Teleconferencing method for a secure key management system.

Background of the Invention

0001The present invention pertains to secure communications among users of a key management system and more particularly to a method for providing secure teleconferencing through the use of an unforgeable certification process.

0002In modern day telephony, specialized circuitry within the switching system provides for teleconferencing. Teleconferencing is simultaneous voice or data exchange between three or more users of a communication system. Typically, specialized networks and circuitry within the network of a switching system provide for interconnecting a number of users in a teleconference.

0003In secure communications systems, the problems of establishing connections among multiple users for voice or data exchange is further complicated by various security protocols. Establishing a secure teleconference between multiple users via a secure network of a switching system is very difficult. This is due to the fact that the terminal devices in a secure teleconferencing system may demand point-to-point, on-line establishment of traffic keys.

0004Further, establishing a session key between terminal devices connected to a secure network of a switching system requires transmission of many messages between the system and the users. This is very cumbersome and time consuming.

0005An object of the present invention is to provide a method to enable any number of terminal devices connected to a switching network to establish a secure conference arrangement by establishing an unforgeable certification method to provide a single session traffic key with a minimal number of point-to-point message exchanges between the terminal devices.

Summarv of the Invention

0006In accomplishing the object of the present invention, a novel teleconferencing method for a secure key management system is shown.

0007A secure teleconferencing method is achieved among a number of interconnected terminals via a switching network. Each of the terminals have previously been certified by a common authority. The secure teleconferencing method first establishes that all the teleconferenced terminals have previously been certified. Each terminal established that all the remaining terminals have been previously certified by a common authority. Then, the terminals collectively develop a unique, single session, session key for providing secure voice and data communication among the interconnected terminals.

Brief Description of the Drawing

0008<ul id="ul0001" list-style="none"><li>FIG. 1 is a block diagram of a teleconferencing arrangement via a switching network.</li><li>FIGS. 2A and 2B are flow charts of a secure teleconferencing method embodying the principles of operation of the present invention.</li><li>FIGS.3A and 3B are timing diagrams of inter-terminal message transmission.</li><li>FIG. 4 is a block diagram depicting the exclusive-OR operation of each terminal.</li></ul>

Description of the Preferred Embodiment

0009FIG. 1 depicts a number of terminals (A, B, C and D) connected to a switching network. This switching network includes circuitry which can provide for teleconferencing among these four terminals. Four terminals are shown here by way of example and not by way of limitation. Almost any number of terminals may be connected in the teleconferencing arrangement. The only limitations are the number of terminals which may physically be interconnected by the switching system.

0010In the present system, each of the terminals is a secure voice and data terminal. Each terminal includes the present secure teleconferencing method. This method may typically be implemented in software. However, hardware implementations are also possible.

0011The secure terminals A-D are designed for both full duplex and half duplex voice and data communication. Half duplex voice and data communication takes place over two wire communication networks. Full duplex voice and data communication occurs over four-wire telecommunication networks. For full duplex transmission, information is received and transmitted over two separate pairs of wires (channels) simultaneously. In half duplex transmission, a terminal puts energy on the line only when transmitting information. When its transmission is complete, its energy is removed allowing another terminal to transmit on the same channel. Teleconferencing is much more straight forward in the half duplex configuration, since energy is on the line only when a particular terminal is transmitting.

0012The present method pertains more readily to the half duplex communication mode. However, via time sharing existing channels or establishing a new channel, one per added terminal, this method is also applicable to the full duplex transmission mode. Each of the terminals A-D include a half duplex modem. Each terminal includes input devices such as a push button for indicating that the teleconferencing mode is selected and that the secure mode is selected. Each terminal also includes the ability to display the identify of each of the other teleconferencers. These identities may be scrolled on a visual display for the user of each terminal.

0013For security, each of the terminals employ a type of asymmetric RSA public encryption and private decryption process. The encryption process is noted by EM and the decryption process is noted by DM.

0014Turning now to FIGS. 2A and B, the secure teleconferencing method is shown. First, the users go off hook, block 1. Next, all the users are established in an unsecured teleconferenced arrangement via a usually switching system procedures, block 2. This is accomplished in the typical manner through conference circuitry in the switching network. The call or data transmission at this point is unsecured.

0015To accomplish the teleconference, each of the users of terminals A-D selects the conference mode via pushing a conference push button on his or her terminal. Next, one of the users presses the secure key on his or her terminal, block 3. This serves two functions. First, that terminal becomes the master terminal for controlling the teleconference function. Since all of the terminals include the present method, any of the terminals may be the master for the teleconference. The determination of the master terminal for teleconference is determined by the first to press the secure key on his terminal.

0016The second function of the master terminal is to transmit its access domain message in half duplex to each of the other conferenced terminals. To accomplish this, the master transmits P1800 data, block 3, stops transmitting for 256 bits (block 4), and sends P1800 data of 776 bits (block 5) followed by SCR1 data (scrambled ones for synchronization), block 6. Meanwhile, the slaves are set in the half duplex mode, block 7. The slave confirms the stop gap, block 8. The slaves detect the P1800 data, block 9, and train their modem to receive, 10. The master sends the access domain message (ADM), 11. This is followed by an end of message (EOM) indication, block 12.

0017Next, each slave receives the access domain message and stores it for later processing, block 13. The access domain messages of the master and other slave terminals are processed when the master's authentication message has been received.

0018The transmission of master's access domain message occurs as shown in FIG. 3A. The master first sends a P1800 protocol, followed by a gap, followed by another P1800 protocol, an SCR1 (scrambled 1's and 0's for modem synchronization) bit stream and then the bit stream of the access domain message, followed by an end of message (not shown). Each slave then responds in an appropriate time slot, is determined by a queue index function. This queue index function may be implemented in a number of ways, such as, a strictly arbitrary sequence.

0019The master terminal then sets its modem in the receive mode, blocks 14 and 15.

0020Next, each slave in the teleconference responds with its access domain message in the selected time slot. Since the slave terminals have received the master's access domain message, the NO path has been followed from block 16 to block 17. Next, it is the slave's turn to send its access domain message, the YES path is followed from block 17 to block 18. In FIG. 3A, slave 1 responds with its P1800 protocol (block 18) followed by SCR1 (block 19), and its access domain message ADM block 20 followed by EOM, block 21. Similarly, each slave responds in its time slot. Slave 2 responds in time slot 2. Slave n responds in time slot n. The master terminal receives all the access domain messages of each of the slaves, block 22. Block 23 determines whether all the slaves have responded. If all the slaves have not responded, block 23 transfers control to block 14 to wait for a response. When all the slave terminals respond, the master (block 24) processes the access domain messages of each of the slaves.

0021Next, the master sends its authentication message to each of the slaves, blocks 25 through 28. The authentication message of the master terminal is given by equation (1). AM_m = [(ID_master,EM_master)]xDM_kca      Equation (1) The Authentication Message identifies the particular master terminal, its public or encrypt key and the private decrypt key of its certification authority (KCA). This message is valid only if it is certified by the key certification authority. Further, each of the other terminals may decode this message only if they are certified under the same key certification authority.

0022Block 16 determines for each slave terminal that this is the authentication message. As a result, block 16 transfers control to block 29 which processes the master access domain message (ADM_m) and the other slaves' access domain messages (ADM_sn′S). Once the access domain messages have been thereby exchanged, each terminal can determine the identity of the terminal to which it is connected via the teleconference and that they are certified by a common certifying authority such as a KCA. FIG. 3B depicts this transmission of the master. This transmission consists of the 1800 protocol followed by the SRC1 bits, followed by the authorization message and an EOM (not shown). The master terminal sets itself to receive the authentication messages of the slaves, blocks 31 and 32. Next, each slave terminal processes the authentication message of the master (AM_m), block 30. Each slave terminal generates a unique random number (RN_s)and a random component message (RCM) from the random number, block 33. The RCM is made by encrypting the slave's unique random number in the master's public key as shown in Equation (2). (RN_s(n)) X EM_master = RCM_s(n)      Equation (2) Each slave, then, stores a copy of its unique random number for later use in the final stages of this process. Each slave then responds in its appropriate time slot with its authorization message followed by its random component message. The slave authorization message is similar to the master's authorization message. It is shown in Equation (3). AM_s = [(ID_slave,EM_slave)]XDM_kca      Equation (3) It includes the identity of the slave, the public key of the slave covered by the private key of the certifying authority or KCA.

0023When each slave determines that its time slot is present, that slave "broadcasts" its authorization message (AM_s), block 34. The slave terminal initialize its modem, blocks 35 and 36. Next, the authorization message (AM_s) and random component message (RCM_s) of each slave is transmitted to each of the other slaves, including the master terminal, block 37, followed by an end of message (EOM), block 38. Block 38 transfers control to block 34. Since this slave terminal has transmitted its AM_s and RCM_s, block 34 transfers control to block 39.

0024The transmission for slave 1 is shown in FIG. 3B. First the 1800 protocol is sent, then the SCR1 bits, the authorization message of slave 1, followed by approximately 200 milliseconds of filler which is a predefined binary bit pattern and finally the random component message of the slave 1 and end of message indication (not shown). Each slave similarly responds in its assigned time slot.

0025Each slave terminal sets its modem to receive the other terminals AM_s and RCM_s messages, blocks 40 and 41. Each slave terminal receives the other slave terminals, AM_s and RCM_s, block 42. Also, simultaneously the master terminal receives each of the slave terminals' AM_s and RCM_s messages, block 43.

0026When all the slave terminals have responded, each block 39 of each slave terminal transfers control to block 44. Simultaneously, when the master terminal has received all the AM_s and RCM_s messages, block 45 transfers control to block 46.

0027The random component message of each terminal includes a unique random number generated by that terminal. This random number is transmitted to the master terminal and is also stored within the terminal. The random number is encrypted with the master's public key to form the random component message. Therefore, even through all of the other slave terminals in the teleconference may receive the RCM, they will not be able to decrypt it.

0028As a result of these transmissions, each slave terminal receives all the authorization messages of the other terminals. Each slave then processes each of the authorization messages of the other slaves, block 44. This includes extracting the identity of each of the other slaves which have been covered by the private key of the certifying authority. The identity of each of the slaves is then scrolled on a viewing device of the terminal in the order in which they are received. As a result, each terminal has established the certified identity of each of the other conferees.

0029The master terminal has also received each of the authentication messages of the other terminals and processed them similarly to that described above for the slave, block 46. Next, the master decrypts each of the slave's random component message, block 46. Since each of the slave's random component message has been covered with the master's public key, the master may decrypt them by applying his private key since this is an asymmetric or one-way function. Each slave sets its modem to receive a message from the master terminal, blocks 47 and 48.

0030When the master terminal receives each of the authorization messages, it decodes each of the other terminals' identification. In addition, the master terminal then scrolls each of the identities of the slave terminals on each visual display on the terminal. Next, the master terminal decrypts each of the slave's random component message. Since each of the slaves have covered their random component with the master's public encrypt key, only the master is able to decode each of the random component messages by application of his private decrypt key. The random component is a predetermined number of bits of information. The random component is, as its name implies, generated randomly by each of the terminals. The larger in terms of bits the random component is, the more difficult this component is to determine for purposes of listening. As a result, the master terminal obtains each of the slave terminals random components. In addition, the master terminal has stored its previously generated random component.

0031Next, the master terminal generates a set of random numbers based upon each of the slave's random numbers and the master's random component or number, block 49. The equations for generating this set of random numbers are shown by Equation Set (4).<maths id="math0001" num=""><img file="EP0402083A2_D0001.tif" /></maths>

0032Examining the last equation of the sets of equations in Equation Set (4) as an example, it can be seen that the new random number for slave 1 RNs(1)′ is comprised of each of the random numbers of the other slaves and the master. These random numbers are exclusive-ORd. Each of the + symbols represents an exclusive-OR operation. The new random number for slave 2 RNs(2)′ is an exclusive-OR of each of the other terminal's random numbers except terminal 2. As a result, the master terminal generates one new random number for each of the terminals in the teleconference which is an exclusive-OR of all the other terminals in that teleconferenced random number except the terminal to which that random number is to be sent.

0033The terminal to which the random number is to be sent has previously stored its own random number. Therefore in FIG. 4, when a given terminal receives the new random number from the master terminal, it must simply exclusive-OR 99 its own stored random number with the primed random number sent from the master terminal. As a result, the terminals have developed a session key. This session key is useful only for this session and is generated for each teleconference call. The generation and transmission of the session key also obeys the "two-man rule", in that interception of a single message will not result in obtaining the random component and hence, the session key.

0034The master terminal covers each of the newly generated primed random numbers with the public encrypt key of the slave terminal to receive that random number as shown in Equation Set (5):<maths id="math0002" num=""><img file="EP0402083A2_D0002.tif" /></maths> Again, since the new primed random number is covered by the slave public encrypt key, only that particular slave may decode using his private decrypt key the new primed random component message. The master generates one such random component message for each slave, covering that message with the slave's public encrypt key.

0035Next, the master generates a crypto synchronization pattern (CS), block 50. The master terminal now formats the final secure call setup message. This message is formulated with the P1800 protocol (block 51) followed by the SCR1 information (block 52), followed by the random component message of each slave in the appropriate time slot. A predetermined amount of filler bits follow the last random component message. After the filler bits, the crypto synchronization message is sent to all terminals by the master. The crypto synchronization message indicates the point at which the symmetric encrypt/decrypt function is to begin. The master terminal then transmits the formatted message (corresponding primed random number) followed by crypto synchronization message, block 53, followed by an EOM, block 54.

0036Each slave terminal then receives its new primed random component message, block 55. In addition, each terminal receives the crypto synchronization message, block 56. Next, each terminal processes the new primed random component message by exclusive-ORing the primed random component which it received with its own previously stored random component, block 57. This exclusive-OR produces a session key. Each of the terminals now have this session key. Therefore, this session key may be used by each terminal's symmetric encrypt/decrypt process to provide secure communications between each of the terminals in a teleconferencing arrangement. A minimal number of messages has been used to achieve security.

0037Lastly, each terminal processes the crypto synchronization message, block 58. This starts each of the symmetric encrypt/decrypt functions at a predetermined point. Therefore, since each terminal has the key and the starting point, it can encode and decode information transmitted to and received from the other terminals in a secure manner.

0038Although the preferred embodiment of the invention has been illustrated, and that form described in detail, it will be readily apparent to those skilled in the art that various modifications may be made therein without departing from the spirit of the invention or from the scope of the appended claims.