FIELD OF THE INVENTION:
The present invention relates to data transmission and in particular to an
easier data transmission with text for simple receivers and text plus data for
more complex receivers, wherein both types of receivers can evaluate the
data stream.
BACKGROUND OF THE INVENTION:
By now, a plurality of individual receiver terminals exist, which can all receive
transmitted information in a mobile manner. It is characteristic for this large
spectrum that the processing and output capacities of such different
information receivers vary strongly. A notebook, for example, has a very high
processing and output capacity as a mobile receiver of transmitted
information, since a notebook has significant processor resources and
memory resources. On the other hand, for example, a mobile phone that also
participates in an information service, such as a broadcasting data channel,
has only very limited processing resources and output resources. A small
mobile broadcasting receiver that is not even intended as a mobile phone but
merely as a mobile broadcasting receiver for receiving data, such as soccer
league results or other sports results, newspaper headlines, weather news,
etc., is even more limited in its processing and output resources if a
respective data broadcasting service is used.
Such a text-based information service for digital broadcasting, which is
suitable for simple data collection and reuse as well as for very efficient
broadcasting transmission, exists under the designation "Journaline". This
data service supports a very broad range of receiver types, ranging from
cost-effective solutions having a small text display to high-end receivers with
graphical user interface and optional text-speech reproduction.
The user can process all information provided by the radio station
immediately and interactively. In this respect, the service can be compared to
videotext for television. The basic information is provided in a simple text
form, wherein, however, at the same time, the option for more complex

graphical representations including an extension to multimedia elements such
as image or video sequences and other functional extensions is to be
enabled.
On the other hand, it is important that a data stream, for example broadcast
by a DAB transmitter, is backward compatible, which means that the same
can be read and processed equally by simple receivers and by more complex
receivers or by base receivers which are simple receivers, and by extension
receivers which are more complex receivers.
WO 02/052857 relates to the invention is essentially characterized in that two
different schema branch codes are used, whereby one of the two is utilized
far more frequently and, as a consequence, effects a compression. The
schema branch code and position code are combined, and the bit length for
the schema branch code is transmitted therewith. This results in simplifying
the search function in such a manner that the first part, alone, specifies the
type of the referenced element and that an improvement in the extensibility
is achieved based on a schema version number to be transmitted and on
firmly prescribed extension strategies that are also familiar to the decoder.
WO 2002008948 relates to a method and system are provided for tagging,
indexing, searching, retrieving, manipulating, and editing video images on a
wide area network such as the Internet. A first set of methods is provided for
enabling users to add bookmarks to multimedia files, such as movies, and
audio files, such as music. The multimedia bookmark facilitates the searching
of portions or segments of multimedia files, particularly when used in
conjunction with a search engine. Additional methods are provided that
reformat a video image for use on a variety of devices that have a wide range
of resolutions by selecting some material (in the case of smaller resolutions)
or more material (in the case of larger resolutions) from the same multimedia
file.; Still more methods are provided for interrogating images that contain
textual information (in graphical form) so that the text may be copied to a tag

or bookmark that can itself be indexed and searched to facilitate later
retrieval via a search engine.
WO 2002052857 relates to the invention is essentially characterized in that
two different schema branch codes are used, whereby one of the two is
utilized far more frequently and, as a consequence, effects a compression.
The schema branch code and position code are combined, and the bit length
for the schema branch code is transmitted therewith. This results in
simplifying the search function in such a manner that the first part, alone,
specifies the type of the referenced element and that an improvement in the
extensibility is achieved based on a schema version number to be transmitted
and on firmly prescribed extension strategies that are also familiar to the
decoder.
OBJECTS OF THE INVENTION:
It is the object of the present invention to provide a concept for generating
and processing a data stream that can be processed equally by simple and
complex receivers, and that allows high flexibility of data transmission and
processing, in addition to the text data that are to be interpretable equally by
both simple receivers and complex receivers.
SUMMARY OF THE INVENTION:
This object is solved by an apparatus for generating a data stream according
to claim 1, an apparatus for reading a data stream according to claims 15 or
17, as method for generating a data stream according to claim 28, a method
for reading a data stream according to claims 29 or 30, or a computer
program according to claim 31 or a data stream according to claim 32.
Apart from text data, a data stream comprises an escape-start-sequence
defining a first number of data units that are to be skipped by a base decoder
and to be interpreted by an extension decoder, the first number of data units,
an escape-continuation-sequence defining a second number of data units that
are again to be skipped by a base decoder and to be interpreted by an

extension decoder together with the first number of data units, as well as the
second number of data units and finally, if necessary, text data.
In this way, it is ensured that, on the one hand, short start sequences can be
used, since the number of data units to which an escape-start-sequence is
related is, at the most, as high as or smaller than the maximum number of
data units that can be signalized by the escape-start-sequence. For the
usually very few cases where the number of data units considered merely for
higher-quality decoders is higher than the number of data units interpretable
by the escape-start-sequence, further, an escape-continuation-sequence is
provided that defines a number of data units that are to be interpreted
together with the first number of data units by an extension decoder. Thus,
for shorter data blocks with a smaller number of data units, always merely
the escape-start-sequence is required, which is a short code, since the same
is not intended for encoding an arbitrarily long length, but merely a limited
length of the data units. Flexibility is obtained in that, on the other hand,
arbitrarily long data introductions are possible that can be introduced into the
data stream by an arbitrarily frequent repetition of escape-continuation-
sequences and second data units after an escape-continuation-sequence.
In other words, due to the flexible data stream the data amount to be
introduced into the text for an extension decoder is unlimited. However, this
does not affect the length of the escape-start-sequence, since the task of
signalizing the length of very long data insertions is actually divided into
several sequences, namely into the escape-start-sequence and an escape-
continuation-sequence occurring later in the data stream and possibly further
escape-start-sequences, while only a very short escape-start-sequence is
required for short data insertions occurring relatively frequently. Thus, when
the extension decoder finds an escape-continuation-sequence in the data
stream, he knows that the data units displayed with this sequence belong to
the first data units. With regard to further flexibilization, when using the data
stream, a data type indicator is located in the data units which states the type
of data and thus the processing to be performed with these data both of the
first data units and the second data units. The data type indicator is located,

for example, in front of the first data units referenced by the escape-start-
sequence, and in front of the second data units referenced by the escape-
continuation-sequence.
Preferred embodiments of the present invention will be discussed below with
reference to the accompanying drawings. They show:
BRIEF DESCRIPTION OF THE ACCOMPANYING
DRAWINGS:
Fig. 1 a block diagram of an apparatus of a data stream;
Fig. 2a a block diagram of a base receiver;
Fig. 2b a block diagram of an extension receiver;
Fig. 3a . an illustration of a data stream;
Fig. 3D an enlarged illustration of an escape-start-sequence and an
escape-continuation-sequence according to one aspect;
Fig. 3c an illustration of the data referenced by the escape-start-
sequence and the escape-continuation-sequence of Fig. 3a or
Fig. 3b, together with a data type indicator interpretable by an
extension decoder;
Fig. 4a an example data stream without continuation code;
Fig. 4b an encoding example with continuation code;
Fig. 5 a flow diagram of a method for generating a data stream;
Fig. 6A a flow diagram executed by a base receiver;
Fig. 6B a flow diagram executed by an extension receiver; and
Fig. 7 different data types as can be indicated by a data type indicator
and as can be processed.
DETAILED DESCRIPTION OF THE PREFERRED
EMBODIMENT OF THE INVENTION:
Fig. 1 shows an apparatus for generating a data stream comprising a means
10 for entering information into the data stream, which is output at an output
12. The data stream is fed into a means 14 for transmitting and/or storing
the data-stream for being transmitted to a receiver or a decoder of the data
stream, respectively, via a free space

transmission path 16 in the example of a broadcasting transmission Alternatively, the
output 16 of the means 14 for transmitting or storing is connected to a computer-readable
memory medium, such as a memory card connected to the output of the means or
provided at the other end of a transmission path, such as in a receiver, when this receiver
merely stores the data but does not process the same or stores and processes them
simultaneously A computer-readable memory medium can thus be a non-volatile
memory card in a receiver or a decoder, or a hard disc of a notebook, or a working
memory of a decoder, which retains its data as long as the same is supplied with current
or voltage, respectively
The means 10 for entering includes text data 11a and data units lib Preferably,
calculations are made in bytes Thus, one data unit is 8 bits or 1 byte long This
granularity is preferred, since both text and data can be easily handled in this granularity.
In this context, it is preferred to use the UTF8 encoding format for text encoding, wherein
typical ASCII characters are encoded with one byte each, while, for example, German
umlauts are encoded with 2 bytes and, for example, Chinese characters are encoded with
3, 4 or more bytes It follows that a base receiver is able to decode an UTF8 encoding,
which can be performed, for example, by depositing an UTF8 decoding table Depending
on the decoding, the means for entering obtains text data and data units m parallel In this
case, the means for entering also receives a time control signal lie, which determines at
what time text data and at what time data units are to be fed, for example, into the serial
data stream 12 Alternatively, the means 10 for entering can already receive text data and
data units via a single input, which already provides text data and data units in the correct
desired temporal or bit stream order
The means for entering is implemented to enter text data into the data stream When data
units to be skipped by a simple decoder are to be entered, while the data units are to be
read or processed by a more complex decoder, an escape-start-sequence is entered in the
data stream such as is indicated, for example, in Fig 3 a The escape-start-sequence 31
defines a first number of data units to be skipped by a base decoder and to be interpreted
by an extension decoder This first number of data units is then also entered in the data

stream, as can be seen in Fig 3a at 32 If the data block to be entered includes more data
units than are to be defined by the escape-start-sequence, the means 10 will enter an
escape-continuation-sequence 33 defining a second number of data units to be skipped by
a base decoder, but, however, to be interpreted by an extension decoder together with the
first number of data units This second number of data units is then also entered in the
data stream, as can be seen in Fig 3a at 34
Depending on the implementation, 1 e when the data block is a large data block that has
even more data than can be defined by the escape-continuation-sequence, a further
escape-continuation-sequence is written into the data stream, etc, until all the data units
to be entered at this position into the text data that are to be interpreted together by an
extension decoder have been entered
Then, the means 10 for the entering of Fig 1 will typically enter text data again, that are
indicated in Fig 3a, for example, as second text data 35, such that a data stream results
wherein first text data 30 and second text data 35 are present, that, in a way, enclose data
to be skipped by the base decoder, wherein the base decoder is at least able to read the
escape-start-sequence and particularly the number of data units as well as the escape-
contmuation-sequence and at least the number of included data units in order to be able to
skip the correct number of data units in the data stream
Fig 3b shows an example of an escape-start-sequence 31 comprising an escape start code
31a as well as a downstream length code 31b The escape-continuation-sequence 33 is
built analogously, comprising a separate continuation code 33a and a downstream length
code 33b For example, all codes 31a, 31b, 33a, 33b are each 1 byte or, respectively, one
data unit long, which has the effect that 256 data units can be coded by the length code
This means that, when a data block is more than 256 data units long, the continuation
code 33a is to be written or entered into the data stream after the first 256 data units in
order to encode the remaining data units of the data block with the subsequent length
code 33b

If a data unit were respectively longer, a length code that is one data unit long could
encode a larger number of data units, which would have the effect that in such a data
stream the overall number of continuation codes becomes smaller If, however, the length
code of less than 8 bits is selected, the maximum number of data that can be encoded by
the length code is respectively smaller, so that with otherwise the same circumstances the
number of continuation codes in the data stream will increase correspondingly Typically,
an optimum exists in that a certain length of the data units exists for a certain average
length of the data blocks If the length code were made too long, this procedure would be
inefficient, since then the whole long length code would have to be written into the data
stream, even for only a very short data insertion If, on the other hand, the length code
were made too short, a continuation code would be written for the very large number of
data blocks, which would not be required in particular m the implementation shown in
Fig 3 b if a longer length code had been used right from the beginning
Fig 4a shows an exemplary data stream resulting in a display "This is a great test'", as
indicated at 40 in Fig 4a The associated data stream has initially first text data 30 with
the text "This is a " This requires 10 bytes, since one byte each is required for every
letter and for the blank spaces in the UTF8 encoding Thus, the first text data are 10 bytes
long
Then, data are to be inserted For signalizing this, the escape start code 31a, which is 1A
in the example shown in Fig 4a, can be found in the byte index (zero-based) 10, wherein
the prefix "Ox" indicates a hexadecimal representation Of course, any other code that
differs from the text code could be inserted for the escape start code In other words, the
escape start code 31a, which is 1 byte long m the example in Fig 4a, is to differ from the
code that defines characters (numbers, letters, ) displayable on the screen in a UTF8
encoding Thus, the escape codes are, for example, UTF8 control characters or other
fixed characters that differ from the text codes for characters displayable on the screen
Subsequent to the escape start code 31a is a length code 31b, which indicates how long
the data field following the length code 31b is The length field includes the code 4, so

that, consequently, the first data units 32 in Fig 4a comprise 5 bytes The background for
this is that the length field implicitly includes the term "-1", since a length of zero would
not make sense
Then, another text field follows, namely the word "great" with 5 bytes, whereupon again,
at byte index 22, an escape start code 31a is written, followed by a length code 31b
having a value of 5, since then 6-byte data are written in the byte indices 24 to 29
following the length code On the output side, another text block 35 is present, which is 6
bytes long, since 1 byte is required for every letter of "test" and since 1 byte is also
required for the exclamation mark
This results in a 3 6-byte long data stream including 11 bytes of inserted additional data at
two different positions
A base decoder will read in and represent the data and will then, when he encounters the
escape start code 31a, interpret the same to the effect that he will search for a length code
associated to the escape start code Then, an interpretation of this length code will cause
the base decoder to skip the data units referenced by the escape start code and the length
code, l e to ignore the same and to consider them not any further Then, the base decoder
will read in the text data in the byte indices 17 to 21 to then recognize again a code 31a as
an escape start code and to search for an associated length indication in the code 31b, in
order to skip again the number of bytes signalized in the code 31b
Thus, it is ensured in a simple manner that a base decoder can even read a data stream
written for newer or more complex extension decoders The fact that the base decoder is
not able to deal with the data but can correctly interpret the escape-start-sequence with
the code and length indication ensures the backward compatibility
However, an extension decoder is not only able to interpret the start sequence, 1 e the
codes 31a and 31b, but also not to just skip the data but to process the same in oidei to be

able to perform, apart from simple text display, additional functions that can be
controlled via the data stream and the data associated to an escape sequence
Fig 4b shows an encoding example with continuation code, wherein it is assumed in the
example shown in Fig 4b that one data block is longer than 256 bytes If such a large
data block is to be inserted, first, an escape-start-sequence, which again consists of an
escape start code 31a and a downstream length code 31b, is inserted at the data byte
index 0 However, now, the length value m the length code 31b is not any number as in
the byte index 11 or the byte index 23, but the maximum, namely "FF" Then, following
the length code 31b, the 256 first data umts 32 are written into the data stream in the byte
indices 2 to 257 Since several further data umts are present, namely 262 bytes all in all,
the data stream generator will determine that it has not yet written all data units into the
data stream Consequently, the same writes a continuation code 33a into the data stream,
which will differ from the escape start code 31a, so that the decoder knows that the data
units belonging to the continuation code, which, for example, immediately follow the
continuation code, still belong to the data units that have been processed or fed in prior to
the continuation code In the example shown in Fig 4b, six more data units have to be
entered so that the length code 33b following the continuation code signalizes a length of
6 (code 0x05) Then, following this length code, the insertion into the data stream is
terminated with writing the second data units 34, wherein then, starting from the byte
index 266 shown in Fig 4b, e g normal text 35 can follow
It should be noted that, depending on the implementation, the escape continuation code
33a does not necessarily have to differ from the escape start code 31a This is not
important for the base decoder The same merely has to interpret both the escape start
code 31a and the escape continuation code 33a and to read the downstream length code
following this interpretation, m order to know how many data are to be skipped In this
case, an extension decoder would be implemented such that the same assumes
automatically, when an escape start code, followed by a length code with a maximum
length, I e , for example, FF, occurs for the first time, that the subsequent data units still
belong to the data umts 32 and thus have to be processed together The meaning of the

escape start code would thus be "switched" in that the length code 31b following the
escape start code 31a indicates a maximum length When the escape start code 31a and
the continuation code 33a have the same value, the decoder would, when the length code
31b after the first escape start code 31a does not have the maximum values, I e FF,
interpret the continuation code following the data units not as a continuation code but as a
new start code, so that the data following the second start code are not considered as a
continuation of the data following the first start code but as new data of a new data block
that can be processed differently The interpretation whether data are continuation data or
no continuation data is of particular importance when a data type indicator is located in
every data block consisting of several data units at a certain position, such as at the
beginning, which is read when the data are no continuation data, or, when the data aie
continuation data, is not expected by a decoder and also not read out, as will be discussed
below with reference to Fig 3 c
In the following, with reference to Fig 5, an exemplary sequence of steps for generating
a data stream is illustrated, as it can be performed by a data stream generator, which is
structured as shown m Fig 1
The flow diagram in Fig 5 begins with a step of entering text 50 In a step 51, it is
checked whether, after entering text, data exist which still have to be entered If the
answer to this question is no, the next text object is entered into the data stream, as
illustrated by loop 52 If, however, the answer to the question is yes, the means 10 for
entering enters an escape-start-sequence, as is illustrated in step 53 The escape-start-
sequence includes information about the length of the data, I e the number of data units
After entering the escape-start-sequence into the data stream, the first data units are
entered, as indicated at 54 In a step 55, it is checked whether all data belonging to a data
block are entered in the data stream If the answer to this question is no, I e no more data
exist, a text object or an escape code will be entered again, as is illustrated by a loop 56
It should be noted here that the escape-start-sequences could stand in front of a text, after
a text or in between a text Alternatively, however, an escape-start-sequence can also be

in front of or after anothei escape sequence, wherein another escape sequence can
indicate things other than specific data
If, however, it is determined that data still exist, an escape-continuation-sequence is
entered, as shown in step 57 In this escape-continuation-sequence, also the length of the
data units is determined which will then be entered in a step 58 In a step 59, it is checked
whether further data are present If this is the case, I e further data of the data block are
entered, a further escape-continuation-sequence is written, as shown at 60 If, on the other
hand, all data are entered, text is again entered basically always in the same manner as
shown at step 50 For purposes of clarity, this entering of text after a data block is
indicated by step 61 in Fig 5
It should be noted that the check m step 55 as to whether data are still present or in the
step 59 as to whether data are still present is not to be performed necessarily when the
length of the data to be entered is determined in parallel or in a separate process, as
shown in Fig 5 at block 62 In block 62, it is determined how long a data block to be
inserted is Based on the number of data units to be maximally encoded in the escape-
start-sequence, block 52 knows immediately how many continuation sequences are
required In the embodiment shown in Fig 4b, block 62 determines that the maximum
length is to be entered into the escape-start-sequence, and that then a continuation
sequence is entered whose length is also determined, as indicated by the control arrows
63a and 63b in Fig 5 In this case, the check in block 55 or block 59 is not to be
performed, as indicated by a dotted connecting arrow 64 This alternative already
represents a solution where the escape-start-sequence is already finished prior to entenng
the data units, while, in the second alternative, the length is only to be entered into the
escape-start-sequence after the first data units or the second data units, respectively, have
been entered
In the following, with reference to Fig 2a and Fig 2b, receivers for receiving or
decoding the data stream will be discussed, wherein this data stream can be structured
basically as shown in Fig 3a, to the effect that the same includes text data, an escape-

start-sequence, a number of data units, then an escape-continuation-sequence and
following the escape-continuation-sequence again a number of data units which can be
followed by text data, or by a further escape-continuation-sequence
The base receiver for reading the data stream comprises an input interface 70 for
obtaining the data stream 16 The data stream is then transmitted to a piocessor 71, which
is coupled to a text display 72 for reading the text from the data stream and displaying the
same, wherein the processor determines the length or the number of data units when the
same encounters an escape-start-sequence, for skipping this number of data units, and
wherein the processor further, when the same encounters an escape-continuation-
sequence, also skips the number of data units belonging to the continuation sequence, as
is shown m block 71 of Fig 2a
An extension receiver as shown in Fig 2b comprises, in addition to the elements of the
base receiver of Fig 2a, a module or a functionality of the processor 71 that the data units
are not simply skipped but are executed together
In addition to the text display 72, that can or will function in the same manner in the
extension receiver as in the base receiver, the extension receiver will peiform an
interpretation of the data units after the escape-start-sequence and the escape-
continuation-sequence
In the following, with reference to Fig 6A and Fig 6B, a comparison of the
functionalities of a base receiver without data processing option and an extension receiver
with data processing option is made
When both the base receiver and the extension receiver read text data, the processor 71 of
Fig 2a or Fig 2b will process the text data, decode the same and then provide them to the
text display 72 and display them, as is shown in the step 80 of Fig 6A and Fig 6B If the
receivers encounter an escape-start-sequence, this escape-start-sequence will be read, as
illustrated in a step 82 In particular, the base receiver will see in step 82 how many data

units are indicated by the escape-start-sequence, 1 e , for example, by the length code 31b
or Fig 3b In a step 83, the base receiver will then skip the number of data units that it
has determined in step 82 In contrast, in a step 84, the extension receiver will not skip
the data units following the escape-start-sequence but read the same Therefore,
preferably, the extension receiver also requires the length information If the data were,
however, encoded in a different manner than text, the length information would not
necessarily be required
In a step 85, both the base receiver and the extension receiver read the escape-
contmuation-sequence following the first data units 32 of Fig 3a Here, the base receiver
is mainly interested in the length code m order to find out how many data are to be
skipped, which is then performed by the base receiver in a step 86 On the other hand, the
extension receiver will not skip the second number of data umts in a step 87, but read the
same, as is illustrated at 87 Then, m a step 88, for which there is no parallel base
receiver, the extension receiver will activate the functionality or the module 73 of Fig 2b,
and achieve a common execution of the first and second data units When all data units
have been processed, both the base receiver and the extension receiver will display the
second text data in a step 89, I e the data 35 following the data stream shown in Fig 3a
If, however, the second length code 33b also has the maximum length, I e in this
example "FF", the extension receiver can read in a further escape continuation code,
interpret the subsequent length code and then also add those referenced data units to the
first and second data units in order to again allow a common execution
Subsequently, with reference to Fig 3 c, it will be discussed in more detail what the
extension receiver performs in step 88, or how the data are interpreted that have been
extracted after the escape start code and the continuation code
In one example, the first number of data units composes a data type indicator 90, which
is, for example, 1 byte long This data type indicator 90 can be found only in a certain, for
example the first byte of the first number of data units, which is referenced by an escape-
start-sequence 31 In comparison, no such data type indicator is present in the second

number of data units, but the second number of data units 34 fully contributes to the data
units to be executed together, or the so-called "payload" When a data type indicator 90 is
used, an extension decoder will interpret the data included after the continuation code as
belonging to the same type as the data that are included in the first number of data umts
This enables the fact that no signahzations are required for the second number of data
units following the continuation code in the data stream, which include information about
whether these are continuation data or which data type the second number of data units
has Instead, the data type indicator m the first number of data units is simply also used or
applied for the second number of data units, or the data in the second number of data
units are simply added to the first number of data units, respectively, as if there had never
been a separation, so that they will then be executed or processed together
It is an advantage of the inventive scalable data stream that the same is based on general
standards Thus, objects can be imported and further processed in XML format
The data stream is particularly suitable for digital broadcasting systems, as an additional
data channel providing an additional value for the listeners, since they now have
immediate access to text information wherever they are, wheiein receivers for digital
broadcasting are either simple receivers displaying at least text information, or
particularly complex and thus, of course, more cost-intensive receivers that can perform
any data processing of the first number of data units and the second number of data units
It follows that receivers having a text display that are both inexpensive and thus available
as a mass product can generate an additional value for the listener However, the data
stream is also suitable for high-end receivers with a graphical user interface and an
optional speech reproduction All this is obtained by simple implementations, even in
inexpensive receivers and by a particularly simple method of utilization where the user
does not have to take care of which data are currently valid Instead, deleting the data or
executing the data, depending on which receiver a user has, is performed fully
automatically without the user having to take care of it

Further, in certain examples, the text data are illustrated in an object-oriented manner,
wherein these objects are all independent and self-contained units Thus, no global data
structures have to be assembled or maintained in the receiver Objects aie transmitted in
the form of a data carousel, and data caching is advantageously used in the receiver Text
data that can, for example, be menu designs, news articles or ticker news are transmitted
as so-called NML objects, wherein NML means News Service Mark Language and is
similar to XML-based binary encoded content representations
Preferably, the hexadecimal code 1A is used as start-escape-code 31a, and the
hexadecimal code IB is, for example, used as continuation code The data type indicator
90 of Fig 3c can indicate several data types, as will be illustrated below The following
data type indicator values are merely exemplary An indicator of "00" indicates padding
The included data carry padding bytes, wherein this content is ignored both by the
extension decoder and the base decoder An alternative type of data set has an indicator
"01" and represents an absolute timeout The absolute presentation timeout period of an
NML object is defined When this timeout has passed, the NML object will no longer be
displayed This escape-sequence-code representation is required for marking individual
NML objects A general timeout for the whole service is alternatively illustrated The
payload after the data type indicator includes the number of 15 minutes since 1 1 2000, as
a 24-bit integer number without signs, which covers more than 450 years A data type
indicator of the type "02" indicates a relative timeout Here, the relative presentation
timeout period of an NML object is defined When this period has passed, the NML
object is no longer displayed The timeout begins with every reception of an NML object,
even when the object is already stored m the cache This escape-sequence-code serves for
marking individual NML objects The payload data include the number of minutes as a
16-bit integer number without signs (unsigned), which covers more than 45 days
Generally, the timeout data are processed by the timeout control 100 of Fig 7
The data type indicator "03" relates to a general link target A general link target is a
target that is presented or activated, for example, by the connection control 101, when the
user explicitly demands the execution of an action, l e when a "hot button" functionality

is illustrated or provided, respectively, for user interaction General link targets can be
defined for all types of NML objects The availability of a general link target for a
currently displayed NML object is communicated to the user in some manner or other
when he has an extension receiver The payload data that are shown, for example, in Fig
3 c have the following format 1 byte represents a connection type and n bytes represent a
link address The following link-type values, for example, are available When the link
type has a value, for example of "00", the subsequent2 bytes are an object ID of another
NML object in the same data service
Another link-type byte, for example of "01", followed by a URI string, points to different
DAB/DRM multiplexes, services or service elements
Another connection type, such as "02" showing a URL string, points to an Internet
address or a document
Another link type, such as "03" followed by a telephone number, points to a voice service
that can be reached by telephone The number starts here, for example, with an
international prefix, for example "+[international country code]"
Generally, an extension receiver is designed to ignore unknown link-type values
A further data type indicator "FF" is considered, for example, as a propnetaiy data type
indicator, preceding data concerning only certain extension receivers that can evaluate
these proprietary data
Contrary to the above object management types, the data units can also include content
management types A data type indicator of "20", also referred to as keyword, marks a
following keyword together with an optional keyword description The keyword can be
used, for example, for generating a receiver-based search index, as shown by the query
generation means 102 in Fig 7 The payload data section has the following format First
comes a length code having the keyword length, wherein the value equal to the length -1

is indicated as unsigned integer number The number of visual text characters following
this data section that should be treated as a keyword is identified Then, a description
with n bytes follows, wherein the optional description can be indexed in addition to the
keyword and/or can be displayed for the user
Another data type indicator, for example of "21", represents a macio definition A macro
allows the definition of text sections including optional escape sequences that can be
inserted several times anywhere in the content section with a simple reference For
example, the macro can define speech descriptions that are to be illustrated in addition to
textual text elements The data section has a format where initially a macro ID (0 to 255)
of one byte identifies the following macro definitions Then a macro definition with n
bytes follows The text (including escape sequences), that is always to be inserted when
this macro is referenced by its ID, is included m the macro definition with n bytes It
should be noted that the macros are not necessarily to be used for essential information,
since receivers can also ignore the same Another data type indicator '22' stands, foi
example, for a macro reference The macro definition referenced by its ID is virtually
introduced at this position of the escape sequence for displaying to the user The data
section includes a 1-byte macro ID (0 to 255) referencing a macro definition Macros are
generally processed by the macro processor 103 in Fig 7
Another group of data types can compose speech support types A data type indicator of,
for example "A0", defines a standard language or a preset language Here, the preset
language of the NML object is described or referenced, respectively The data section, I e
the payload of Fig 3c, carries a 3-letter ISO language code of small letteis
Another speech support type, for example referenced by "Al" is the language section
The same defines the language of a specified number of a text section or a specific part of
an NML object The payload section has the following format A text length having one
byte is present, whose value equals the value of the number of data units of this text
length -1, as unsigned integer number This identifies the number of visual text
characters that follow this data section for which the language definition applies Then, a

group of 3 bytes follows with a language definition, which carries a 3-letter ISO language
code of small letters
Another language support type is indexed, for example, by the data type indicator "A2"
It relates to speech phonemes and defines a phoneme description of a text section using
the international phonetic alphabet (IPA) The payload section has a format having first a
text length with one byte whose value is an integer number This byte identifies the
number of visual text characters following this data section, which are to be represented
by a phonetic definition Then a group of n bytes with IPA phonemes follows This group
mcludes the phoneme definitions of the phonemes as IPA notations
Another speech support type, for example referenced by the data type indicator "A3",
comprises a speech break, which defines a break for text-to-speech processors that is to
be inserted at the position of the escape sequence The data section carries 1 byte as an
unsigned integer number, wherein this byte defines a speech period in units of 0 1
seconds
Another speech support type that can, for example, have the data type indicator "A4"
defines characters and particularly a number of visual text characters following the
escape sequence that are to be treated as individual characters or numbers instead of
continuous words or numbers by the text-to-speech processor The payload section
carries 1 byte defining the number of visual text characters as unsigned integer number
having a respective value
It should be noted that all these data types, depending on the maximum number that can
be represented by the length code, can either be represented in the first number of data
units 32 of Fig 3a or, when the number of data units is not sufficient for writing all the
data, are included in the second number of data units that can each be started with an
escape-continuation-sequence It follows that not necessarily all data types that are
referenced by a data type indicator need to have data units both after an escape-start-
sequence and an escape-continuation-sequence Instead, even short data types can include

merely data requiring no continuation since the number of data units of these data is
smaller than the maximum number of data units representable by the length code 31b
Depending on the circumstances, every inventive method can be implemented in
hardware or m software The implementation can be perfoimed on a digital memoiy
medium, in particular a disc or CD with electronically readable control signals that can
cooperate with a programmable computer system such that the method is performed
Thus, generally, the invention also consists of a computer program product with a
program code stored on a machine-readable earner for performing an inventive method
when the computer program product runs on a computer In other words, the invention
can be realized as a computer program with a program code for performing the method
when the computer program runs on a computer
According to an embodiment of the apparatus for generating, the data stream is a serial
data stream and the means 10 for entering is implemented to enter the escape-start-
sequence 31 in the data stream after or in front of the text data 30 and to enter the escape-
continuation-sequence 33 after the first number 32 of data units into the data stream
In a preferred embodiment of the apparatus for generating, the means 10 for entering is
implemented to place first the text data 30, then the escape-start-sequence 31, then the
first number of data units, then the escape-contmuation-sequence 33 and then the second
number 34 of data umts into the data stream
In a preferred embodiment of the apparatus for generating, the length code 31b is enteied
into the data stream immediately after the escape start code 31a, or the length code 33b is
entered into the data stream immediately after the continuation code 33a
In a preferred embodiment of the apparatus for generating, the length code 33b entered
into the data stream after the continuation code 33a and the length code 31b entered into
the data stream after the escape start code 31a originate from the same encoding table for
length encoding

In a preferred embodiment of the apparatus for generating, the data units in the data
stream are equal and comprise a predefined plurality of bits each
In a preferred embodiment of the apparatus for generating, the predefined plurality of bits
equals 8, such that one data unit is 1 byte
In a preferred embodiment of the apparatus for generating, the data stream is to be
readable by a first type of base receivers and by a second type of extension receivers,
wherein the means 10 for entering is implemented to use text data codes that can be
detected by the base receivers and the extension receivers
In a preferred embodiment of the apparatus for generating, the escape start code 31a has
the same length as a continuation code 33a within an escape-continuation-sequence 33,
wherein this length is equal to the length of one data unit
In a preferred embodiment of the apparatus for generating, the length code is
implemented to encode a content length of 1 to 256 bytes
In a preferred embodiment of the apparatus for generating, the data type indicator 90 is
such that the same does not have to be interpretable by a base decoder, but is
interpretable by an extension decoder
In a preferred embodiment of the apparatus for generating, the means 10 for entering is
implemented to enter, following the data type indicator, data having the data type
indicated by the data type indicator 90
In a preferred embodiment of the apparatus for generating, the means 10 for entering is
implemented to enter the data type indicator 90 into the data stream in the first numbei of
data units 32 immediately after a length code 31b of the escape-start-sequence 31

In a preferred embodiment of the apparatus for reading, the data stream is latched in a
buffer and the processor 71 is implemented to delete the first number of data units 32 and
the second number of data units 34 from the buffer and to continually read out the buffer,
or to control the buffer such that an area of the buffer where the first number of data units
32 and the second number of data units 34 are stored is skipped when reading out the
buffer
In a preferred embodiment of the apparatus for reading, the first number 32 of data units
comprises a data type indicator 90 for indicating a data type of the first number of data
units 32 and the second number of data units 34, and the processor 71 is implemented to
read out the data type indicator 90 and to process the first number of data units 32 and the
second number of data units 34 together according to the data type indicator 90
In a preferred embodiment of the apparatus for reading, the target is an object with text
data in the data stream, a DAB/DRM multiplex, a service or service element, an Internet
address, a document on the Internet or a telephone number
In a preferred embodiment of the apparatus for reading, the processor 71 is implemented
to ignore the first data units 32 or the second data units 34 when the same define a non-
interpretable target
In a preferred embodiment of the apparatus for reading, the first numbei of data units 32,
the second number of data units 34 or further data units have a data type indicatoi 90
comprising speech support data for a speech processor 73, that indicate a language of a
text section, comprise speech phonemes, are related to a speech break or have speech
characters, and the processor 71 is implemented to output speech support data and to
supply the same to the speech processor for generating or influencing a speech output
In a preferred embodiment, the data stream is stored on a computer-readable medium

WE CLAIM:
1. An apparatus for generating a data stream, comprising:
a means (10) for entering text data (30) into a data stream, for
entering an escape-start-sequence (31) into the data stream, wherein
the escape-start-sequence defines a first number of data units to be
skipped by a base decoder and to be interpreted by an extension
decoder, for entering the first number (32) of data units into the data
stream, for entering an escape-continuation-sequence (33) into the
data stream, wherein the escape-continuation-sequence defines a
second number of data units to be skipped by a base decoder and to
be interpreted by an extension decoder together with the first number
of data units, and for entering the second number (34) of data units
into the data stream, wherein the escape-start-sequence (31) has an
escape start code (31a) and a length code (31b), or wherein the
escape-continuation-sequence (33) has a continuation code (33a) and
a length code (33b), or wherein the escape start code and the
continuation code are identical to each other and wherein the escape
start code and the continuation code are identical to each other and
wherein the a first length code (31b) having a maximum length signals
that a code following the first length code in the data stream is a
continuation code.
2. The apparatus as claimed in one of the previous claims, wherein the
means (10) is implemented to enter, in a predetermined relation to the
escape-start-sequence (31), a data type indicator (90) indicating a
data type of both the first number of data units and the second
number of data units (34) into the data stream.
3. An apparatus for reading a data stream having text data (30), an
escape-start-sequence (31) defining a first number of data units (32),
the first number of data units (32), an escape-continuation-sequence
(33) defining a second number of data units, and the second number

of data units (34), comprising:
a means (72) for displaying text data (30); and
a processor (71) for interpreting the escape-start-sequence (31) such that the first number of data units is determined from the escape-
start-sequence for skipping the first number of data units and for
interpreting the escape-continuation-sequence (33) such that the
second number of data units (34) is determined from the escape-
continuation-sequence for skipping the second number of data units
(34).
4. An apparatus for reading a data stream having text data (30), an
escape-start-sequence (31) defining a first number of data units (32),
the first number of data units (32), an escape-continuation-sequence
(33) defining a second number of data units, and the second number
of data units (34), comprising:
a means (72) for displaying the text data (30);
a processor (71) for interpreting the escape-start-sequence (31) such
that the first number of data units (32) is read in, for interpreting the
escape-continuation-sequence (33) such that the second number of
data units (34) is read in, and for jointly processing (73) the first
number of data units (32) and the second number of data units (34) in
addition or instead of displaying the text data (30), wherein the
escape-start-sequence (31) has an escape start code (31a) and a
length code (31b), or wherein the escape-continuation-sequence (33)
has a continuation code (33a) and a length code (33b), wherein the
escape start code (31a) or the continuation code (33a) differ from the
text codes, or wherein the escape start code and the continuation code
are identical to each other and wherein the a first length code (31b)
having a maximum length signals that a code following the first length
code in the data stream is a continuation code.
5. The apparatus as claimed in claim 4, wherein the first number of data
units (32) and the second number of data units (34) define a

connection target, and wherein the processor (71) is implemented to
execute a data connection to the target.
6. The apparatus as claimed in one of claims 4 and 5, wherein the first number of data units (32) or the second number of data units (34) or a further number of data units define an absolute or relative timeout
period for an object, and wherein the processor (71) is implemented to
perform timeout control (100) having a comparison of the timeout time
and a processor time for displaying the text data and extended
multimedia content only when the timeout time has not been reached
by the processor time.
7. The apparatus as claimed in one of claims 4 to 6, wherein the first
number of data units (32) and the second number of data units (34) or
further data units define text characters that are to be interpreted as
keywords, and wherein the processor (71) has a search-query
generation means (102) for performing a database query in the
apparatus for decoding based on the keyword.
8. The apparatus as claimed in one of claims 4 to 6, wherein the first
number of data units (32) or the second number of data units (34) or
further data units have macro data, and wherein the processor (71) is
implemented to perform a macro processor functionality (103) for
executing a macro defined by the macro data.
9. The apparatus as claimed in one of claims 4 to 6, wherein the first
number of data units (32), the second number of data units (34) or a
further number of data units have a data type indicator (90),
representing a speech description, wherein the processor (71) has a
speech output.
10. A method for generating a data stream, comprising:
entering text data (30) into a data stream;

entering an escape-start-sequence (31) into the data stream, the
escape-start-sequence defining a first number of data units that are to
be skipped by a base decoder and to be interpreted by an extension
decoder,
entering an escape-continuation-sequence (33) into the data stream,
the escape-continuation-sequence defining a second number of data
units that are to be skipped by a base decoder and to be interpreted
by an extension decoder together with the first number of data units;
and
entering the second number (34) of data units into the data stream,
wherein the escape-start-sequence (31) has an escape start code
(31a) and a length code (31b), or wherein the escape-continuation-
sequence (33) has a continuation code (33a) and a length code (33b),
wherein the escape start code (31a) or the continuation code (33a)
differ from the text codes, or wherein the escape start code and the
continuation code are identical to each other and wherein the a first
length code (31b) having a maximum length signals that a code
following the first length code in the data stream is a continuation
code.
11. A method for reading a data stream having text data (30), an escape-
start-sequence (31) defining a first number of data units (32), the first
number of data units (32), an escape-continuation-sequence (33)
defining a second number of data units, and the second number of
data units (34), comprising:
displaying (72) text data (30);
interpreting the escape-start-sequence (31) such that the first number
of data units is determined from the escape-start-sequence;
skipping the first number of data units based on the step of
interpreting the escape-start-sequence;

interpreting the escape-continuation-sequence (33) such that the
second number of data units (34) is determined from the escape-
continuation-sequence, wherein the escape-start-sequence (31) has an
escape start code (31a) and a length ode (31b), or wherein the
escape-continuation-sequence (33) has a continuation code (33a) and
a length code (33b), wherein the escape start code (31a) or the
.continuation code (33a) differ from the text codes, or wherein the
escape start code and the continuation code are identical to each other
and wherein the a first length code (31b) having a maximum length
signals that a code following the first length code in the data stream is
a continuation code.
12. A method for reading a data stream having text data (30), an escape-
start-sequence (31) defining a first number of data units (32), the first
number of data units (32), an escape-continuation-sequence (33)
defining a second number of data units, and the second number of
. data units (34), comprising:
interpreting the escape-start-sequence (31) such that the first number
of data units (32) is determined,
reading in the first number of data units based on the step of
interpreting the escape-start-sequence;
interpreting the escape-continuation-sequence (33) such that the
second number of data units (34) is determined;
reading in the second number of data units based on the step of
interpreting the escape-continuation-sequence; and
jointly processing (73) the first number of data units (32) and the
second number of data units (34) in addition to or instead of
displaying the text data (30), wherein the escape-start-sequence (31)
has an escape start code (31a) and a length code (31b), or wherein
the escape-continuation-sequence (33) has a continuation code (33a)

and a length code (33b), wherein the escape start code (31a) or the
continuation code (33a) differ from the text codes, or wherein the
escape start code and the continuation code are identical to each other
and wherein the a first length code (31b) having a maximum length
signals that a code following the first length code in the data stream is
continuation code.

ABSTRACT
When generating a data stream, text data, an escape-start-sequence, a first
number of data units, an escape-continuation-sequence and a second
number of data units are entered into a data stream A base decoder displays
only the text data and skips the data units referenced by the escape-start-
sequence and the escape-continuation-sequence, while an extension decoder
reads those data units and processes them together.