JPH1032573A - Receiving protocol equipment, multi-address message transmitter and transmission and reception protocol equipment - Google Patents

Abstract

PROBLEM TO BE SOLVED: To receive an impartial communication service even at places different in transmission delay by providing a means setting a transmitting time to a previously consented time and a means holding a message until the time set before transmitting this message. SOLUTION: In order to process the same messages transmitted from a transmitter by simultaneously transmitting them to plural high-order equipments, a timer function 301 setting the transmitting time to the time previously consented with another receiving protocol equipment is provided. A message storing function 303 holding this message until the time set by the timer function 301 before transmitting the message from the transmitter to the plural high-order execution equipments is provided. Thereby, messages can impartially be transferred within the range of the time synchronizing error of a terminal through a network by setting the absolute time of a client and making the message transferring time to the client coincident with it.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a transmission / reception protocol device, and more particularly, to a method of broadcasting one message to a plurality of devices via a communication network, the fairness of message delivery time between the transmission / reception protocol devices. , Equipment for ensuring fairness.

[0002]

2. Description of the Related Art Broadcasting is a communication system useful for broadcasting using a communication network and distributed information processing applications.
A highly reliable broadcast communication system is required for computer applications and information providing services. Also,
In order to make the most of the characteristics of broadcast communication, there is a case where the same result can be obtained in all receiving apparatuses. This property can be applied to, for example, a method of maintaining data identity when storing data in a distributed manner, and a method of providing information fairly among subscribers.

[0003] A protocol which guarantees the following properties for these requests has been proposed and realized. Here, a set of receiving apparatuses that receive the same broadcast message is called a broadcast group.

[0004] Regarding the delivery of a (atomic) message, the message is not delivered unless it is received by all receiving devices. That is, if a broadcast message is received by a certain receiving device, it is guaranteed that the message is delivered to all other receiving devices belonging to the same broadcast group.

(Order Equivalent) The received information of a message is the same in all the receiving devices of the broadcast group. The major factors that reduce the reliability of broadcast communication are bit errors in message transmission and discard during transmission. Bit errors are caused by transmission line errors, and discards are caused by congestion (here, congestion refers to congestion in packet networks and ATM networks). The error correction technique is required between terminals for bit errors, and retransmission is performed for discards, so that the probability of message delivery failure can be reduced. Of course, retransmission is also effective for message delivery failure due to bit errors, but error correction techniques are not effective for congestion over a time period such that a message is completely lost. To ensure reliability against congestion, a retransmission technique is indispensable.

On the other hand, in order to realize the property that the same reception result can be obtained in all receiving apparatuses, failure in message delivery and non-uniformity of transmission delay become obstacles. Although the failure of message delivery can be recovered by retransmission as described above, the retransmission causes the message delivery time to be sent. Also, the difference in transmission delay affects the message delivery time.

[0007] A conventional technique which realizes the broadcast communication satisfying the above two properties will be described below. A device that manages the transmission and reception of messages in a broadcast group is called a master. A device that receives a message is called a client. A device that sends a message is called a sender. The master, the client, and the sender are each conceptually divided into a functional part for processing a broadcast protocol and a functional part for performing a broadcast application using the function of the protocol processing.

[0008] Here, taking the provision of stock price information by broadcast as an example, an application is a part that broadcasts stock price information in a sender, and a receiving side processes electronic information based on the information, for example, based on the information. It is in the part that performs a structured transaction.

Although the protocol processing function and the application function are conceptually separated, their implementation is generally not clearly separated. In many personal computers, even though the protocol processing function and the application function are separated as functional parts of the software,
Both functions are executed by the same processor on the same memory space.

The procedure of broadcast reception will be described below with reference to FIG. The master 701 transmits a message to the clients A 702 and B 703 through the broadcast connection.
The message is given an identifier having a continuous order relationship. The message with the identifier p is denoted as M _p .

[0011] client A702, B703 acknowledgment respectively upon receipt of the _{M p ACK p (A),} ACK p
(B) is transmitted to the master 701. The acknowledgment includes the message identifier p and the client identifier. This response need not be broadcast. When the master 701 confirms the acknowledgment ACK _p (X) from all the clients, here the clients A 702 and B 703, it broadcasts the message release permission REL _p (i). i is an identifier of a message for which release permission has been issued, and is assigned in the order of release permission, such as i, i + 1,.

Next, a procedure in the case where a broadcast reception message is lost will be described with reference to FIG. The notation here is the same as in FIG. Master 701 sends a message M _p to the client A702, B 703 through the broadcast connection. The master 701 sets the timer T1 simultaneously with the transmission. Client A 702 receives M _p at time t ₀ . On the other hand, it is assumed that the message to the client B 703 has been lost. At this time, the server sets the timer T1
Detects that the confirmation response from the client B703 corresponding to the message Mp at the time of the time-out has not arrived, the retransmission of the M _p of the message. Upon receiving M _p , client B 703 sends an acknowledgment ACK _p (B). When receiving the ACK _p (B), the master 701 broadcasts the message release permission REL _p (i) because the acknowledgments of all the clients are ready.

In the retransmission of a message, the message is broadcast to all the clients, and the client that has received the message transmits an acknowledgment again to the message that has already acknowledged. As a result, even when the acknowledgment ACK _p (X) is lost, the master 701 can know that all the clients have received the message by the acknowledgment for the retransmission after the timeout.

The procedure in the case where the release permission message for broadcast reception is lost will be described below with reference to FIG. The notation is the same as in FIG. Here, a case is shown where the release permission message REL _p (i) to the client B 703 for the message M _p has disappeared.
Client A receives the release permission message normally, has released a message M _p to time t _2.

In the client B 703, the message Mp
Is not released, the reception procedure of the next message M _{p + 1} is normally performed, and the release permission message RE
Lp _{+ 1} is received by the client B 703. The client B 703 receives the identifier p + 1 of the release permission message immediately before receiving the identifier p + 1 of the release permission message.
And discontinuity is detected. Upon detecting the discontinuity, the client B 703 determines that the release permission message has disappeared, and releases the corresponding message. In this case, the message M corresponding to p between the identifiers p-1 and p + 1
Release _p .

Since the message is released upon detecting the discontinuity of the identifier of the release permission message, the message M _p is released before the message M _{p + 1} and the receiving order is maintained.

The procedure in the case of message delivery failure will be described below with reference to FIG. The notation is the same as in FIG. Master 701 sends a message M _p to the client A702, B 703 through the broadcast connection. The master 701 sets the timer T1 simultaneously with transmission.

Client A 702 receives M _p at time t ₀ . On the other hand, the message to the client B 703 disappears. The server detects that an acknowledgment response from the client B 703 corresponding to the message M _p has not arrived at the time when the timer T1 times out, and the message M _p
Is resent. If a reception acknowledgment is not obtained from the client B 703 after repeating this procedure a predetermined number of times n times, the message delivery is determined to have failed and the master 701 broadcasts a transmission failure message FAIL _p .
Clients and the A702, which has received the transmission failure message is cancel the delivery of the message M _p.

The above protocol has the following two problems. 1. Unfairness due to Transmission Delay In FIG. 16, there is a time difference t _d between the time t _{3 at} which the client A 702 receives the release permission and the time t ₂ at which the client A 702 receives the release permission due to the difference in transmission delay. This difference is unfair with the time the message is passed to the application. 2. Early up client A702 of the message can be will read the message at the time t ₀ immediately receive a message M _p. Of course, this is not allowed under the protocol rules, but it is impossible to check whether or not such a violation has been performed from another network device unless the implementation of the protocol is checked.

Further, even if the protocol implementation of the client device is legitimate, if a device on the network or a function in the client device intercepts the message, the message can be sent earlier than other client devices. It is possible to read. In the case of the personal computer described above, the application function can intercept by reading the working memory of the protocol processing,
If the receiving device is a device on the user side, it is easy to make such a change.

Next, a description will be given of a conventional technique for matching the message receiving order of all clients in a broadcast network having a plurality of senders.

The operation will be described with reference to FIG. Here, the broadcast group includes a master 801, clients A802 and B803, senders a804 and b805. Sender a804 is to broadcast a message M _p ^a to time t _0. M _p ^a denotes that the message having an identifier p sent from the sender A 804. p client A802 is a unique identifier for the sender a, B 803 message M _p each receives the ^{_{a ACK p a (A),}} ACK p a
(B) is transmitted to the sender a 804. Sender a804 When receiving the ACK from all the client sends a message release permission request RELREQ _p ^a master 801.

Message release permission request RELREQ
master 801, which has received the _p ^a (i) is to broadcast a message release permission request RELREQ _p ^a (i). The client receiving the release permission releases the message.

Next, the procedure for transmitting a message from a plurality of senders will be described. Sender a 804 sends message M at time t _{1
Broadcast p + 1} ^a . Thereafter, the sender b 805 sets the time t2
The message M _q ^b broadcasts to. Clients A 802 and B 803 that have received these messages respectively receive acknowledgments ACK _{p + 1} ^a (A), ACK _q ^b (A), and ACK
_{p + 1} ^a (B) and ACK _q ^b (B) are converted to senders a804 and b
Sent to 805, and transmits the release request RELREQ _{p + 1} ^a, the RELREQ _q ^b to the master 801 at the sender with uniform the ACK.

The order in which the messages are received by the clients A 802 and B 803 does not always match the order of transmission due to transmission delay, delay fluctuation, and the like. Master 801
Broadcasts release permission from the message that received the release request. The release permission is given a unique identifier i, i + 1,... For the broadcast group. Therefore, even if the clients A 802 and B 803 fail to receive the release permission or the receiving order is reversed, these clients can detect it, and all the clients maintain the release order of the message determined by the server. It is.

Here, in client A 802, sender b
The message from 805 arrives first, but the message from sender a 804 arrives first at client B 803. The sender a 804 has to wait for all the acknowledgments of all the clients, and the transmission of the release request is later than that of the sender b 805.

Since the master 801 issues a release permission in accordance with the order in which the release requests are accepted, a message having the latest arrival of the release request to the master 801 has the slowest release.

Here, the message transmitted by the sender a 804 at the time t ₁ is the master 801 and the client B 803.
In has arrived earlier than messages sent by the sender b805 is delayed by the time t _2. However, client A8
02, the message of sender b 805 is sent to sender a 804
Arrived earlier. Eventually, even if the ACK arrives at the master 801 late even if it has been transmitted earlier, the sender a
The release of the message 804 is later than the release of the message of the sender b 805 transmitted later.

Further, when the implementation of the protocol is illegal, the release time of the message can be delayed by not intentionally returning an acknowledgment of the message. If the sender and the client are one, by transmitting the message after receiving the message from the other sender and then transmitting the ACK for the received message, it is as if the own device has already transmitted the message to the other client. The receiving order can be changed as if the message was sent.

As described above, in the broadcast protocol having a plurality of senders, although the order relation of the received messages is maintained in all the clients, the time relation of the sending side message transmission between different senders is not necessarily preserved. is not.

In particular, when a communication network having fluctuation in transmission delay is used, it is difficult to correct these effects. Also, when retransmission occurs, the same effect as if the arrival of the message was delayed is obtained.

Broadcast protocols with multiple senders include:
Although there is unfairness in betting due to differences in transmission delays and stochastic delays in message delivery due to retransmission, it is in principle impossible to solve this problem on a protocol based on message identification numbers. It is possible. As one solution to this problem, there is known a method in which a sender assigns a message transmission time and the receiving side evaluates the time to determine a message reception order (Birman, K., Schiper,
A., Stephenson, P.: "Lightweight Causal and Atomic Gr
oup Multicast ", ACM Trans.Computer systems, 9 (3):
However, this method cannot prevent the sender from manipulating the message order by falsifying the time.

As described above, even if the atomicity and the order equivalence are guaranteed by the protocol, it is necessary to consider the time required for the application to read the message and the possibility of performing the illegal protocol processing. It was not necessarily fair. This is an obstacle to using broadcast for e-commerce and news distribution affecting transactions.

[0034]

That is, the prior art first has a problem of unfair message release time. That is, unfairness occurs at the time when the message becomes available due to the difference in transmission delay. Second, there is a possibility of a protocol implementation violation. In other words, the reception of another receiving terminal can be sent by falsifying the message reception failure. Further, it is possible to cause other clients to determine that the message transmitted by the own device after reading the received message is the reverse message order. Third, there is the problem of message interception. That is, there is a possibility that the application may read the message in a state where the other recipients have not received the message (reception confirmation state).

The present invention has been made in view of such a problem, and an object of the present invention is to provide a method which is not affected by a difference in transmission delay or a stochastic message delivery delay due to retransmission.
In addition to matching the reception time of the broadcast message reception time, the message transmission time as seen from the user, that is, the time at which the application passed the message to the broadcast protocol device, is correctly reflected in the reception order in other reception devices. Another object of the present invention is to provide a reception protocol device, a broadcast message transmission device, and a transmission / reception protocol device that can guarantee the fairness and fairness of message delivery time between transmission / reception devices.

[0036]

According to a first aspect of the present invention, there is provided a broadcast communication for transmitting the same message from one or more transmitting devices to a plurality of predetermined receiving devices. In the receiving protocol device in the system, in order to simultaneously transmit the same message transmitted from the transmitting device to a plurality of higher-level execution devices for processing, the transmission time is set to a time agreed in advance with another receiving protocol device. Time setting means for setting, and maintenance means for holding the message from the transmitting apparatus until the time set by the time setting means before the message is transmitted to the plurality of higher-level execution apparatuses.

[0037] In the second invention, an identifier storage unit in which an identifier for identifying the reception protocol device from another reception protocol device is stored, and an identifier stored in the identifier storage unit is stored in a third party. A first prohibiting unit for prohibiting access from a third party, and a second prohibiting unit for prohibiting a third party from intercepting the identifier when transmitting the identifier stored in the identifier storage unit to the transmitting device for authentication. Prohibiting means, an encryption key holding unit for holding the same encryption key as an encryption key used for encrypting a message in the transmission device, and an encryption message transmitted from the transmission device for encrypting the encrypted message. Decryption means for decrypting using the encryption key stored in the key storage unit, wherein the storage means stores the decrypted message until the time set by the time setting means is reached. To hold.

According to a third aspect of the present invention, there is provided a broadcast message transmitting apparatus for transmitting a same message from one or more transmitting apparatuses to a plurality of predetermined receiving protocol apparatuses. Authentication means for authenticating whether or not the device has a predetermined identifier, distribution means for pre-distributing a predetermined encryption key only to the receiving protocol device authenticated by the authentication means,
A transmission means for encrypting a transmission message using the same encryption key as the encryption key distributed to each reception protocol device and transmitting the encrypted message to each reception protocol device.

Further, in the fourth invention, the authentication means communicates with each of the receiving protocol devices by storing an identifier for storing an identifier for identifying each of the receiving protocol devices. Confirming means for confirming whether or not the identifier is included in the identifier storage unit, and when the identifier is included in the identifier storage unit, recognizes the reception protocol device as a legitimate reception protocol device, When the identifier is not included in the identifier storage unit, a certification unit that certifies that the reception protocol device is not a legitimate reception protocol device is provided. Authentication is performed using an identifier for receiving protocol device identification stored in an identifier storage unit provided outside the broadcast message communication device.

A sixth invention is a transmission / reception protocol device comprising the reception protocol device according to the first or second invention, and the broadcast message transmission device according to the third, fourth, or fifth invention.

The seventh invention is a transmission / reception protocol device in which the reception protocol device according to the second invention and the broadcast message transmission device according to the fourth invention are constituted by a single device.

Further, the present invention has the following features. That is, the message transmitted from the transmitting device includes information on a unique transmission time or transmission order, and the reception protocol device transmits the received message to the higher-level execution device according to the information on the transmission time or transmission order.

Further, the message is a message transmitted from a plurality of transmitting devices, and the message includes information on a unique transmission time or transmission order. Further, the encryption key is changed at a predetermined timing.

Further, the first message is a message transmitted from a plurality of transmitting devices, and the receiving protocol device delays a transmission timing according to each transmitting device when transmitting the message to the higher-level execution device. . Further, the reception protocol device has storage means for recording the failure when the failure of the lower communication network is detected.

[0045]

DESCRIPTION OF THE PREFERRED EMBODIMENTS First, the outline of an embodiment of the present invention will be described. To solve the problem of the message release time, if the absolute time of the client is adjusted and the message transfer time to the client is matched, a fair message transfer within the time synchronization error of the terminal through the network becomes possible. However, the problem of protocol implementation violations is never solved by the broadcast protocol itself. For example, adding a past time stamp to a transmitted message cannot be prevented as long as there is a possibility that the message is lost.

Even if the protocol processing is completely correct, it is possible to quickly read a message that is not in the receiving state according to the protocol by directly reading the memory for storing the message. Remains. This problem can be solved by performing encryption so that a message cannot be obtained unless proper protocol processing is performed, and at the same time, preventing unauthorized reading of internal information of the decryption device. it can.

Therefore, in the present embodiment, first, a clock synchronized with the broadcast group is provided in the protocol device, and agreement is made on the time at which the message is passed to the application in the broadcast group. Incorporating means for determining whether or not a device has been authenticated by a trusted authority into a protocol and a device that is an implementation of the protocol;
In addition, the broadcast message is encrypted, the protocol device accumulates the decryption result of the message, and solves the above-mentioned problem by outputting the message at the agreed time, thereby enabling the targeted fair protocol processing. It is.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a diagram showing a basic configuration of the first embodiment. In FIG. 1, 1 is a broadcast transmitting device, 2 is a receiving device identifier table, 11-1 to n are broadcast receiving client devices (client devices), and 13-1 to 13-1.
n is a network side interface device (network interface device), 12-1 to n are broadcast receiving protocol devices (receiving protocol devices), 11-1 to n are broadcast receiving client devices (client devices), 14-1 to n Is an upper interface point, 15-1 to n are lower interface points, 1
6-1 is a network-side interface point, 17-1 to 17-n are reception protocol device identifiers (reception device identifiers), and 31 is a communication network. 32-1 is a client device 11-1
A point-to-point connection 33 is set between the broadcast transmission device 1 and the broadcast transmission device 1. Reference numeral 33 denotes a broadcast connection composed of the broadcast transmission device 1 and the client devices 11-1 to 11-n.

The network interface devices 13-1 to 13-n and the broadcast transmitting device 1 and the communication network 31 to which they are connected are connected if the identifiers of the network interface devices 13-1 to n and the broadcast transmitting device 1 are known. Any device can be used as long as it can be set and has a broadcast function.
In this case, each of the network interface devices 13-1 to 13-n and the broadcast transmitting device 1 is uniquely identified by E.E. It is assumed that an ATM interface specification (ATM forum UNI 3.1 specification, The ATM Forum, 1994) having an address of 164 format and having a point-multipoint connection setting means is used. Here, a group of devices that receive the broadcast of the message is collectively referred to as a broadcast group. FIG. 2 is a flowchart schematically illustrating the operation of the reception protocol devices 12-1 to 12-n. S101 represents a broadcast group subscription procedure, S111 represents a broadcast message reception procedure, S121 represents a broadcast end determination procedure, 141 represents a reception error processing procedure, S151 represents a broadcast group reception termination procedure, and S131 represents a key change processing.

FIG. 3 shows the receiving protocol devices 12-1 to 12-n
6 is a flowchart showing details of a broadcast group joining procedure S101. S102 is a connection setting procedure, S10
3 is a receiving protocol device identifier certifying procedure, S104
Is an encryption key sharing procedure with the broadcast transmitting apparatus 1, and S105 is a broadcast decryption key acquisition procedure.

FIG. 4 is a flowchart showing details of the broadcast message receiving procedure S111 of the receiving protocol devices 12-1 to 12-n. S122 is a message receiving step;
S123 is a confirmation response transmission step, S124 is a release permission reception timeout determination step, S125 is a release permission reception step, S126 is a release permission confirmation response transmission step, S127 is a message output time determination step, S128 is a release time re-reception step, S13.
0 is a control variable j increment step, S131 is a control variable determination step, S132 is an error process, and S140 is a message output step.

FIG. 5 shows a message sequence when the broadcast message receiving procedure is normal. Reference numeral 601 denotes a transmitting device;
Reference numerals 2 and 603 denote receiving apparatuses A and B, respectively. FIG. 6 shows a message sequence when the transfer of the message release permission in the broadcast message receiving procedure fails. Reference numeral 601 denotes a transmitting device, and 602 and 603 denote receiving devices A and B, respectively.

FIG. 7 shows a broadcast message sequence for obtaining agreement on the message order of a plurality of senders. 611 is a master, 612 and 613 are clients A, B, 61
4,615 represents senders a and b, respectively.

FIG. 8 is a functional block diagram of the broadcast transmitting apparatus 1 shown in FIG. 1 is a broadcast transmission device, 2 is a reception protocol device identifier table, 3 is a broadcast message input terminal,
202 is a broadcast message input function, 203 is a broadcast protocol state management function, 204 is a broadcast message encryption function, 205 is a message authenticator addition function, 206 is a message authentication function, 211 is a clock function, and 212 is a broadcast group. A management function, 213 is a connection setting function, 214 is a device identifier function, and 221 is a network interface function. FIG. 9 shows the reception protocol device 12-1 shown in FIG.
1 to n are functional block diagrams. 14 is an upper interface point, 15 is a lower interface point, 17 is a receiving protocol device identifier, 301 is a clock function, 302 is an upper interface function, 303 is a message storage function, 304
Is a message decryption function, 305 is a broadcast protocol state management function, 306 is a message authenticator addition function, 307
Is a message authenticator inspection function, 308 is a device interface function described below, 309 is a connection setting function, 310
Represents a received message processing scheduling function.

FIG. 10 is a schematic flowchart of the broadcast group management operation of the broadcast transmitting apparatus 1. S401 is a broadcast group initialization procedure, S411 is a broadcast group joining / withdrawing request procedure, S42.
1 is a broadcast group subscription acceptance procedure, S431 is a broadcast group withdrawal procedure, S441 is a broadcast group encryption key change procedure, and S45.
Reference numeral 1 denotes a broadcast group communication end determination procedure.

FIG. 11 is a detailed flowchart of the broadcast group joining request receiving operation S421 of the broadcast transmitting apparatus 1.
S422 represents a receiving device authentication procedure, S423 represents a communication key sharing procedure with the receiving device, S424 represents a broadcast decryption key distribution procedure, S425 represents a broadcast group setting procedure, and S426 represents a broadcast connection setting procedure.

FIG. 12 shows a flowchart of the broadcast transmission operation of the broadcast transmitting apparatus 1. S501 is an initialization step of the control variable s, S502 is a message transmission step, S50
3 is an acknowledgment reception timeout determination step, S504
Is an acknowledgment receiving step, S505 is a control variable w initialization step, S506 is a message release time determination procedure, S507 is a release permission message transmission procedure,
S508 is a release permission acknowledgment reception time-out determination procedure, S509 is a release permission acknowledgment reception procedure,
S510 is a step of increasing the control variable s, S511 is a step of determining the control variable s, S512 is error processing, and S52.
0 is a step of increasing the control variable w, and S521 is a control variable 2
Step S522 represents error processing.

FIG. 13 is a diagram showing a configuration using the authentication server 20. 1 is a broadcast transmitting device, 901 is a receiving device 1,
Reference numeral 902 denotes the receiving device 2. FIG. 14 shows an embodiment of a system that combines news delivery and automatic stock trading. 1001 is a news agency, 1002 is a news delivery system, 1003 is a broadcast transmitting device, 1103 is a broadcast connection, and 1101-1 to 1101-1n are securities investment systems that receive news from a news company through the broadcast connection 1103, respectively. The transmission / reception devices 1102-1 to 110n are devices connected to the securities investment systems 1101-1 to 110n to guarantee the fairness of the broadcast.

Reference numerals 1201-1 to 120m denote securities trading systems on the securities company side, reference numerals 1202-1 to 120m denote corresponding transmission / reception devices,
203-1 is a transmitting / receiving device 1 of the securities investment system on the customer side
This is a many-to-one connection type connection for connecting the receiving devices 1202-1 of the securities trading system on the securities company side with 102-1 to 102-n.

Reference numerals 1201-1 to m indicate securities trading systems of the securities companies 1 to m, 1202-1 to m indicate medical examination apparatuses of the respective securities trading systems, and 1203-1 to m indicate the securities companies 1 to m and customers. It is a many-to-one connection type connection.

Reference numeral 1301 denotes a stock exchange;
Represents a dedicated line connecting the stock exchange and the securities trading system of each securities company. Hereinafter, the operation of one embodiment of the present invention will be described with reference to FIGS.

In the following, the present invention is broadly divided into three parts, namely, a procedure for confirming that the receiving apparatus is a reliable apparatus through a network, a configuration of the apparatus, a procedure for matching the release time of a message in the receiving apparatus, and the present invention. The application system will be described in three order.

Hereinafter, the word message indicates one data unit having a meaning exchanged between a transmitting apparatus or a receiving protocol apparatus and a peer transmitting apparatus or a protocol higher-level apparatus that guarantees a message receiving time.
The term packet means a group of data transmitted from a transmitting device or a receiving protocol device to a physical network. In this embodiment, an AAL5 packet is specifically shown. A message is transmitted as one or more packets when transmitting over a network.

First, the joining of the broadcast group will be described. The broadcast group has a unique broadcast group identifier defined in advance, and the E.E. Along with the 164 address, the information can be disclosed and known by a telephone directory or a directory service, for example.

The client device 11-1 that intends to receive a broadcast group first executes a broadcast group joining procedure 101. FIG. 3 shows a more detailed flowchart of the broadcast group joining procedure 101. First, the receiving apparatus sets a point-to-point connection with the broadcast transmitting apparatus 1 and executes step 102 for starting a broadcast group joining procedure. The client device 11-1 communicates with the broadcast transmission device corresponding to the identifier of the subscribed broadcast group. Prepare a broadcast group subscription setup request including 164 addresses,
Handover to 12-1 reception protocol device through 4-1.

When the message passed from the client device is a broadcast group join request,
The connection setting function 309 in FIG. 9 passes connection setting messages 13-1 to 13-n to the network-side interface device. The network interface device transmits a setup message to the communication network 31 according to a signaling protocol.

The client device 11-1 communicates with the broadcast transmission device E.E. A connection setup message to the broadcast transmitting device is generated from the 164 address. This message includes an identifier indicating a broadcast group, and is transmitted toward the communication network 31 from the 16-1 network side interface point. This message format conforms to the ATM interface specification (ATM forum
An identifier indicating a broadcast group shall be added to that specified by the UNI 3.1 specification, TheATM Forum, 1994).

The setup message establishes a point-to-point connection 32-1 between the receiving protocol device 12-1 and the broadcast transmitting device 1 in the communication network 31. At the same time, the broadcast group identifier included in the setup message is passed to the broadcast transmitting apparatus 1, and the broadcast transmitting apparatus 1 starts the broadcast group joining procedure S421 in FIG.

Next, the broadcast transmitting apparatus 1 recognizes whether or not the identifier 17 of the broadcast receiving protocol apparatus 12-1 exists in the identifier table 2 of the broadcast transmitting apparatus 1 through the point-to-point connection 32-1. . This corresponds to step S422 in FIG. 11 in the broadcast transmitting apparatus 1 and step S103 in FIG. 3 in the receiving protocol apparatus 12-1.

Each of the broadcast receiving protocol devices 12-1 to 12-n is guaranteed in advance by a reliable public organization that the protocol implementation is valid at the time of manufacture and shipment, and a secret receiving protocol given as a proof thereof. Device identifier 1
7 is stored. Since the identifier 17 is used by a public organization to prove that the value of the identifier is valid, the value of the identifier is transmitted from the upper interface 14 and the lower interface 15 in order to prevent unauthorized use of the value. Protected from direct reading.

Further, the device identifier 1 of the reception protocol device
In order to prevent unauthorized reading of 7, the housing of the device is provided with a function of detecting destruction. Identifier is EEP
It is stored in a ROM (Electric Erasable Programmable ROM), and is erased when the destruction of the housing is detected. Such techniques are already known. (Mori, R., Kawahara, M ： "Sup
erdistribution: The Concept and Architecture ",
IEICE transaction 73 (7), 1990).

The broadcast transmitting apparatus 1 has a receiving apparatus identifier table 2 which is guaranteed in advance by a public organization to have a correct protocol implementation. This table is maintained by a trusted authority.

The receiving protocol device proves that the broadcast transmitting device 1 has the device identifier by communication without directly including the device identifier as a part of the message on the network, and that the implementation of the protocol processing is valid. Proof that there is. This is to prevent a device identifier from being stolen due to interception of communication, and to prevent a device other than the transmitting device that authenticates the receiving device from falsifying the transmitting device and stealing the device identifier.

Such a protocol is well known as a so-called zero knowledge proof. Representative algorithms include Fiat and Shamir (Fiat, A., Shami
r, A.: "How to prove yourself: practical solution t
o identification and signature problems ", Proc.of C
RYPTO 86, Springer-Verlag, Berlin, 1987).

In this embodiment, the Fit and Shami are used.
Based on the algorithm r, the receiving protocol device has a device identifier s and a large integer n. On the other hand, the identifier table 2 of the receiving protocol device stores the square of the device identifier s in the modulus n, v: s = v ² (mod n).

The broadcast transmitting apparatus confirms that the receiving protocol apparatus has s by a zero knowledge proof method,
Also, confirm that the value v: s = v ² (mod n) obtained for the proof is present in the identifier table 2 of the receiving protocol device, and confirm that the identifier is given by a trusted authority. Confirm. As long as the factorization of n is difficult, s is safe because it is difficult to know s even if v: s = v ² (mod n) is known.

Each transmitting device does not wait for the receiving device identifier table 2, and as shown in FIG.
Authentication may be performed by making an inquiry to the authentication server 20 having. When performing the authentication, the broadcast transmitting apparatus relays the message exchange between the authentication server and the receiving apparatus, and obtains the final authentication result from the server. With this method, it is not necessary for each transmitting device to have a receiving device identifier table.

Next, the receiving device shares a secret key between the broadcast transmitting device 1 and the receiving protocol device 12 (step S1).
04 is executed. The transmitting device corresponds to step S423. Well-known methods exist for this procedure (Diffie, W., Hellman, H .: "New directions in crypto"
graphy "IEEE transactrion of information theory6
: 644-645,1976). In this method, it is necessary that a sufficiently large prime number p and an atomic root α on GF (p) are shared in advance between the transmitting device and the receiving device. Since these values are safe to publish, the broadcaster forwards these values to the receiving protocol device at the beginning of execution of this step.

An impersonation attack is conceivable in which the receiving protocol device to which the regular identifier is given performs the steps up to step S103, and the secret key sharing procedure S104 immediately after that is performed by an unauthorized device. To prevent this, the broadcast transmitting device 1
Is a secret key sharing step 10 after it has been confirmed in step S103 that the receiving protocol device has a proper identifier.
Also in 4, it is desirable that the receiving protocol device continue authentication with a proper identifier.

Thereafter, the secret key sharing means 104 and 423 terminate and the broadcast transmitting apparatus 1 and the receiving protocol apparatus 12-1
After the common secret key K1 has been generated, the receiving protocol device adds an authenticator to the transmitted message using the secret key K1 to prevent the message from being tampered with and spoofed. The broadcast transmitting apparatus checks the authenticator added to the message using the common key, and discards the packet if the authenticity is found.

Next, the broadcast transmitting apparatus transmits the decryption key K of the broadcast transmission message encrypted using the secret key K1 to the receiving protocol apparatus 12-1, and the receiving protocol apparatus receives the key. Execute S105. The encryption algorithm in this case is a normal secret key encryption, for example, D
ES can be used. The transmitting device corresponds to step S424.

The following two steps are taken for the following reason. The key K1 to be shared first is the broadcast transmitting device 1
It is common only between the receiving protocol device 12-1 and the receiving protocol device 12-1. Other receiving protocol devices 12- belonging to the same broadcast group
2 to 12-n have keys K2,..., Kn, respectively. In order to perform broadcast encryption with one key, it is necessary that all the receiving protocol devices 12-1 to 12-n have a common key.

The broadcast group joining procedure viewed from the broadcast transmitting apparatus 1 will be described below with reference to FIG. In the broadcast transmitting device 1, the receiving device authentication step S422 corresponds to the illumination step S103 of the identifier of the receiving protocol device, and the key sharing step S423 with the receiving protocol device is the key sharing step S104 with the transmitting device of the receiving protocol device.
, And the broadcast group key transmission step S105 corresponds to the broadcast group key acquisition step S105 of the receiving protocol device, respectively.

After these steps are sequentially performed in the broadcast transmitting apparatus 1, a step 425 of registering the receiving protocol device 12-1 in the broadcast group is executed by the broadcast group management function 212 in FIG.

Broadcast connection authorization function 2 in FIG.
The interface device 13-1 in which the receiving protocol device 12-1 is connected to the broadcast connection 33 through the interface 13
Add. This setting is performed based on the E.E. of the interface device 13-1 obtained in the broadcast group joining procedure performed in step S102. This is performed by a signaling procedure using information of the 164 address. In step S426, the reception protocol device 12 is added to the broadcast connection, and the broadcast is started.

The above is the procedure for indicating that the receiving device is a reliable device when the terminal joins the broadcast group. Next, a procedure for controlling the order of messages and matching the release times will be described.

An object of the present invention is to release a message at the same time in all receiving apparatuses included in a broadcast group. There are several methods for sharing the message release time. Although details will be described in each embodiment, there are six combinations depending on the delivery method of the message body and the delivery time. FIG. 27 shows the relationship between the receiving apparatus and the minimum number of processing packets required for delivering one message in each system. Hereinafter, each method will be described. Here, the case where the transmitting device determines the release time of the message is described as an example,
The device that transmitted the message whose release time is to be determined and the device that notifies the determined release time may be different.

First, "confirmation in message delivery" means that the message body has been correctly delivered to each terminal, and thereafter (in the case of a plurality of transmitting devices, the release order), the release time is determined. 2 to be notified
This is a phase method. Since an acknowledgment is returned for each packet constituting the message body, if the message body is composed of m packets, at least 2m packets are required for transmission of the body. This method is further subdivided according to the method of notifying and confirming the release time. In any method, the number of repetitions of the release time notification is determined in advance, the release time is set after the repetition time, and the packet for notifying the release time is repeatedly transmitted to the receiving device, so that even in an environment where packet delay and loss occur. The feature is that all the receiving devices succeed in sharing the release time with a sufficiently high probability.

In the case of "release time confirmation response", the receiving device returns a confirmation response to the transmitting device for each message notifying the release time. The transmitting device retransmits the release time notification to the terminal that does not return an acknowledgment. In addition, if the terminal does not return an acknowledgment, that is, if there are many terminals that failed to receive, it is possible to make a correction to delay the release time and transmit the release time notification again.

In the case of "with release time confirmation", the notification of the release time is repeatedly performed at predetermined intervals. The receiving terminal does not make an acknowledgment every time, but instead notifies the transmitting device that the message was released at the correct time after releasing the message. The transmitting device sends a confirmation response to this notification.

In the case of "no confirmation of release result", confirmation of the release result to the transmitting device is not performed. Even if the release time match fails, the transmitting device cannot know it.

"No confirmation in message delivery" is a method for simultaneously delivering the message body and delivering the release time. This provides a time matching means which is simple for a short message in which the message can be accommodated in one packet, and which has a particularly low delay compared to other methods. It should be noted that in this method, the delivery of the message body is not confirmed, so that the order of the messages cannot be controlled based on the transmission time among a plurality of transmission devices.

In the case of “release time acknowledgment received”, the packet including both the release time and the message is transmitted, omitting the phase of the message transfer. 'Confirmed' As described, the receiving device returns an acknowledgment to the transmitting device, and the transmitting device retransmits the release time notification to the terminal that does not return the acknowledgment. It should be noted that when the number of packets constituting a message increases, the number of packets increases due to retransmission.

In the case of "with release time confirmation", the notification of the release time is repeatedly performed at predetermined intervals. The receiving terminal does not make an acknowledgment every time, but instead notifies the transmitting device that the message was released at the correct time after releasing the message. The transmitting device sends a confirmation response to this notification.

'No release result confirmation' No confirmation of the release result to the transmitting device is performed. Even if the release time match fails, the transmitting device cannot know it. Next, transmission and reception of a broadcast message will be described with reference to FIGS. 4 is a flowchart of a reception procedure, FIG. 12 is a flowchart of a transmission procedure,
FIG. 5 shows a broadcast message sequence.

Various attempts have been made to reduce the number of acknowledgments for performing retransmission control for broadcast communication requiring a retransmission procedure. A typical example is a piggyback that reduces the number of packets by incorporating an acknowledgment message into a transmission message or a negative acknowledgment only when a packet loss is detected (Takizawa, M: "Cluster con
trol protocol for Highly reliable broadcast commun
ication ", Proc. of the IFIP Conf. on Distributed P
rocessing, 1987; Internet RFC1301 MTP, 1992; Mellia
r-Smith et al., "Reliable broadcast protocol", USP 5,2
16,675,1990).

In the present embodiment, the first goal is to make the time for delivering a message to the client device fair, so that the packet use efficiency of the broadcast message transfer is not discussed, and the simplest message unit A description will be given of a method of performing an acknowledgment (handshake method) as an example. The number of packets can be further reduced by combining this embodiment with a piggyback or negative acknowledgment technique.

Before the detailed description, the definition of constants used for the description will be described in advance. From one device i to another device j
T _{d at which} there is a delay time t _ij when transferring a packet to
The probability larger than P is expressed by P (t _ij > T _d ). When the maximum value of the delay time is _Td , a packet that is delayed beyond that can be regarded as lost. The maximum value of the delay time is T
_{Let d loss.ij} = P (t _ij > T _d ) be the message loss probability from device i to device j, where _{d is d} .

Similarly, when T _d is given, the maximum value of the loss probability per single packet is P _loss (T _d ) = max {P _loss.ij } for an arbitrary combination of devices constituting the broadcast group.
Is defined. Conversely, when the loss probability is given as P _loss , the maximum value (T _max ) of the delay time corresponding to the loss probability can be obtained.

The time RTT _ij from when a message is sent from the device i to the device j, where the processing time T _P is processed and the response is returned is represented by the transfer time from i to j being t _ij , j
_Assuming that the transfer time from i to i is t _ji , RTT _ij = t _ij +
T _P + t _ji , which has a distribution of P (RTT _ij > T). When a certain timeout probability P _tout is defined, T is set so that P _tout > P (RTT _ij > T _s ).
Determine _s .

The probability of _loss P _{loss for} each packet defined above
(T _d ) applies well to the probability of loss of a packet transmitted with a sufficiently large time interval. However, the probability of loss of packets transmitted consecutively or at close intervals is not considered to be independent. This is because errors occurring in the network tend to concentrate within a certain time. For example, there is a possibility that an error occurs in consecutive packets due to loss of ATM cells due to network congestion, a failure in path control, or a bit error due to path switching of a transmission link. This phenomenon is hereinafter referred to as a burst error.

A probability distribution in which a certain burst error state continues for a longer period than the parameter T is _defined as P _Burst.ij (T), and its maximum value is defined as P _Burst (T _i ), similarly to P _loss . When a burst error is present, if two packets are transmitted in a short period of time, they may be lost together by a single burst error. You can avoid it with a long interval.

By determining the packet transmission interval T _i so that P _Burst (T _i ) <P _loss (T), the loss probability of each message is independent and its value is at most P
can be regarded as _loss .

As a constant that varies depending on the processing capability of the receiving device, the time required from when a release message or a message including a message and a release time is received until the message can be released to a higher-level processing device is _Tdec . .

FIG. 28 summarizes the definitions of the constants. First, an example of “confirmation of message delivery” and “confirmation of release time” will be described.

The message transmission sequence starts with the broadcast transmitting device transmitting the message M ₁ to the broadcast connection. The message contains the following information: -Broadcast message body-Message type = message-Broadcast message identifier-Message authenticator The broadcast message body includes a message input from the broadcast message input function 202 to the broadcast protocol management function 203 or control information. The message type indicates whether the message includes a message or control information. The broadcast message identifier is a serial number indicating the order of messages, and has a sufficiently long cycle.

The message authenticator is an authenticator for confirming that these messages are not falsified.
The message excluding the message authenticator, that is, the broadcast message body before being encrypted, the broadcast message type identifier, and the broadcast message identifier are encoded by a secret key K for broadcasting, such as MD5 and DES. (J. Kohl, "RFC1510:
The Kerberos Networks Out
Hentication Service ”, 199
The value encoded by 3) is used. The key of the message authenticator used for authentication of the broadcast communication is a key K common to the broadcast group.
Note that this is different from the authenticator of the key distribution message because it uses. The coding scheme of the authenticator itself may be the same.

These fields are all used for the encryption function 20.
4, encrypted with the key K and broadcast. The receiving side calculates the authenticator of the corresponding part of the message using the key K,
If the values match, it is determined that the message and the protocol information have not been falsified, and the subsequent processing is performed. In the following description, the addition of the authenticator at the time of transmission and the inspection of the authenticator at the time of reception are omitted. If there is no authenticator, if a part of the message, for example, the position on the message corresponding to the encryption corresponding to the value of the specific message identifier is known, the protocol information can be falsified by exchanging only the part. . However, partial tampering of the message can be detected by including information that cannot be known unless it has information about the entire message and key information, that is, an authenticator.

At step 501, before transmitting the message, the broadcast transmitting apparatus increments the message identifier p by one as pp + 1. Then, the control variable s of the message P is initialized to 0. The control variable s is distinguished for each message, but is simply expressed as s here to avoid complexity.

[0110] Broadcast transmission device in step 502 in FIG. 12 broadcasts a message M _p with identifier number p, a message type = message. The broadcast transmitter simultaneously operates the timer T1
_Is set T _s after the current time.

The receiving protocol device determines in step 12 of FIG.
2 receives the message with the identifier number M _p, is set to 0. Similarly control variable j and s of the sender. Next, an acknowledgment ACK _p (A) is transmitted to the master. This is step 123.

The acknowledgment ACK _p (A) contains at least the following information, and is encrypted with the key K in the same manner as the transmission message. The message contains at least the following information: -Message type = confirmation-Message identifier-Receiving device identifier-Message authenticator The transmitting device receives a confirmation response in step 504. If the acknowledgment of all the broadcast groups can be received for the message with the identifier p, the next step 505 is executed. Timer T
If all acknowledgments have not been received by the time set in 1, the transmitting device determines that a timeout has occurred. This is step 503. The control variable s is a predetermined value S
It is determined whether it is equal to or less than _max (step 511).
Is _smaller than _Smax , the value of S is increased by 1 (step 5).
10) Retransmit the message.

The receiving side also transmits an acknowledgment again when receiving a message determined to have been received by the message identifier. This is because even if the receiving side succeeds in receiving the message, the transmitting side may determine that the reception has failed due to the loss of the acknowledgment. In the transmission side, the probability of acknowledgment times out because the time T1 is T _s, equal to or less than P _tout. The probability that all failed S _max transmission terminals will not receive the message is (P _tout ) S _max .
When the number of terminals is n, the probability that one or more terminals fail to receive is (1- ( _Ptout ) _Smax ) ⁿ . Therefore,
By increasing S _max , the probability of failure can be arbitrarily reduced. However, increasing S _max increases the delay time of message delivery, and the two have a trade-off relationship. FIG. 20 shows n when P _loss = 10 ^−3.
7 is a diagram plotting the number of retransmissions and the probability of a node not receiving a message at all for = 3, 5, 10, 1000. If the probability of delivery failure is 10 ^-12 or less when the message loss probability is 10 ^-3 and the number of nodes is 1000, the number of retransmissions may be set to 5 or more.

If s exceeds _Smax , an error processing step 512 is executed. There are two types of error processing: a method in which the transmission of the message fails, and a method in which a device for which an acknowledgment is not obtained is determined to be a failure and the device is separated from the broadcast group. These will be described in detail later.

Next, the relationship between the error processing and the host application will be described. There are three ways to deal with an error. -The higher-level group is notified of the occurrence of the error and the broadcast group is maintained.

Distribution of materials for the electronic conference system-The terminal that failed to receive is separated from the broadcast group and the broadcast group is maintained. News delivery ・ Broadcasting of broadcast groups ended.

Canceling the application. Bank account decentralized management system In a processing step 505 after confirmation has been received from all devices, the control variable w is initialized to 0. If there is a message of the last release having an identification number younger than this message, the determination of the release time is waited until the release time notification of the time of the last release message is completed. When the waiting state ends, in step 506, the release time of the message is set. The release time is determined as follows.

The time required from when the receiving device receives the release message until the packet can be output to the host device is t _dec , and the maximum number of retransmissions is W _max . Assuming that the timeout period of the acknowledgment is T _s and the time of the release message transmission is T, the release time of the message is _Tr.
Is represented by the following equation.

T _r (p) = T + W _max · T _s + t _dec W _max is (1− (P _tout )) as in the case of the message.
W _max ) ⁿ is determined to be less than the allowable value of the failure probability of time delivery. Here, it is assumed that the timeout time T _s is longer than the bucket sending interval T _{i that} can avoid the burst error time, that is, the relationship of T _s > t _i holds. If T _s <t _i , the set value T _{i of the} time T3 used in the following description is used. If the release time of the message is determined by the above method, the release order is the order of the release message issuance time T, and the release order of the message is correctly maintained.

When the time is determined, a scheduling condition for outputting a message to the client device of the receiving device described below may be added. If the receiving device accommodates a plurality of broadcasts, each broadcast group may compete for the use of resources such as an upper interface and a message decoding function. To avoid this conflict,
At the time of transmitting the acknowledgment in step 123 in FIG. 4, information indicating which time is available is added to the acknowledgment message, and based on the information, sufficient resource allocation can be performed by all the receiving protocol devices of the broadcast group. If the time at which a message can be output to the client device is set as the scheduled message release time, the release time can be guaranteed by all receiving protocol devices even when a plurality of broadcast groups are handled. The transmitting device acquires scheduling information in an acknowledgment receiving step 504. In this case, it is necessary to select a release time so that the release order of the messages does not reverse.

When the scheduling condition is considered, the acknowledgment message includes at least the following information. -Message type = acknowledgment-Message identifier-Receiving device identifier-Message scheduling information-Message authenticator When the receiving time is determined, the transmitting device broadcasts a release permission message REL _p (i). p is the corresponding transmission message, and i is the identifier of the release permission message. The release permission message includes at least the following information.

Message type = Release Message identifier Release message identifier Scheduled message release time = T _r (i) Message authenticator The transmitting device sets the timer T3 after Ts from the current time at the same time as transmitting the release permission message. . This is step 507.

At step 125, the receiving device receives the release permission message REL _p (0). The processing steps 131, 130, and 132 when the release permission message cannot be received until the timer T2 times out are the same as the processing of the timeout of the acknowledgment response on the transmission side.

When receiving the release permission, the receiving apparatus executes step 126 to execute the release permission acknowledgment RACK.
Transmit _p (i, A) to the transmitting device. Here, i is a release message identifier, and A is a receiving device identifier. Then, the timer T4 is set to the scheduled release time included in the release permission message.

The release permission acknowledgment includes at least the following information. -Message type = Release confirmation-Message identifier-Release message identifier-Receiving device identifier-Message authenticator In addition to these, even if message scheduling information updated in preparation for failure to receive release permission acknowledgment is added to release permission acknowledgment good. In this case, the receiving device acquires the scheduling information again in step 126 and transmits a release permission acknowledgment.

When the transmitting device receives the release permission acknowledgments from all the medical examination devices before the timeout of the timer T3, the transmitting device normally ends the transmitting process. If the receiving apparatus does not receive the release permission REL _p (i + 1) of the message from the transmitting apparatus again before the timeout time of the timer T4, that is, the release time in step 127, the receiving apparatus executes step 140 to release the release time TR (i ), And outputs the decrypted message to the client device through the upper interface. This completes the reception of the message p.

The message may be decrypted at any time from the time when the message is first received until the time when the message is output, as long as the release time is reached. The arrangement of the message decoding means 303 and the message storage means 304 in FIG. 9 must be changed in accordance with this order.

The first purpose for the receiving device to perform the release permission acknowledgment is to reduce the useless retransmission by the transmitting device selectively retransmitting the release permission message. By performing another acknowledgment, the transmitting device knows how much the receiving device can receive the release permission message, and the release time can be postponed if there are few devices that can correctly receive the release permission. . This is performed by notifying the receiving device that has already received the release permission of the corrected release time.

The condition for postponing the release time is that a predetermined ratio, for example, half or more of the receiving devices can receive the release permission message when the transmitting device completes the acknowledgment receiving process of half of W _max times. If not,
It can be. If the modified release time is notified at this point, it can be expected that the receiving device that has already received the old release permission time can also receive the modified message release time before the message release time has arrived.

However, when the release time is corrected, there is a possibility that the release order of the messages is reversed. This problem can be solved in one of the following two ways. One is to wait for the release time of the next message to be notified until the release time is completely determined. Another method is to modify the release time of one message, and then modify the release time of subsequent messages.

Next, a case where a plurality of transmitting apparatuses exist in one broadcast group will be described. When there are a plurality of transmitting apparatuses, the conventional technique can be used as it is to make the reception order of the messages transmitted by the broadcasting groups coincide. A method of determining the release time of a message by the method described above after the reception order is determined can be easily analogized. FIG. 7 shows a plurality of transmitting devices (senders).
4 shows a message sequence when one master determines the receiving order. Here, the transmitting device and the receiving device are separately described for simplicity. However, the transmitting device and the receiving device have the same network address or are stored in the same housing, and a method of simultaneously performing device authentication is adopted. Is also good. The joining procedure is performed for a master that exists in only one broadcast group, and the members and keys of the broadcast group are centrally managed by the master.

Here, the operation of the receiving order due to the illegal protocol operation, which has been a problem in the prior art, does not occur because the correct operation of the protocol is guaranteed by the authentication of the device.

When it is required to reflect the message transmission time of the transmitting apparatus in the message receiving order to ensure fairness, the receiving order is determined by the absolute time attached to the transmission message. Is also good.

Next, a description will be given of how a receiver leaves a broadcast group. When a receiving device leaves the broadcast group, the device has a decryption key for the broadcast, so the decryption key must be changed.

Next, the change of the encryption will be described. Distribution of the changed key is performed by transmitting an encrypted message using a key distribution key for each receiving device through a connection for each receiving device. This corresponds to steps 431 and 441 in FIG.

Various time synchronization methods are already known. For example, as a standard for the Internet, which is mainly a packet switching network, "Mills, DL:" Network
TimeProtocol "; IETF RFC1059,1988" is known.
Also, the frame cycle of the physical line connected to the public network is very accurate because it is synchronized with the atomic clock, and high accuracy can be obtained if used for time synchronization. Also, GPS (Global Positioning System) is installed inside the tamper-resistant device.
It is also possible to obtain the correct time by placing the receiving device of m).

It is desirable that the time period be synchronized before the start of the broadcast communication so as to satisfy the requirement of accuracy, and that the time period be constantly synchronized even while the broadcast communication is continued. Next, a penalty for calculating the release time will be described. The order of responses to a broadcast can be important. For example, a bid. For such applications,
Fairness can be ensured by setting the release time of the message so as to guarantee the response delay time, instead of setting the delivery time to be the same.

Next, the storage of the failure history will be described.
Although the device of the present invention is suitable for a field where a high degree of fairness is required for message delivery time such as electronic commerce, if the device is put to practical use, the effect of device failure has a great effect.

For example, when the apparatus of the present invention is used for buying and selling stocks, if the apparatus cannot be traded due to a failure of the apparatus, the user may suffer a great deal of damage. A similar situation occurs when a network failure occurs.

The result of the self-diagnosis of the apparatus is stored in the apparatus, and if a reliable third party inspects the result and determines that the apparatus has failed, the damage caused by the failure is checked. The institution guarantees that the device of the present invention can be operated more safely. Of course, the failure information cannot be rewritten by the user.

Further, regarding a failure in the public network, failure information on the network side can be proved by receiving failure information from the public network and storing it in the apparatus. In the transfer of the failure information, the ATM can use the overhead information of the OAM cell or the frame of the physical layer. The network side can prevent forgery of the failure information by providing information such as the failure device and the time by encrypting the information.

On the other hand, it is possible to know that the apparatus was intentionally inoperable, for example, in a situation such as a power-off or a line disconnection, by checking the records of the apparatus and the network side. No warranty for damages caused by unavailability is required. Of course, it is impossible to detect and record all faults, but by combining these methods for detectable faults, it is possible to clarify fault responsibilities and reduce the burden on the user.

(Second Embodiment) In the first embodiment, it is confirmed that the transfer of a message has been successfully performed by all the receiving apparatuses in the broadcast group, and the procedure of distributing the release time to the receiving apparatuses in consideration of the message order is adopted. I was

In the second embodiment, a description will be given of a simple time guarantee procedure in a case where a message is sufficiently small and the message body and the specified release time can be stored in a single packet.
This corresponds to "no confirmation in message delivery" and "no confirmation of release result" in the classification of message delivery methods shown in Table 1. The purpose of this embodiment is only to match the message reception time of each receiving device with respect to a message transmitted from a certain transmitting device. Although the order control of the messages based on the transmission time is not performed among the plurality of transmission devices, the reception times coincide with each other, so that the order of the messages is guaranteed as a result. However, it should be noted that when a message with the same release time specified from a plurality of transmitting devices arrives, how it is handled depends on the processing of each receiving device.

[0145] The message transmitted from the transmitting device includes the following information. -Message body-Message type = message-Release time-Broadcast message identifier-Message authenticator In this embodiment, by sending this message multiple times, the receiving side receives the message correctly, and Even when a message is received, the release time is specified with a margin enough to release the message at the specified release time, so that each receiving device can release the message at the same time.

The definitions of the constants are the same as those shown in Embodiment 1 and Table 2. As a constant, the maximum message loss probability P _loss (t _d ) of the broadcast group when the delay time is set to t _d ,
The packet transmission interval t _i release processing time t _dec when the timeout time between the devices ij is T _s and the probability of _loss due to burst is P _loss is used.

[0147] release time of the message is T _r is determined as follows: the time when the message is sent as T. T _r = T + t _r = t _i · k + t _d + t _{dec The} receiving side executes a decoding procedure after receiving the message,
The message is released at the specified release time _Tr . No acknowledgment is given.

FIG. 21 shows a message transmission procedure of this embodiment, FIG. 22 shows a message reception procedure, and FIG. 23 shows a message sequence. (Embodiment 3) Acknowledgment Response In Embodiment 2, simple time matching was realized by sending the message body and the release time as one message. However, in the method of the second embodiment, the transmitting device could not know whether the terminal succeeded in receiving. In the third embodiment, the receiving apparatus returns a result to the transmitting apparatus after releasing the message, so that the transmitting apparatus can know the reception result. This corresponds to “No confirmation in message delivery” and “Confirm release result” in the classification of Table 1.

The message transmission procedure is shown in FIG.
FIG. 26 shows a message receiving procedure, and FIG. 26 shows a message sequence. The receiving device that has released the message at the designated time sends an acknowledgment message to the transmitting device.

The confirmation response message contains at least the following information. -Message identifier-Message type = confirmation-Receiving device identifier-Message authenticator The transmitting device that has received the acknowledgment from the receiving device records "Received" in the entry of the receiving device in the acknowledgment reception table, and sends the acknowledgment received message Send to receiving device.

The acknowledgment response received message contains at least the following information. -Message identifier-Message type = acknowledgment received-Message authenticator When the receiving device receives the acknowledgment received message related to the message from the transmitting device, the receiving device completes the acknowledgment sending operation for the message. If If the acknowledgment is not received to retry until the number of retries W _max which is determined in advance at intervals T _s. The method of determining the parameter W _max is as described in the first embodiment.

In the above-described first to third embodiments, three of the six combinations shown in FIG. 27 have been specifically described as the procedure for matching the message release times. The remaining three combinations can be easily analogized from the above description, and the description is omitted here.

Embodiment 4 which is a more specific embodiment to which the present invention is applied will be described with reference to FIG. 14, 1001 is a news agency, 1002 is a news delivery system, 1003 is a broadcast transmission device, 1103 is a broadcast connection, and 1101-1 to 110n are broadcast connections 1103.
Respectively represent securities investment systems that receive news from news agencies through The transmission / reception devices 1102-1 to 110n are connected to the securities investment systems 1101-1 to 110n, and represent devices that guarantee the fairness of the broadcast.

Reference numerals 1201-1 to 120m denote securities trading systems on the securities company side, reference numerals 1202-1 to 120m denote corresponding transmission / reception devices,
203-1 is a transmitting / receiving device 1 of the securities investment system on the customer side
This is a many-to-one connection type connection for connecting the receiving devices 1202-1 of the securities trading system on the securities company side with the securities 102-1 to 102-1-n. 1201-1 to m are securities trading systems of the securities companies 1 to m, 1202-1 to m are receiving devices of the respective securities trading systems, and 1203-1 to m are securities companies 1
To m and the customer. Reference numeral 1301 denotes a stock exchange, and reference numerals 1204-1 to 1204-1m denote dedicated lines for connecting the stock exchange and a securities trading system of each securities company.

The news agency provides a service for delivering information that may affect the stock market to customers. The information includes, for example, employment statistics by public institutions, information on the number of housing starts, exchange rates, and the occurrence of international disputes.

Information is distributed fairly and fairly from the broadcast transmitting apparatus 1003 to the receiving apparatuses 1102-1 to 110-n of each customer. Therefore, all clients' securities investment systems 1101-
1 to n simultaneously receive information. The system determines the investment based on the information received and places an order with a securities company. The order is sent to the securities trading system 1201-1 of the securities company through the many-to-one communication path 1203-1. At this time, an absolute time is added to the transmission message, and the securities company processes the first-come-first-served order based on the time. Each securities company and the stock exchange are connected by dedicated lines 1204-1 to 120m, and trading is performed.

[0157] Since the simultaneity and the time order are guaranteed, all customers can receive the message from the communication company without being affected by the transmission delay and the occurrence of retransmission. Also,
Orders to securities companies are processed in the order based on the time of transmission, so that transmission delays and retransmissions are not affected here. This mechanism provides fair business opportunities to many customers regardless of the distance or communication network.

In the following, the present invention is roughly divided into three parts, namely, a procedure for confirming that the receiving apparatus is a reliable apparatus through the network, a configuration of the apparatus, a procedure for matching the release time of the message in the receiving apparatus, and the present invention. The application system will be described in three order.

[0159] Simultaneity and time order are guaranteed so that all customers can receive messages from the telecommunications company without being affected by transmission delays or retransmissions. Also,
Orders to securities companies are processed in the order based on the time of transmission, so that transmission delays and retransmissions are not affected here. This mechanism provides fair business opportunities to many customers regardless of the distance or communication network.

[0160]

According to the present invention, a fair communication service can be received even in a place where there is a difference in transmission delay. Also, since the possibility of fraudulent activities due to protocol violations can be eliminated, fair transactions can be realized.

[Brief description of the drawings]

FIG. 1 is a diagram showing a basic configuration of a transmission / reception protocol device according to an embodiment of the present invention.

FIG. 2 is a flowchart illustrating an outline of an operation of the reception protocol device.

FIG. 3 is a broadcast group joining procedure S10 of the receiving protocol device.
3 is a flowchart showing details of a first example.

FIG. 4 is a flowchart showing details of a broadcast message receiving procedure S111 of the receiving protocol device.

FIG. 5 is a diagram showing a message sequence when a broadcast message receiving procedure is normal.

FIG. 6 is a diagram showing a message sequence when transfer of a message release permission in a broadcast message receiving procedure fails.

FIG. 7 is a diagram showing a broadcast message sequence for obtaining an agreement on the message order of a plurality of senders.

FIG. 8 is a functional block diagram of a broadcast transmitting apparatus.

FIG. 9 is a functional block diagram of a broadcast receiving protocol device.

FIG. 10 is a flowchart showing an outline of a broadcast group management operation of the broadcast transmitting apparatus.

FIG. 11 is a flowchart showing details of a broadcast group joining request accepting operation S421 of the broadcast transmitting apparatus.

FIG. 12 is a flowchart showing a broadcast transmission operation of the broadcast transmitting apparatus.

FIG. 13 is a diagram showing a configuration of the present embodiment using an authentication server.

FIG. 14 is a diagram showing a configuration of a system that combines news delivery and automatic stock trading.

FIG. 15 is a diagram showing a conventional broadcast message sequence.

FIG. 16 is a diagram showing a broadcast message sequence when a message disappears in the related art.

FIG. 17 is a diagram showing a broadcast message sequence when a release permission message disappears in the related art.

FIG. 18 is a diagram showing a broadcast message sequence when a release permission message disappears in the related art.

FIG. 19 is a diagram showing a broadcast message sequence for obtaining a message order agreement of a plurality of senders in the prior art.

FIG. 20 is a graph showing the relationship between the error probability, the number of terminals, the number of retransmissions, and the probability of successful message reception.

FIG. 21 is a flowchart of a message transmission procedure according to the second embodiment. S2101 the transmission procedure of the broadcast message, S2102 delay of T _i time, S2103 represents each of the k iterations procedures.

FIG. 22 is a flowchart of a message receiving procedure according to the second embodiment. S2201 is a release time determination procedure,
S2202 is a received message presence / absence determination procedure, S22
03 denotes a received message duplication determination procedure, S2204 denotes a procedure for adding to a release schedule list, S2205 denotes a message decryption procedure, S2206 denotes a received message discard procedure, and S2207 denotes a message release procedure.

FIG. 23 is a message sequence of the second embodiment. 2
Reference numeral 301 denotes a transmitting device, and 2302 denotes a receiving device.

FIG. 24 is a flowchart of a message transmission procedure according to the third embodiment. S2401 is a broadcast message sending procedure, S2402 waits of T _i time, S2403 is k times of repeated procedures, S2404 acknowledgment decision procedure, S
Reference numeral 2405 denotes a time-out determination procedure, S2406 denotes a reception failure device determination procedure, S2407 denotes an acknowledgment response recording procedure, and S2408 denotes an acknowledgment received message sending procedure to the receiving device.

FIG. 25 is a flowchart of a message receiving procedure according to the third embodiment. S2501 is a release time determination procedure,
S2502 is a procedure for determining the presence or absence of a received message, S25.
03 is a received message duplication determination procedure, S2504 is a procedure for adding to a release schedule list, S2505 is a message decryption procedure, S2506 is a received message discard procedure, S2507 is a message release procedure, A
Reference numeral 2508 denotes an acknowledgment response sending procedure, S2509 denotes an acknowledgment response received message determination procedure, and S2510 denotes a timeout determination procedure.

FIG. 26 is a message sequence according to the third embodiment. 2
Reference numeral 601 denotes a transmitting device, and 2602 denotes a receiving device.

FIG. 27 is a diagram illustrating a relationship between a message delivery method and a required number of packets.

FIG. 28 is a diagram summarizing the definitions of constants.

[Explanation of symbols]

1 ... broadcast transmission device, 2 ... reception device identifier table,
11-1 to n: broadcast receiving client (client device); 12-1 to n: broadcast receiving protocol device (receiving protocol device); 13-1 to n: network side interface device (network interface device); 1n: upper interface point; 15-1 to n: lower interface point; 16-1 to n: network side interface point;
-1 to n: Receiver device protocol device identifier, 31: Communication network.

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ Identification number Agency reference number FI Technical indication location H04L 9/32 H04L 9/00 601B H04M 11/08 673B 675C

Claims (19)

[Claims] 1. A receiving protocol device in a broadcast communication system for transmitting the same message from one or more transmitting devices to a plurality of predetermined receiving devices, wherein a plurality of the same messages transmitted from the transmitting device are simultaneously transmitted. Time setting means for setting a transmission time to a time agreed in advance with another reception protocol device in order to transmit to a higher-level execution device for processing, and executing the message from the transmission device in the plurality of higher-level execution devices Holding means for holding the message until the time set by the time setting means before transmitting the message to the device; and a receiving protocol device. 2. An identifier storage unit storing an identifier for identifying the reception protocol device from another reception protocol device, and prohibiting a third party from accessing the identifier stored in the identifier storage unit. A first prohibition unit; a second prohibition unit that prohibits a third party from intercepting the identifier when transmitting the identifier stored in the identifier storage unit to the transmitting device and performing authentication; An encryption key holding unit for holding the same encryption key as the encryption key used for encrypting the message in the device; and an encryption message sent from the transmitting device held by the encryption key holding unit. Decryption means for decrypting using the encryption key, and holding the decrypted message in the holding means until the time set by the time setting means is reached. Receiving protocol apparatus according to claim 1, wherein the a. 3. A broadcast message transmitting apparatus in a broadcast communication system for transmitting the same message from one or more transmitting apparatuses to a plurality of predetermined receiving protocol apparatuses, wherein each receiving protocol apparatus has a predetermined protocol. An authentication means for authenticating whether or not it has an identifier; a distribution means for previously distributing a predetermined encryption key only to the reception protocol device authenticated by the authentication means; A transmission means for encrypting a transmission message using an encryption key and transmitting the encrypted transmission message to each receiving protocol device. 4. The authentication means includes: an identifier storage unit storing an identifier for identifying each reception protocol device; and communicating with each reception protocol device, whereby the identifier of each reception protocol device is stored in the identifier storage unit. A confirmation unit for confirming whether the received protocol device is included in the identifier storage unit, if the identifier is included in the identifier storage unit, the reception protocol device is recognized as a proper reception protocol device, and the identifier is the identifier 4. The broadcast message transmitting apparatus according to claim 3, further comprising: a certifying unit for certifying that the receiving protocol device is not a legitimate receiving protocol device when the receiving protocol device is not included in the storage unit. 5. The authentication unit according to claim 1, wherein the authentication unit performs authentication using an identifier for identifying a reception protocol device stored in an identifier storage unit provided outside the broadcast message transmitting device. 5. The broadcast message transmitting device according to 3 or 4. 6. A transmission / reception protocol device comprising the reception protocol device according to claim 1 or 2 and the broadcast message transmission device according to claim 3, 4, or 5. 7. A transmission / reception protocol device, wherein the reception protocol device according to claim 2 and the broadcast message transmission device according to claim 4 are constituted by a single device. 8. The transmission / reception protocol device according to claim 3, wherein the broadcast message is transmitted to a receiving device of a predetermined broadcast group as a means for time distribution concerning delivery of the broadcast message and release thereof. a broadcast message reception confirmation step of receiving an acknowledgment from the receiving apparatus and transmitting step, in the verification step, after the message transmission, the receiver does not return an acknowledgment within a certain time T _s to a predetermined, preset A retransmission step of attempting to retransmit the message by one-to-one transmission up to a certain number of times _Smax; a message release time determining means for determining a message release time of the message; and reception of a predetermined broadcast group Release time notification step and release time notification that send a release time notification to the device And release time confirmation step of receiving an acknowledgment from the receiver for, at the confirmation step, after the message transmission, the receiver does not return an acknowledgment within a certain time T _s of predetermined certain predetermined number of times A transmission / reception protocol apparatus, comprising: a retransmission step of attempting to retransmit the release time notification by one-to-one communication up to _Wmax . 9. The receiving protocol device according to claim 1, wherein the receiving a broadcast message broadcast from the transmitting device and transmitting an acknowledgment to the message by one-to-one communication. Transmitting the release time notification broadcast from the transmitting device, returning an acknowledgment, and storing the release time included in the message for each corresponding received message; and A transmission / reception protocol device, comprising a step of transmitting a corresponding message to a host execution device at a release time stored for each. 10. The transmission / reception protocol device according to claim 3, wherein the broadcast message for transmitting a message to a reception device of a predetermined broadcast group is transmitted as a means for distributing a broadcast message and agreeing a time on release thereof. a transmission step, in the verification step and the broadcast message reception confirmation step of receiving an acknowledgment from the receiver, after the message transmission, the receiver does not return an acknowledgment within a certain time T _s to a predetermined, preset A retransmission step of attempting to retransmit the message by one-to-one communication up to a given number of times _Smax; a message release time determining means for determining a message release time; at a predetermined interval in over time T _i the release time notification including, in advance Receiving protocol and wherein the step of transmitting a number of times S _max because was, that the step of receiving a release success message from the receiving apparatus, comprising a. 11. The receiving protocol apparatus according to claim 1, wherein the step of receiving a broadcast message broadcasted from the transmitting apparatus, and transmitting an acknowledgment to the message by one-to-one communication. Transmitting to the device; receiving the release time notification broadcast from the transmitting device; and storing the release time included in the message for each corresponding received message; and Transmitting and receiving a corresponding message to a higher-level execution device; and transmitting a release success message to the transmission device when the release is successful. 12. The transmission / reception protocol apparatus according to claim 3, wherein a broadcast message for transmitting a message to a receiver of a predetermined broadcast group is transmitted as a means for time distribution concerning delivery of the broadcast message and release thereof. a transmission step, a broadcast message reception confirmation step of receiving an acknowledgment from the receiving device, in the confirmation step, after the message transmission, the receiver does not return an acknowledgment within a certain time T _s to a predetermined pre A retransmission step of attempting to retransmit the message by one-to-one communication up to a predetermined number of times _Smax ; message release time determination means for determining a message release time; at a predetermined interval in over time T _i the release time notification including, pre Sending a number of times S _max of a defined, transceiver protocol apparatus characterized by comprising. 13. The receiving protocol device according to claim 1, wherein the receiving device receives a broadcast message broadcasted from the transmitting device and transmits an acknowledgment to the message by one-to-one communication. Receiving the release time notification broadcast from the transmitting device, and storing the release time included in the message for each corresponding received message, and corresponding to the release time stored for each message. A transmission / reception protocol device, comprising a step of transmitting a message to a host execution device. 14. The transmission / reception protocol apparatus according to claim 3, wherein a step of determining a release time of a broadcast scheduled message is performed as a means of time distribution regarding delivery of a broadcast message and release thereof. A broadcast message transmitting step of transmitting a message including the release time to the broadcast group receiving apparatus; a broadcast message receiving confirming step of receiving a confirmation response from the receiving apparatus; a receiving device that does not return an acknowledgment within the time T _s has a retransmission step of attempting to resend the one-to-one communication with the message until the number S _max of predetermined characterized by comprising Transmit / receive protocol device. 15. The receiving protocol device according to claim 1, wherein the step of receiving a broadcast message broadcast from the transmitting device and transmitting an acknowledgment to the message by one-to-one communication. Receiving the message and the release time notification broadcast from the transmitting device, returning an acknowledgment, and storing the release time included in the message for each corresponding message. Transmitting a corresponding message to a higher-level execution device at the release time stored for each message. 16. A transmission / reception protocol apparatus according to claim 3, wherein a step of determining a release time of a broadcast scheduled message is performed as a means for time distribution concerning delivery of a broadcast message and release thereof. A notification of the release time including the release time is transmitted to the information group receiving device by a predetermined time T.
a predetermined number of times S _max at intervals of _s or more
A transmission / reception protocol device, comprising: a broadcast message transmission step of broadcasting only a message; and a switch for receiving a release success message from a reception device. 17. The receiving protocol apparatus according to claim 1, wherein the step of receiving the broadcast message broadcasted from the transmitting apparatus and the step of transmitting an acknowledgment to the message by one-to-one communication. Receiving the message and the release time notification broadcast from the transmitting device, returning an acknowledgment, and storing the release time included in the message for each corresponding message. A transmission / reception protocol device, comprising: transmitting a corresponding message to a higher-level execution device at a release time; and transmitting a release success message to a transmission device when the release is successful. 18. The transmission / reception protocol apparatus according to claim 3, wherein a step of determining a release time of a broadcast scheduled message as a means of time distribution regarding delivery of a broadcast message and release thereof, and reception of a broadcast group. The apparatus further comprises a broadcast message transmitting step of broadcasting a release time notification including the release time at a predetermined time T _i or more at a predetermined number of times S _max. Transmission and reception protocol device. 19. The receiving protocol device according to claim 1, further comprising the step of receiving a broadcast message broadcast from the transmitting device, and receiving a message body and a release time notification broadcast from the transmitting device. Storing a release time included in the message for each corresponding received message; and transmitting a corresponding message to the higher-level execution device at the release time stored for each message. Receiving protocol device.