USP 7181758
`U.S. Patent Application No. 08/224,280
`
`EFFICIENT CRYPTOGRAPHIC METHODS AND DEVICES
`
`Background of the Invention
`Description
`
`This application is a continuation of application Ser. No. 10/079,257 filed Feb. 19, 2002, now
`abandoned, which is a continuation of application Ser. No. 09/699,022 filed Oct. 27, 2000, now
`abandoned, which is a continuation of Ser. No. 09/480,226 filed Jan. 10, 2000, now U.S. Pat. No.
`6,347,215, which is a continuation of Ser. No. 08/939,368 filed Sep. 29, 1997, now U.S. Pat. No.
`6,021,307, which is a continuation in part of application Ser. No. 08/644,838 filed May 10, 1996,
`now abandoned, which is a continuation in part of application Ser. No. 08/279,424 filed Jul. 25,
`1994, now abandoned. All these patent applications are incorporated herein by reference.
`
`The presentThis invention relates generally to information processingdistribution, and more
`specifically to an efficient cryptographic system that is designed to process a large amount of
`information by selectively encrypting a portion of the information.particularly to distributing
`information using a broadcast channel and a bi-directional communication channel.
`It has long been recognized that information needs to be encrypted under some
`circumstances. For example, military messages are typically encrypted in order to prevent
`enemies from obtaining information contained in the message. Recently, commercial
`applications of encryption have been proposed. For example, several commercial software
`products were developed to encrypt all or a portion of the data stored on a storage device (e.g.,
`hard disk and floppy diskette). Some of the communication software packages also allow users
`the option of encrypting their electronic mails. Recently, new methods for marketing information
`have been proposed. In one of the proposed methods, encrypted information is distributed (e.g.,
`using CD·ROMs or via digital highway, such as the Internet) by a vendor to potential customers
`either free or at nominal cost. However, the customers are not provided with the key to decrypt
`the information. When the customers wish to purchase the information, they need to contact the
`vendor to obtain the necessary decryption key, typically upon payment of fees. In another
`proposed method, a secure hardware is connected to the computer of a potential customer.
`Encrypted information (e.g., software) is distributed by a vendor to the customer either free or at
`nominal cost. The secure hardware contains the necessary key for decrypting the information.
`When the customer uses the information, the secure hardware decrypts the information and
`measures one or more parameters (e.g., time) relating to the usage of the information. The
`charges incurred by the customer is based on these parameters. Because the decryption key and
`the decrypted information is never exposed to the customer, the customer has to pay each time
`the information is used.
`
`One of the major differences between commercial applications of encryption and other
`applications (e.g., military applications) is the amount of information needs to be encrypted.
`Typically, commercial information, such as a database or a digitized movie, may contain more
`than several hundred million bytes of information. On the other hand, military messages are
`relatively short. Another major difference is that commercial applications typically use low cost
`general purpose computers (such as microcomputers) to decrypt the encrypted data while
`
`Ex. 1024 - Page 1 of 12
`
`Groupon, Inc.
`Exhibit 1024
`
`

`

`military applications use high power computers specially designed for cryptographic
`applications. Even though the processing power of low cost computers is expected to increase at
`a fast pace, these computers take on more and more demanding tasks (e.g., video processing and
`high speed communication). As a result, the amount of processing power devoted to encryption
`in a commercial system may not increase with time, and in some situations, even decrease.
`
`Prior art cryptographic systems are mainly developed and designed for military
`applications, i.e. they are designed to use high power computers to handle a small amount of
`information. These systems may not be suitable for commercial applications. As a result, there is
`a need to develop an efficient method to handle commercial cryptographic needs.
`
`SummaryBACKGROUND OF THE INVENTION
`Recent advancements in modem and computer technology allow large amount of digital data to
`be transmitted electronically. A number of information providers (such as newspaper and
`magazine publishers) and on-line information distributors have formed partnerships to deliver
`newspaper and other information on-line. In this system, a subscriber uses a computer and a
`modem to connect, through a regular phone line, to the computer of an on-line information
`provider. The subscriber can retrieve information, including newspaper articles, stored in the
`computer of the information provider.
`Broadly stated. the present invention is a method and a signal processing device for encrypting
`information. Only a portion of the information needs to be encrypted because the information
`would not be useful if this portion is not decrypted. When the information has a low degree of
`correlation and error tolerance, even encrypting a small portion would be sufficient to render the
`information useless unless this small portion is decrypted. For information that is highly
`correlated, it would be difficult to take advantage of the correlation if substantially all the
`correlated portions are encrypted. As a result, the correlative nature of the information cannot
`be used to bypass decrypting the encrypted portion.
`On-line delivery of newspaper has many advantages. For example, the information can be
`updated throughout the day while the printed version is printed only once or twice a day. Further,
`it is possible to do text-based searches on the information. However, it is found that on-line
`deliver of newspaper and other information is slow. For example, a subscriber has to wait many
`seconds for a newspaper article to be delivered. The quality of the electronic newspaper is low.
`For example, in order to reduce storage and communication requirements, graphic images
`appeared in the printed version are not universally supplied in the on-line version of newspaper.
`One of the reasons for such poor performance is the limited bandwidth of communication
`channels used by on-line information distributors. Another reason is that information is centrally
`processed by the computer at the site of the information distributor, with the result that each
`subscriber only gets a small slice of the time of the computer.
`SUMMARY OF THE INVENTION
`The present invention uses two channels to deliver digital information: a broadcast channel and a
`bi-directional channel. The broadcast channel is used to deliver the bulb of the digital
`information to subscribers. The amount of information delivered is preferably sufficient to
`satisfy the needs of a large number of subscribers so that they do not have to obtain additional
`information using the bi-directional channel. The broadcasted information is stored on fast
`storage media located at subscriber sites. As a result, search and retrieval of the broadcasted
`information is quick. Further, the broadcasted information is processed locally using a dedicated
`on-site processor instead of relying on the computers of the information distributors. As a result,
`
`Ex. 1024 - Page 2 of 12
`
`

`

`the load on the computers of the information distributors is reduced. If the subscribers desire to
`receive additional information relating to the broadcasted information, the bi-directional
`communication channel is used to transmit the request and the requested information.
`
`The distribution costs of broadcast channels are typically much lower than that of a bi-directional
`communication channel. Consequently, the major portion of information is delivered using low
`cost distribution channels. For a large number of subscribers, the broadcasted information will
`provide all the information they normally need. Thus, expensive bi-directional communication
`channels are used only occasionally.
`
`These and other features and advantages of the present invention will be fully understood by
`referring to the following detailed description in conjunction with the accompanying drawings.
`BRIEF DESCRIPTION OF THE DRAWINGS
`FIG. 1 is a schematic drawing showing one embodimentan information distribution system of the
`present invention.
`
`FIG. 2 is A drawing showing an embodiment of the present invention as applied to videoshows
`a newspaper article as displayed on a monitor of the information distribution system shown in
`FIG. 1.
`
`FIG. 2B shows the contents of the broadcast information which corresponds to the newspaper
`article of FIG. 2A.
`
`FIG. 3 is a drawing showing anshows another embodiment of the information distribution
`system of the present invention where information is stored in a plurality of files.
`DETAILED DESCRIPTION OF THE INVENTION
`It has been observed that information generally has a certain degree of redundancy. For
`example, the structure of many languages dictates that redundant words or letters be used at
`predetermined positions of a sentence. A reader can still understand the sentence if some of these
`words and letters are removed. For example, the grammar of the English language imposes a set
`of rules which includes putting the letter “s” at the end of a noun to designate plural quantity. In
`many sentences, the noun is not the only place where plural quantity is indicated. For example,
`the sentence “there are two birds” uses the word “two” to indicate the existence of more than
`one bird, in addition to the letter “s” attached to the word “bird.” Thus, the letter “s” at the end
`of the word “bird” can be omitted without affecting the meaning of the sentence. In fact, there
`are languages in this world in which noun does not have a plural form.
`
`The degree of redundancy in a certain piece of information depends on the nature of the
`information. At one extreme, information may contain many redundant parts. An example is
`video information which consists of a series of pictures depicting time progression of a scene.
`Each picture typically differs slightly from an adjacent picture in the series because the time
`difference in the scene depicted by adjacent pictures is typically less than 0.1 second. As a result,
`video information contains many pictures which are substantially the same. Consequently, it is
`easy to recreate a picture missing from the series by interpolating from the pictures prior and
`subsequent to the missing picture. This type of information is considered to have a high degree
`of temporal correlation. Video information also has another type of redundancy. The spatial
`variation of a picture is typically very gentle. For example, if the picture is a human swimming in
`
`Ex. 1024 - Page 3 of 12
`
`

`

`water, there is little variation (in terms of color and intensity) in the portion of the picture relating
`to water. Consequently, it may be easy to recreate a missing portion of a picture by interpolating
`from portions of the picture surrounding the missing portion. This type of information is
`considered to have a high degree of spatial correlation.
`
`At the other extreme is information in which it is difficult to create a missing portion
`from other portions. This type of information has a low degree of correlation. An example of this
`type of information is the binary code of a piece of software. Typically, it is difficult to recreate
`a missing byte (or a series of missing bytes) from other bytes in the binary code.
`
`Information can also be classified according to its effect on intended uses if a portion of
`the information is missing. At one extreme is information which would be useless if a small
`portion is missing. An example is the binary code of a piece of software. A computer is unlikely
`to successfully execute the software if the binary code has a few erroneous bytes. This type of
`information is considered to be error intolerant. At the other II extreme is information that
`degrades gracefully. For example, when noise of TV signal increases (i.e., portions of video
`information is missing or has erroneous values), color TV pictures often become monochrome.
`However, it is still possible to watch and comprehend the TV pictures, even though they are less
`pleasing to the eyes. This type of information is considered to be error tolerant. Error toleration
`can also be different depending on 1 spatial or temporal types of errors.
`
`It should be pointed out that even though the degree of error tolerance has some
`relationship to the degree of correlation of information, it does not depend solely on the degree
`of correlation. For example, a reader is likely to tolerate a large number of missing words in a
`newspaper article. On the other hand, the same reader probably would not tolerate the same
`percentage of missing words in a poem. Thus, even though the degree of correlation of the
`newspaper article and poem may be the same (because they follow essentially the same grammar
`rules), the degree of error tolerance is different. This is because error tolerance depends, to a
`certain extent, on subjective considerations. As another example, a person typically does not
`tolerate a small distortion in a familiar song while may tolerate a large distortion in a new song,
`even though the degree of correlation of these two songs are the same.
`
`In the prior art cryptographic systems, every bit of information is encrypted. One aspect
`of the present invention is the realization that it may not be necessary to completely encrypt the
`information, especially when it has a low degree of correlation. This is because it is often
`difficult to recreate the encrypted portion of information (if it cannot be decrypted) based on the
`clear portion of information. In addition, information that has a low degree of error tolerance
`may only need to be encrypted at a few critical places (e.g., the destination address of a jump op
`code, or the last name field of a customer database). This is because a user would not accept the
`information if a small portion is missing or erroneous. Thus, even though most of the
`information is in the clear, it is still not commercially useful if isolated portions are unavailable
`because they cannot be decrypted (i.e., people are still willing to pay a high price to obtain the
`full information, even though they already have percent of the information). Since only a small
`portion of information need to be decrypted, the amount of computation power required to
`decrypt the information is reduced.
`
`Ex. 1024 - Page 4 of 12
`
`

`

`Encrypting a portion of information may also help to reduce the computation power
`required for other signal processing tasks. As an example, information which is massive and
`highly correlated (e.g., video information) is often compressed in order to reduce the amount of
`memory space used for storing and the bandwidth used for transmitting the information. Many
`compression methods make extensive use of the correlative nature of information. However,
`many encryption methods have a tendency to randomize information. For example, if the plain
`text is a string of identical letters, the encrypted text using algorithms such as DES may be a
`string of letters in which every letter is different. Consequently, it may be more difficult to
`compress the encrypted text.
`
`The word “encryption” is used broadly in the present invention to include different ways
`of transforming information so that it is difficult for an 1 unauthorized person to correctly
`understand the information. It includes 1 transformation in which a key is required, such as
`public key and secret key encryption methods. It also includes scrambling information according
`to a secret algorithm without using a particular parameter which may be classified as a “key.”
`
`The word “information” is used broadly in the present invention to include data that is
`organized in some logical manner. Examples of information include video signal, audio signal,
`picture, computer software, text, database, and multimedia composition.
`
`FIG. 1 is a block diagram showing one embodiment in whichof an information having a low
`degree of correlation is processed with a signal processor ofdistribution system 100 in
`accordance with the present invention. Fig. 1 shows the portion of information 1 containing
`sixteen pieces of plain text which are grouped into four blocks B1, B2, B3 and B4. Each block
`contains four pieces of plain text labeled as P1, P2, P3, and P4. Each of these sixteen pieces
`could contain a single bit or a plurality of bitsIn this embodiment, system 100 is designed to
`electronically distribute newspaper. It should be pointed out that the above grouping is
`exemplary, and any number of pieces could be grouped into any number of blocks. In addition, it
`is not necessary to require every block to have 1 the same number of pieces.system 100 can also
`be used advantageously to distribute other types of information. System 100 contains a plurality
`of subscriber units (such as units 102 and 104) each connected to a bi-directional communication
`channel (e.g., telephone connections 106 and 108 coupled to units 102 and 104, respectively) and
`a satellite transponder 110 for broadcasting digital data to these subscriber units. Telephone
`connections 106 and 108 (which could be line-based or wireless) are coupled to a central
`database 109. In system 100, satellite transponder 110 is used to broadcast the content of a
`newspaper to the subscriber units while telephone connections 106 and 108 are used to provide
`additional information (stored in central database 109) to subscriber units 102 and 104,
`respectively, on a demand basis.
`Processor 20 contains a program stored in memory, such as ROM 22. This 1 program is
`executed by a CPU 24. The program stored in ROM 22 could 1 be any program that can
`selectively encrypt some of the pieces of plain text in information 12. Preferably, an algorithm
`should be chosen such that it is difficult for an unauthorized person to determine which pieces
`are selected for encryption. Any encryption method can be used by processor 20.
`
`An exemplary program in ROM 22 is described below. The pieces in the blocks having
`an odd numeral in their label, e.g., B1 and B3, are not encrypted. They provide the “base” for
`selecting one or more pieces in the subsequent blocks (i.e., blocks B2 and B4) for encryption.
`
`Ex. 1024 - Page 5 of 12
`
`

`

`For example, the selected piece in block B2 can be calculated from the pieces in block B1 by
`adding the number of bits in all the pieces of block B1, and takes the remainder after dividing by
`four. It should be pointed out that the blocks used as base do not have to be adjacent to the
`blocks having pieces selected for encryption (e.g., the block B1 can be used to determine the
`selected piece in block B4). In addition, the pieces of more than one blocks can be used as base
`(e.g., blocks B1 and B3 can be used to determine the selected piece in block B2).
`
`In Fig. 1, it is assumed that the third piece of block B2 is selected (using the above
`described formula) for encryption. This piece is encrypted and is labeled as E3 in the encrypted
`information 40. Similarly, the fourth piece of block B4 is assumed to be selected (using the
`above described formula) for encryption and is labeled as E4. The selection of the third piece in
`block B2 and the fourth piece in block B4 should appear as random selection to persons who do
`not known the above formula.
`
`The decryption process is now described. Since blocks B1 and B3 are not encrypted, the “base”
`is not change by processor 20. Consequently, the above formula can be used by processor 30 to
`determine which pieces in blocks B2 and B4 are encrypted. As a result, processor 30 can decrypt
`the selected pieces and the original information 12 can be recovered.
`Because information 12 has a low degree of correlation, it would be difficult to create
`pieces P3 in block B2 and P4 in block B4 by interpolating information contained in other pieces
`if E3 and E4, respectively, are not decrypted. However, if information 12 is highly correlated, it
`may be possible to create pieces P3 in block B2 and P4 in block B4 by using information
`contained in the other pieces without decrypting E3 and E4. Methods to render such creation
`difficult even for highly correlated information will be discussed below.
`
`The structure of these subscriber units are substantially identical; consequently, only one of these
`units, such as unit 102, is described in detail. Unit 102 contains an antenna 116 for receiving
`broadcast signals from satellite transponder 110, a signal/data processor 118 for performing
`signal and data processing functions, a monitor 120 for displaying the electronic newspaper, and
`an input device 122 (such as a keyboard and/or a mouse).
`The ratio of the number of encrypted pieces to the number of plain text pieces is determined by
`the error tolerance of the information. If the error tolerance is low (i.e., it is unacceptable to miss
`a small amount of 1 information), only a small number of pieces need to be encrypted. On the
`other hand, if the error tolerance is high, it may be necessary to encrypt a larger number of plain
`text pieces.
`Signal/data processor 118 contains a transponder interface 132 for processing transponder signal
`received from antenna 116. Transponder interface 132 typically contains a low noise receiver for
`receiving high frequency (e.g., C or Ku band) transponder signal and a “universal data interface”
`for converting the transponder signal to digital data. The retrieved data is stored in nonvolatile
`storage 134, such as a hard disk or solid state flash memory. Preferably, satellite transponder 110
`broadcasts the newspaper data at predetermined times. Thus, a real-time dock 136 is preferably
`used to turn on interface 132 at the predetermined times. Processor 118 contains a
`microcomputer 140 which coordinates the operation of dock 136, nonvolatile storage 134, and
`interface 132. Processor 118 also contains a communication interface 142 for sending and
`receiving digital data from central database 109 through telephone connection 106.
`
`Ex. 1024 - Page 6 of 12
`
`

`

`The time for broadcast is preferably chosen when communication load of transponder 110 is at a
`low level (e.g., around mid-night). As a result, the cost of information delivery is low.
`Alternatively, the time of broadcasting is chosen by transponder 110 because it knows when
`communication load is light. In this case, transponder 110 first sends a signal to signal/data
`processor 118 for alerting processor 118 to receive and process the newspaper information.
`
`A user can use the input device 122 and monitor 120 to read the content of the electronic
`newspaper stored in nonvolatile storage 134. In this embodiment, the complete content of the
`newspaper is stored in nonvolatile storage 134. The term “complete content” means that the user
`is able to read the newspaper without relying on information stored in central database 109
`(although other embodiments may deliver less than the complete content). In this aspect, system
`100 functions in a similar way as the distribution of a conventional printed newspaper. However,
`the digital data of the electronic newspaper delivered by satellite transponder 110 preferably
`contains linkage reference which allows fast retrieval of additional information from central
`database 109.
`
`If the newspaper information received from satellite transponder 110 is sufficient to satisfy the
`needs of a user, signal/data processor 118 will not activate telephone connection 106. However,
`if the user wishes to receive additional information relating to an item mentioned in the
`electronic newspaper (e.g., by selecting at the item using the input device), process 118 will
`retrieve the information stored in central database 109 using the embedded linkage reference.
`
`In system 100 of the present invention, the complete content of the electronic newspaper
`(including graphics and other multimedia contents, if delivered) is stored in nonvolatile storage
`134, which has fast access time. Further, a dedicated processor (i.e., microcomputer 140) is used
`to process newspaper information. On the other hand, prior art on-line newspaper distribution
`systems rely on modem to deliver the content of the newspaper stored in a central site. Further,
`the processor in the central site has to serve many users in delivering the newspaper. As a result,
`system 100 has superior performance compared to the prior art on-line newspaper delivery
`systems.
`
`If it is desirable to limit circulation of the newspaper to a certain class of subscribers only (e.g.,
`paid subscribers), the data transmitted by transponder 110 could be encrypted. As a result, only
`subscribers who have a decryption key are able to read the newspaper. In the case,
`microcomputer 140 also performs decryption functions.
`
`FIG. 2A shows an example of a portion of a newspaper article as seen on monitor 120. In FIG.
`2A, the terms which a user may obtain additional information are underlined (or highlighted in
`other ways, such as setting in different colors, depending on the choice of the publishers). If
`desired, the user may select these terms using a pointing device, such as a mouse, and signal/data
`processor 118 will obtain the additional information from central database 109.
`
`FIG. 2B shows the same portion in FIG. 2A as transmitted by transponder 110 (for simplicity,
`the embedded formatting codes, such as center, bold, etc., are not shown). Each of the terms
`underlined in FIG. 2A are enclosed by a special symbol (e.g., the “Y” symbol) and followed by a
`
`Ex. 1024 - Page 7 of 12
`
`

`

`linkage reference enclosed by another special symbol (e.g., the “X” symbol). These symbols are
`invisible to the users and is recognizable only by microcomputer 140.
`
`When an underlined term in FIG. 2 is A diagram showing one embodiment in which information
`having a high degree of correlation, such as video information 110, is processed with a signal
`processor 130 of the present invention. Only three frame 112, 114, and 116 of video information
`110 are shown, although video information 110 typically contains a large number of frames.
`Frames 112, 114, and 116 have a centrally located region 122, 124, and 126, respectively, which
`are of the same size and shape. Only these regions are encrypted while the rest of the frames are
`in the clear. Since the area of each of these regions are small compared to the size of a full frame,
`the amount of encryption and decryption is reduced.is selected by a user, microcomputer 140
`extracts the linkage reference and transmits it to central database 109. The linkage reference
`allows central database 109 to retrieve the necessary information quickly without doing
`extensive searches. As a result, the response time of system 100 is fast. The retrieved information
`can itself contains linkage references and can be searched.
`The frames 112, 114, and 116 are processed by signal processor 130 which contains
`memory (such as ROM 132) and a CPU 134 for storing and processing the algorithm of the
`present invention. Signal processor 130 generates encrypted video information 150 with three
`frames 152, 154, and 156, each corresponding to frames 122, 124, and 126, respectively. Frames
`152, 154, and 156 contain encrypted regions 122, 124, and 126, respectively, which are resulted
`from encrypting regions 122, 124, and 126, respectively.
`
`The encrypted video information 150 can be decrypted using a decryption processor 140.
`Because the location, size and shape of the encrypted 1 region in each frame is predetermined,
`processor 140 is able to find the encrypted regions and decrypt them accordingly.
`
`In this embodiment, the same region of all the frames (i.e., for all times from beginning to
`end) are encrypted. Thus, it is not possible for an unauthorized person to take advantage of
`temporal correlation because there is no unencrypted data of the region for all times. It is also
`difficult to take advantage of spatial correlation near the center of the frame because there is
`little unencrypted data available at areas surrounding the center. The reason for positioning the
`encrypted region at the center is to take advantage of the fact that the most informative area of a
`picture is usually the center.
`
`The method disclosed in Fig. 2 encrypts substantially all the correlated portions of the
`information. Since almost none of the correlated portions are in the clear, it is impossible to
`bypass the correlated portions by using techniques such as interpolation.
`
`The size of the regions 122, 124 and 126 depends on the degree of error tolerance. If
`spatial error tolerance is low (i.e., a user will not tolerate even a small amount of erroneous
`spatial information), the size of the regions 122, 124 and 126 could be small. Further, for
`situations where temporal error tolerance is low, it is not necessary to encrypt all the frame.
`
`It may be advantageous to encrypt several regions (instead of only one centrally located
`region). This is because the most informative region may not be located at the center. Further,
`the size and shape of the encrypted regions may vary from frame to frame. The only requirement
`is to incorporate the same or a similar set of region selection rules in processor 130 and processor
`
`Ex. 1024 - Page 8 of 12
`
`

`

`140. In digital applications, information is preferably organized into one or more files. If the size
`of information is small, it could be stored in one file.
`
`However, if the size of information is large, it may be stored in a plurality: of files. Fig. 3
`shows a plurality of files 212-217 in a medium 210. Examples of medium 210 are magnetic
`storage, optical storage, RAM, and communication channel. Files 212-217 could be organized
`serially or in other formats. The files relating to the information may be linked in well known
`manners. For example, file 212 is linked to file 215 through a link 222 ; file 215 is linked to file
`216 through a link 223; and file 216 is linked to file 213 through a link 225. Alternatively, no
`link is needed if the files are organized sequentially. Links 222, 223 and 225 could be stored
`inside files 212, 215, and 216, respectively. Alternatively, information relating to the links can be
`stored in other locations, such as a separate file 230.
`
`If one or more of the files 212, 213, 215 and 216 are totally or partially encrypted, the
`information stored in these files may be commercially useless. Consequently, there may not be a
`need to encrypt all the files. Even if files 212, 213, 215 and 216 are in the clear, the information
`may be useless if the links are encrypted. This is the case if the size of each file is small and the
`number of links is large, i.e., it would be very difficult to 0 organize the information without the
`links.
`
`While the invention has been particularly described, it will be apparent to those skilled in
`the art that the disclosed invention may be modified in numerous ways and may assume many
`embodiments other than that specifically set out and described above. It is intended by the
`appended claims to cover all modifications which falls within the true spirit and scope of the
`invention.
`
`If the speed of searching and retrieving data by central database 109 is fast, it may not be
`necessary to include linkage reference in the information broadcasted by transponder 110. In this
`case, the user selects (e.g., using the mouse) words and terms he/she is interested in. Signal/data
`processor 118 transmits the selected items to central database 109, which searches for matches in
`its database. Matched information is sent to subscriber unit 102 for processing.
`
`The bi-directional channel also allows updating of the broadcasted information. There is
`typically a time difference between the broadcast and display of information. New information
`gathered during this time difference can be stored in central database 109 and later transmitted to
`signal/data processor 118.
`
`In this embodiment of the present invention, satellite transponder 110 is used as the vehicle to
`electronically broadcast newspaper. However, other broadcast distribution methods can be used.
`In the present invention, broadcast is defined as one-to-many distribution of information. The
`broadcast distribution channels do not have to be electrical. For example, the present invention
`allows the distribution of CDROMs encoded with digital information to the subscriber sites. In
`the case of electrical broadcast communication channels, both