Field of the invention

The present invention relates to a method for compressing electronic program guide data to be transmitted in a one-way broadcast network.

Prior art

Electronic program guide data is gathered from various sources, and combined in a broadcast television signal. The EPG data is received by a set top box connectable to a television set, and processed to allow viewing of the electronic program guide on the television set.

American patent application US-A-6, 005,561 discloses an interactive information delivery system, more in particular a head end broadcasting system broadcasting a data stream of media objects. The data is used to disseminate an electronic program guide EPG and is in some embodiments compressed, e.g. using run length encoded data fields or LZW encoding, or by using a code dictionary stored in the end user equipment. In the latter option, frequently occurring phrases are encoded into a dictionary to take advantage of redundant use of program titles and descriptions.

American patent application US-B-7,574,719 discloses a program guide data compression method and system. The method uses encoding scheme values as compression index values where the encoding scheme values are part of the encoding scheme, but are not used to encode the data. Each word or text part of an EPG is encoded in a dictionary using e.g. an 8-bit encoding scheme wherein words/ text parts are indexed using only two bytes.

Summary of the invention

The present invention seeks to provide a broadcast metadata compression method and system which has an improved efficiency.

According to the present invention, a method for compressing electronic program guide data to be transmitted in a one-way broadcast network is provided, comprising:

- receiving the program guide data as input data having text items; - applying a first layer of compression of data by preparing a list of unique entries and associated identifications, the unique entries comprising text items which are mutually different;

- and applying a second layer of compression of data by providing a compressed list by repeating the following steps for each text item in the list of unique entries:

o retrieving a next text item from the list of unique entries;

o determining whether the next text item in the list of unique entries

number of shared consecutive characters at the start of the next text item in common with any one compressed text item from the compressed list, identifying said compressed text item as a source of shared consecutive characters and determining the number of shared consecutive characters; o storing for the next text item an identification, a reference to the source of shared consecutive characters, the number of shared consecutive characters, and the remaining non-shared text in the compressed list.

By using this two layer compression technique, the EPG data to be transmitted using a broadcast network is reduced in volume, allowing more efficient transfer of these data. The identifications used in the method may be implemented as actual fields having a number stored therein, or alternatively may be implemented as a position indicator (index) in the relevant list.

In a further aspect of the present invention, a broadcast server is provided for transmitting electronic program guide data to a plurality of set top boxes via a broadcast network to which the broadcast server is connected in operation, the broadcast server being further connected in operation to one or more listing providers to receive

the program guide data as input data having text items, and being arranged to apply a first layer of compression of data by preparing a list of unique entries and associated identifications of the text items received, the unique entries comprising text items which are mutually different, and to apply a second layer of compression of data by providing a compressed list to the plurality of set top boxes (4),

wherein provision of the compressed list is accomplished by the broadcast server (8) by repeating the following steps for each text item in the list of unique entries:

- retrieving a next text item from the list of unique entries; - determining whether the next text item in the list of unique entries has a number of shared consecutive characters at the start of the next text item in common with any one compressed text item from the compressed list, identifying said compressed text item as a source of shared consecutive characters and determining the number of shared consecutive characters;

storing for the next text item an identification, a reference to the source of shared consecutive characters, the number of shared consecutive characters, and the remaining non-shared text in the compressed list.

Such a broadcast server can be efficiently implemented in a general computer system, thereby providing a very efficient hardware implementation for broadcasting electronic program guide data.

In an even further aspect the present invention relates to a set top box for decoding data received from a broadcast server according to an embodiment of the present invention. The set top box comprises a processor arranged to retrieve data from the received data for an entry in the electronic program guide from a compressed list prepared using a method embodiment of the present invention, the entry in the program guide comprising an identification referring to an entry in the compressed list, wherein the processor is further arranged to recursively retrieve from the compressed list the non-shared text, a reference to a source and a number of shared consecutive characters associated with the identification, and to add in front of the non-shared text an additional text string comprising the number of shared consecutive characters of the non-shared text retrieved from an entry in the compressed list associated with the reference, until an entry has been retrieved for which the number of shared consecutive characters is equal to zero.

Short description of drawings

The present invention will be discussed in more detail below, using a number of exemplary embodiments, with reference to the attached drawings, in which

Fig. 1 shows a schematic diagram of a system for collecting, processing and transmitting electronic program guide data;

Fig. 2 shows a flow diagram of an embodiment of a first layer of compression according to the present invention; Fig. 3 shows a flow diagram of an embodiment of a second layer of compression according to the present invention; and

Fig. 4 shows a schematic presentation of a decompression method according to an embodiment of the present invention.

Detailed description of exemplary embodiments

In present day television broadcast systems, often use is made of Electronic Program Guides (EPG), which provide to a user information concerning TV programs. EPG-data is transmitted together with TV-signals, nowadays more and more in the form of digital communication.

EPG's or EPG applications in general are designed to allow a user to watch television programs whilst browsing available channel and program information, e.g. for up to eight days ahead. This information can be presented in a multi-channel grid view or a single channel list view on a television screen. The application will allow (where hardware and middleware permits) the setting of reminder memos, recording memos and ongoing recordings for a series of related programmes. Other functionality may include locking/unlocking of channels, setting/unsetting of favourites and signalling of specific channel and program properties (e.g. HD or copy protection).

In some implementations the EPG application has the capability to cache itself on persistent flash storage (e.g. provided as part of a set top box 4, see below), so that it does not need to be loaded from in-band every time it is launched. The EPG application may also do some shorter term caching of its data files on a volatile RAM file system, e.g. again implemented as part of the set top box 4.

In Fig. 1 a schematic diagram is shown of elements of a system to deliver an EPG to the end user. The end user is watching a television screen on a television set 2, which is connected to a set top box 4. A set top box 4 is a low CPU end device, which can transform a signal received from a broadcast network 6 into a signal suitable for the television set 2 (i.e. audio and video data, e.g. in the form of an analogue or digital signal). The set top box 4 may be integrated with and form a part of the television set 2.

In general, the broadcast network 6 is a one-way network, sending out a signal, or a set of signals, from a central (broadcast) server 8 to a multitude of set top boxes 4 connected to the broadcast network 6. The broadcast signal comprises the audio and video information of a (large) number of television channels, and also additional data. The additional data may comprise data relating to an EPG, i.e. data which is

transformed by the set top box 4 from binary data to program guide data to be displayed on the television set 2.

The transfer of the EPG data requires a part of the bandwidth available in the broadcast network 6. The broadcast network 6 may be based on radio transmission (VHF/UHF band analogue transmission, digital transmission), a cable network, or a data communication network, e.g. based on the internet protocol IP. Efficient use of data transfer may reduce the amount of bandwidth required for transfer of the EPG data.

Furthermore, as shown in the exemplary embodiment of Fig. 1, the central server

8 is connected to a number of providers of listing data 10. As an example, such a provider 10 may be a television broadcast company which makes available their listing of TV programs to the TV broadcast server 8 in a computerized manner. E.g. the providers of listing data 10 deliver their data in the form of raw XML formatted data to the central server 8 using FTP. The data from the listing providers 10 is processed by the central server 8 and added to the broadcast signal for delivery via the broadcast network 6 to the (multitude of) set top boxes 4. E.g. the central server 8 combines the received XML data with other data (advertisements, menu data, etc.) and transforms it into a binary data file for transmission.

A major part of the EPG data comprises text for titles and synopses of scheduled

TV events. A TV event is being defined as a scheduled program with a channel identification, a start and finish time, or a start time and scheduled duration, and further data, e.g. a code for categories. To minimise the required bandwidth and memory requirements the titles and synopses data will undergo one or two layers of

compression according to embodiments of the present invention. This may be executed in addition to other data compression or signal modulation techniques applied in broadcast of TV signals.

In a first layer of compression, executed e.g. by the central server 8 as depicted in Fig. 1, first the listing data is received as input data having text items, e.g. in the form of consecutive entries of events with associated data such as title text and synopsis text. The text items are strings of characters, e.g. provided in XML format. From the listing of events, a title list and a synopsis list is composed (by the central server 8), each comprising unique entries and associated identifications, as depicted in the flow diagram of Fig. 2. A unique entry is defined as an entry having characters different from all the other entries. Each event can then be referenced by using the identification.

Starting from the first entry of the event listing (title list) a next entry is retrieved in block 20, and a check is done whether or not the text of that next entry is already present in the title list (block 21). If the text is already in the title list, the relevant title identification is used for that event, and the flow progresses to block 24. If not, the title text is added to the title list, with a new title identification in block 23, after which the flow progresses to block 24. In block 24 the synopsis for that event is retrieved from the event listing, and a check is made whether or not that synopsis text is already in the synopsis list. If this is the case, again, the associated synopsis identification is used in the event listing (block 25) after which the flow progresses to block 27. If not, the synopsis text is added to the synopsis list with a new synopsis identification in block 26, after which the flow progresses to block 27. In block 27 a check is performed to see whether further events exist in the event listing. If so, the flow returns to block 20, if not, the flow stops at block 28.

At the end of the first compression step, a title list and a synopsis list is provided each with identifications (ID numbers), and unique text entries, and in the event listing, the title text and synopsis text is replaced by the associated identification numbers. Already, the amount of EPG data can have decreased by deleting copies of identical texts in the titles and synopsis, even though identification numbers are added.

Especially when the EPG data comprises a lot of series events (with same title text) a decrease in EPG data amount can be achieved. The identifications (ID numbers) may be explicit, i.e. present as an actual number, or implicit, e.g. using an index pointing to a position in a list.

In a second layer of compression, possible duplication of text parts is exploited to reduce the amount of EPG data to be transmitted, but only for the longest common prefix, i.e. the start of the text entry. Text entries in both the title list as in the synopsis list are checked for the presence of duplicated text. This is depicted schematically in the flow chart of Fig. 3, which is used both for the title list with unique entries and for the synopsis list with unique entries.

In general terms, a compressed list is provided by repeating the following steps for each text item in the list:

- retrieving a next text item from the list of unique entries; - determining whether the next text item in the list of unique entries has a number of shared consecutive characters at the start of the next text item in common with any one compressed text item from the compressed list, identifying said compressed text item as a source of shared consecutive characters and determining the number of shared consecutive characters;

- storing for the next text item an identification, a reference to the source of shared consecutive characters (this reference can be same as identification, when zero characters of shared text exist), the number of shared consecutive characters (or copy count), and the remaining non-shared text in the compressed list.

This time, the event listing is unaffected (and maintained in the compressed list), as this event listing only comprises references to the identification numbers used in the title list and synopsis list. However, in the title list and in the synopsis list, two columns are added, i.e. source (reference to the source of shared consecutive characters) and copy count.

In the source field, an identification number is given, which identifies an earlier entry in the compressed list which shares a prefix of the title or synopsis. If no text is shared (which is e.g. the case for the first entry in the list), the source field refers to the corresponding identification number of that entry. Copy count is a value, which indicates the number of the consecutive characters shared between this entry and the entry referred to. If no text is shared, the copy count is set to zero. In the following this is explained in more detail for entries in the synopsis list, however, similar steps are also applied in the title list.

To take advantage of duplication, text entries make reference to other text entries to copy characters from. For instance:

Synopsis 1 : "Friends is a sitcom based in New York. Today Joey discovers his feelings for Rachel."

Synopsis 2: "Friends is a sitcom based in New York. Ross loses his pet monkey."

In this case, Synopsis 2 would store a reference to Synopsis 1 , a copy count of 39, and the text "Ross loses his pet monkey." It is important to note that this is not full tokenization of the strings like LZW compression; it is only the prefix of the text string that is shared. The algorithm used furthermore assures that there are no circular references between shared text items. The flow of the algorithm is depicted graphically in the flow diagram of Fig. 3. From either the title list or the synopsis list, a next item to compress is retrieved (block 31), after which two test variables (SHARE AMOUNT and BORROWED ITEM) are initialized to 0 and NONE, respectively in block 32. From the part of the title or synopsis list already compressed, a next item to compare the retrieved item with is retrieved in block 33. A check is performed in block 34 ('does this share more than the value of SHARE AMOUNT') and if not, the flow continues to a next decision block 38 ('more items to compare with?' (in compressed list)), after which the flow returns to block 33 (in case affirmative) or block 39 (in case answer is no). It is noted that of course a first item in the list can not be subjected to this steps, as it is the first item and cannot be compared to another item.

If the check in block 34 is positive, the flow progresses to decision block 35, in which a check is performed to determine if the item checked against is borrowing text from the item being compressed. If affirmative, the flow progresses to block 38 (see before, the item checked against will be compressed in a later cycle of the algorithm), and if not, the algorithm determines in block 36 the value of SHARED AMOUNT and the value of BORROWED ITEM, and in block 37 reduces the text for the present item with the BORROWED ITEM (i.e. takes away the first 'SHARE AMOUNT' characters from the present entry text). After this, the flow again continues with block 38 to make further comparisons.

In block 38, it is checked whether there is a next item to compare with from the compressed list. In an embodiment, the compressed text item is retrieved from the compressed list in sequence of first to last compressed text item. This assures an efficient compression, also with regard to preventing redundant steps in a subsequent decompression of the data.

In block 39, the list is checked whether or not there are more entries to be compressed. If there are, the flow returns to block 31 and all steps are repeated, if not, the algorithm is ready. .

As a result of the two-step compression, a compressed title/synopsis list is obtained with three columns of numbers and one column of text, e.g. as follows:

ID Source Copy Count Non-shared Text

0 0 0 Ross loses his pet monkey.

1 0 5 falls in love with Rachel.

2 0 5 and Monica go dancing. In this example, the second and third entry both share the first five characters of entry 'Ο', i.e. the character string 'Ross '. The first column may be actually present, or an implicit reference may be implemented, e.g. using an index number to refer to a position in the compressed title/synopsis list.

To decompress the text once all EPG data is received on the set top box 4, a recursive algorithm will be used to build it up piece by piece, starting at the end of a text string. As a further example the following compressed list is given:

The decompression is depicted graphically in Fig. 4. A first call 41 (Get Text) gets the text string as available in the last entry with the proper identification (in this example ID=2, 'the lazy dog'). This part is written in a temporary buffer, and a next call 42 is made using the source (1) and copy count data of the entry. This retrieves the text string 'jumped over ' from entry with ID=1, which is added in front of the text already present in the buffer. Then a further call 43 is made using source ID=0, and the text string 'The quick brown fox ' is added in front of the already present text in the buffer. As the source ID referred to in this entry is the same as the entry ID, the algorithm stops. Again, the first column (ID) may be actually present, or an implicit reference may be implemented, e.g. using an index number to refer to a position in the compressed list.

In more general terms, decompression relates to retrieving data for an entry in the electronic program guide from a compressed list prepared using an embodiment of the compression method as described above. The entry in the program guide comprises an identification referring to an entry in the compressed list. In a recursive manner, the non-shared text, a reference to a source and a number of shared consecutive characters associated with the identification are retrieved from the compressed list. In front of the non-shared text an additional text string is added comprising the number of shared consecutive characters of the non-shared text retrieved from an entry in the compressed list associated with the reference. This is executed until an entry has been retrieved for which the number of shared consecutive characters is equal to zero. It is noted that there are no redundant steps in the decompression. For any text item, it's source text item must have inside its non-borrowed section of text at least one character that is used by the first text item.

Example of a possible problem:

As you can see, there is no benefit to have item 2 borrowing from item 1 since the text it needs from item 1 is in fact entirely borrowed from item 0. A better result is to have item 2 borrowing the text directly from item 0. The correct compressed list should thus be:

When following the algorithm as shown graphically in Fig. 4, this situation should not happen, as a present entry (e.g. ID=2) is first compared with the first item (ID=0) of the already compressed list. After the first run, there are more items to compare with (block 38, ID=1), but at that moment ID=2 and ID=1 do not share any text anymore.

Again, in the embodiment shown, the identification ID is present in the list, however, in a further embodiment this identification can be made implicit, e.g. using an index reference.

As an alternative, the compressed list may be rechecked for possible occurrence of redundant steps. This may be accomplished by having a further run through the compressed list to check whether such duplications exist, and to change the reference to an earlier identification if necessary.

Based on tests performed with near to real data the compression algorithms according to the present invention embodiment achieve between 66% and 71% compression. Data sets used for testing range between 2.3 and 5 megabytes of data, and after compressing them they become 0.8 and 1.4 megabytes respectively. Since the compression algorithms according to the present invention leaves the data in a usable form, it means that all of the EPG data can be stored in memory on the set top box 4 instead of loading modules in the set top box 4 on an as needed basis. This is especially important for the Video Window as the EPG application will need to tune to the last in-list station and stay there for the duration of the Video Window showing (which is to say for the duration of the EPG application).

The text lists for both the titles and for the synopses may be structured in the form shown in the table above, i.e. for each entry, an ID is given (explicit or implicit), a source reference, a copy count value, and an unshared text field. As the unshared text may vary over a large range of number of characters, an alternative structure may be provided, where all text is at the end of a file as a sequence of text items separated by a NULL character, and the unshared text field is replaced by a pointer to or an offset of the relevant position in the sequence of text items. More in general, the compressed list is a binary file which comprises at its end a sequence of non-shared text entries separated by a NULL character, and at its start a sequence of entries with a fixed length format for identification, reference to the source of shared consecutive characters, number of shared consecutive characters, and a pointer (e.g. a text reference or an offset value indicating how many NULL characters should be skipped) to the associated non-shared text entry.

The EPG data is transferred via the broadcast network 6, e.g. in the form of binary files. The following details for the binary structures are used in the files. All numbers are encoded as big endian, and all strings/text data are single byte arrays, terminated with a NULL character. This includes text data at the end of the binary files.

The text items are grouped in text lists, both for a Title Table and a Synopsis Table. A text list comprises the definition of one piece of text in a compressed format. These items are e.g. sorted by their whole text (combination of all borrowed text plus its own text element), in alphabetical order. This is summarized in the following structure: Field Description Data Type Bytes Name

Sourcelndex Index (0 based) of Text Item to borrow characters from. If not Unsigned 2 copying characters from anywhere then it should refer to itself. short

CopyCount Number of bytes to copy If not copying bytes from anywhere then Unsigned 2 this must be zero. short

Text Unshared text, to add to the end of the copied characters. Stored as Unsigned 4 an offset into the file, pointing towards the first character of the NULL integer terminated string.

Total 8

A header may precede the list of Text Items, e.g. using the following format:

In all, structuring the text data for an EPG application in the manner described provides a very efficient transfer of EPG data, both in the bandwidth needed to transmit the EPG data over the broadcast network 6, as in retrieving the EPG data in the set top box 4 for display on the television set 2.

Aspects of the present invention may be implemented with a distributed computer system operating environment, together providing e.g. the functionality of the central or broadcast server 8. In a distributed computing environment, tasks may be performed by remote computer devices that are linked through communications networks. The distributed computing environment may include client and server devices that may communicate cither locally or via one or more computer networks. Embodiments of the present invention may comprise special purpose and/or general purpose computer devices that each may include standard computer hardware such as a central processing unit (CPU) or other processing means for executing computer executable instructions, computer readable media for storing executable instructions, a display or other output means for displaying or outputting information, a keyboard or other input means for inputting information, and so forth. Examples of suitable computer devices include hand-held devices, multiprocessor systems, microprocessor-based or programmable consumer electronics, networked PCs, minicomputers, mainframe computers, and the like.

Some of the invention embodiment have been described above in the general context of computer-executable instructions, such as program modules, that are executed by a processing device, which is part of e.g. the set top box 4. Generally, program modules include routines, programs, objects, components, data structure definitions and instances, etc, that perform particular tasks or implement particular abstract data types. Typically the functionality of the program modules may be combined or distributed as desired in various environment.

Embodiments within the scope of the present invention also include computer readable media having executable instructions. Such computer readable media can be any available media that can be accessed by a general purpose or special purpose computer. By way of example, and not limitation, such computer readable media can comprise RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired executable instructions and which can be accessed by a general purpose or special purpose computer. Combinations of the above should also be included within the scope of computer readable media. Executable instructions comprise, for example, instructions and data which cause a general purpose computer, special purpose computer, or special purpose processing device to perform a certain function or group of functions.

The present invention has been described above using a detailed description of embodiments, with reference to the attached drawings. In these embodiments, elements may be replaced by equivalent elements providing a similar functionality. The scope of the invention is determined by the language of the claims as attached and its equivalents. The used reference signs are referring to the embodiments described above and are not intended to limit the scope of the claims in any manner.