METHOD FOR MANAGING SPEECHING ON A COMMUNICATION CHANNEL IN THE CONTEXT OF ALTERNATE COMMUNICATIONS

专利摘要:
In the context of alternate communications in a communication network composed of a plurality of subnetworks and respectively associated with super-node devices (110, 120), a receiving super-node device (302) a request for making speech: transmits (303) a pre-authorization message to speak; determines (308) a delay time as a function of a distribution law fτ (t) of the time period between a speech release and a next speech, and a distribution law fδ (d) of network latency , so that the probability of subsequent dismissal of the speech is below a predefined threshold; and broadcasts (318) the data transmitted as part of the speaking, when said timer has elapsed. When a request for more priority speech is detected before the expiration of the timer, said super-node device does not broadcast said data and notifies that the speech is finally denied.
公开号:FR3021482A1
申请号:FR1454670
申请日:2014-05-23
公开日:2015-11-27
发明作者:Jaume Joan Ventura；Laurent Franck
申请人:Astrium SAS；
IPC主号:

专利说明:

[0001] The present invention relates to talk management over a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks.
[0002] PTT (Push-To-Talk) communication systems are widely used, particularly in the context of communications in professional circles by security services (police, firefighters), building and public works companies , and the industry. PTT-based communication systems implement a half-duplex conversation method on a shared communication channel, that is, communication goes in both directions, but not simultaneously . PTT type communication systems implement Link Level (MAC) mechanisms ("Medium Access Control" in English) to implement conversations between users having interconnected terminal devices. by a communication network, so that only one user speaks at a time, the other users then listening to the one who spoke. In order to ensure that a single communication device speaks on the shared communication channel, various mechanisms are known. It is notably known a mechanism called "Floor Control" in English, implemented at the level of the application layer ("application layer" in English). Such a mechanism of "Floor Control" is thus applicable to audio data transmission, but also to a transmission of video data or even audio-video data or data of another type. Such a "Floor Control" mechanism typically relies on the implementation of a server from which each terminal device must request authorization before being able to speak. This server is then in charge of managing the conflicts of requests to speak, that is to say, to manage the situations in which several terminal devices require respective speech-taking.
[0003] In practice, a terminal device wishing to take the floor sends the server a request to speak ("Floor Request" in English). This desire to speak out typically emanates from a user by pressing a dedicated button of the terminal device. The server receives said request and selects the terminal device to which the speech is granted among possible competing requests addressed to it. The server sends a Floor Grant confirmation message to the terminal device that has been selected and a notification message representative of said Floor Taken in the United States. other terminal devices, which then listen to the terminal device to which speaking has been granted. The terminal device to which the speaking has been granted can then take the floor and the other terminal devices are invited to wait for the release of the call to issue a request to speak. When the terminal device to which the speech has been granted has ended with the communication, said terminal device sends the server a request for release of speech ("Floor Release" in English). The server then warns the other terminal devices that the communication has been released, and that a new speech can be implemented. The communication network may be composed of a plurality of subnetworks interconnected by a communication link, e.g. a satellite link, involving a high transmission latency between said subnetworks. This transmission latency is not negligible, especially given the transmission latency within the sub-networks. Because of this transmission latency, the server can then typically have to wait for long periods of time after receiving a request to speak to determine if one or more concurrent requests exist, including from another subnet than the one from which the request for speech received is received. If the server does not wait for such periods of time before deciding which terminal device to grant speaking, the terminal devices closest in terms of server latency are favored and the mechanism of "Floor Control" is unfair ( Unfair in English). Subnetworks may further involve very different transmission latencies. It is therefore desirable, in these circumstances, to provide a solution that is fair regardless of the location, in the communication network, of the terminal device transmitting a request to speak. It is further desirable to provide a solution that reduces the access time to speaking, that is to say the period of time between the time at which a terminal device requires speaking and the instant at which said terminal device actually obtains the authorization to speak, while limiting the risk of not respecting the chronology of attempt to speak effectively (ie the risk that a user X having chronologically tried to speak before a user Y not be heard by at least one other user Z after the user Y). It is also desirable to provide a solution in which, when a terminal device has actually obtained the authorization to speak, the data transmitted by said terminal device in the context of the speech is received substantially simultaneously by the terminal devices. listening. It is also desirable to provide a solution that is simple to implement and low cost. The invention relates to a speech management method on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks, each subnetwork being associated with a communication network. super-node device of the communication network, the method being implemented by a said super-node device. When a first talk request is received from a terminal device of the subnet to which said super-node device is associated and said super-node device is not aware of at least one second request for decision of more priority speech, said super-node device performs the following steps: propagate the first request for speaking to each other super-node device; transmitting a pre-authorization message to said terminal device to allow said terminal device to speak; transmitting a pre-authorization notification message to each other terminal device of the subnet to which the super-node device is associated, so as to place each said other terminal device listening to the shared communication channel; determining a duration of a delay according to a distribution law fr (t) of period of time between a speech release and a next speech, and a distribution law f8 (d) of network latency between an instant at which a request for speech is transmitted by a terminal device and a time at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from the speaking is below a first predefined threshold; storing in a buffer memory data received from said terminal device in the context of speaking, and propagating said data to each other super-node device; enable read buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer is elapsed. In addition, when the first polling request is received from a terminal device of another subnet than that to which said super-node device is associated and said super-node device is not aware of the less a second priority request for speaking, said super-node device performs the following steps: transmitting a pre-authorization notification message to each terminal device of the subnet to which the super-node device is associated so as to place each terminal device of said subnet listening to the shared communication channel; determining the duration of the time delay according to the distribution law Mt) and the distribution law f8 (d), so that the probability Pfau of removing said terminal device from the speech is below the first predefined threshold; storing in the buffer data propagated by another super-node device as part of the speaking session; enable read buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer is elapsed. In addition, when said super-node device detects at least a second request for priority speech before expiration of said timer, said super-node device releases the buffer memory and, when said first request has been received from a terminal device of the subnet which is associated with said super-node device, said super-node device notifies said terminal device that the speech is denied to said terminal device. According to a particular embodiment, said super-node device transmits the pre-authorization message for speaking after the expiry of another time delay determined according to the distribution law fr (t) and the distribution law f8 (d), so that the probability of removing said terminal device from speaking is below a second predefined threshold strictly greater than said first threshold. According to a particular embodiment, the distribution law fr (t) is defined by a model of log-normal type and expressed as follows: 1 (iog (0 -, u) 2 fr (t) = 202 t 0 where a represents the standard deviation of the distribution law Mt) and kt represents the average of the distribution law fr (t) According to a particular embodiment, the distribution law f8 (d) is defined by a model under the form of a mixture of Weibull distribution distributions with three parameters and expressed in the following way: f8 (d) => qi1.8i (d) where i is an index on the components of the mixture and where qi represents coefficients such as: Ri = 1 and where: si (d - fsi (d) = e-1 ri) for cl> li where si is a shape parameter, ri a scale parameter and 1i a location parameter. According to a particular embodiment, the distribution laws fr (t) and f8 (d) are predetermined by said super-node device as a function of long-term statistics previously collected during the exchange of messages in the context of the communications in alternate mode. . According to a particular embodiment, the Pfail probability of removing said terminal device from the speaking space is expressed as follows: N-1 Pfa1 = 1 - (N-1 k Ppushk (1 - Ppush) N-1-k (1 - do) kk = 0 where Ppush represents the probability for a terminal device to request the talk determined by said super-node device by counting the number of concurrent speech requests detected for each given speech, or Where Rq represents a quantity of concurrent requests received for each granted speech q with Rq> 1 and where Q represents a quantity of allocated speech slots considered; and where: Prob {T1 + d1 + Tb <T2 + d2 & T1> T2} fact = Prob {Ti + <T2 + d2> T2} + Prob {Ti + <T2 + d2 <T2} where: Prob {x} represents the probability that an expression represented by x is verified and where T1 is representative of a time at which said expression request for speaking has been transmitted, T2 is representative of the moment at which a second request for more priority speaking has been transmitted, di is representative of a transmission latency experienced by said first request to speak in the communication network, d2 is representative of a transmission latency experienced by said second request to speak in the communication network, and Tb is representative of the duration of the timer whose expiration triggers the reading of said buffer memory.
[0004] According to a particular embodiment, each request for speaking includes information representative of a period of time elapsed between an instant at which a speech release message has been received as part of the last speech by the terminal device having transmitted said request for speaking and the time at which said terminal device transmitted said request.
[0005] According to a particular embodiment, said super-node device considers that a said second request for speaking is temporally more priority than said first request for speaking when the period of time represented by said information included in said second request for taking of speech is less than said period of time represented by said information included in said first request for speaking. According to a particular embodiment, in order to determine the duration of said delay, the super-node device performs the following steps: determining a period of time t elapsed between an instant at which a speech release message has been received in the context of the last speech by the terminal device having transmitted said first request to speak and the time at which said terminal device has transmitted said first request to speak; determining a latency c / suffered by said first request to speak between the transmission of said first request for speaking by said terminal device and the reception of said first request for speaking by said super-node device; and determining the duration of said timer from a look-up table which provides a delay time as a function of a set of predefined pairs (t, cl), the contents of the correspondence table being pre-defined according to the distribution law Mt) and the distribution law f8 (d) concerned so that the probability Pfau of removing a terminal device of the speech is below the first predefined threshold. According to a particular embodiment, to fill the correspondence table, a device performs the following steps: determining parameters for representing the distribution law fr (t); determining parameters to represent the applicable distribution law f8 (d); determining a quantity of terminal devices present in the communication system; define minimum and maximum values of time period t between a speech release and a next speech, and latent network cl between an instant at which a request for speech is transmitted by a terminal device and a time at which said request is actually received by said super-node device; defining a step between two successive time period values t to be considered and a step between two successive latent network latency values to be considered, so as to define each pair (t, cl) to be represented in the correspondence table; set a maximum acceptable value for the duration of the timer; define the value of said first threshold; define an acceptable margin of error E; determine by bisection, according to distribution laws Mt) and f8 (d), a delay time for each pair (t, cl) to be represented in the correspondence table, so that the probability Pfau of destitute a speech terminal device is at a distance from the first predefined threshold lower than said acceptable error margin E and that the delay time is less than said maximum acceptable value. According to a particular embodiment, when a speech release message is received from said terminal device and said terminal device is on the subnet to which said super-node device is associated, said super-node device performs the following steps: propagate said speech release message received at each other super-node device; transmitting a speech release notification message to each other terminal device of the subnet to which the super-node device is associated, after the buffer has been emptied; and release the buffer memory; and when a speech release message is received from said terminal device and said terminal device is on another subnet than that to which said super-node device is associated, said super-node device performs the following steps: transmit a message speech release notification to each terminal device of the subnet to which the super-node device is associated, after the buffer has been emptied; and release the buffer. According to a particular embodiment, various configurations of the communication network being possible, said super-node device selects the distribution law fr (t) and the distribution law f8 (d) from among a predefined set of distribution laws, as a function of an effective configuration of the communication network. The invention also relates to a speech management method on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks, each subnetwork being associated with a super-node device of the communication network, the method being implemented by a said super-node device. When a first talk request is received from a terminal device of the subnet to which said super-node device is associated and said super-node device is not aware of at least a second request for decision of more priority speech, said super-node device performs the following steps: propagate the first request for speaking to each other super-node device; transmitting a pre-authorization message to said terminal device to allow said terminal device to speak; transmitting a pre-authorization notification message to each other terminal device of the subnet to which the super-node device is associated, so as to place each said other terminal device listening to the shared communication channel; determining a duration of a delay according to a distribution law fr (t) of period of time between a speech release and a next speech, and a distribution law f8 (d) of network latency between an instant at which a request for speech is transmitted by a terminal device and a time at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from the speaking is below a predefined threshold; storing in a buffer memory data received from said terminal device in the context of speaking, and propagating said data to each other super-node device; and enable read buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer is elapsed. In addition, when said super-node device detects at least a second priority request for speech priority before expiration of said timer, said super-node device performs the following steps: release the buffer; and notifying said terminal device that speaking is denied to said terminal device. The invention also relates to a speech management method on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks, each subnetwork being associated with a super-node device of the communication network, the method being implemented by a said super-node device. When a first speaking request is received from a terminal device of another subnet than that to which said super-node device is associated and said super-node device is not aware of at least one second request for more priority speaking, said super-node device performs the following steps: transmitting a pre-authorization notification message to each terminal device of the subnet to which the super-node device is associated, of so as to place each terminal device of said subnet listening to the shared communication channel; determining the duration of the delay according to a distribution law fr (t) of the period of time between a speech release and a next talk, and a distribution law f8 (d) of network latency between a instant at which a request to speak is transmitted by a terminal device and a time at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from the speaking is below a predefined threshold; storing in a buffer memory data propagated by another super-node device as part of speaking; enable read buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer is elapsed. In addition, when said super-node device detects at least a second priority request for speaking priority before expiration of said timer, said super-node device releases the buffer memory.
[0006] The invention also relates to a super-node device for managing speech on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks, each sub-network. network being associated with a super-node device of the communication network. Said super-node device is such that, when a first request to speak is received from a terminal device of the subnet to which said super-node device is associated and that said super-node device is not aware of at least one second request for more priority speaking, said super-node device includes: means for propagating the first request for speech to each other super-node device; means for transmitting a pre-authorization message to said terminal device to allow said terminal device to speak; means for transmitting a pre-authorization notification message to each other terminal device of the subnet to which the super-node device is associated, so as to place each said other terminal device listening to the shared communication; means for determining a duration of a delay according to a distribution law fr (t) of the period of time between a speech release and a next speech, and a distribution law f8 (d) of network latency between an instant at which a request for speech is transmitted by a terminal device and a time at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from the speaking below a first predefined threshold; means for storing in a buffer memory data received from said terminal device in the context of speaking, and for propagating said data to each other super-node device; means for activating the read buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer has elapsed. In addition, when the first polling request is received from a terminal device of another subnet than that to which said super-node device is associated and said super-node device is not aware of the less a second request for more priority speaking, said super-node device includes: means for transmitting a pre-authorization notification message to each terminal device of the subnet to which the super-node device is associated so as to place each terminal device of said subnet listening to the shared communication channel; means for determining the duration of the delay as a function of the distribution law fr (t) and of the distribution law f8 (d), so that the probability Pfau of removing said terminal device from the speaking space is below the first predefined threshold; means for storing in the buffer memory data propagated by another super-node device as part of the speaking process; means for activating the read buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer has elapsed. In addition, when said super-node device detects at least a second priority request for a speech priority before expiration of said timer, said super-node device includes means for releasing the buffer memory and, when said first request has been received in from a terminal device of the subnet to which said super-node device is associated, said super-node device includes means for notifying said terminal device that the speech device is denied to said terminal device. The invention also relates to a super-node device for managing speech on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks, each sub-network. network being associated with a super-node device of the communication network. When a first talk request is received from a terminal device of the subnet to which said super-node device is associated and said super-node device is not aware of at least a second request for decision of more priority speech, said super-node device includes: means for propagating the first speech request to each other super-node device; means for transmitting a pre-authorization message to said terminal device to allow said terminal device to speak; means for transmitting a pre-authorization notification message to each other terminal device of the subnet to which the super-node device is associated, so as to place each said other terminal device listening to the shared communication; means for determining a duration of a time delay according to a distribution law Mt) of time period between a speech release and a next speech, and a network latency distribution law f8 (d) between an instant at which a request to speak is transmitted by a terminal device and a time at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from the speaking is in below a predefined threshold; means for storing in a buffer memory data received from said terminal device in the context of speaking, and propagating said data to each other super-node device; and means for enabling read buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer has elapsed. In addition, when said super-node device detects at least a second priority request for speech timeout before said timer expires, said super-node device includes: means for releasing the buffer memory; and means for notifying said terminal device that speaking is denied to said terminal device. The invention also relates to a super-node device for managing speech on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks, each sub-network. network being associated with a super-node device of the communication network. When a first speaking request is received from a terminal device of another subnet than that to which said super-node device is associated and said super-node device is not aware of at least one second request for more priority speaking, said super-node device includes: means for transmitting a pre-authorization notification message to each terminal device of the subnet to which the super-node device is associated, of so as to place each terminal device of said subnet listening to the shared communication channel; means for determining the duration of the delay as a function of a distribution law fr (t) of the period of time between a speech release and a next speech, and a distribution law f8 (d) of latency network between an instant at which a request for speech is transmitted by a terminal device and a time at which said request is actually received by said super-node device, so that the probability Pfau to remove said terminal device from the speaking stand is below a predefined threshold; means for storing in a buffer memory data propagated by another super-node device in the context of speaking; means for activating the read buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer has elapsed. In addition, when said super-node device detects at least a second request for more priority speaking before expiry of said timer, said super-node device includes means for releasing the buffer memory.
[0007] The characteristics of the invention mentioned above, as well as others, will emerge more clearly on reading the following description of an exemplary embodiment, said description being given in relation to the attached drawings, among which: Fig. 1 schematically illustrates a communication system in which the present invention can be implemented; FIG. 2 schematically illustrates an example of hardware architecture of at least one interconnection device of the communication system; FIG. 3 schematically illustrates first exchanges in the communication system to allow a terminal device to speak in the context of a PTT type mechanism; FIG. 4 schematically illustrates second exchanges in the communication system to allow the terminal device to voluntarily release the speech in the context of the PTT type mechanism; FIG. 5A schematically illustrates an algorithm, implemented by a first interconnection device of the communication system, to allow the terminal device to speak in the context of the PTT type mechanism; FIG. 5B diagrammatically illustrates an algorithm, implemented by a second interconnection device of the communication system, to enable the terminal device to speak in the context of the PTT type mechanism; FIG. 6A schematically illustrates an algorithm, implemented by the first interconnection device, to allow the terminal device to voluntarily release speech in the context of the PTT type mechanism; FIG. 6B schematically illustrates an algorithm, implemented by the second interconnection device, to enable the terminal device to voluntarily release the speech in the context of the PTT type mechanism; and - FIG. 7 schematically illustrates an algorithm implemented by a communication system interconnection device for managing a speech conflict in the context of the PTT type mechanism.
[0008] To make it possible to reduce the access time for speaking on a communication channel in a communication network, it is proposed to organize the communication network into interconnected subnetworks and to implement a device, called super -Node SN, by subnet. A subnetwork is a subset of the communication network, the same subnetwork can implement different communication technologies and thus form a hybrid subnetwork. The subnets being interconnected, the SN devices of the communication network are then able to communicate with each other through the interconnections between the subnetworks. The communication network comprises a plurality of terminal devices to share a communication channel, each subnet having at least one such terminal device. The SN devices are coordinated to manage speaking on the shared communication channel and thus serve as a relay between the terminal devices. It is then proposed that an SN device receiving a request to speak from a terminal device of the subnet to which said SN device is associated transmits an authorization for early speaking, called pre-authorization ("Floor Pre Grant "in English). Following this pre-authorization, said terminal device transmits data as part of the speaking session; these data are stored by said SN device in a buffer before being broadcast to terminal devices of said listening subnet. When said device SN receives said data, said device SN further propagates said data to the other SN devices of the communication network, which also implement a temporary storage buffer. This temporary storage entails placing, at the level of each SN device, the broadcasting of said data for a period of time which is determined by each SN device according to a distribution law of the period of time between a speech release. and a next talk (ie the speech that sequentially follows said speech release) and a network latency distribution law between an instant at which a request to speak is transmitted by a terminal device and a time at which said request is actually received by said SN device. Storing the data in the buffer is managed by a timer which, once passed, triggers the reading of said buffer to enable the actual transmission of said data to said listening end devices. This approach makes it possible to shorten the access time for speaking and to maintain an equitable character in the speaking authorization process, regardless of the location in the communication network of the terminal device requiring the speech, while by ensuring that the risk that a terminal device to which speech has been granted is removed is statistically below a predefined threshold and while ensuring the best timing of speaking. It is said that a terminal device is removed when termination of the speech by said terminal device without the will of said terminal device and while speaking by said terminal device had been confirmed.
[0009] Thus, the waiting time of the data in the buffer memory before broadcasting to the listening terminal devices aims to maintain the probability of removing a terminal device below a predefined threshold. An error case therefore occurs when a request to speak is received after expiry of the timer triggering the reading of the buffer associated with another request to speak which should have been temporally less priority, that is, ie when: + di + Tb <T2 + d2 & T1> T2 where T1 is representative of the moment at which the request to speak which should have been less priority was transmitted, T2 is representative of the moment at which the other request for speaking has been transmitted, di is representative of a transmission latency experienced by the request to speak which should have been less priority in the communication network, d2 is representative of a transmission latency undergone by the other request for speaking in the communication network, and Tb is representative of the duration of the timer whose expiration triggers the reading of the buffer associated with the request speaking which should have been less priority. The probability that such an error case occurs can then be expressed as follows: Pifail = Prob {Ti + di <T2 + d2> T2} + Prob {Ti + di <T2 + d2 <T2} where: Prob {x} represents the probability that an expression represented by x is satisfied, and where: T1 +00 Mt) f8 (d) dd dt Prob {Ti + + Tb <T2 + d2 & Ti> T2} = f + f co Ti + Cli + Tb-t T1 + co Prob {Ti + <T2 + d2 & Ti> T2} = ff fr (t) fs (d) dd dt o + co + co Prob {T1 + d1 + Tb <T2 + d2 & T1 <T2} = ff fr (t) fs (d) dd dt T1 Prob {T1 + di + Tb <T2 + d2 & T1> T2} with fr (t) a distribution law of time period t between a speech release and a forthcoming speech and f8 (d) a network latency distribution law c / between an instant at which a request to speak is transmitted by a terminal device and a time at which said request is actually received by a SN device considered. Finally, according to a first embodiment, considering two competing requests to speak, the probability Pf that such an error case occurs is expressed as follows: Pfail = fail According to a second embodiment considering that all terminal devices may seek to speak concurrently, the probability that such an error occurs is expressed as follows: Pf ail = 1 - (1 - fau) N- Where N represents the quantity of terminal devices present in the communication network. According to a third embodiment, considering that all terminal devices will not seek to speak competitively, the probability Pfau that such an error case occurs is expressed as follows: N-1 Pf garlic = 1- (N k-1) D k (1 push P push) N-1-k (1 fact) k = 0 where Ppush represents the probability for a terminal device to require speech. The probability Ppush can be determined by each SN device by counting a quantity of concurrent speech requests detected for each granted speech, ie: 7 12 Ra-11 Ppush = N 1 / Q q = 1 where Rq represents the the quantity of concurrent requests received for each granted speech q (Rq> 1) and where Q represents a quantity of allocated speech tapers considered.
[0010] Thus, by virtue of these distribution laws fr (t) and f8 (d), it is possible to determine a delay time Tb making it possible to statistically ensure that the probability Pfau that a request to speak which should have been more priority is received after expiry of the delay time Tb is less than or equal to a predefined threshold P - target - The proposed approach also makes it possible to expire substantially simultaneously the readout time of the buffer memory of each SN device implemented. It is a spoken word, offering all listening users the opportunity to hear the communication at the same time. Fig. 1 schematically illustrates a communication system in which the present invention can be implemented.
[0011] The communication system comprises a first interconnection device 100a and a second interconnection device 100b. The first 100a and second 100b interconnection devices are interconnected by a communication link 100. The communication link 100 may be a physical link or a logical link. The communication link 100 can be part of a communication network, such as the Internet. The communication link 100 may be a terrestrial communication link or a satellite communication link. The first interconnection device 100a is in charge of interconnecting a first subnet 101 to the communication link 100, and the second interconnection device 100b is in charge of interconnecting a second subnet 102 to the communication link 100. Once interconnected by the communication link 100, the first 101 and second 102 subnets thus form a communication network, so that one or more terminal devices 111a, 111b, 111c of the first sub-network 101 can indirectly communicate with one or more end devices 121a, 121b, 121c of the second subnet 102.
[0012] In order to allow said terminal devices to share a communication channel in the context of a PTT type mechanism, the communication system includes SN devices 110 and 120. Each sub-network 101 and 102 is associated with an SN device to which each terminal device of said subnetwork refers to speak on the communication channel. Thus, in FIG. 1, the first sub-network 101 is associated with a first SN device 110 and the second sub-network 102 is associated with a second SN device 120. The SN device functionalities described below are preferably implemented at the level of the application layer of the Open Systems Interconnection (OSI) model. Thus, each SN device can be implemented on an independent machine, for example according to the architecture presented below in relation with FIG. 2. Each SN device may also be incorporated in the interconnection device which connects the subnet to which said SN device is associated with the communication link 100. It should also be noted that each interconnection device may be a single piece of equipment. in charge of implementing the interconnection of a sub-network and the communication link 100, as well as the implementation of said sub-network. Each interconnection device may also be a set of equipment, one or more equipment being in charge of implementing the interconnection of a sub-network and the communication link 100 and one or more other equipment being in communication. implementation of said subnetwork. Each subnetwork may be a wireless communication cell in which the interconnection device of said cell has access point or base station functionality. For example, each cell conforms to the IEEE 802.11 family of standards, also known as Wi-Fi (registered trademark). Each sub-network may also be a wired Local Area Network (LAN) or a LTE (Long Term Evolution) network. The first 101 and second 102 subnets may be different in nature, i.e., the first 101 and second 102 subnets may conform to different communication technologies. Such a communication system is particularly useful in emergency and emergency situations. Indeed, emergencies and relief situations in the context of natural disasters (eg avalanches, earthquakes, floods) or human-made disasters (eg bombings, industrial disasters) are often accompanied by disruption or even a temporary destruction of telecommunication cellular network infrastructures previously deployed on site and / or terrestrial access networks ("backhaul network" in English) for interconnecting such infrastructures. It is then necessary to temporarily deploy cells on site and to interconnect these cells together to allow rescuers to communicate and thus coordinate. Satellite links are then frequently used to interconnect these deployed cells temporarily on site. Such a communication system can also be advantageously implemented in the context of maintaining in operational condition a private communication network intended for example for management and security operations of critical infrastructures (eg nuclear power plant, site sensitive industrial). In this context, the use of a communication system, insensitive to the deployment of a backup satellite link to interconnect these private networks in the event that the terrestrial access networks are inoperative, offers a guarantee of resilience of the means. Communication. In the context of the PTT mechanism, only one terminal device at a time can speak on the shared communication channel, the other terminal devices then being placed in a listening situation of the terminal device having spoken. As already mentioned, a terminal device that has spoken on the shared communication channel transmits data to the other terminal devices then placed in a listening situation. This data is for example voice type audio data. Such a situation occurs when the communication system is a half-duplex telephone communication system, that is, communication is going both ways, but not simultaneously. The data transmitted during the speech and broadcast to the listening end devices may be of another type, such as video or audiovisual data. For example, the terminal devices include a camera and, when a predetermined event occurs on a terminal device, said terminal device requires speech to broadcast to the other terminal devices the video data captured by the camera of said terminal device. The communication system may include a larger amount of subnetworks. Each subnetwork includes an interconnection device for connecting said subnetwork to the other subnetworks. These interconnection devices are then interconnected via one or more communication links, so that each terminal device of a subnetwork is able to access, after having been authorized by an SN device associated with said subnetwork, to said channel. of transmission. For the sake of simplicity, the implementation of the invention is detailed in a communication system including only two subnetworks, but the principles described here are applicable identically in the presence of a communication system including a larger one. amount of subnets. The communication system can be adapted to manage a plurality of groups of terminal devices, the terminal devices of a group to share a dedicated communication channel to said group. Such groups are typically defined according to a logical organization reflecting the operational reality of teams using the communication system. It follows that each group may be composed of terminal devices distributed in different subnetworks of the communication system and not necessarily present within the same sub-network. It should also be noted that a terminal device can belong to several groups. When several groups coexist in the communication system, the mechanisms described below are implemented independently for each group.
[0013] In the communication system, the first 110 and second 120 SN devices share the same time reference. In a first example, the first 110 and second 120 SN devices rely on GPS ("Global Positioning System") technology to share the same time reference. According to a second example, the first 110 and second 120 SN devices rely on the Network Time Protocol (NTP) as described in the normative document RFC 5905. Preferably, each terminal device shares the same time reference as the SN device associated with the subnet to which said terminal device belongs.
[0014] In a preferred embodiment, the terminal devices and the SN devices communicate within the communication network by means of IP format messages ("Internet Procotol" in English), and in a particular embodiment, the data transmitted by said terminal devices on the shared communication channel are of the VoIP ("Voice over IP") type.
[0015] Fig. 2 schematically illustrates an example of hardware architecture of the first SN device 110 and / or the second SN device 120. Consider later that the example of hardware architecture shown schematically in FIG. 2 corresponds to the first device SN 110. The first device SN 110 then includes, connected by a communication bus 210: a processor or CPU (Central Processing Unit in English) 201; Random Access Memory (RAM) 202; a ROM (Read Only Memory) 203; a storage unit 204 or a storage medium reader, such as a SD (Secure Digital) card reader or a hard disk drive (HDD); and a set of interfaces 205.
[0016] The set of interfaces 205 allows the first SN device 110 to communicate with the second SN device 120 via the communication link 100, as well as with the terminal devices of the first subnet 101. Considering that the example of hardware architecture shown schematically in FIG. 2 corresponds to the second SN device 120, the set of interfaces 205 allows the second SN device 120 to communicate with the first SN device 110, as well as with the terminal devices of the second subnet 102. The processor 201 is capable of executing instructions loaded into the RAM 202 from the ROM 203, an external memory, a storage medium, or a communication network. When the first SN device 110 is turned on, the processor 201 is able to read instructions from RAM 202 and execute them. These instructions form a computer program causing the processor 201 to implement all or part of the algorithms and steps described below.
[0017] All or some of the algorithms and steps described below can thus be implemented in software form by executing a set of instructions by a programmable machine, such as a processor or a DSP (Digital Signal Processor in English) or a microcontroller , or be implemented in hardware form by a machine or a dedicated component, such as a Field-Programmable Gate Array (FPGA) or an Application-Specific Integrated Circuit (ASIC). Fig. 3 schematically illustrates first exchanges in the communication system to allow a terminal device to speak in the context of the PTT type mechanism. In the context of FIG. 3, it is considered that the first sub-network 101 comprises the terminal devices 111a and 111b, and the second sub-network 102 comprises the terminal device 121a. It is further considered that the terminal device 111a attempts to speak in the context of the PTT type mechanism. In a step 301, the terminal device 111a transmits to the first SN device 110 a request for speaking ("Floor Request" in English).
[0018] In a step 302, the first SN device 110 receives the request transmitted in step 301 by the terminal device 111a and verifies that there is no request for speaking which is concurrent and more priority. A higher priority speaking request determining mechanism is described below in the context of the algorithm of FIG. 7. The first device SN 110 extracts from the received request information representative of a time at which the terminal device lila transmitted said request. In a preferred embodiment, said information corresponds to a period of time elapsed between an instant at which the speech release message was received by the terminal device 111a in the context of the last speech (a reference time common to the communication network in the case of the very first speech) and the time at which the terminal device lila transmitted said request. In this case, said information is representative of the reactivity of the user to want to access the shared communication channel with respect to the last speech. It is thus possible to consider as a priority a terminal device seeking to speak in a shorter period of time after the last speech release, despite the fact that the terminal devices are respectively notified of the speech release at different times. . In another embodiment, said information corresponds to the instant at which the terminal device 111a transmitted said request according to the shared time reference.
[0019] In a step 303, when there is no such competing request and more priority, the first device SN 110 transmits to the terminal device 111a a pre-authorization to speak ("Floor Pre Grant" in English). In a particular embodiment, the transmission of this pre-authorization can be delayed, as detailed below in relation with FIG. 5A.
[0020] In a step 304, when there is no such competing request and more priority, the first SN device 110 transmits to the terminal device 111b a notification message representative of the pre-authorization granted to the terminal device 111a. Said notification message may further include an identifier of the terminal device 111a so as to indicate to the terminal device 111b which terminal device the pre-authorization has been granted. The terminal device 111b then starts to listen to the shared communication channel. In a preferred embodiment, step 304 is carried out before step 303. This makes it possible to limit the risk that the terminal device 111b will require speaking while a request to speak has already been received. from the terminal device 111a. In a step 305, when there is no such competing request and more priority, the first SN device 110 propagates, in the direction of the second SN device 120, the request received in step 302 from the terminal device 111a. In a preferred embodiment, step 305 is performed as soon as possible after receiving the request to speak. This makes it possible to limit the risk of subsequent conflict by notifying the second SN device 120 as soon as possible. In a step 306, the second SN device 120 receives the request propagated by the first SN device in step 305. The second device SN 120 check then that there is no request for speaking which is competing and more priority. A higher priority speaking request determining mechanism is described below in the context of the algorithm of FIG. 7. The second device SN 120 extracts from the received request information representative of a time when the terminal device 111a has transmitted said request. In a preferred embodiment, said information corresponds to a period of time elapsed between an instant at which the speech release message was received by the terminal device 111a in the context of the last speech (a reference time common to the communication network in the case of the very first speech) and the time at which the terminal device 111a transmitted said request. In another embodiment, said information corresponds to the instant at which the terminal device 111a transmitted said request according to the shared time reference. In a step 307, when there is no such competing request and more priority, the second device SN 110 transmits to the terminal device 121a a notification message representative of the pre-authorization granted by the first device SN 110 to the terminal device 111a, the propagation of the request by the first SN device 110 to the second SN device 120 implicitly indicating that the first SN device 110 has granted such a pre-authorization to the terminal device 111a. The terminal device 121a then starts to listen to the shared communication channel.
[0021] In a step 308, when the first SN device 110 determines that there is no concurrent and more priority request in step 302, the first SN device 110 sets up a first buffer memory dedicated to data that is subsequently transmitted by the terminal device 111a in the context of speaking. The first buffer temporarily stores data that the terminal device 111a wishes to transmit via said shared communication channel as part of the speaking. The first device SN 110 also activates a first read delay of the first buffer memory, the reading of the first buffer memory being deactivated as long as the first read timer has not elapsed. The duration of the first delay is determined by the first device SN 110 according to the distribution laws fr (t) and f8 (d) so as to ensure that the probability Pfau is less than or equal to the predefined threshold P - target, such as as detailed later in connection with FIG. 5A.
[0022] In a step 309, when the second SN device 120 determines that there is no competing and higher priority request in step 306, the second SN 120 device sets up a second buffer dedicated to the data that is subsequently transmitted. by the terminal device 111a in the context of speaking. The second buffer temporarily stores data that the terminal device 111a wishes to transmit via said shared communication channel as part of the speaking. The second SN 120 device further activates a second read timeout of the second buffer, the reading of the second buffer being disabled until the second read timer has elapsed. The duration of the second timer is determined by the second device SN 120 according to the distribution laws Mt) and f8 (d) so as to ensure that the probability Pfau is less than or equal to the predefined threshold P - target, as detailed subsequently with reference to FIG. 5B. It should be noted that the distribution law Mt) can be the same for each of the devices SN, or be different. In practice, since only the accepted calls are taken into account in the parameters of the distribution law Mt), the instances of the distribution law fr (t) used by the devices SN are identical since these calls are shared by all SN devices. It should also be noted that the instances of the distribution law f8 (d) are coherent with each other. This means that, depending on the structure of the communication network, the instances of the distribution law f8 (d) used by the SN devices can be identical; we then face a homogeneous communication network where the latencies induced in each of the sub-networks are equivalent. The instances of the distribution law f8 (d) used by the devices SN may be different, but remain coherent with each other because the said instances are representative of the latencies induced in the communication network from the point of view of the device SN considered; we then face a heterogeneous communication network where the latencies induced in each of the subnetworks are substantially different and can evolve according to the operational context of this communication system. The instances of the distribution law f8 (d) used by the SN devices may be different, while remaining coherent with one another, when the users who take the floor are not evenly distributed within the communication network. Following the pre-authorization transmitted by the first device SN 110 in step 303, the terminal device 111a takes the floor and begins transmitting, in a step 310, data to be broadcast to all the terminal devices at the same time. listen, ie the terminal devices 111b and 121a. In a step 311, the first SN device 110 receives this data and stores said received data in the first buffer memory. In a step 312, the first SN 110 propagates said received data to the second interconnect 120, i.e., the first SN 110 does not use the first buffer to propagate said data to the second device SN 120. In a step 313, the second SN 120 receives the data propagated by the first SN 110 in step 312 and stores them in the second buffer. This is the case for all the data subsequently transmitted by the terminal device 111a in the context of speaking, ie storing the data by the first device SN 110 in the first buffer memory, parallel propagation of said data by the first device SN 110 to the second SN 120 device and storing said data by the second SN 120 device in the second buffer memory. These subsequent transmissions of data in the context of speaking are not illustrated in FIG. 3 for the sake of simplification. In a step 314, the first timer that has been activated by the first SN device 110 in step 308 expires. The first device SN 110 then transmits to the terminal device 111a, in a step 315, a confirmation message of permission to speak ("Floor Grant" in English). The first device SN 110 also transmits, in a step 316, the terminal device 111b a notification message ("Floor Taken" in English) representative of the confirmation of permission to speak to the terminal device 111a. The first SN device 110 can also transmit, in a step 317, to the second SN device 120 a notification message representative of the confirmation of permission to speak given to the terminal device 111a. The second SN device 120 is then able to indicate, by acknowledgment of said notification message, whether the second SN device 120 has detected a conflict situation requiring the removal of the terminal device 111a or whether the second device SN 120 also confirms the authorization of speech given to the terminal device 111a. The first device SN 110 then activates the reading of the first buffer memory so as to transmit, in a step 318, the data stored in the first buffer memory to each terminal device which is listening to the shared communication channel and which is present in the first sub-network 101, ie the terminal device 111b. In a step 319, the second timer that has been enabled by the second SN 120 in step 309 expires. The second device SN 120 then transmits, in a step 320, the terminal device 121a a notification message representative of the confirmation of permission to speak given to the terminal device 111a. The second device SN 120 then activates the reading of the second buffer memory so as to transmit, in a step 321, the data stored in the second buffer memory to each terminal device which is listening to the shared communication channel and which is present in the second subnet 102, ie the terminal device 121a. It is considered in the context of FIG. 3 that there is no request for speaking which is concurrent and more priority. When the first SN device 110 or the second SN device 120 detects a competing and higher priority call request, said SN device refuses the speech by the device 111a. This aspect is detailed later in connection with Figs. 5A, 5B and 7. FIG. 4 schematically illustrates second exchanges in the communication system to allow a terminal device voluntarily release speech in the context of PTT type mechanism. In the context of FIG. 4, it is considered that the first sub-network 101 comprises the terminal devices 111a and 111b, and the second sub-network 102 comprises the terminal device 121a. It is further considered that the terminal device 111a has spoken in the context of the PTT type mechanism. In a step 401, the terminal device 111a transmits to the first device SN 110 a speech release message ("Floor Release" in English). In a step 402, the first SN device 110 receives the message transmitted in step 401 by the terminal device 111a. The first SN device 110 extracts from the received message information representative of a time at which the terminal device 111a has transmitted said message.
[0023] In a step 403, the first SN device 110 propagates, in the direction of the second SN device 120, the message received at the step 402 from the terminal device 111a. In a step 404, the first SN 110 disables writing to the first buffer and continues to transmit the data contained in the first buffer until the first buffer is empty. Since the terminal device 111a is not expected to continue transmitting data as part of speaking after transmitting the speech release message, the first SN device 110 may also refrain from disabling the write in the first buffer. Once the first buffer memory is empty, the first SN device 110 releases the first buffer memory and transmits, in a step 405, the terminal device 111b a notification message representative of the release of speech by the terminal device 111a.
[0024] In a step 406, the second SN device 120 receives the message propagated by the first SN device 110 in step 403. The second SN device 120 extracts from the received message information representative of a time at which the terminal device 111a has transmitted said message. In a step 407, the second SN 120 disables writing to the second buffer and continues to transmit the data contained in the second buffer until the second buffer is empty. Since the terminal device 111a is not expected to continue transmitting data as part of speaking after transmitting the speech release message, the second SN device 110 may also refrain from disabling write in the second buffer. Once the second buffer memory is empty, the second SN device 120 releases the second buffer memory and transmits to the terminal device 121b a notification message representative of the speech release by the terminal device 111a. Fig. 5A schematically illustrates an algorithm, implemented by the first SN device 110, to allow a terminal device present in the first sub-network 101 to speak in the context of the PTT type mechanism. The same principles are applicable by the second device SN 120, to allow a terminal device present in the second subnet 102 to speak in the context of the PTT type mechanism.
[0025] In a step 501, the first SN device 110 receives a request to speak from a terminal device present in the first subnet 101. In a next step 502, the first device SN 110 determines the priority of the request. received in step 501 vis-à-vis any other request for speaking which the first device SN 110 would know. The first device SN 110 extracts from the received request information representative of a time when said terminal device has transmitted said request. In a preferred embodiment, said information corresponds to a period of time elapsed between an instant at which the speech release message was received by said terminal device in the context of the last speech (a reference time common to the network in the case of the very first speech) and the time at which said terminal device transmitted said request. In another embodiment, said information corresponds to the instant at which said terminal device transmitted said request according to the shared time reference.
[0026] In a next step 503, the first SN device 110 checks whether the request received at step 501 can be taken into account or if another request that would be competing and more priority should be processed. If no other request, concurrent and higher priority, exists, a step 505 is performed; otherwise, a step 504 is performed.
[0027] In step 504, the first SN device 110 transmits, to the terminating device having transmitted the request received in step 501, a refusal notification message with respect to the request received in step 501. The device terminal having transmitted the request received in step 501 can not then take the floor. In step 505, the first SN device 110 propagates the request received in step 501 to the second SN device 120. In a next step 506, the first SN device 110 sets up a buffer memory dedicated to data that is subsequently transmitted, as part of the speaking, by the terminal device that transmitted the request received in step 501. Any data subsequently transmitted by said terminal device in the context of speaking is then stored in said buffer memory. The first device SN 110 also activates a reading delay of said buffer memory, the reading of said buffer being deactivated as long as the read timer has not elapsed. The first interconnection device 110 determines the duration of the reading delay of said buffer memory according to the distribution laws Mt) and f8 (d) so as to ensure that the probability Pf garl is less than or equal to the predefined threshold P In a preferred embodiment, the distribution laws Mt) and f8 (d) are predetermined by predefined models.
[0028] When the distribution law Mt) is defined by a model, the distribution law fr (t) is preferentially of log-normal type and expressed in the following way: (log (0-, u) 2 e 202 t6 2rc where a represents the standard deviation of the distribution law and u represents the mean of the distribution law 10 a and u can be determined using an estimation algorithm When the distribution law f8 (d) is defined by a model, the distribution law f8 (cl) is preferentially expressed in the form of a mixture of laws of the highest likelihood ("maximum likelihood estimation"). Weibull distribution with three parameters, as follows: f8 (d) => qi f8i (d) where i is an index on the components of the mixture and where qi represents weighting coefficients such that: Ri = 1 and where: if (cl - 1. e ri) for cl> li 8 i (d) = ri ri Each component of the mixture has parameters si, ri and I, different, if being a shape parameter, a scale parameter in English, and a location parameter ("location"). " in English). These parameters can be determined by an expectation maximization algorithm. Preferably, the first device SN 110 pre-fills a look-up table (LUT) which provides a delay time as a function of the input torque (t, cl). This approach is particularly advantageous when the first device SN 110 does not have sufficient fr (t) = 1 computational resources so that the calculation time of the delay time is negligible in view of the delay time itself, or even when the calculation time of the delay time is greater than the delay time itself. The content of this correspondence table LUT is then determined in the following manner for each SN device: 1. Determination of parameters making it possible to represent the distribution law Mt), that is to say the standard deviation a and the mean kt by means of a maximum likelihood estimation algorithm in the case where the distribution law Mt) is of log-normal type; 2. Determination of parameters to represent the distribution law f8 (d) applicable, that is to say the parameters si, ri and li, and the weighting coefficients qi when the distribution law f8 (d) is a mixture of three-parameter Weibull distribution laws; 3. Determination of the quantity of terminal devices present in the communication system; 4. Determination of Ppush, when Pfaii is defined according to Ppush 5. Definition of minimum and maximum values of time period t between a release of speech and a next speech, and network latency c between a time when a request speaking is transmitted by a terminal device and a time at which said request is actually received by said SN device; 6. Definition of a step ("step" in English) between two successive values of time period t to be considered and of a step between two latent network latent values ci to be considered, so as to define each pair (t, cl) to be represented in the LUT correspondence table; 7. Definition of a maximum acceptable value for the timer Tb, for example as a function of the memory capacities of said SN device; 8. Definition of the P-target threshold 9. Definition of an acceptable margin of error E between Pfail and Ptarg and, 10. Determination by bisection, according to the distribution laws Mt) and f8 (d), of a value timing Tb for each pair (t, cl) to be represented in the correspondence table LUT, such that Ptarget Pfail E and that the timing value Tb is less than the maximum acceptable value for the timer Tb; it is also possible to consider that the margin of error E applies around the value P - target and in this case is determined by a timeout value Tb for each pair (t, d) to be represented in the table. Corresponding LUT, such that I P - target Pf E and that the timing value Tb is less than the maximum acceptable value for the timer Tb.
[0029] Another correspondence table LUT can also be constructed to be able to determine the delay time possibly used for sending the pre-authorization message to speak. The same filling approach is used for this other correspondence table LUT, except that another predefined threshold P - target, such that P - target is strictly greater than P - target, is used instead of Ptar g and - Thus, from the information representative of the instant at which the terminal device lila transmitted said request, the first device SN 110 is able to determine the latency c / suffered by said request to arrive at the first device SN 110 (latency c / undergone by said request to arrive at the second SN device 120 in the context of Fig. 5B), as well as the time period t between the last speech release and the time at which said request was transmitted. From this pair (t, d) determined for the speech request received in step 501, the first SN device 110 finds the corresponding timer value Tb contained in the correspondence table LUT for this pair (t, c /).
[0030] Since a correspondence table LUT may not be representative of all the pairs (t, d) because of the quantization related to the definition of the steps, the first device SN 110 may interpolate between several pairs (t, d) approaching the representative torque (t, d) of the received speech request. In an alternative embodiment, the distribution laws fr (t) and f8 (d) are predetermined by each of the SN devices as a function of long-term statistics previously collected over the course of message exchanges in the context of the type mechanism. PTT. A correspondence table LUT which provides a delay time as a function of the input torque (t, d) can also be previously constructed on the basis of these statistics.
[0031] In a particular embodiment, instances of distribution laws Mt) and f8 (d) are predetermined for each configuration of a set of configurations of the communication network, each of said configurations responding to a distinct operational context (eg a first configuration when the communication link 100 is a terrestrial access network and a second configuration when the communication link is a satellite link). For example, each device SN has instances of first distribution laws Mt) and f8 (d) (for example implemented via first respective LUT correspondence tables) representative of a homogeneous communication network using a communication link. 100 of terrestrial access network type, and each SN device has instances of second distribution laws fr (t) and f8 (d) (for example implemented via second respective LUT correspondence tables) representative of a heterogeneous communication network using a communication link 100 of the satellite type. In the event of a change in the configuration of the communication network, the SN devices respectively select the distribution law instances Mt) and f8 (d) applicable to the actual operating context, e.g. the LUT correspondence tables corresponding to the new network configuration. In other words, different configurations of the communication network being possible, each SN device selects the distribution law MT) and the distribution law f8 (cl) to be applied among a predefined set of distribution laws, depending on a effective configuration of the communication network. Such a configuration change of the communication network can be detected by each SN device by receiving a message from a PTT-type manager application at the scale of the communication network and having knowledge of network topology changes. Communication. When each SN device is integrated or connected to an interconnection device (interconnecting the subnet to which said SN device is associated and each alternative implementation of the communication link 100), the interconnection device can prevent said SN device a change of interface (eg satellite link, land access network) implementation of the communication link 100. In a next step 507, the first device SN 110 also activates a pre-authorization timeout speaking. The first interconnection device 110 determines the duration of the pre-authorization speaking delay according to the distribution laws fr (t) and f8 (d) so as to ensure that the probability Pfau is lower than or equal to the said another predefined threshold P target, which implies that the duration of the pre-authorization time of speech is strictly less than the duration of the read time of said buffer. The duration of the pre-authorization timeout is thus determined in the same manner as the duration of the read delay of said buffer, but using a different predefined threshold. In a next step 508, the first SN device 110 determines whether another request that would be concurrent and higher priority is received before the speaking pre-authorization timer expires. If such another request is received, step 504 is performed; otherwise, when the pre-authorization timer has expired, a step 509 is performed. In step 509, the first SN device 110 transmits to each terminal device of the first subnet 101 listening, other than the terminal device having transmitted the request received in step 501, a notification message representative of a pre-authorization of speech granted to the terminal device having transmitted the request received in step 501. In a following step 510, the first SN device 110 transmits, to the terminal device having transmitted the request received in step 501, a pre-authorization message to speak. In a particular embodiment, as already described with reference to FIG. 3, the transmission of the pre-authorization message is not delayed. Delaying the transmission of the pre-authorization message for speaking, however, makes it possible to limit the risks of mistakenly accepting a request to speak, especially with respect to competing requests originating from terminal devices present on the first sub-authorization. network 101. In a next step 511, the first SN device 110 activates the write buffer to allow the storage of data transmitted by said terminal device in the context of speaking. The first device SN 110 also activates a mechanism for propagating the data transmitted by said terminal device in the context of speaking. The data propagation mechanism is such that the data received by the first SN device 110 from said terminal device in the context of speaking are propagated by the first SN device 110 to the second SN device 120, without having to go through said buffer memory.
[0032] In a next step 512, the first SN device 110 determines whether another competing and higher priority request is received before the read timer of said buffer expires. If such another request is received, step 504 is performed, wherein the first SN device 110 releases said buffer memory and cancels the read timer of said buffer memory; otherwise, when the read timing of said buffer has expired, a step 513 is performed. In step 513, the first SN device 110 transmits, to the terminal device that has transmitted the request received in step 501, a confirmation of authorization to speak message. The first device SN 110 transmits to each terminal device of the first subnet 101 listening, ie other than the terminal device having transmitted the request received in step 501, a notification message representative of the authorization confirmation of speaking. In a next step 514, the first SN device 110 activates said read buffer. In other words, the first SN device 110 starts transmitting to each terminal device of the first subnet 101 listening, ie other than the terminal device having transmitted the request received in step 501, the data contained in said buffer memory. The data contained in said buffer memory is extracted from said buffer memory and transmitted in the first subnet 101 substantially at the rate at which the first SN device 110 received said data from the terminal device having transmitted the request received at step 501 Fig. 5B schematically illustrates an algorithm, implemented by the second SN device 120, to allow a terminal device present in the first sub-network 101 to speak in the context of the PTT type mechanism. The same principles are applicable by the first device SN 110, to allow a terminal device present in the second subnet 102 to speak in the context of the PTT type mechanism. In a step 551, the second device SN 120 receives a request to speak from the first device SN 110. Said request for speaking is initially transmitted by a terminal device present in the first sub-network 101 and was propagated by the first SN device 110. In a next step 552, the second SN 120 device determines the priority of the request received in step 551 vis-à-vis any other request for speech including the second device SN 120 would know. The second device SN 120 extracts from the received request information representative of a time when said terminal device has transmitted said request. In a preferred embodiment, said information corresponds to a period of time elapsed between an instant at which the speech release message was received by said terminal device in the context of the last speech (a reference time common to the network in the case of the very first speech) and the time at which said terminal device transmitted said request. In another embodiment, said information corresponds to the instant at which said terminal device transmitted said request according to the shared time reference. In a next step 553, the second SN device 120 checks whether the request received at step 551 can be taken into account or if another request that would be concurrent and higher priority must be processed. If no other request, concurrent and higher priority, exists, a step 555 is performed; otherwise, a step 554 is performed. In step 554, the second SN device 120 transmits to the first SN device 110 a refusal notification message with respect to the request received in step 551. The terminal device at the origin of the request received at step 551 can not then speak.
[0033] In step 555, the second SN device 120 transmits, to each terminal device of the second subnet 102 listening, a notification message representative of a pre-authorization of speech granted to the terminal device to the origin of the request received in step 551. In a subsequent step 556, the second device SN 120 sets up a buffer memory dedicated to data that is subsequently transmitted, as part of the speaking, by the terminal device at the origin of the request received in step 551. Any data subsequently transmitted by said terminal device in the context of speaking is then stored in said buffer memory. The second device SN 120 also activates a reading delay of said buffer memory, the reading of said buffer being deactivated as long as the read timer has not elapsed. The second device SN 120 determines the duration of the read delay of said buffer memory according to the distribution laws fr (t) and f8 (d) so as to ensure that the probability Pfau is less than or equal to the predefined threshold P - target - The second device SN 120 determines the duration of the read delay of said buffer in the same manner as the first device SN 110 does in the context of the algorithm of FIG. 5A. This makes it possible to ensure that the readings of the buffers respectively implemented by the first 110 and second 120 SN devices are substantially triggered at the same time, since the distribution law Mt) is in practice identical for all the devices. SN and that the instance of the distribution law f8 (d) used by each of the devices SN is consistent with the instances of the distribution law f8 (d) respectively used by the other devices SN.
[0034] In a next step 557, the second device SN 120 activates the write buffer to enable storage of data propagated by the first device SN 110 as part of the speaking by the terminal device at the origin of the request received. In step 551, in a next step 558, the second SN 120 determines whether another competing, higher priority request is received before the read timer of said buffer expires. If such another request is received, step 554 is performed, wherein the second SN 120 releases said buffer and cancels the read timer of said buffer; otherwise, when the read timing of said buffer has expired, a step 559 is performed. In step 559, the second SN device 110 transmits, to each terminal device of the second subnet 102 listening, a notification message representative of a confirmation of authorization to speak by the terminal device to the terminal. origin of the request received at step 551.
[0035] In a next step 560, the second SN device 120 activates said read buffer. In other words, the second SN device 120 starts transmitting, to each terminal device of the second subnet 102 listening, the data contained in said buffer memory. The data in said buffer is extracted from said buffer and transmitted in the second subnet 102 substantially at the rate at which the second SN 120 received said data from the first SN 110. FIG. 6A schematically illustrates an algorithm, implemented by the first SN device 110, to allow a terminal device present on the first sub-network 101 to voluntarily release the speech in the context of the PTT type mechanism. The same principles are applicable by the second device SN 120, to allow a terminal device present in the second sub-network 102 to voluntarily release the speech in the context of the PTT type mechanism. In a step 601, the first SN device 110 receives a speech release message from a terminal device present in the first subnet 101 and to which the speech has been previously granted. The first SN device 110 extracts from the received message information representative of a time at which the terminal device 111a has transmitted said message. In a subsequent step 602, the first SN 110 propagates the message received in step 601 to the second SN 120. In a subsequent step 603, the first SN 110 disables write to the buffer whose reading has been read. activated in step 514. Since the terminal device having transmitted the message received in step 601 is not supposed to continue transmitting data as part of the speaking after transmission of the speech release message, the first device SN 110 may also refrain from deactivating the writing in said buffer. In a next step 604, the first SN device 110 checks whether the buffer whose reading has been activated in step 514 is empty. The reading of the buffer is maintained until said buffer memory is empty, so as to transmit the data stored in said buffer memory to each terminal device of the first subnet 101 listening, ie with the exception the terminal device having transmitted the message received in step 601. When said buffer memory is empty, a step 605 is performed; otherwise, step 604 is repeated. In step 605, the first SN device 110 releases said buffer.
[0036] In a next step 606, the first SN device 110 transmits, to each terminal device of the first subnet 101 listening, ie with the exception of the terminal device having transmitted the message received in step 601, a message of representative notification of the speech release by the terminal device having transmitted the message received in step 601.
[0037] Fig. 6B schematically illustrates an algorithm implemented by the second SN 120 to allow a terminal device present on the first sub-network 101 to voluntarily release the speech in the context of the PTT type mechanism. The same principles are applicable by the first device SN 110, to allow a terminal device present in the second sub-network 102 to voluntarily release speech in the context of the PTT type mechanism. In a step 651, the second SN device 120 receives a speech release message from the first SN device 110. Said speech release message is originally transmitted by a terminal device present in the first subnet 101 and wherein the speaking has been previously granted, and has been propagated by the first SN device 110. The second SN 120 device extracts from the received message information representative of a time at which the terminal device 111a has transmitted said message. In a next step 652, the second SN 120 disables writing to the buffer whose reading has been activated in step 560. Since the terminal device at the origin of the message received in step 651n It is not expected to continue transmitting data as part of the speaking after transmitting the speech release message, the second SN device 120 may also refrain from disabling writing in said buffer.
[0038] In a next step 653, the second SN device 120 checks whether said buffer memory is empty. The reading of the buffer is maintained until said buffer is empty, so as to transmit the data stored in said buffer to each terminal device of the second subnet 102 listening. When said buffer is empty, a step 654 is performed; otherwise, step 653 is repeated. In step 654, the second SN 120 device releases said buffer. In a next step 655, the second SN device 120 transmits to each terminal device of the second subnet 102 listening a notification message representative of the release of speech by the terminal device at the origin of the message of release of received at step 651. FIG. 7 schematically illustrates an algorithm, implemented by the first SN device 110 and the second SN 120 device for managing a speech conflict in the context of the PTT type mechanism. Consider that the case where the algorithm of FIG. 7 is implemented by the first device SN 110. The same principles apply in the case where the algorithm of FIG. 7 is implemented by the second device SN 120. In a step 701, the first device SN 110 detects a conflict of requests to speak. Such a conflict is detected when a request for speech, whose origin is a first terminal device, is received by the first interconnection device 110 while speaking has already been granted to a second terminal device. Such a conflict is also detected when the first interconnection device 110 has to process several speech requests, the respective origins of which are separate terminal devices. Such a conflict is also detected when the first SN device 110 is informed by the second SN device 120 that such a conflict exists. When the first interconnection device 110 has to process several speech requests whose respective origins are distinct terminal devices, the first SN device 110 determines which speech request has the highest priority. In a particular embodiment, each request for speaking is associated with an explicit or tacit priority level. Said priority level is tacit when, without mentioning said priority level in said request to speak, the first interconnection device 110 is able to determine the priority level of said request. For example, the terminal devices may be associated with a predefined priority level, eg as part of a deployment of the communication system in emergency and emergency situations, terminal devices held by first responders in charge of coordinating field actions may be associated with higher priority levels than first responders in charge of performing field actions. In a preferred embodiment, the first device SN 110 determines which request to speak is temporally the highest priority, so as to limit the risk of not respecting the chronology of attempt to speak (ie the risk that a user X having chronologically tried to speak before a user Y is heard by at least one other user Z after the user Y). Each request for speaking preferably includes information of time period elapsed between an instant at which the speech release message has been received by the terminal device concerned as part of the last speech (a reference time common to the network in the case of the very first speech) and the time when said terminal device concerned has transmitted said request to speak. In this case, said information is representative of the reactivity of the user to want to access the shared communication channel with respect to the last speech. It is thus possible to consider temporally a higher priority terminal device seeking to speak in a shorter period of time after the last release of speech, despite the fact that the terminal devices are respectively notified of the release of speech to different moments. Alternatively, each request for speaking includes information representative of the time when the terminal device concerned has transmitted the request to speak according to the shared time reference. It is thus possible to consider temporally more priority a terminal device seeking to speak as soon as possible in view of the shared time reference.
[0039] In a particular embodiment, the first device SN 110 seeks to determine the priority level, explicit or tacit, associated with each request to speak. The first device SN 110 then considers the request associated with the highest priority level as being the highest priority request. And, when the first device SN 110 is faced with two requests for speaking which are explicitly or tacitly associated with the same level of priority, the first device SN 110 seeks to determine which request is temporally more priority. In a step 702, the first SN device 110 checks whether the detected conflict requires removing from a terminal device the speech that had been previously pre-tuned to said terminal device or whose speech had been confirmed to said terminal device. This is particularly the case when the first device SN 110 is informed by the second device SN 120 that a conflict exists. If the detected conflict necessitates removing a speaking device from a terminal device, a step 704 is performed; otherwise, a step 703 is performed.
[0040] In step 703, the first device SN 110 refuses any request to speak which, in the context of the conflict, is not the highest priority. When said request for speaking comes from a terminal device of the first subnet 101, the first device SN 110 transmits to said terminal device a notification message representative of a refusal to speak. When said request for speech comes from a terminal device of the second sub-network 102, that is to say that said request for speech has been propagated by the second device SN 120, the first device SN 110 transmits the second interconnection device 120 a notification message representative of a refusal to speak. In step 704, the first device SN 110 deactivates the buffer memory that has been previously set up, in step 511 or step 557, for the speech of said terminal device. When said terminal device is present in the first sub-network 101, the first device SN 110 also deactivates the propagation mechanism, in the direction of the second device SN 120, of the data transmitted by said terminal device as part of said speaking. .
[0041] In a next step 705, the first SN device 110 sends to each terminal device that is present on the first subnet 101 a voice-out notification message that has been pre-tuned or confirmed to said terminal device. When said terminal device is present in the first sub-network 101, said terminal device receives said notification message and stops the transmission of data accordingly. Thus, when said terminal device is present on the first sub-network 101, said terminal device is notified that speaking is denied to said terminal device. In a subsequent step 706, when the first SN device 110 has detected the conflict itself, i.e. the first SN device 110 has not been informed of the conflict by the second SN device 120, the first device SN 110 informs the second device SN 120 of the detected conflict.

权利要求:
Claims (2)
[0001]
CLAIMS1 / Speech management method on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks (101, 102), each subnetwork being associated a super-node device (110, 120) of the communication network, the method being implemented by a said super-node device, characterized in that, when a first request for speech is received (302; 501 ) of a terminal device of the subnet to which said super-node device is associated and that said super-node device is not aware of at least one second priority request for speech, said super-node device performs the following steps: propagate (305; 505) the first request to speak to each other super-node device; transmitting (303; 510) a pre-authorization message for speaking to said terminal device so as to allow said terminal device to speak; transmitting (304; 509) a pre-authorization notification message to each other terminal device of the subnet to which the super-node device is associated, so as to place each other terminal device listening the shared communication channel; determining (308; 506) a duration of a delay as a function of a distribution law fr (t) of the period of time between a speech release and a next talk, and a distribution law f8 ( d) network latency between a time at which a request to speak is transmitted by a terminal device and a time when said request is actually received by said super-node device, so that the probability Pfau to remove said terminal device from speaking is below a predefined first threshold; storing (311; 511) in a buffer the data received from said terminal device as part of the speaking, and propagating (312; 511) said data to each other super-node device; - activate (314; 514) the read buffer memory to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer is elapsed; and in that, when the first request for speech is received (306; 551) from a terminal device of another subnet than that to which said super-node device is associated and that said super-node device is not aware of less a second priority request for speaking, said super-node device performs the following steps: - transmitting (307; 555) a pre-authorization notification message to each terminal device of the subnet to which the super-node device is associated, so as to place each terminal device of said sub-network listening to the shared communication channel; determining (309; 556) the duration of the delay as a function of the distribution law fr (t) and of the distribution law f8 (d), so that the probability Pfau of removing said terminal device from the speech below the first predefined threshold; storing (313; 557) in the data buffer propagated by another super-node device as part of the speaking session; - Activate (319; 560) read the buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer is elapsed; and in that, when said super-node device detects (701) at least a second request for more priority speaking before expiration of said timer, said super-node device releases (704) the buffer memory and, when said first request has been received from a terminal device of the subnet to which said super-node device is associated, said super-node device notifies (705) said terminal device that the speech is denied to said terminal device.
[0002]
2 / A method according to claim 1, characterized in that said super-node device transmits the pre-authorization message to speak after expiry of another time delay (507) according to the distribution law Mt) and the distribution law f8 (d), so that the probability of removing said terminal device from the speech is below a second predefined threshold strictly greater than said first threshold.3 / Method according to any one of Claims 1 and 2, characterized in that the distribution law Mt) is defined by a model of log-normal type and expressed as follows: (log (0-, u) 2 t VU e 202 where a represents l standard deviation of the distribution law Mt) and u represents the mean of the distribution law fr (t). 4 / A method according to any one of claims 1 to 3, characterized in that the distribution law f8 (d) is defined by a model in the form of a mixture of Weibull distribution laws with three parameters and expressed as the following way: f8 (d) = f8i (d) where i is an index on the components of the mixture and where qi represents weighting coefficients such that: Ri = 1 and where: fs, (d) = si (d - ii) if-i for cl> li ri) ri ri where si is a form parameter, ri a scale parameter and 1i a location parameter. 5 / A method according to any one of claims 1 and 2, characterized in that the distribution laws Mt) and f8 (d) are predetermined by said super-node device based on long-term statistics previously collected over time. exchange of messages in the context of alternate communications. 6 / A method according to any one of claims 1 to 5, characterized in that the probability Pfau of removing said terminal device from speaking is expressed as follows: N-1 (N - 1) N-1 -k Pfcul = 1 k Ppushk (1 - Ppush) (1 - fau) k fr (t) = 1 k = where Ppush represents the probability for a terminal device to require the talk determined by said super-node device in count the number of concurrent speaking requests detected for each granted speech, namely: R-1 Ppush =, N / Q 1, q = 1 where Rq represents a quantity of concurrent requests received for each decision taken. given word q with Rq> 1 and where Q represents a quantity of allocated speech tapers considered; and where: Pifcul = Prob {Ti + d1 <T2 + d2> T2} Prob {Ti + d1 <T2 + d2 <T2} where: Prob {x} represents the probability that an expression represented by x is verified and where T1 is representative of a moment at which said first request for speech has been transmitted, T2 is representative of the moment at which a second priority request has been transmitted, di is representative of a transmission latency experienced by said first request to speak in the communication network, d2 is representative of a transmission latency experienced by said second request to speak in the communication network, and Tb is representative of the duration of the delay of which expiration triggers the reading of said buffer. 7 / A method according to any one of claims 1 to 6, characterized in that each request for speech includes information representative of a period of time elapsed between an instant at which a speech release message was received in the frame of the last speech by the terminal device having transmitted said request for speaking and the time at which said terminal device has transmitted said request. 8 / A method according to claim 7, characterized in that said super-node device considers that a said second request for speaking is temporally more priority than said first request to speak when the time period represented by said information included in said second take request Prob {T1 + di + Tb <T2 + d2 & T1> T2} of speech is less than said period of time represented by said information included in said first speech request. 9 / A method according to any one of claims 1 to 8, characterized in that, to determine the duration of said delay, the super-node device performs the following steps: - determining a period of time t elapsed between a time when a speech release message was received in the context of the last speech by the terminal device having transmitted said first request for speaking and the time at which said terminal device transmitted said first request for speaking; - Determining a latency c / suffered by said first request to speak between the transmission of said first request for speaking by said terminal device and the receipt of said first request for speaking by said super-node device; and - determining the duration of said timer from a look-up table which provides a delay time as a function of a set of predefined pairs (t, d), the contents of the correspondence table being previously defined as a function of the law of distribution Mt) and of the distribution law f8 (d) concerned so that the probability Pfau of removing a terminal device of the speech is below the first predefined threshold. 10 / A method according to claim 9, characterized in that, to fill the correspondence table, a device performs the following steps: - determining parameters for representing the distribution law (t); - determining parameters to represent the distribution law f8 (d) applicable; determining a quantity of terminal devices present in the communication system; define minimum and maximum values of time period t between a speech release and a next speech and network latency c / between a time at which a request to speak is transmitted by a terminal device and a time at which said request is actually received by said super-node device; - defining a step between two successive time period values t to be considered and a step between two successive network latency values ci to be considered, so as to define each pair ( t, d) to be represented in the correspondence table; - define a maximum acceptable value for the duration of the delay; - define the value of said first threshold; - define an acceptable margin of error E; determining by bisection, according to the distribution laws Mt) and f8 (d), a time delay for each pair (t, cl) to be represented in the correspondence table, so that the probability Pfau of removing a terminal device from the speaking stand at a distance from the first predefined threshold lower than said acceptable error margin E and that the delay time is less than said maximum acceptable value. 11 / A method according to any one of claims 1 to 10, characterized in that, when a speech release message is received (402; 601) of said terminal device and said terminal device is on the subnet to which said super-node device is associated, said super-node device performs the following steps: - propagate (602) said received speech release message to each other super-node device; transmitting (606) a speech release notification message to each other terminal device of the subnet to which the super-node device is associated, after the buffer has been emptied; releasing (605) the buffer memory; and when a speech release message is received (406; 651) from said terminal device and said terminal device is on another subnet than that to which said super-node device is associated, said super-node device performs the steps the following: transmitting (655) a speech release notification message to each terminal device of the subnet to which the super-node device is associated, after the buffer has been emptied; and - releasing (654) the buffer memory.12 / Method according to any one of claims 1 to 11, characterized in that, different configurations of the communication network being possible, said super-node device selects the distribution law Mt) and the distribution law f8 (d) among a predefined set of distribution laws, as a function of an effective configuration of the communication network. 13 / Speech management method on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks (101; 102), each subnetwork being associated to a super-node device (110; 120) of the communication network, the method being implemented by a said super-node device (110), characterized in that, when a first request to speak is received ( 302; 501) of a terminal device of the subnet to which said super-node device is associated and that said super-node device is not aware of at least one second priority request for speech, said super device -node performs the following steps: - propagate (305; 505) the first talk request to each other super-node device; transmitting (303; 510) a pre-authorization message for speaking to said terminal device so as to allow said terminal device to speak; transmitting (304; 509) a pre-authorization notification message to each other terminal device of the subnet to which the super-node device is associated, so as to place each other terminal device listening the shared communication channel; determining (308; 506) a duration of a delay as a function of a distribution law fr (t) of the period of time between a speech release and a next talk, and a distribution law f8 ( d) network latency between an instant at which a request to speak is transmitted by a terminal device and an instant at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from the speech is below a predefined threshold; storing (311; 511) in a buffer the data received from said terminal device in the context of speaking, and propagating (312; 511) said data to each other super-node device; and- activate (314; 514) the read buffer memory to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer is elapsed; and in that, when said super-node device detects (701) at least a second request for priority speaking before expiration of said timer, said super-node device performs the following steps: - releasing (704) the buffer memory ; and - notifying (705) said terminal device that speaking is denied to said terminal device. 14 / Speech management method on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of interconnected subnetworks (101; 102), each subnetwork being associated a super-node device (110; 120) of the communication network, the method being implemented by a said super-node device (120), characterized in that, when a first request to speak is received ( 306; 551) of a terminal device of another subnet than that to which said super-node device is associated and that said super-node device is not aware of at least a second request for more speech As a priority, said super-node device performs the following steps: transmitting (307; 555) a pre-authorization notification message to each terminal device of the subnet to which the super-node device is attached ocié, so as to place each terminal device of said subnetwork listening to the shared communication channel; determining (309; 556) the duration of the delay as a function of a distribution law fr (t) of the period of time between a speech release and a next talk, and a distribution law f8 (d ) network latency between an instant at which a request to speak is transmitted by a terminal device and an instant at which said request is actually received by said super-node device, so that the probability Pfau to remove said terminal device from the socket speech is below a predefined threshold; storing (313; 557) in a buffer the data propagated by another super-node device as part of the speaking session; - activating (319; 560) reading the buffer memory to broadcast the stored data to each device; terminal listening on the subnet associated with said super-node device, when said timer is elapsed; and in that, when said super-node device detects (701) at least a second request for more priority speaking before expiration of said timer, said super-node device releases (704) the buffer memory. A super-node device (110; 120) for speech management on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of subnetworks (101; 102) interconnected, each subnetwork being associated with a super-node device of the communication network, characterized in that, when a first request for speech is received (302; 501) from a terminal device of the sub-network, network to which said super-node device is associated and that said super-node device is not aware of at least a second request for priority speaking, said super-node device includes: - means for propagating (305; 505) the first request to speak to each other super-node device; means for transmitting (303; 510) a pre-authorization message for speaking to said terminal device so as to allow said terminal device to speak; means for transmitting (304; 509) a pre-authorization notification message to each other terminal device of the subnet to which the super-node device is associated, so as to place each said other terminal device at listening to the shared communication channel; means for determining (308; 506) a duration of a delay as a function of a distribution law Mt) of a period of time between a speech release and a next speech, and a distribution law f8 (d) network latency between an instant at which a request to speak is transmitted by a terminal device and a time at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from speaking is below a predefined first threshold; means for storing (311; 511) in a buffer memory data received from said terminal device in the context of speaking, and for propagating (312; 511) said data to each other super-node device; means for activating (314; 514) reading the buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer has elapsed; and in that, when the first speaking request is received (306; 551) from a terminal device of another subnet than that to which said super-node device is associated and said super-node device not aware of at least one second priority request for speech, said super-node device includes: means for transmitting (307; 555) a pre-authorization notification message to each device terminal of the subnet to which the super-node device is associated, so as to place each terminal device of said subnet listening to the shared communication channel; means for determining (309; 556) the duration of the delay as a function of the distribution law fr (t) and of the distribution law f8 (d), so that the probability Pfau of removing said terminal device from the speech is below the first predefined threshold; means for storing (313; 557) in the buffer memory data propagated by another super-node device as part of the speaking session; means for activating (319; 560) the read buffer memory to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer has elapsed; and in that, when said super-node device detects (701) at least one second request for priority speech before expiration of said timer, said super-node device includes means for releasing (704) the buffer memory and, when said first request has been received from a terminal device of the subnet to which said super-node device is associated, said super-node device includes means for notifying (705) said terminal device that speaking is refused terminal device. 16 / Super-node device (110) for managing speech on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of subnetworks (101; 102) interconnected, each subnetwork being associated with a super-node device of the communication network, characterized in that, when a first request to speak is received (302; 501) from a terminal device of the sub-network to which said super-node device is associated and that said super-node device is not aware of at least one second request for more priority speech, said super-node device includes: - means for propagating (305; 505) the first request to speak to each other super-node device; means for transmitting (303; 510) a pre-authorization message for speaking to said terminal device so as to allow said terminal device to speak; means for transmitting (304; 509) a pre-authorization notification message to each other terminal device of the subnet to which the super-node device is associated, so as to place each said other terminal device at listening to the shared communication channel; means for determining (308; 506) a duration of a delay as a function of a distribution law Mt) of a period of time between a speech release and a next speech, and a distribution law f8 (d) network latency between an instant at which a request to speak is transmitted by a terminal device and a time at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from speaking is below a predefined threshold; means for storing (311; 511) in a buffer memory data received from said terminal device in the context of speaking, and propagating (312; 511) said data to each other super-node device; and - means for activating (314; 514) reading the buffer to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer has elapsed; and in that, when said super-node device detects (701) at least a second request for more priority speaking before expiration of said timer, said super-node device includes: means for releasing (704) the buffer memory ; and means for notifying (705) said terminal device that speaking is denied to said terminal device. 17 / Super-node device (120) for speech management on a shared communication channel in the context of alternate communications implemented in a communication network composed of a plurality of subnetworks (101; 102) interconnected, each subnetwork being associated with a super-node device (110; 120) of the communication network, characterized in that, when a first speech request is received (306; 551) from an end device of another subnet than that to which said super-node device is associated and that said super-node device is not aware of at least one second priority request for speech, said super-node device includes: means for transmitting (307; 555) a pre-authorization notification message to each terminal device of the subnet to which the super-node device is associated, so as to place each terminal device l of said subnetwork listening to the shared communication channel; means for determining (309; 556) the duration of the delay as a function of a distribution law Mt) of the period of time between a speech release and a next speech, and a distribution law f8 ( d) network latency between an instant at which a request to speak is transmitted by a terminal device and an instant at which said request is actually received by said super-node device, so that the probability Pfau of removing said terminal device from the speech is below a predefined threshold; means for storing (313; 557) in a buffer memory data propagated by another super-node device as part of the speaking session; means for activating (319; 560) the read buffer memory to broadcast the stored data to each terminal device listening on the subnet associated with said super-node device, when said timer has elapsed; and in that, when said super-node device detects (701) at least one second request for priority speaking before expiration of said timer, said super-node device includes means for releasing (704) the buffer memory.

类似技术:

---

公开号 | 公开日 | 专利标题

---

EP3146734B1|2019-10-23|Method for managing speaking turns over a communication channel in the context of simplex calls

---

FR3045996A1|2017-06-23|

---

EP3409054A1|2018-12-05|Synchronisation method for a node in a cellular network

---

EP1199818A1|2002-04-24|Method of control of radio resources in an interactif telecommunication network

---

EP3704888B1|2022-01-05|Primary channel selection method for wireless communications

---

CA2900735C|2021-03-16|Transmission of a multimedia message doubled with the transmission of a text message

---

FR2922398A1|2009-04-17|SYSTEM FOR INTERCONNECTING BETWEEN AT LEAST ONE COMMUNICATION APPARATUS AND AT LEAST ONE REMOTE INFORMATION SYSTEM AND INTERCONNECTION METHOD

---

EP2341728B1|2013-01-23|System and method for controlling communications in a mobile ad-hoc network

---

EP2896268B1|2021-08-18|Management of the use of a gateway by a plurality of terminals

---

EP2141868A1|2010-01-06|Service admission control

---

WO2014006350A1|2014-01-09|Method of processing a request for connection to a wireless access network

---

EP3072355B1|2020-06-17|Quick-start method, terminal and system for radio communication

---

EP2608454A2|2013-06-26|Auslöseverfahren einer durch eine Schnittstelle ausführbaren Funktion, sowie entsprechendes Computerprogrammprodukt, Speichermedium und System

---

WO2020229057A1|2020-11-19|Method for managing a piece of security information in a communication network, corresponding device, equipment for accessing the network and computer programs

---

EP3811679A1|2021-04-28|Method and device for managing an overload of a network core controlling a mobile access network

---

WO2019149679A1|2019-08-08|Method for aggregating a plurality of radio connections in a wireless network

---

FR3100682A1|2021-03-12|PROCESS FOR RELAYING A ROUTE DISCOVERY REQUEST IN A MESH COMMUNICATION NETWORK

---

CA3077313A1|2019-04-04|Method of aiding toggling of a mobile terminal between community wireless local area networks

---

FR3095564A1|2020-10-30|PROCESS FOR SUPERVISING THE CONNECTION BETWEEN TWO NODES OF A COMMUNICATION NETWORK

---

FR3104761A1|2021-06-18|Method for monitoring data passing through user equipment

---

WO2016156386A1|2016-10-06|System for broadcasting audio and/or video content via a local wifi network, and devices implementing the method

---

WO2011023904A1|2011-03-03|Method for the geolocated broadcasting of content in a telecommunication network

---

FR3029725A1|2016-06-10|METHOD OF ESTABLISHING A TELEPHONE LINK BETWEEN A FIRST COMMUNICATION DEVICE AND A SECOND COMMUNICATION DEVICE, AND ASSOCIATED SERVER

---

WO2015007651A1|2015-01-22|User equipment offering a duration to be applied to a time delay to be set by equipment of a core network

同族专利:

---

公开号 | 公开日

---

FR3021482B1|2016-09-09|

---

AU2015261758B2|2019-08-22|

---

US10200830B2|2019-02-05|

---

WO2015177350A1|2015-11-26|

---

AU2015261758A1|2016-12-01|

---

US20170188205A1|2017-06-29|

---

EP3146734B1|2019-10-23|

---

EP3146734A1|2017-03-29|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

US7684818B1|2006-02-23|2010-03-23|Nextel Communications Inc.|System and method for facilitating communications between distinct dispatch communication systems|

---

US20100261494A1|2009-04-13|2010-10-14|Qualcomm Incorporated|Latency improvement methods in native ptt gateway for a group call with dispatch console clients|

---

FI20011263A|2001-06-14|2002-12-15|Nokia Corp|Group call management mechanism|

---

US7688764B2|2002-06-20|2010-03-30|Motorola, Inc.|Method and apparatus for speaker arbitration in a multi-participant communication session|

---

GB0413972D0|2004-06-22|2004-07-28|Nokia Corp|A communication system|

---

GB0500483D0|2005-01-11|2005-02-16|Nokia Corp|Multi-party sessions in a communication system|

---

US20070266077A1|2006-03-31|2007-11-15|Alcatel|Presence and preference-enabled push to talk telephony system|

---

US9369520B2|2012-08-19|2016-06-14|Box, Inc.|Enhancement of upload and/or download performance based on client and/or server feedback information|

---

US9357359B2|2013-02-05|2016-05-31|Qualcomm Incorporated|Dynamic quality of service for services over cellular|US10225226B2|2015-05-11|2019-03-05|Samsung Electronics Co., Ltd.|Terminals and method of communication between same|

---

RU2711023C1|2016-07-15|2020-01-14|Хуавей Текнолоджиз Ко., Лтд.|Method of seeking permission for media transmission and method and device for canceling permission for media transmission|

---

CN108401229B|2017-02-08|2021-03-05|普天信息技术有限公司|Speaking right duration application method and device|

---

US10555119B2|2018-07-06|2020-02-04|Apple Inc.|Ranging priority indication|

法律状态:
2015-05-28| PLFP| Fee payment|Year of fee payment: 2 |

---

2015-11-27| PLSC| Search report ready|Effective date: 20151127 |

---

2016-05-27| PLFP| Fee payment|Year of fee payment: 3 |

---

2016-06-17| CD| Change of name or company name|Owner name: AIRBUS DEFENCE AND SPACE, FR Effective date: 20160512 |

---

2016-06-17| CA| Change of address|Effective date: 20160512 |

---

2017-05-29| PLFP| Fee payment|Year of fee payment: 4 |

---

2018-05-29| PLFP| Fee payment|Year of fee payment: 5 |

---

2019-05-23| PLFP| Fee payment|Year of fee payment: 6 |

---

2020-04-22| PLFP| Fee payment|Year of fee payment: 7 |

---

2021-04-21| PLFP| Fee payment|Year of fee payment: 8 |

优先权:

---

申请号 | 申请日 | 专利标题

---

FR1454670A|FR3021482B1|2014-05-23|2014-05-23|METHOD FOR MANAGING SPEECHING ON A COMMUNICATION CHANNEL IN THE CONTEXT OF ALTERNATE COMMUNICATIONS|FR1454670A| FR3021482B1|2014-05-23|2014-05-23|METHOD FOR MANAGING SPEECHING ON A COMMUNICATION CHANNEL IN THE CONTEXT OF ALTERNATE COMMUNICATIONS|

---

AU2015261758A| AU2015261758B2|2014-05-23|2015-05-22|Method for managing floor control on a communication channel in the context of half-duplex communications|

---

EP15730073.2A| EP3146734B1|2014-05-23|2015-05-22|Method for managing speaking turns over a communication channel in the context of simplex calls|

---

PCT/EP2015/061426| WO2015177350A1|2014-05-23|2015-05-22|Method for managing speaking turns over a communication channel in the context of simplex calls|

---

US15/312,979| US10200830B2|2014-05-23|2015-05-22|Method for managing floor control on a communication channel in the context of half-duplex communications|