKET ENCRYPTION
23 Substantial efficiency in bandwidth utilization can be achieved by use of a
24 selective packet-by-packet dual encryption technique. In this technique,
packets are
selected for encryption based upon their importance to the proper decoding of
the
26 audio and/or video of the program content.
27 This embodiment can reduce the bandwidth requirement compared with full
28 dual carriage of encrypted content by only scrambling a small fraction of
the packets.
29 Clear packets are shared between the two (or more) dual carriage PIDs: In
one
preferred embodiment, as will be disclosed, less that about one percent of the
total
Docket No.: SNY-84646.02 -30- PATENT
CA 02405902 2002-10-O1
1 content bandwidth is used. In a system with a legacy encryption scheme,
clear '
2 program content packets can be received by both legacy and new set-top
boxes.
3 As mentioned before, encrypted packets are dual carried and processed by the
4 respective set-top boxes with the appropriate CA. Each CA system is
orthogonal.
Key sharing is not required and different key epochs may be used by each CA
6 system. For example, a system with Motorola's proprietary encryption can
generate
7 fast changing encryption keys using the embedded security ASIC, while an NDS
8 smart card based system can generate slightly slower changing keys. This
9 embodiment works equallywell for ScientificAtlanta and Motorola legacy
encryption.
Referring now to FIGURE 6, a block diagram of a system consistent with an
11 embodiment of the present invention in which portions of programming are
dual
12 encrypted on a packet-by-packet basis is illustrated as system 300. In this
system,
13 packets of each program are dual encrypted using, for example, legacy CA
system
14 A and CA system B. The packets that are encrypted are selected based upon
their
importance to the proper decoding of the video and/or audio stream.
16 In the system illustrated in FIGURE 6, the cable system headend 322 selects
17 A/V content 304 packets at a packet selector 316 for encryption. Packets
selected
18 for encryption are chosen so that their non-receipt (by a non-paying
decoder) would
19 severely affect the real-time decoding of a program, and any possible post
processing of recorded content. That is, only critical packets are encrypted.
For the
21 video and audio, this can be accomplished by encrypting "start of frame"
transport
22 stream packets containing PES (packetized elementary stream) headers and
other
23 headers as part of the payload, since without this information, the STB
decoder
24 cannot decompress the MPEG compressed data. MPEG2 streams identify "start
of
frame" packets with the "Packet Unit Start Indicator" in the transport header.
26 Generally, packets carrying a payload that contains a group of pictures
header or
27 a video sequence header can be used to effect the present scrambling
technique.
28 MPEG (Moving Pictures Expert Group) compliant compressed video
29 repackages the elementary data stream into the transport stream in somewhat
arbitrary payloads of 188 bytes of data. As such, the transport stream packets
Docket No.: SNY-R4fi46.02 -31- PATENT
CA 02405902 2002-10-O1
1 containing a PES header can be selected for encryption at selector 316 and
dual
2 encrypted by both the CA system A encrypter 318 and the CA system B
encrypter
3 324. Packets to be dual partially encrypted are duplicated and the PIDs of
duplicate
4 packets encrypted by encrypter 324 are remapped at 330 to a secondary PID as
in
the previous embodiment. The remaining packets are passed in the clear. The
6 clear packets, system A encrypted packets, system B encrypted packets and
system
7 information 328 are multiplexed together for broadcast over the cable system
32.
8 As with the previous system, the legacy STB 36 receives clear data and data
9 encrypted under CA encryption system A and transparently passes unencrypted
data combined with data decrypted by CA decryption A 40 to its decoder. In the
new
11 STB 336, the program is assigned to both a primary and a secondary PID. The
clear
12 packets with the primary PID are received and passed to the decoder. The
13 encrypted packets with the primary PID are discarded. Encrypted packets
with the
14 secondary PID are decrypted and then recombined with the data stream (e.g.,
by
rernapping the packets to the primary PID) for decoding.
16 Using video is used as an example, each sample is known as a frame and the
17 sample rate is typically 30 frames per second. If the samples are encoded
to fit into
18 3.8 Mbps, each frame would occupy 127K bits of bandwidth. This data is
sliced for
19 MPEG transport into packets of 188 bytes with the first packets) of each
frame
containing the header used for instructions to process the body, of the frame
data.
21 Dual encrypting just the first header packet (1504 additional bits)
requires only 1.2%
22 (1504/127K) of additional bandwidth. For high definition (19 Mbps) streams
the
23 percentage is even less.
24 As previously stated, transport stream packets containing a PES header are
the preferred target for encryption according to the present embodiment. These
26 packets contain sequence headers, sequence extension headers, picture
headers,
27 quantization and other decode tables that also fall within the same packet.
If these
28 packets cannot be decoded (i.e., by a hacker attempting to view
unauthorized
29 programming without paying the subscription charges), not even small
portions of
the program can be viewed. In general, any attempt to tune to the program will
likely
Docket No.: SNY-84646.02 -32- PATENT
CA 02405902 2002-10-O1
1 be met with a blank screen and no audio whatsoever since known decoder
2 integrated circuits use the PES header to sync up to an elementary stream
such as
3 video and audio in real-time. By encrypting the APES header, the decoding
engine
4 in an un-authorized set-top box cannot even get started. Post processing
attacks,'
e.g. on stored content, are thwarted by critical dynamically changing
information in
6 the packet containing the PES header. Those skilled in the art will
appreciate that
7 for implementation of this embodiment of the invention, other critical or
important
8 packets or content elements may also be identified for encryption that could
severely
9 inhibit unauthorized viewing without departing from the present invention.
For
example, MPEG intra-coded or I frame picture packets could be encrypted to
inhibit
11 viewing of the video portion of the program. Embodiments the present
invention may
12 be used in any combination with other embodiments, e.g. scrambling the
packet
13 containing the PES header as well as random, M~' and N, or data structure
14 encryption of the other packets. Critical packet encryption may be applied
to video
encryption, while a different method may be applied to audio. Audio could be
dual
16 encrypted, for instance. Other variations within the scope of the present
invention
17 will occur to those skilled in the art.
18 FIGURE 7 is a flow chart depicting an exemplary encoding process such .as
19 that which would be used at headend 322 of FIGURE 6. When a transport
stream
packet is received at 350, the packet is examined to determine if it meets a
selection
21 criteria for encryption. In the preferred embodiment, this selection
criteria is the
22 presence of a PES header as a portion of the packet payload. If not, the
packet is
23 passed as a clear unencrypted packet (C) for insertion into the output data
stream
24 at 354. If the packet meets the criteria, it is encrypted under CA
encryption system
A at 358 to produce an encrypted packet EA. The packet is also duplicated and
26 encrypted under CA encryption system B at 362 to produce an encrypted
packet.
27 This encrypted packet is mapped to a secondary PID at 366 to produce an
28 encrypted packet EB. Encrypted packets EA and EB are inserted into the
output
29 data stream along with clear packets C at 354. Preferably, the EA and EB
packets
are inserted at the location in the data stream where the single original
packet was
Docket No.: SNY-84646.02 -33- PATENT
CA 02405902 2002-10-O1
1 obtained for encryption so that the sequencing of the data remains
essentially the
2 same.
3 When the output data stream from 354 is received at an STB compliant with
4 CA encryption system B such as 336 of FIGURE 6, a process such as that of
FIGURE 8 (which is similar to that of FIGURE 5) can be utilized to decrypt and
6 decode the program. When a packet is received having either the primary or
the
7 secondary PID at 370, a determination is made as to whether the packet is
clear (C)
8 or encrypted under system A (EA} at 370 or encrypted under system B (EB) at
374.
9 If the packet is clear, it is passed directly to the decoder 378. In some
embodiments,
the relative position of the primary packet, before or after, to the secondary
packet
11 may be used to signal a primary packet for replacement in the stream. A
check of
12 the scrambling state of the primary packet is not specifically required. If
the packet
13 is an EA packet, it is dropped at 380. If the packet is an EB packet, it is
decrypted
14 at 384. At this point, the secondary PID packets and/or the primary PID
packets are
remapped to the same PID at 388. The decrypted and clear packets are decoded
16 at 378.
17 The dual partial encryption arrangement described above can greatly reduce
18 the bandwidth requirements over that required for full dual carriage.
Encrypting the
19 PES header information can be effective in securing video and audio
content, while
allowing two or more CA systems to independently "co-exist" on the same cable
21 system. Legacy system A set-top boxes are un-affected, and system B set-top
22 boxes require only an minor hardware, firmware, or software enhancement to
listen
23 for two PIDs each for video and audio. Each type of STB, legacy and non-
legacy,
24 retains its intrinsic CA methodology. Headend modification is limited to
selecting
content for encryption, introducing the second encrypter, and providing a
means to
26 mix the combination into a composite output stream.
27 In one embodiment, the headend equipment is configured to opportunistically
28 scramble as much of the content as the bandwidth will allow, and not just
the critical
29 PES headers. These additional scrambled packets would be either in the PES
Docket No.: SNY-84646.02 -34- PATENT
CA 02405902 2002-10-O1
( E
1 payload or other packets throughout the videolaudio frame to provide even
further
2 security of the content.
3
4 SI ENCRYPTION
Turning now to FIGURE 9, one embodiment of a system that minimizes
6 the need for any additional bandwidth is illustrated as system 400. In this
7 embodiment, the system takes advantage of the fact that system information
(S1) 428
8 is required for a set-top box to tune programming. In a cable system, SI is
sent in
9 the out-of-band, a frequency set aside from the normal viewing channels. It
is
possible to also sent it in-band. If sent in-band, the SI 428 is replicated
and sent
11 with each stream. For discussion purposes, assume that the SI delivered to
"legacy"
12 set-top boxes from previous manufacturers is separate from the SI delivered
to set-
13 tops from new manufacturers such as STB 436. Consequently, each version of
the
14 SI can be independently scrambled as illustrated using conditional access
system
A 418 and conditional access system B 424. The clear video 404 and clear audio
16 406 are delivered in the clear, but in order to understand how to find
them, the SI
17 information 428 is needed.
18 The SI delivers information about channel names and program guide
19 information such as program names and start times, etc. ... as well as the
frequency
tuning information for each channel. Digital channels are multiplexed.
together and
21 delivered at particular frequencies. In the embodiment of the invention,
the SI
22 information is encrypted, and only made available to authorized set-top
boxes. If the
23 SI information is not received to allow knowledge of the location of all
the AIV
24 frequencies in the plant, then tuning cannot take place.
To frustrate a hacker who might program a set-top box to trial or scan
26 frequencies, the frequencies for the channels can be offset from the
standard
27 frequencies. Also, the frequencies can be dynamically changed on a daily,
weekly
28 or other periodic or random basis. A typical cable headend may have roughly
30
29 frequencies in use. Each frequency is typically chosen to avoid
interference
between, among other things, each other, terrestrial broadcast signals, and
Docket No.: SNY-84646.02 -35- PATENT
CA 02405902 2002-10-O1
l,
1 frequencies used by clocks of the receiving equipment. Each channel has at
least
2 1 independent alternate frequency that if used would not could not cause
3 interference, or cause the frequency of adjoining channels to be changed.
The
4 actual possible frequency maps are therefore 23° or 1.07 x 109.
However, a hacker
might simply quickly try both frequencies on each tune attempt for each of the
30
6 channels or so. If successful in locating a frequency with content, the
hacker's set-
7 top box can then parse the PSI 429 to learn about the individual PIDs that
make up
8 a program. The hacker will have difficulty learning that "program 1" is
"CNN", and
9 that °program 5" is "TNN", and so on. That information is sent with
the SI, which as
'f 0 stated above is scrambled and otherwise unavailable to the un-authorized
set-top
11 box. However, a persistent hacker might yet figure those out by selecting
each one
12 and examining the content delivered. So in order to frustrate the
identification of
13 channels, the assignment of a program within a single stream can move
around, e.g.
14 program 2 and program 5 swapped in the example above so that "program 1" is
"TNN" and "program 5" is "CNN". Also, it is possible to move programs to
entirely
16 different streams with entirely new program groupings. A typical digital
cable
17 headend can deliver 250 programs of content including music. Each can be
18 uniquely tuned. The possible combinations for re-ordering are 250!
(factorial).
19 Without a map of the content provided by either the delivered SI or by a
hacker, the
user is faced with randomly selecting each program in a stream to see if it is
the one
21 interest.
22 Thus, at headend 422, the video signal 404 and the audio signal 406 are
23 provided in the clear (unencrypted) while the SI 428 is provided to
multiple CA
24 systems for delivery over the cable network. Thus, in the exemplary system
400,
clear SI 428 is provided to an encryption system 428 that encrypts SI data
using
26 encryption system A. Simultaneously, clear SI 428 is provided to encryption
system
27 424 that encrypts the SI data using encryption system B. Clear video and
audio are
28 then multiplexed along with encrypted SI from 418 (S1 A) and encrypted
audio from
29 424 (S1 B) out of band system information 428.
Docket No.: SNY-R464fi.02 -36- PATENT
CA 02405902 2002-10-O1
1 After distribution through the cable system 32, the video, the audio, system
2 information A and system information B are all delivered to set-top boxes 36
and
3 436. At STB 36, the encrypted SI is decrypted at CA system A 40 to provide
tuning
4 information to the set-top box. The set-top box tunes a particular program
to allow
it to be displayed on television set 44. Similarly, at STB 436, the encrypted
SI is
6 decrypted at CA system B 440 to provide tuning information for the set-top
box, allow
7 a particular program to be tuned and displayed on television set 444.
8 An advantage of this-approach is that no additional AIV bandwidth is
required
9 in the content delivery system, e.g. cable system. Only the SI is dual
carried. No
special hardware is required. Any offset frequencies from the standard ones
can be
11 easily accommodated by most tuners. S1 decryption can be performed in
software
12 or can be aided by hardware. For example, legacy Motorola set-top boxes
have an
13 ability to descramble the SI delivered in the Motorola out-of-band using a
hardware
14 decrypter built into the decoder IC chip.
A determined hacker can potentially use a spectrum analyzer on the coax
16 cable to team where the AN channels are located. Also, it may be possible
for the
17 hacker to program a set-top box to auto-scan the frequency band to learn
where the
18 A/V channels are - a relatively slow process. If the A/V channel
frequencies
19 changed dynamically, then that could foil the hackers, since they would
need to be
constantly analyzing or scanning the band. Also, the program numbers and
assigned
21 PIDs can vary. However, dynamically changing frequencies, program numbers,
and
22 PIDs might create operational difficulties to a service provider, e.g.
cable operator.
23
24
GENERALIZED REPRESENTATION
26 Each of the above techniques can be represented generically by the system
27 500 of FIGURE 10. This system 500 has a cable system headend 522 with clear
28 video 504, clear audio 506, Sl 528, and PSI 529 any of which can be
selectively
29 switched through an intelligent processor controlled switch 518, which also
serves
to assign PIDs (in embodiments requiring PID assignment or reassignment), to
Docket No.: SNY-84646.02 -37- PATENT
CA 02405902 2002-10-O1
1 conditional access system A 504 or conditional access system B 524 or passed
in
2 the clear to the cable system 32. As previously, the program or SI encrypted
3 according to the legacy CA system A can be properly decoded by STB 36. The
CA
4 system B encrypted information is understood by STBs 536 and decrypted and
decoded accordingly, as described previously.
6
7 PID MAPPING CONSIDERATIONS
8 The PID mapping concepts described above can be generally applied to the
9 dual partial encryption techniques described herein, where needed. At the
cable
headend, the general concept is that a data stream of packets is manipulated
to
11 duplicate packets selected for encryption. Those packets are duplicated and
12 encrypted under two distinct encryption methods. The duplicated packets are
13 assigned separate PIDs (one of which matches the legacy CA PID used for
clear
14 content) and reinserted in the location of the original selected packet in
the data
stream for transmission over the cable system. At the output of the cable
system
16 headend, a stream of packets appears with the legacy encrypted packets and
clear
17 packets having the same PID. A secondary PID identifies the packets that
are
18 encrypted under the new encryption system. In addition to the PID remapping
that
19 takes place at the headend, MPEG packets utilize a continuity counter to
maintain
the appropriate sequence of the packets. In order to assure proper decoding,
this
21 continuity counter should be properly maintained during creation of the
packetized
22 data stream at the headend. This is accomplished by assuring that packets
with
23 each PID are assigned continuity counters sequentially in a normal manner.
Thus,
24 packets with the secondary PID will carry a separate continuity counter
from those
of the primary PID. This is illustrated below in simplified form where PID 025
is the
26 primary PID and PID 125 is the secondary PID, E represents an encrypted
packet,
27 C represents a clear packet,. and the end number represents a continuity
counter.
28
29 025C04 025E05 125E11 025C06 025C07 025C08 025C09 125E12
Docket No.: SNY-84646.02 -3$- PATENT
CA 02405902 2002-10-O1
1 In this exemplary segment of packets, packets with PID 025 are seen to have
2 their own sequence of continuity counters (04, 05, 06, 07, 08, 09, ...).
Similarly, the
3 packets with secondary PID 125 also have their own sequence of continuity
counters
4 (11, 12, ...).
At the STB, the PIDs can be manipulated in any number of ways to correctly
6 associate the encrypted packets with secondary PID with the correct program.
In
7 one implementation, the packet headers of an input stream segment
illustrated
8 below:
9
025C04 025E05 125E11 025C06 025C07 ~ 025C08 I 025C09 I 025E10 I
11
12 are manipulated to create the following output stream segment:
13
14 125C04 025E11 125E05 125C06 125C07 125C08 125C09 ~ 125E10 1
16 The primary PIDs (025) in the input stream are replaced with the secondary
PID
17 (125) for the clear packets (C). For the encrypted packets; the primary PID
and
18 secondary PID are retained, but the continuity counters are swapped. Thus,
the
19 stream of packets can now be properly decrypted and decoded without errors
caused by loss of continuity using the secondary PID. Other methods for
21 manipulation of the PIDs, e.g. mapping the PID (125) on the scrambled
legacy
22 packet to a NOP PID (all ones) or other PID value not decoded, and the
continuity
23 . counters can also be used in embodiments consistent with the present
invention.
24 The primary and secondary PIDs are conveyed to the STBs in the program
map table (PMT) transmitted as a part of the program system information (PSI)
data
26 stream. The existence of a secondary PID can be established to be ignored
by the
27 STB operating under CA encryption system A (the "legacy" system), but new
STBs
28 operating under CA encryption system B are programmed to recognize that
29 secondary PIDs are used to convey the encrypted part of the program
associated
Docket No.: SNY-84646.02 -39- PATENT
CA 02405902 2002-10-O1
1 with the primary PID. The set-top boxes are alerted to the fact that this
encryption
2 scheme is being used by the presence of a CA descriptor in the elementary
PID "for
3 loop" of the PMT. There typically would be a CA descriptor for the video
elementary
4 PID "for loop", and another one in the audio elementary PID "for loop". The
CA
descriptor uses a Private Data Byte to identify the CA PID as either the ECM
PID
6 or the secondary PID used for partial scrambling, thus setting up the STB
operating
7 under system B to lookfor both primary and secondary PIDs associated with a
single
8 program. Since the PID field in the transport header is thirteen bits in
length, there
9 are 2'3 or 8,192 PIDs available for use, any spare PIDs can be utilized for
the
secondary PIDs as required.
11 In addition to the assignment of a PID for each program component or
12 selected portion thereof, a new PID may be assigned to tag ECM data used in
the
13 second encryption technique. Each PID number assigned can be noted as a
user
14 defined stream type to prevent disrupting operation of a legacy STB. MPEG
defines
a reserved block of such numbers for user defined data stream types.
16 While conceptually the PID mapping at the cable headend is a simple
17 operation, in practice the cable headend equipment is often already
established and
18 is therefore modified to accomplish this task in a manner that is minimally
disruptive
19 to the established cable system while being cost effective. Thus, the
details of the
actual implementation within the cable system headend are somewhat dependent
21 upon the actual legacy hardware present in the headend, examples of which
are
22 described in greater detail below.
23
24
Headend IMPLEMENTATIONS
26 Those skilled in the art will appreciate that the above descriptions as
related
27 to FIGURES 2, 3, 6, 9 and 10 are somewhat conceptual in nature and are used
to .
28 explain the overall ideas and concepts associated with the various
embodiments of
29 the present invention. In realizing a real world implementation of the
present
invention, those skilled in the art will recognize that a significant real
world issue to
Docket No.: SNY-84646.02 -40- PATENT
CA 02405902 2002-10-O1
1 contend with is providing a cost effective implementation of the various
partial
2 encryption methods within existing legacx headend equipment at established
cable
3 providers. Taking two of the primary legacy cable_ systems as examples, the
4 following describes how the above techniques can be implemented at a cable
headend.
6 First, consider a cable system headend using a Motorola brand conditional
7 access system. In such a system the modifications shown in FIGURE 11 can .be
8 done to provide a cost effective mechanism for partial dual encryption
9 implementation. In a typical Motorola system, a HITS (Headend In The Sky) or
similar data feed is provided from a satellite. This feed provides aggregated
11 digitized content that is supplied to cable providers and is received by a
receiver l
12 descrambler I scrambler system 604 such as the Motorola Integrated Receiver
13 Transcoder (18T) models IRT 1000 arid IRT 2000, and Motorola Modular
Processing
14 System (MPS). A clear stream of digitized television data can be obtained
from the
satellite descrambler functional block 606 of the receiver / descrambler /
scrambler
16 604. This clear stream can be manipulated by a new functional block shown
as
17 packet selector / duplicator 610. This new block 610 may be implemented as
a
18 programmed processor or may be otherwise implemented in hardware, software
or
19 a combination thereof.
Packet selector / duplicator 610 selects packets that are to be dual encrypted
21 under any of the above partial dual encryption methods. Those packets are
then
22 duplicated with new PIDs so that they can be later identified for
encryption. For
23 example, if packets at the input of 610 associated with a particular
program have
24 PID A, then packet selector / duplicator 610 identifies packets to be
encrypted and
duplicates those packets and remaps them to PIDs B and C respectively, so that
26 they can be identified later for encryption under two different systems.
Preferably,
27 the duplicate packets are inserted into the data stream adjacent one
another in the
28 location of the originally duplicated packet now with PID C so that they
remain in the
29 same order originally presented (except that there are two packets where
one
previously resided in the data stream). Assume, for the moment, that the new
CA
Docket No.: SNY-84646.02 -41- PATENT
CA 02405902 2002-10-O1
1 system to be added is NDS encryption, In this case, PID A will represent
clear
2 packets, PID B will represent NDS encrypted packets and PID C will represent
3 Motorola encrypted packets. The packets having PID B may be encrypted under
the
4 NDS encryption at this point in 610 or may be encrypted later.
The packets with PIDs B and C are then returned to the system 604 where
6 packets with PID C are encrypted under Motorola encryption at cable
scrambler 612
7 as instructed by the control system 614 associated with the Motorola
equipment.
8 The output stream from cable scrambler 612 then proceeds to another new
device -
9 PID remapper and scrambler 620, which receives the output stream from 612
and
now remaps the remaining packets with PID A to PID C and encrypts the PID B
11 packets under the NDS encryption algorithm under control of control system
624.
12 The output stream at 626 has clear unencrypted packets with PID C and
selected
13 packets which have been duplicated and encrypted under the Motorola
encryption
14 system with PID C along with encrypted packets under the NDS encryption
system
with PID B. This stream is then modulated (e.g., Quadrature Amplitude
Modulated
16 and RF modulated) for distribution over the cable system. The preferred
17 embodiment maps the unencrypted packets on PID A to match the scrambled
18 packets on PID C because the audio and video PIDs called out in legacy
program
19 specific information (PSI) is correct that way. The control computer, the
scrambler,
and legacy set-top boxes only know about PID C. Alternatively, the scrambled
21 packets on PID C could be mapped back to PID A, but this would likely mean
editing
22 the PSI, that was automatically generated, to map the PID numbers from PID
C back
23 to PID A in the PID remapper and scrambler 620.
24 In the above example, the PID remapper and scrambler620 may also be used
to demultiplex PSI information, modify it to reflect the addition of the NDS
encryption
26 (through the use of CA descriptors in the PMT) and multiplex the modified
PSI
27 information back into the data stream. The ECMs to support NDS encryption
may
28 also be inserted into the data stream at PID remapper and scrambler 620 (or
could
29 be inserted by packet selector / duplicator 610).
Docket No.: SNY-84646.02 -42- PATENT
CA 02405902 2002-10-O1
(.
1 Thus, in order to add NDS encryption (or another encryption system) to a
2 cable system headend using Motorola equipment, packets are duplicated and
PIDs
3 are remapped in the data stream from the satellite descrambler. The remapped
4 PIDs are then used to identify packets that are to be scrambled under each
CA
system. Once the legacy system encryption has taken place, the clear PID is
then
6 remapped so that both clear and encrypted packets in the legacy system share
the
7 same PID (or PIDs). PID remapping as in 620 and packet selection and
duplication
8 as in 610 can be implemented using a programmed processor or using custom or
9 semi-custom integrated circuitry such as an application specific integrated
circuit or
a programmable logic device or field programmable gate array. Other
11 implementations are also possible without departing from the present
invention.
12 FIGURE 12 depicts a similar equipment configuration such as that used in
13 implementing the partial dual encryption of the present, invention in a
Scientific
14 Atlanta based cable headend. In this embodiment, the HITS feed or similar
is
received at IRD 704 which incorporates a satellite descrambler 706. This may
be
16 a Motorola IRT or MPS with only the satellite descrambler function enabled.
The
17 output of the satellite descrambler 706 again provides a clear data stream
that can
18 be manipulated by a new packet selector l duplicator 710 which selects
packets to
19 be encrypted, duplicates them and maps the PIDs of the duplicate packets to
new
PIDs. Again, for example, packets to remain in the clear are assigned PID A,
21 packets to be encrypted under the new system (e.g., NDS) are assigned PID B
and
22 packets to be encrypted under the Scientific Atlanta encryption system are
assigned
23 PID C. The packets with PID B may be encrypted at this point under the NDS
24 encryption system.
The stream of packets is then sent to a multiplexes 712 (e.g., a Scientific
26 Atlanta multiplexes) where the packets having PID C are encrypted under the
27 Scientific Atlanta encryption system at 714 under control of control system
718
28 associated with multiplexes 712. The stream of data is then supplied
internal to
29 multiplexes 712 to a QAM modulator 720. In order to properly remap
the,packets,
the QAM modulated signal at the output of multiplexes 712 is provided to a new
Docket No.: SNY-84646.02 -43- PATENT
CA 02405902 2002-10-O1
1 processor system 724 where the QAM modulated signal is demodulated at a QAM
2 demodulator 730 and the clear PID A packets are remapped to PID C at PID
3 remapper 734 under control of a control system 738. Encryption under the NDS
4 encryption algorithm can also be carried out here rather than in 710. The
data
stream with remapped PIDs and dual partial encryption is then QAM and RF
6 modulated at 742 for distribution over the cable system.
7 In the above example, the PID remapper and scrambler 734 may also be used
8 to demultiplex PSI information, modify it to reflect the addition of the NDS
encryption
9 (adding the CA descriptors to the PMT) and multiplex the modified PSI
information
back into the data stream. The ECMs to support NDS encryption may also be
11 inserted into the data stream at PID remapper and scrambler 734 (or could
be
12 inserted by packet selector I duplicator 710). PID remapping and or
scrambling as
13 in 734 along with QAM demodulation and QAM modulation as in 730 and 742
14 respectively, and packet selection and duplication as in 710 can be
implemented
using a programmed processor or using custom or semi-custom integrated
circuitry
16 such as an application specific integrated circuit or a programmable logic
device or
17 field programmable gate array. Other implementations are also possible
without
18 departing from the present.invention.
19 The above embodiments of the present invention allow legacy scrambling
equipment to scramble only the packets desired in an elementary stream instead
of
21 the entire elementary stream. The scrambling of certain packets of an
elementary
22 stream is accomplished by using a PID. number for packets that are not
going to be
23 scrambled, e.g., PID A. Packets that will be scrambled will be placed on
PID C. The
24 scrambling equipment will scramble the packets on PID C (the ones that have
been
selected for scrambling). After the scrambling has taken place, the
unscrambled
26 packets have the PID number mapped to the same as the scrambled packet -
PID
27 A becomes PID C. The legacy set-top boxes will receive an elementary stream
with
28 both scrambled and un-scrambled packets.
29 The packets in these embodiments are handled as a stream. The entire
stream is sent to the legacy scrambling equipment for scrambling. This keeps
all of
Docket No.: SNY-84646.02 -44- PATENT
CA 02405902 2002-10-O1
1 the packets in exact time synchronous order. If packets were extracted from
a
2 stream and sent to the legacy scrambling equipment, time fitter might be
introduced.
3 The present embodiment avoids that problem by keeping all the packets in a
stream.
4 The embodiment does not require cooperation from the legacy scrambling
equipment provider because that equipment is not involved in the remapping of
6 packets- from PID A to PID C. This remapping is preferable because the PID
called
7 out by the PSf generated by the legacy scrambling system does not need to
change.
8 The legacy system knows about PID C, but not PID A. The entire elementary
stream
9 to be scrambled by the legacy scrambling equipment is found on a single PID
that
the scrambling system has been instructed to scramble.
11 In the above examples, the use of NDS as the second encryption system
12 should not be considered limiting. Moreover, although two widely used
systems -
13 Motorola and Scientific Atlanta have been depicted by way of example,
.similar
14 modifications to legacy systems to permit PID remapping and dual~partial
encryption
can be used. In general, the technique described above involves the process
16 generally described as 800 in FIGURE 13. A feed is received at 806 which is
17 descrambled as it is received at 810 to produce a clear data stream of
packets. At
18 814, packets are selected according to the desired partial dual encryption
technique
19 (e.g., audio only, packets containing PES header, etc.). At 818, the
selected
packets are duplicated and the duplicate pairs are remapped to two new PIDs
(e.g.,
21 P1D B and PID C). The duplicated packets are then encrypted based upon PID
(that
22 is, PID C is encrypted according to legacy encryption and PID B is
encrypted
23 according to the new encryption system) at 822. The clear packets (e.g.,
PID A) are
24 then remapped to the same PID as the legacy encrypted PID (PID C) at 826.
The order in which some of the elements of the process of FIGURE 13 are
26 carried out can vary according to the particular legacy system being
modified to
27 accommodate the particular dual-encryption arrangement being used. For
example,
28 encryption under a new encryption system can be carried out either at the
time of
29 duplication or later at the time of remapping the legacy packets, as
illustrated in
FIGURE 11 and 12. Additionally, various demodulation and re-modulation
Docket No.: SNY-84848.02 -45- PATENT
CA 02405902 2002-10-O1
1 operations can be carried out as needed to accommodate the particular legacy
2 system at hand (not shown in FIGURE 13).
3
4 SET-TOP BOX IMPLEMENTATIONS
Several set-top box implementations are possible within the scope of the
6 present invention. The method used at the headend to select packets for
encryption
7 is irrelevant to the STB.
8 One such implementation is illustrated in FIGURE 14. In this embodiment,
9 packets from a tuner and demodulator 904 are provided to a decoder circuit
908's
demultiplexer 910. The packets are buffered into a memory 912 (e.g., using a
11 unified memory architecture) and processed by the STB's main CPU 916 using
12 software stored in ROM memory 920.
13 Selected PIDs can be stripped from the incoming transport via the STB's PID
14 filter, decrypted and buffered in SDRAM, similar to the initial processing
required in
preparation for transfer to an HDD in a PVR application. The host CPU 916 can
then
16 "manually" filter the buffered data in SDRAM for elimination of the packets
17 containing unneeded PIDs. There are some obvious side effects to this
process.
18 The host overhead is estimated to be about 1 % of the bandwidth of the CPU.
19 In the worst case, this is equivalent to 40K byteslSecond for a 15 Mbit/S
video
stream. This reduction is possible since at most only 4 bytes of each packet
is
21 evaluated and the location is on 188 byte intervals so the intervening data
does not
22 have to be considered. Each packet header in SDRAM can therefore be
directly
23 accessed through simple memory pointer manipulation. Additionally, Packets
are
24 cached in blocks and evaluated en masse to reduce task switching of the
host. This
would eliminate an interrupt to other tasks upon the reception of each new
packet.
26 This may produce a increased latency for starting decode of a stream upon
channel
27 change to allow time for cache fill. This may be negligible depending upon
the
28 allocated SDRAM cache buffer size.
29 The host filtered packets in the SDRAM buffer are then transferred to the
A/V
Queue through existing hardware DMA processes and mimics a PVR
Docket No.: SNY-84646.02 -46- PATENT
CA 02405902 2002-10-O1
t
1 implementation. The filtered packets are then provided to the decoder 922
for
2 decoding.
3 A second technique for implementation in a set-top box is illustrated in
4 FIGURE 15. Since RISC processor AIV decoder module in 930 processes the
partial transport PIDs and strips/concatenates for decode, the firmware within
6 decoder IC 930 'can be altered to exclude individual packets in a partial
transport
7 stream based upon criteria in each packet header. Alternatively, the
demultiplexer
8 910 can be designed to exclude the packets. Legacy scrambled packets) pass
9 through the CA module still encrypted. By using the decoder IC 930 to
perform the
removal of the legacy scrambled packets and assuming that the packets
encrypted
11 under the new encryption algorithm (e.g., NDS) is immediately adjacent the
legacy
12 encrypted packet (or at least prior to next primary stream video packet)
then the
13 pruning of the legacy packet in effect accomplishes the merging of a
single, clear
14 stream into the header strip and video queue.
A third technique for implementation of partial decryption in a set-top box is
16 illustrated in FIGURE 16. In this embodiment, the PID remapping is carried
out
17 either within a circuit such as an ASIC, Field Programmable Gate Array
(FPGA), or
18 a programmable logic device (PLD) 938 or other custom designed circuit
placed
19 between the tuner and demodulator 904 and the decoder IC 908. In a
variation of
this embodiment, the decoder IC 908 can be modified to implement the PID
21 remapping within demultiplexer 940. In either case, the legacy encrypted
packets
22 are dropped and the non-legacy packets re-mapped either in circuit 938 or
23 demultiplexer 940.
24 This third technique can be implemented in one embodiment using the PLD
depicted in FIGURE 17. This implementation assumes that there will be not be
more
26 than one encrypted packet of a particular PID appearing in a row, thus, the
27 implementation could be modified to accommodate bursts of encrypted packets
such
28 as with the M and N~' encryption arrangement described above (as will be
explained
29 later). The input stream passes through a PID identifier 950 which serves
to
demultiplex the input stream based upon PID. Primary PID packets are checked
for
Docket No.: SNY-84646.02 -47- PATENT
CA 02405902 2002-10-O1
1 continuity at 958. If a continuity error is detected, the error is noted and
the counter
2 is reset at 960.
3 The original input packet stream contains packets tagged with many PIDs.
4 The PID identifier 950 separates packets with the two PIDs of interest
(primary and
secondary PIDs) from all other packets. This capability can be scaled to
process
6 multiple PID pairs. These other packets are bypassed directly to the revised
output
7 stream. This processing results in a three or four byte clocking delay.
8 Packets with the secondary PID are routed by the PID identifier 950 to a
9 continuity count checker 954 which verifies sequence integrity for this PID.
Any
errors are noted at 956, but specific handling of errors is not relevant to
11 understanding the present invention. The packet's continuity value is
preserved for
12 use in checking the sequence of packets to follow. A corresponding
continuity check
13 958 is done for packets with the primary PID using the independent primary
counter,
14 and again any errors are noted at 960.
The secondary packet is checked for a secondary flag at 962. This Boolean
16 indicator is used to remember if a secondary packet has been processed
since the
17 last clear packet. More than one secondary packet between clear packets is
an
18 error in this embodiment and is noted at 964. Presence of a secondary
packet is
19 remembered by setting the secondary flag at 966.
The continuity counter of the secondary packet is changed at 968 to fit into
21 the sequence of the clear packets. Data for this substitution comes from
the value
22 used to verify continuity of the primary stream at 958. The revised packet
is sent out
23 from 968 and merged into the revised stream forming the output stream.
24 After packets with primary PIDs have had their continuity checked at 958,
they
are differentiated at 970 by the scrambling flags in the header. If the packet
is
26 scrambled, the primary flag is queried at 974. This primary flag Boolean
indicator
27 is used to remember if a primary encrypted packet has been processed since
the
28 last clear packet. More than one encrypted primary packet between clear
packets
29 is an error in this embodiment and is noted at 976 before the packet is
discarded at
978. Presence of a encrypted primary packet is remembered by setting the
primary
Docket No.: SNY-84648.02 -48- PATENT
CA 02405902 2002-10-O1
1 flag at 980. If there is no downstream consumer for the primary encrypted
packet,
2 it can be discarded at 978. In some cases it may be necessary for the packet
to
3 continue on (in which case its continuity counter can use the discarded
secondary
4 continuity value).
If the primary PID scramble test at 970 detects a clear packet, the state of
the
6 secondary and primary flags is tested at 984. Valid conditions are neither
set and
7 both set, since encrypted packets should come in matched pairs. A sequence
of one
8 without the other should be noted as an error at 988. However, the order of
9 appearance is inconsequential in this embodiment. It should be noted that
there
may be other ways to flag a primary packet for deletion other than the
scrambling
11 bits in the transport header, e.g. the transport_priority bit. Also, it is
possible not to
12 use any bits what-so-ever, e.g. using the primary packet's simple
positional
13 information, before or after the secondary packet, as an indicator for
replacement.
14 Clear packets with the primary PID then have their PID value changed at 992
to the secondary PID before being output in the revised output stream.
Alternatively,
16 the secondary PID packets can be remapped to the primary PID value. The
content
17 can be decoded when the decoder is provided with the correct PID for
decoding the
18 content (whether the primary or secondary PID). Presence of a clear packet
also
19 clears the primary and secondary Boolean flags.
In all the embodiments proposed, the secondary packet can be inserted
21 adjoining the primary packet to be replaced even when a series of primary
packets
22 are tagged for replacement. However, in some instances, it may facilitate
headend
23 partial scrambling if multiple encrypted packets can be inserted into the
stream
24 without the intervening secondary packets. In order to accommodate multiple
consecutive encrypted packets (such as with the M~" and N partial encryption
26 method), the use of primary and secondary flags can be replaced with a
counter
27 matching test function. Thus, in place of elements 962, 964 and 966, a
secondary
28 encrypted packet counter can be incremented. In place of elements 970, 974,
976
29 and 980, a primary encrypted packet counter can be incremented. Element 984
can
be replaced with a comparison of the primary and secondary encrypted packet
Docket No.: SNY-84646.02 -49- PATENT
CA 02405902 2002-10-O1
1 counters to assure that the same number of encrypted packets are received in
both
2 the primary and secondary paths. Instead of clearing flags at 992, the
counters are
3 cleared. Using this variation, multiple encrypted packets may be
consecutively
4 received and the number received are compared to monitor the integrity of
the data
stream. Other variations will occur to those skilled in the art.
6 The function described above in connection with FIGURE 17 can be
7 integrated into an A/V decoder chip that functions similar to that of the
commercially
8 available Broadcom series 70xx or 71 xx decoder used in commercial set-top
boxes.
9 FIGURE 18 illustrates a block diagram for such a decoder chip where the
functions
already provided in the commercial chip are essentially unchanged. Normally,
11 commercial decoder chips expect there to be a one-to-one correspondence
between
12 the PIDs and program components (e.g., audio or video).
13 The decoder illustrated in FIGURE 18 permits multiple PIDs to be
14 programmed into the decoder via a connection to the STB central processor
so that
both primary and secondary PIDs can be handled for main audio, main video and
16 a secondary video used for picture-in-picture (PiP) functions. In this
embodiment,
17 the raw data stream is received- by a Packet sorter 1002 that provides a
function
18 similar to that described in connection with FIGURE 1T above to demultiplex
the
19 stream of packets based upon PID. Preferably, the decoder of FIGURE.18
carries
out the PID sorting function of 1002 using hard wired logic circuitry rather
than
21 programmed software. Program guide and stream navigation information is
output
22 for use by an STB's main processor, for example. The packets associated
with the
23 main audio program are buffered in a FIFO 1006, decrypted in a decrypter
1010 and
24 then buffered at 1014 for retrieval by an MPEG audio decoder 1018 as
needed.
Decoded MPEG audio is then provided as an output from the decoder.
26 In a similar manner, packets associated with the main video program are
27 buffered. in a FIFO 1024, decrypted in a decrypter 1028 and then buffered
at 102
28 for retrieval by an MPEG video decoder 1036 as needed. Decoded MPEG video
for
29 the main channel is then provided to a compositer 1040 and then provided as
an
output from the decoder. Similarly, packets associated with picture-in-picture
video
Docket No.: SNY-84646.02 -50- PATENT
CA 02405902 2002-10-O1
1 are buffered in a FIFO 1044, decrypted in a decrypter 1048 and then buffered
at
2 1052 for retrieval by an MPEG video decoder 1056 as needed. Decoded MPEG
3 video for the picture-in-picture channel is then provided to the composites
1040
4 where it is combined with the main channel video and then provided as a
decoded
video output from the decoder. Other packets not associated with the main or
6 picture-in-picture channel are discarded. Of course, other functions may be
7 incorporated in the decoder chip or deleted without departing from
embodiments of
8 the present invention.
9
CONCLUSION
11 As previously mentioned, in order to thwart a persistent threat by hackers,
12 several of the above partial encryption arrangements can be combined to
further
13 enhance security. For example, the critical packet encryption can be used
in any
14 combination with SI encryption, M~' an N, random encryption, time slice and
other
techniques to further enhance security. In one embodiment, as many packets
would
16 be encrypted as bandwidth is available. The amount of encryption might
depend on
17 whether the content was a regular program or premium (such as a pay-per-
view or
18 VOD), whether it was an adult program or a regular movie, and the security
level that
19 the various cable operators feel comfortable operating. Those skilled in
the art will
appreciate that many other combinations are possible to further enhance the
21 security of the encryption without departing from the present invention.
22 The present invention, as described above in its various embodiments, has
23 been described in terms of a digital AIV system using MPEG 2 coding. Thus,
the
24 various packet names and protocol specifically discussed is related the
MPEG 2
coding and decoding. However, those skilled in the art will appreciate that
the
26 concepts disclosed and claimed herein are not to be construed in such a
limited
27 scope. The same or analogous techniques can be used in any digital cable
system
28 without limitation to MPEG 2 protocols. Moreover, the present techniques
can be
29 used in any other suitable content delivery scenario including, but not
limited to,
terrestrial broadcast based content delivery systems, Internet based content
Docket No.: SNY-84646.02 -51- PATENT
CA 02405902 2002-10-O1
1 delivery, satellite based content delivery systems such as, for example, the
Digital
2 Satellite Service (DSS) such as that used in the DirecTVT"" system, as well
as
3 package media (e.g. CDs and DVDs). These various alternatives are considered
4 equivalent for purposes of this document, and the exemplary MPEG 2 cable
embodirnent should be considered to be an exemplary embodiment presented for
6 illustrative purposes.
7 In addition, the present invention has been described in terms of decoding
8 partially encrypted television programs using a television set-top box.
However, the
9 present decoding mechanism can equally be implemented within a television.
receiver without need for an STB, or music player such as an MP3 player. Such
11 embodiments are considered equivalent.
12 Also, while the present invention has been described in terms of the use of
13 the encryption techniques described to provide a mechanism for dual partial
14 encryption of a television program, .these partial encryption techniques
could be
used as a single encryption technique orfor multiple encryption under more
than two
16 encryption systems without limitation. More than two encryption systems
would be
17 accommodated with additional duplicated packets that are encrypted.
Alternatively,
18 the encryption key for one of the duplicated packets may be shared amongst
the
19 multiple encryption systems. Additionally, although specifically disclosed
for the
purpose of encryption of television programming, the present inventions can be
21 utilized for single or dual encryption of other content including, but not
limited to
22 content for download over the Internet or other network, music content,
packaged
23 media content as well as other types of information content. Such content
may be
24 played on any number of playback devices including but not limited to
personal
digital assistants (PDAs), personal computers, personal music players, audio
26 systems, audio I video systems, etc. without departing from the present
invention.
27 Those skilled in the art will recognize that the present invention has been
28 described in terms of exemplary embodiments that can be realized by use of
a
29 programmed processor. However, the invention should not be so limited,
since the
present invention could be implemented using hardware component equivalents
Docket No.: SNY-84646.02 -52- PATENT
,. . CA 02405902 2002-10-O1
1- such as special purpose hardware and/or dedicated processors which are
2 equivalents to the invention as described and claimed. Similarly, general
purpose
3 computers, microprocessor based computers, micro-controllers, optical
computers,
4 analog computers, dedicated processors andlor dedicated hard wired logic may
be
used to construct alternative equivalent embodiments of the present invention.
6 Those skilled in the art will appreciate that the program steps and
associated
7 data used to implement the embodiments described above can be implemented
8 using disc storage as well as other farms of storage such as for example
Read Only
9 Memory (ROM) devices, Random Access Memory (RAM) devices; optical storage
elements, magnetic storage elements, magneto-optical storage elements, flash
11 memory, core memory and/or other equivalent storage technologies without
12 departing from the present invention. Such alternative storage devices
should be
13 considered equivalents.
14 The present invention, as described in embodiments herein, can be
implemented using a programmed processor executing programming instructions
16 that are broadly described above in flow chart form that can be stored on
any
17 suitable electronic storage medium or transmitted over any suitable
electronic
18 communication medium. However, those skilled in the art will appreciate
that the
19 processes described above can be implemented in any nuri~ber of variations
and in
many suitable programming languages without departing from the present
invention.
21 For example, the order of certain operations carried out can often be
varied,
22 additional operations can be added or operations can be deleted without
departing
23 from the invention. Error trapping can be added andlor enhanced and
variations can
24 be made in user interface and information presentation without departing
from the
present invention. Such variations are contemplated and
26 considered equivalent. .
27 While the invention has been described in conjunction with specific
28 embodiments, it is evident that many alternatives, modifications,
permutations and
29 variations will become apparent to those skilled in the art in light of the
foregoing
description. Accordingly, it is intended that the present invention embrace
all such
Docket No.: SNY-84646.02 -53- PATENT
CA 02405902 2002-10-O1
alternatives, modifications and variations as fall within the scope of the
appended
2 claims.
3
Docket No.: SNY-84646.02 -54- PATENT
