Selective data distribution method using unidirectional broadcast or multicast transmission.

0001The invention relates to a method for transmitting data over a unidirectional broadcast or multicast transmission medium. More specifically it relates to a method for transferring digitally encoded data from one or more information-providers, acting as the sources of information, to selected end-users or groups of end-user through a single broadcaster.

0002US-A-4,829,569 discloses a subscriber television system in which individual text messages may be directed to individual subscribers mixed with the TV programs. US-A-4,700,184 relates to a one-way data transmission system. A field of useful data is split into a number of basic entities, called "data groups". Each data group is, in turn, split into smaller basic entities or "data blocks". Each resulting data block is inserted in a data packet to be transmitted by means of a teletext like technique.

0003The cited prior art does not teach a method for simultaneous, concurrent transmission of data originating from a plurality of information-providers and directed to a broadcaster which, without using a central data base, retransmits such data to different end-users or different groups of end-users. Prior art suggests only a distinction between a transmission toward a single addressee and a broadcast transmission. According to the prior art, individual subscribers in a subscription television system can be addressed for transmission of either billing information or individual messages. The same scheme may be applied to non-subscription television system, substituting in the subscriber TV set the decoder with a special purpose device. However, in the prior art, the source of data to be distributed to individual addressees is unique; this is the case in which a central data base, updated by a plurality of information-providers, sends messages to different subscribers about their credit/billing status.

0004However prior art does not teach a method for implementing multicast transmissions of data provided by different information-providers, over a single broadcast system, for grouping the end-users in groups simultaneously receiving the same data and for dynamically modifying the end-users groups. There is the need of distributing data originated from different, unrelated sources that share the access to the data distribution system through a single broadcaster.

0005The present invention as claimed allows that several information-providers provide data to a plurality of end-users, each being able to select the desired service. In fact any end-user receives only the data that it is enabled to receive and no data discrimination has to be done by the end-user at its receiving station.

0006It is also desirable that the information-providers do not arrange the data content of the transmission in a centralised data base storage located in the broadcaster premises as it is the case for example, when the data are extremely variable in time and not worth of being deposited in a centralised data base for a successive transmission, or when there are high security requirements or extremely high volumes of data. It is evident that in the above situations each information-provider should be able to organise its transmission, select the end-users to be addressed and organise the addressees in dynamically arranged groups.

0007One advantage of the invention is that each information-provider is able to simultaneously distribute different data to different end-users, or groups of end-users. It is further desirable that each end-user be guaranteed of the identity of the information-provider, i.e. the source of the data from which it is currently receiving data. Another advantage of the invention is that the method may be used in a television network since the data can be intermixed with TV programmes using the standard teletext protocol.

0008For a better understanding of the present invention, together with other and further advantages and features thereof, reference is made to the following description and to the accompanying drawings.

0009Figure 1 represents a schematic view of the data distribution system between the information-providers and the end-users according to the present invention. Figure 2 shows the general packet structure. Figure 3 describes a block diagram of the apparatus implementing the transmission protocol. Figure 4 shows the preferred embodiment packet structure.

0010With reference to Figure 1, a plurality of information-providers (1a,..1i,..1N) are connected to a broadcaster (2) having the function of transmitting the received data through a broadcasting medium (3) to a plurality of end-users (4a,..4k,..4M). The broadcaster (2) is a specific unit having the function of collecting and transmitting the data received from the information-providers. However also the broadcaster may transmit data originated by itself, acting in this case as an information-provider.

0011The invention implements a selective distribution mechanism ensuring N to M unidirectional connections over broadcast medium between N information-providers and M end-users. Data is structured as digitally encoded packets having either data transfer or various control purposes, as it will be described in further detail. Each end-user has a unique identifier representing a permanent attribute. Selective transmission is achieved by establishing a connection between an information-provider and a specific end-user, each end-user being identified by a different unique identifier. It has to be noted that each information-provider may concurrently transmit, in time division, different messages or data arranged in packets to an end-user or to different end-users or groups of end-users; in each of these transmissions the information-provider may be referenced as "calling terminal". Conversely each end-user may be engaged simultaneously in different concurrent connections with one or more information-providers; in each of these connections the end-user may be referenced as "called terminal". Once a connection is established, it is possible to transmit data from the information-provider to the connected end-user. A connection may be terminated as a result of a specific command issued by the calling terminal or may automatically terminate after a given timeout; during the connection establishment phase, the calling terminal specifies the duration of the connection to the called terminal(s). This scheme has the advantage of ensuring that the connection will be automatically terminated without relying on recovery actions to be executed by either the calling terminal or the called terminal.

0012The parameters identifying a connection between a calling terminal and a called terminal are here reported: <ul id="ul0001" list-style="none"><li>1) 〈calling-terminal-address〉 constituted by: <ul id="ul0002" list-style="none"><li>* 〈information-provider-id〉, a unique identifier of the information-provider, and</li><li>* 〈multiplex-id〉, identifying a particular connection between an information-provider and one (or a group) of end-users; it is used to multiplex different connections originating from the same information-provider.</li></ul></li><li>2) 〈called-terminal-address〉 constituted by: <ul id="ul0003" list-style="none"><li>* 〈unique-id〉, a unique identifier of the called terminal; it is used only during the establishment of the connection.</li></ul></li></ul>

0013Once a connection is established the called terminal discriminates the received data packets only on the basis of 〈calling-terminal-address〉 parameter; this allows up to N simultaneous connections between one calling terminal and N called terminals, since all data packets originated from the same calling terminal include the same 〈calling-terminal-address〉. In such a case the calling terminal may sequentially establish N connections with N called terminals, one connection at a time; it can then proceed to multicast distribution of data to all connected called terminals. Called terminals can also be aggregated in groups so that the same connection and disconnection procedure may be applied to a group of called terminals instead of a specific called terminal at a time.

0014The method allows that a calling terminal splits up the called terminals into groups, so that a "single" connection procedure establishes the connection with "all" the members of a group. Groups may be static or dynamic entities; the capacity of creating or deleting a group, of adding or deleting a member in a group, resides entirely in the information-provider that is considered as a group owner. Group membership relations are status information maintained in a private storage in each end-user terminal but are not modifiable by the end-user. In each called terminal the membership relation is represented by a membership tuple (i.e. a combination of n elements). The maximum number of membership tuples that can be obtained is only a design issue.

0015According to the method of the present invention a central authority, namely the broadcaster, distributes periodically a clock message so that time alignment between the clocks of the information-providers and the end-users is maintained; this time alignment is necessary for terminating connections and membership relations.

0016Transmitted messages carrying data and control information are structured as digitally encoded packets. The following structure of the packet represents a "high level" representation of the required information to be carried by every packet. Depending on the specific embodiment, other fields (error correction, framing, masking fields) may be added into this packet but this is outside the scope of the present invention. Packets have the structure shown in Figure 2. A framing-code field providing packet delimitation is put before each packet. In the following the symbol " ::= " means "is composed of ": 〈packet〉 ::= 〈calling-terminal-address〉〈continuity-index〉〈info〉 where:   〈calling-terminal-address〉 ::=    〈information-provider-id〉〈multiplex-id〉 previously defined. 〈info〉 ::= 〈info-header〉〈info-body〉

0017The 〈calling-terminal-address〉 represents the address of the calling terminal and an identifier of a particular established connection. The 〈continuity-index〉 is a progressive index of the packets and allows to recognise if a packet is repeated. In fact, in this case, the 〈continuity-index〉 is kept constant; this allows any decoder of the called terminals to recognise between a new packet and a copy of the same packet. 〈info-header〉 is an identifier for specifying the packet type as indicated in the following. 〈info-body〉 is the information carried by a specific packet type; of course this field will have a structure that depends on the packet type. In addition to the data transfer process described above that requires the establishment of a connection between the calling terminals and the called terminals, another data transfer process is provided by the method specifically aimed at the transfer of short, high priority, messages called datagrams that can be contained in a single packet; said messages can be addressed either to a single called terminal or to a group of called terminals. The two processes can coexist simultaneously over the same transmission system.

0018The following are some types of 〈info-body〉: <ul id="ul0004" list-style="none"><li>1) Network Connection Request (NCR) packet. It establishes a connection between a calling terminal and a specific called terminal. NCR packet parameters are: 〈info-body〉 ::= 〈unique-id〉〈connection-duration〉 Where: 〈unique-id〉 is the address of the called terminal. 〈connection-duration〉 indicates the time after which the connection is automatically terminated. In addition to the above elements the preferred embodiment may also define a 〈class-of-service〉 parameter that indicates a particular category of data transfer and error detection and recovery algorithms to be applied during data transmission. </li><li>2) Network Disconnection Request (NDR) packet. It terminates a connection between a calling terminal and a specific called terminal. NDR packet parameters are: 〈info-body〉 ::= 〈unique-id〉 Where 〈unique-id〉 is the address of the called terminal.</li><li>3) Network Group Connection Request (NGCR) packet. It establishes a connection between a calling terminal and a group of called terminals; the group address of the called terminals is contained in a parameter field. NGCR packet parameters are: 〈info-body〉 ::= 〈group-id〉〈connection-duration〉 Where: 〈group-id〉 univocally identifies the group of called terminals. 〈connection-duration〉 indicates the time after which the connection is automatically terminated. With reference to the NCR and NGCR packets, in addition to the above parameters a preferred embodiment also includes another parameter called 〈class-of-service〉, that indicates a particular category of data transfer and error detection and recovery algorithms to be applied during data transmission. </li><li>4) Network Group Disconnection Request (NGDR) packet. It terminates a connection between a calling terminal and a group of called terminals. NGDR packet parameters are: 〈info-body〉 ::= 〈group-id〉 Where 〈group-id〉 univocally identifies the group of called terminals.</li><li>5) Network Information (NI) packet. It transports data for a pre-established connection between a calling terminal and one or more called terminals or groups of called terminals. NI packet parameters are: 〈info-body〉 ::= 〈data-unit〉 Where 〈data-unit〉 is the unit of information to be transferred from a calling terminal to one or a group of specific called terminals.</li><li>6) Network Exception (NE) packet. This packet will be used to convey errors detected at the calling terminal site. NE packet parameters are: 〈info-body〉 ::= 〈exception-message〉 Where 〈exception-message〉 indicates the related code for the signalled exception or any other useful information for exception handling.</li><li>7) Network Unsolicited Information (NUI) packet. This packet will be used to convey a datagram to a specific called terminal. NUI packet parameters are: 〈info-body〉 ::= 〈unique-id〉〈message-length〉〈message〉 Where: 〈unique-id〉 is the address of the called terminal. 〈message-length〉 identifies the length of the field 〈message〉. 〈message〉 is the unit of information to be transferred from a calling terminal to a specific called terminal.</li><li>8) Network Group Unsolicited Information (NGUI) packet. This packet will be used to convey a datagram to a group of called terminals. NGUI packet parameters are: 〈info-body〉 ::= 〈group-id〉〈message-length〉〈message〉 Where: 〈group-id〉 univocally identifies the group of called terminals. 〈message-length〉 identifies the length of the field 〈message〉. 〈message〉 is the unit of information to be transferred from a calling terminal to a specific group of called terminals.</li><li>9) Network Group Membership (NGM) packet. This packet will be used to grant a membership relation to a specific called terminal. NGM packet parameters are: 〈info-body〉 ::= 〈unique-id〉〈number-of-membership-tuple〉 〈membership-tuple〉 Where: 〈unique-id〉 is the address of the called terminal. 〈number-of-membership-tuple〉 indicates the number of group membership relations contained in the NGM packet. 〈membership-tuple〉 ::= 〈group-id〉〈information-provider-id〉 〈membership-duration〉 Where: 〈membership-tuple〉 represents the group membership relation. 〈information-provider-id〉 identifies the group owner. 〈group-id〉 univocally identifies a specific group within the totally of groups owned by a single information-provider. 〈membership-duration〉 indicates the time after which the membership relation expires.</li><li>10) Network Group Membership Deletion (NGMD) packet. This packet will be used to revoke a membership relation to a specific called terminal. NGMD packet parameters are: 〈info-body〉 ::= 〈unique-id〉〈number-of-membership-tuples〉 〈membership-tuple〉 〈membership-tuple〉 ::= 〈group-id〉〈information-provider-id〉 〈membership-duration〉 Where: 〈unique-id〉 is the address of the called terminal. 〈number-of-membership-tuples〉 indicates the number of group membership relations contained in the NGM packet. 〈membership-tuple〉 represent the group membership relation. 〈information-provider-id〉 identifies the group owner. 〈group-id〉 univocally identifies a specific group within the totally of groups owned by a single information-provider. 〈membership-duration〉 indicates the time after which the membership relation expires.</li><li>11) Network Time Service (NTS) packet. This packet will be used to distribute a time stamp. NTS packet parameters are: 〈info-body〉 ::= 〈time-stamp〉 Where 〈time-stamp〉 is a time expression having a defined granularity. This packet type is created by the broadcaster for synchronisation purposes.</li></ul>

0019Hereinafter a detailed description of the data distribution method is given with reference to previously defined packet structures and types. The selective connection between a calling terminal and a called terminal over the unidirectional broadcast medium is achieved by means of the NCR packet that carries the address of the called terminal. The calling terminal sends an NCR packet to all end-users. At each terminal upon the reception of the NCR packet, the 〈unique-id〉 contained in the packet is compared with its own address; if they match, a connection between the calling terminal and the called terminal is established. Every other terminal having different 〈unique-id〉 ignores the NCR packet. Once a connection is established, the 〈connection-duration〉 is stored in the called terminal and the expiration time is computed through the internal clock, that is kept aligned with a central clock via the NTS packet.

0020Then the calling terminal sends the NI packets. Data to be transmitted are sent as a sequence of NI packets each containing a 〈data-unit〉. NI packets will be accepted by the called terminal only after a check of the 〈calling-terminal-address〉 to verify that they belong to an already established connection; no check of 〈unique-id〉 is needed on these NI packets once the connection was established and the called terminal remains in the connected status. Any end-user may be concurrently engaged in several connections. Every active connection is identified by the end-user by means of 〈information-provider-id〉 and 〈multiplex-id〉 parameters. Every information-provider can maintain multiple connections with different called terminals; this is referred to as the "instantiation" of many calling terminals within the information-provider, each calling terminal tied to a different connection and univocally identified by the pair of parameters: 〈information-provider-id〉 and 〈multiplex-id〉; all connections originating from a single information-provider will share the same 〈information-provider-id〉 but have different 〈multiplex-id〉. Connections may be mapped from 1 to N; in this case NCR packets are sent from the calling terminal up to N called terminals. A connection can be terminated either explicitly at calling terminal request via an NDR packet, or automatically after the expiration of a period of time specified in 〈connection-duration〉 parameter contained in the NCR packet. Reception of the NDR packet by the called terminal will cause a connection termination provided that the 〈unique-id〉 carried by the NDR packet matches the address of the called terminal. Each information-provider may enable a specific called terminal to be included in a specific group of which it is the "owner". Membership tuples are distributed from such information-provider on the NGM packet addressed to said specific terminal. Membership can either expire after a given period of time or be explicitly revoked at any time by the information-provider that is the group owner, sending an NGMD packet. The same value of 〈group-id〉 can represent different groups provided that they are owned by different information-providers having different 〈information-provider-id〉; this prevents conflicts in the attribution of 〈group-id〉. Group membership relation is represented by the following tuple: 〈group-id〉〈information-provider-id〉〈membership-duration〉 whereas 〈information-provider-id〉 univocally identifies the group owner that splits the plurality of the called terminals into groups; 〈group-id〉 univocally identifies a specific group within the totality of groups owned by a single information-provider; no agreement is needed among the information-providers to ensure univocity of the 〈group-id〉 since it is always referred to the specific information provider. 〈membership-duration〉 is a required attribute of each membership tuple; it is a multiple of a given unit of time and specifies the duration of the membership relation for a specific called terminal. After that time the membership tuple is removed from a table present in the private storage of the called terminal. This means that the specific terminal is no longer a member of that group. A membership tuple can be removed from the membership table of a specific called terminal also before the expiration of 〈membership-duration〉; this is obtained by a NGMD packet sent from the information-provider, owner of the group, and addressed to the specific called terminal. The effect of such packet is to remove a specific terminal from a specific group. A connection between a calling terminal and a group of called terminals can be achieved by means of an NGCR packet that carries the 〈group-id〉 of the called terminals. A group connection is accepted if the end-user is currently a member of the group specified in the 〈group-id〉 carried by the NGCR packet. Once a connection is established, its expiration time is computed according to the 〈connection-duration〉 parameter.

0021Also during a multiple connection data is organised as a sequence of NI packets, each containing a 〈data-unit〉; NI packets will be accepted by each called terminal only after a check of the 〈calling-terminal-address〉 to verify that they belong to an already established connection; no check of the 〈group-id〉 is needed on these NI packets once the connection has been established. Every active connection is identified by the following parameters: 〈information-provider-id〉, 〈multiplex-id〉. Each information-provider can maintain multiple connections with different groups of end-users; all connections originating from a single information-provider will share the same 〈information-provider-id〉 but have different 〈multiplex-id〉. This may be again referred to as the "istantation" of many calling terminals within the information-provider as already said in the case of single end-users (not grouped). The activation of 1 to N connections can be obtained via a single NGCR packet provided that all called terminals belong to the same group. Connection can be terminated either explicitly at calling terminal request (via a NGDR packet) or automatically after the expiration of a period of time that is a parameter contained in the NGCR packet. This value is referenced to as 〈connection-duration〉. Reception of the NGDR packet by the called terminal will cause a connection termination provided that 〈group-id〉, 〈information-provider-id〉 and 〈multiplex-id〉 of the group address match one of the currently active group connections. A successful disconnection can happen simultaneously at many called terminals. A connection activated with an NGCR packet can be terminated with an NDR packet; this means removing a specific called terminal from a group of called terminals.

0022The data distribution method here described requires a clock alignment maintained in every called terminal with the broadcaster and is used to enforce expiration of single connection, group connection and membership relation. An NTS packet is used for giving the clock time to every end-user.

0023The data distribution method here described comprises also the possibility of selective distribution of high priority, asynchronous messages contained in a single NUI packet to a single end-user. This NUI packet contains both the message and the 〈unique-id〉 of the addressee. Data contained in the NUI packet is acquired by the end-user if the 〈unique-id〉 in the NUI packet matches the unique identifier of the called terminal. No connection activation phase is required in this case. The data distribution method here described allows also the selective distribution of high priority, asynchronous messages contained in a single NGUI packet to a group of end-users. The NGUI packet contains both the messages and the 〈group-id〉 of the addressees. Data contained in the NGUI packet is received by the end-user if both 〈group-id〉 and 〈information-provider-id〉 in the NGUI packet match one of the membership tuples currently maintained in that called terminal.

0024An apparatus for implementing the transmission protocol is described in Fig. 3. A command, for example specifying the request to send a message or data to an end-user, is issued from the calling terminal to a Command Interpreter through an external interface. The decoded command is sent to a Communication Supervisor that controls all transmission functions, selectively activating the different logic blocks each of them implementing the specific transmission protocol function ( i. e. connection, disconnection, data transfer, etc.). The Calling Terminal Address Generator block provides the couple of parameters 〈information-provider-id〉 and 〈multiplex-id〉 forming the address of the calling terminal. This couple of parameters is put in every packet pertaining to a specific connection by the Packet Formatter Sequencer block. The Distribution List Manager block is responsible for updating the Distribution List Database containing the called terminal address and the group membership relations of the called terminal enabled to receive. It is also responsible for sending a request for the transmission of the NGM and NGMD packets to the Packet Formatter/Sequencer block. All the queries and the updating requests directed to the Distribution List Database block, are handled by the Database Manager indicated in the figure with the Distribution List Database Interface block. In the same manner, all queries directed to the Data/Message Database block are handled by the Data/Message Interface block. The Connection/Disconnection Manager block receives the called terminal address from the Distribution List Database block and issues requests for transmission of NCR, NDR, NGCR, NGDR packets to the Packet Formatter/Sequencer block. The messages and the data to be transmitted reside in the Data/Message Database. The Data Transfer Manager block controls the segmentation of the data performed by the Data/Segmentation block. The Data transfer Manager block handles also the Exception events through the Error Handler block that, in turn is responsible for sending an NE packet request to the Packet Formatter/Sequencer block. Error Handler block also receives input from the Communication Supervisor and the Unsolicited Message Handler block. This block is responsible of managing the transmission of NUI, NGUI packets. To do this the called terminal address is requested to the Distribution List Database. A Message Preparation block keeps the message from the Data/Message Database, assembles it with the called terminal address and send it to the Packet Formatter/Sequencer block. A Time Handler block is present only if the information-provider acts also as the broadcaster; in this case it sends a time stamp to all end-users for remote clock synchronisation purposes. The Packet Formatter/Sequencer block implements the algorithms for assembling the packet according to the structure defined for every type of packet. This block may also have storage capacity.

0025In a preferred embodiment the method for the unidirectional transmission of data is used for transmitting messages using as a support a TV signal according to CCIR specifications. In such embodiment the broadcaster receives messages originated from different information-providers and organises them so that they can be transmitted intermixed with conventional transmissions of TV programmes, for example during the field blanking interval or alone in a full-channel mode preserving the structure of the television synchronising signal. Information-providers transfer data to the broadcaster using TP links or on a magnetic media support as disks or tapes. In the preferred embodiment the above digitally encoded packets are transmitted concurrently with TV transmissions and this requires a fixed packet structure, such as the Teletext standard (Broadcast Teletext Specification ISBN 0 563 17261 4). The preferred embodiment of the invention is described with reference to the transmission method of ITA - 1215752 to RAI - Radio Televisione Italiana. Such type of packet has a fixed length of 42 bytes plus 3 bytes for clock synchronisation and packet delimitation. The following Table is referenced to the structure set forth in Fig. 4. <tables id="tabl0001" num="0001"><img file="EP0491069A1_D0001.tif" /></tables>

0026It has to be noted that the fields (MRAG + PA) also include the 〈calling-terminal-address〉 parameter, the fields (CI + CO) include the 〈continuity-index〉 parameter, and 〈info-header〉 and 〈info-body〉 parameters are embodied in the DATA BLOCK field.

0027The broadcaster also act as an authentication authority having a "secure" mean to ascertain the identity of the information-provider from which it receives data. Once the identity of the information-provider has been ascertained, the broadcaster checks the Packet Address (PA) of every packet presumably coming from an information-provider and verifies if the 〈information-provider-id〉 field is in accordance with the ascertained identity. This embodiment strives to efficiently convey data or control information over previously defined teletext packet structure as shown in Fig. 4.

0028The end-user is provided with a decoder (or receiving adapter) with the following features: <ul id="ul0005" list-style="none"><li>1) Its own 〈unique-id〉 which represents the unique identifier referenced above. The specific 〈unique-id〉 is stored in an inalterable storage of the receiving adapter.</li><li>2) A group-membership table containing all the groups of which the end-user is a member and the membership duration.</li><li>3) A clock to be synchronised with the clock of the broadcaster.</li></ul>

0029An implementation of such adapter is described in co-pending application titled "Selective data broadcasting receiver adapter for personal computers".

0030None of the above features are instrumental in order to implement the described method but rather they rely with robustness of the implementation in terms of protection against non authorised end-users. The practical implementation may include some or all the above features. The information-providers should have a system to organise their transmissions in data packets, to aggregate end-users into groups, to enable end-users to the reception of data and to disable end-users from the reception of data. Conversely each end-user should have a receiving adapter and a system to recognise packets, to filter packets, to reassemble data.