Note: Within nine months of the publication of the mention of the grant of the European patent in the European Patent Bulletin, any person may give notice to the European Patent Office of opposition to that patent, in accordance with the
2
023
338
B1
TEPZZ Z ¥¥¥8B_T
(11)EP 2 023 338 B1
(12)
EUROPEAN PATENT SPECIFICATION
(45) Date of publication and mention of the grant of the patent:
28.05.2014 Bulletin 2014/22
(21) Application number: 08161276.4 (22) Date of filing: 28.07.2008
(51) Int Cl.:
G10L 15/26(2006.01)
(54) Data attribute detection method and system and data attribute analysis device Datenattribut-Detektionsverfahren und -system und Datenattribut-Analysevorrichtung
Procédé de détection d’attribut de données et système et dispositif d’analyse d’attribut de données (84) Designated Contracting States:
AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MT NL NO PL PT RO SE SI SK TR
(30) Priority: 30.07.2007 CN 200710130070
22.05.2008 PCT/CN2008/071048
(43) Date of publication of application:
11.02.2009 Bulletin 2009/07
(73) Proprietor: Huawei Technologies Co., Ltd.
Longgang District, Shenzhen Guangdong 518129 (CN)
(72) Inventors:
• Yu, Zhou,
Huawei Administration Building, Shenzhen Guangdong 518129 (CN)
• Meng, Yuetao,
Huawei Administration Building, Bantian Shenzhen Guangdong518129 (CN) • Xu, Ji,
Huawei Administration Building, Bantian Shenzhen Guangdong 518129 (CN)
(74) Representative: Isarpatent
Patent- und Rechtsanwälte Friedrichstrasse 31
80801 München (DE)
(56) References cited:
US-A1- 2004 260 560
• RUNSHENG WANG ET AL: "VoIP Development in China" COMPUTER, IEEE SERVICE CENTER, LOS ALAMITOS, CA, US, vol. 37, no. 9, 1 September 2004 (2004-09-01), pages 30-37, XP011118740 ISSN: 0018-9162
5 10 15 20 25 30 35 40 45 50 55 Description
Field of the Invention
[0001] The present invention relates to the field of
com-munication and in particular to a data attribute detection method and system and a data attribute analysis device.
Background of the Invention
[0002] Security of some important data information is
critical to the survival and development of an enterprise, and therefore it is necessary to take measures to reduce security risks of the data information.
[0003] Prior to the birth of the Voice over IP (VoIP)
technology, a non-voice data file could be transmitted over a voice channel. For example during a facsimile, a sender modulates a non-voice data file of a user into an analog signal transmitted by a modem to a telephone line for transmission, and a modem of a receiver upon recep-tion of the analog signal demodulates the signal into a digital signal and transmits the signal to a fax machine. The fax machine recombines the received digital signal into a non-voice data file and provides the file to a user. Obviously, prior to the birth of the VoIP technology, de-vices of a telephone line, a modem or a fax machine, etc. are needed in transmitting the non-voice data file over a voice channel, and therefore a better management of these devices might result in an increased management effort although the better management of these devices can reduce to some extent risks in transmitting the non-voice data file containing important data information to the outside over a voice channel.
[0004] After the birth of the VoIP technology, the
trans-mission of a non-voice data file over a voice channel has been made more convenient. For example, a data infor-mation file can be packed by a PC into packets with a voice data attribute, which are transmitted to an opposite party over a voice channel established by using an IP telephone, especially a soft telephone which is installable on the PC. However, the transmission of a non-voice data file containing important data information can not be controlled with this convenience, which may endanger the security of the important data information.
[0005] US 2004/260560 A1 discloses a VoIP security
intelligence method through detecting voice data and identifying the voice data as significance or importance. RUNSHENG WANG ET AL: "VoIP Development in Chi-na" (COMPUTER IEEE SERVICE CENTER, LOS ALAMITOS, CA, US, vol. 37, no.9, 1 September 2004 (2004-09-01), pages 30-37, XP011118740 ISSN:00189162) provides a method of differentiating
voice and fax traffic automatically through differen-tiating encoding standards.
Summary of the Invention
[0006] Embodiments of the invention provide a data
attribute detection method and system and a data at-tribute analysis device to improve the security of impor-tant data information.
[0007] To attain this, an embodiment of the invention
provides a data attribute detection method including: ob-taining a media stream transmitted over a voice channel established through a voice switch; recognizing the ob-tained media stream to obtain a recognition result; and determining whether the recognition result is above a pre-set recognition success threshold; if the recognition result is above a preset recognition success threshold, deter-mining that the data attribute of the media stream trans-mitted over the voice channel is voice data; otherwise, determining that the data attribute of the media stream transmitted over the voice channel is non-voice data; if the data attribute of the media stream is non-voice data, determining a corresponding control mode for a commu-nication process of transmitting the media stream with the data attribute of non-voice data in accordance with preconfigured policy information; if the determined con-trol mode is a disturbance concon-trol mode, instructing the voice switch to perform a disturbance control on the cur-rent communication process according to the distur-bance control mode, so that the voice switch transmits a corresponding instruction to a media server to instruct the media server to add trashy data, which is noise or the like, in currently transmitted media stream and trans-mit the media stream with the trashy data to receiver, wherein all media streams exchanged between both communication parties are forwarded via the media serv-er.
[0008] An embodiment of the invention further
pro-vides a data attribute detection system including: a rec-ognition server, a data attribute analysis device and a voice switch; the recognition server is adapted to obtain a media stream transmitted over a voice channel estab-lished through the voice switch, to recognize the obtained media stream to obtain a recognition result; and the data attribute analysis device is adapted to determine whether the recognition result is above a preset recognition suc-cess threshold; if the recognition result is above a preset recognition success threshold, determining that the data attribute of the media stream transmitted over the voice channel is voice data; otherwise, determining that the data attribute of the media stream transmitted over the voice channel is non-voice data; if the data attribute of the media stream is non-voice data, is further adapted to determine a corresponding control mode for a com-munication process of transmitting the media stream with the data attribute of non-voice data in accordance with preconfigured policy information, and if the determined control mode is a disturbance control mode, is further adapted to instruct the voice switch to perform a distur-bance control on the current communication process ac-cording to the disturbance control mode; the voice switch is adapted to establish the voice channel, and is further adapted to transmit a corresponding instruction to a me-dia server according to instruction of the data attribute
5 10 15 20 25 30 35 40 45 50 55 analysis device to instruct the media server to add trashy data, which is noise or the like, in currently transmitted media stream and transmit the media stream with the trashy data to receiver, wherein all media streams ex-changed between both communication parties are for-warded via the media server.
[0009] An embodiment of the invention further
pro-vides a data attribute analysis device including: an ob-taining unit adapted to obtain a recognition result of a media stream transmitted over a voice channel estab-lished through a voice switch; and a detection unit com-prising a comparison sub-unit and a detection sub-unit, wherein, the sub-unit adapted to compare the recognition result with a preset recognition success threshold; and the detection sub-unit adapted to determine that the data attribute of the media stream transmitted over the voice channel is voice data when a comparison result is that the recognition result is above the preset recognition suc-cess threshold, or to determine that the data attribute of the media stream currently transmitted over the voice channel is non-voice data when a comparison result is that the recognition result is below or equal to the preset recognition success threshold; a control mode determi-nation unit adapted to determine a corresponding control mode for a communication process of transmitting the media stream with the data attribute of non-voice data in accordance with preconfigured policy information if the data attribute of the media stream is non-voice data; and, if the determined control mode is a disturbance control mode, instruct the voice switch to perform a disturbance control on the current communication process according to the disturbance control mode, so that the voice switch transmits a corresponding instruction to a media server to instruct the media server to add trashy data, which is noise or the like, in currently transmitted media stream and transmit the media stream with the trashy data to receiver, wherein the media server is used to forward media streams exchanged between both communication parties.
[0010] As apparent from the above solutions, the
em-bodiments of the invention can obtain a recognition result of a media stream transmitted over a voice channel and determine a data attribute of the media stream transmit-ted over the voice channel in accordance with the rec-ognition result. Therefore, the embodiments of the inven-tion can detect whether a media stream transmitted over a voice channel is non-voice data, and particularly can prevent the non-voice data from being transmitted to the outside over the voice channel, thereby improving the security of important data information.
Brief Description of the Drawings
[0011] Fig.1 illustrates a structural diagram of a first
embodiment of the invention;
[0012] Fig.2 illustrates a flow chart of a second
embod-iment of the invention; and
[0013] Fig.3 illustrates a structural diagram of a third
embodiment of the invention.
Detailed Description of the Invention
[0014] A first embodiment of the invention provides a
data attribute detection system as structured in Fig.1, which includes a voice switch, a media server, a recog-nition server and a data attribute analysis device.
[0015] Particularly, the voice switch accomplishes a
complete call process between communication parties (e.g., a user A and a user B in Fig.1). In this call process, both communication parties negotiate about media stream information, including information of a coding mode, a compression mode, and number of a commu-nication port, etc., to be used, and media commucommu-nication connection channels between both communication par-ties and the media server in the system are established.
[0016] The media server forwards a media stream
ex-changed between both communication parties and buff-ers a certain number of media streams.
[0017] The recognition server performs voice
recogni-tion for a media stream in the media server and obtains a corresponding recognition result (taking a voice recog-nition result as an example in the embodiments of the invention here and below), for example, the number of voice words contained in the media stream transmitted within unit time or the number of successfully recogniz-able data packets contained in the media stream trans-mitted within unit time or the percentage of the number of successfully recognizable data packets contained in the media stream transmitted within unit time relative to the total number of data packets contained in the media stream, etc., and transmits the recognition result to the data attribute analysis device.
[0018] The data attribute analysis device determines
whether the data attribute of the currently communicated media stream is voice or non-voice data in accordance with the voice recognition result and a preset recognition success threshold in a way that the received voice rec-ognition result transmitted from the recrec-ognition server is compared with the preset voice recognition success threshold; and if the comparison result exceeds (or is greater than) the preset recognition success threshold, the data attribute of the currently communicated media stream is determined as voice data; otherwise the data attribute of the currently communicated media stream is determined as non-voice data.
[0019] When the data attribute of the currently
com-municated media stream is determined as voice data, it is determined that the media stream will not be proc-essed; and when the data attribute of the currently com-municated media stream is determined as non-voice da-ta, a control mode for the current communication is de-termined in accordance with policy information precon-figured for different communication processes, and the control mode may include but not limited to an interruption control mode in which the current communication proc-ess is interrupted, a disturbance control mode in which
5 10 15 20 25 30 35 40 45 50 55 the current communication process is disturbed, for ex-ample, by adding trashy data such as noise or the like in the currently transmitted media stream to make it difficult for a receiver to recover the original initial data, an alarm control mode, etc.
[0020] If the control mode for the current
communica-tion is determined as the interrupcommunica-tion control mode, the data attribute analysis device notifies the voice switch about the interruption control mode. The voice switch per-forms an interruption control on the current communica-tion in accordance with the control interrupcommunica-tion mode. In other words, the voice switch interrupts the media com-munication connection channel between both communi-cation parties in accordance with the received notifica-tion.
[0021] If the control mode for the current
communica-tion is determined as the disturbance control mode, the data attribute analysis device notifies the voice switch of the disturbance control mode. The voice switch performs a disturbance control on the current communication in accordance with the disturbance control mode. In other words, the voice switch transmits a corresponding in-struction to the media server in accordance with the dis-turbance control mode, and the media server adds trashy data such as noise or the like in the currently transmitted media stream and transmits the media stream to the re-ceiver in accordance with the received instruction from the voice switch.
[0022] If the control mode for the current
communica-tion is determined as the alarm control mode, the data attribute analysis device can directly generate an alarm signal to notify a network administrator or can notify the voice switch in accordance with the alarm control mode, and the voice switch can generate an alarm signal to notify the network administrator in accordance with the received notification.
[0023] The data attribute analysis device can be
phys-ically separate or arranged on another device, e.g., a voice switch.
[0024] The recognition server can be physically
sepa-rate or arranged on another device, e.g., a media server or a voice switch.
[0025] The above descriptions have been given taking
voice recognition of a media stream in the media server by the recognition server as an example. Additionally, the recognition server can also recognize a media stream in the media server by a human voice detection method. For example, a spectrum model method is employed, a spectrum profile of a media stream is compared with a general human voice spectrum profile; and if the error of the spectrum profile is within an allowable range, the data attribute of the media stream can be determined as voice data, and if the error of the spectrum profile is beyond the range, the data attribute of the media stream can be determined as non-voice data. Obviously, the data at-tribute of the transmitted media stream can be deter-mined as voice or non-voice data in accordance with an obtained human voice detection result. Thereafter, the
human voice detection result is taken as a recognition result of the recognition server, which is transmitted to the data attribute analysis device, and the data attribute analysis device can determine whether the data attribute of the media stream currently transmitted over a voice channel is voice or non-voice data in accordance with the recognition result of the recognition server.
[0026] Alternatively, only the recognition server and
the data attribute analysis device are provided in the above embodiment. The recognition server is adapted to obtain and recognize a media stream transmitted over the voice channel and to obtain a recognition result, and the recognition server can acquire a media stream trans-mitted over the voice channel through an acquisition de-vice or can obtain a media stream transmitted over the voice channel by taking other methods.
[0027] The data attribute analysis device is adapted to
determine the data attribute of the media stream trans-mitted over the voice channel in accordance with a rec-ognition result of the recrec-ognition server. Processing tails thereof are analogous to those in the related de-scriptions in the above embodiment and will not be re-peated here.
[0028] The second embodiment of the invention
re-lates to a data attribute detection method in which a user A calls a user B, and a media communication connection channel between the users and a media server, i.e., a voice channel, is established through voice switches serving the users. In a communication process through the media communication connection channel, a media stream currently transmitted over the voice channel is obtained and voice recognition of the media stream is performed to obtain a corresponding voice recognition result. The voice recognition result is compared with a preset recognition success threshold to detect the data attribute of the media stream transmitted over the voice channel, and if the voice recognition result is below or equal to the preset recognition success threshold, the data attribute of the media stream is non-voice data, and the communication process of transmitting the media stream with the data attribute of non-voice data is con-trolled in accordance with policy information preconfig-ured for different communication processes, that is, firstly a corresponding control mode is determined for the com-munication process of the media stream of non-voice da-ta, and then the communication process of the current media stream of non-voice data is controlled in accord-ance with the determined corresponding control mode. A specific implementation flow as illustrated in Fig.2 in-cludes:
[0029] In the step S101, the user A initiates a request
of calling the user B to the voice switch currently serving him. The request carries media description information of the user A, including information on a coding mode and a compression mode supported by the user A, a number of a communication port of the user A, etc.
[0030] In the step S 102, upon reception of the request,
5 10 15 20 25 30 35 40 45 50 55 media description information of the user A to media de-scription information of the media server and forwards the request to the user B.
[0031] In the step S103, upon reception of the request,
the user B generates a success reply and transmits the success reply to the voice switch currently serving him.
[0032] In the step S104, upon reception of the success
reply, the voice switch currently serving the user B mod-ifies media description information of the user B to the media description information of the media server and then forwards the success reply to the user A. The media description information includes information of a coding mode, a compression mode and/or a number of a com-munication port, which can be available, etc.
[0033] After the above steps, both call parties finish
the negotiation about the media information, including the information of a coding mode, a compression mode and/or a number of a communication port, which can be available, etc.; and establish a media communication connection channel, i.e., a voice channel used for the media communication, respectively with the media serv-er. Then all media streams exchanged between both call parties are forwarded via the media server in the com-munication process of the user A and the user B, details are as followed.
[0034] In the steps S105a and S105b, media streams
exchanged between both communication parties in the communication process reach the media server, and the media server buffers a certain number of media streams and then transmits these media streams to a recognition server.
[0035] In the steps S106a and S106b, the recognition
server recognizes the obtained media stream, obtains a corresponding recognition result and transmits the rec-ognition result to a data attribute analysis device.
[0036] The recognition result is a voice recognition
re-sult which is obtained from voice recognition of the media stream by the recognition server using a voice recogni-tion method, typically the number of voice words con-tained in the media stream transmitted within unit time or the number of successfully recognizable data packets contained in the media stream transmitted within unit time or the percentage of the number of successfully recog-nizable data packets contained in the media stream transmitted within unit time relative to the total number of data packets contained in the media stream, etc.
[0037] The voice recognition method used by the
rec-ognition server may be an existing voice recrec-ognition tech-nology, for example, an acoustic feature analysis in which a segment of the voice media stream transmitted within unit time is divided into syllables (a syllable, i.e., a seg-ment of consecutive data packets, is typically success-fully recognizable) according to an acoustic feature, and in the process of division into syllables, a transition voice segment between syllables will be discarded (the transi-tion voice segment, also a segment of consecutive data packets, is typically unrecognizable), so that a recogni-tion result is obtained as data packets occupied by
suc-cessfully recognizable syllables contained in the media stream transmitted within a unit time and the percentage of data packets occupied by successfully recognizable syllables contained in the media stream transmitted with-in a unit time relative to the total number of data packets.
[0038] A match is conduct according to an acoustic
model after the acoustic feature analysis is finished, that is, the successfully recognized syllables are associated with words. Typically, a match of a syllable and a word can be marked in a voice recognition process and will be accepted if the mark is above a threshold, otherwise the recognition is disabled, and thus the second recognition result can be obtained as successfully recognized voice words contained in the media stream transmitted within unit time.
[0039] The above successfully recognized voice words
are corrected by a linguistic model, and a corresponding recognition result can be obtained as the number of suc-cessfully recognizable voice words contained in the me-dia stream transmitted within unit time.
[0040] The recognition server transmits the above
voice recognition results or one of the voice recognition results to a data attribute analysis device.
[0041] In the step S107, the data attribute analysis
de-vice receives the voice recognition result and determines the data attribute of the media stream currently transmit-ted over the voice channel in accordance with the voice recognition result and a preset recognition success threshold; and if the data attribute of the currently com-municated media stream is voice data, the data attribute analysis device performs no processing, and if the data attribute of the currently communicated media stream is non-voice data, the communication process of transmit-ting the media stream with the data attribute of non-voice data is controlled in accordance with policy information preconfigured for different communication processes, that is, firstly a control mode for the communication proc-ess of the media stream of non-voice data is determined, and the communication process of the current media stream of non-voice data is controlled in the determined corresponding control mode. For example, the step S108 is executed to notify a voice switch of the control mode, and the voice switch issues a corresponding instruction to a media server according to the notification (i.e. step 109).
[0042] The voice switch performs processing
accord-ing to the received determined control mode or performs corresponding processing by itself according to the de-termined control mode (in this embodiment, as an exam-ple, the voice switch is instructed to perform correspond-ing processcorrespond-ing accordcorrespond-ing to the determined control mode.).
[0043] In the step S107, the preset recognition success
threshold can be preconfigured in the data attribute anal-ysis device or in configuration information of a system where the data attribute analysis device is located. The preset recognition success threshold can be the number of voice words contained in the media stream transmitted
5 10 15 20 25 30 35 40 45 50 55 within unit time or the number of successfully recogniz-able data packets contained in the media stream trans-mitted within unit time or the percentage of the number of successfully recognizable data packets contained in the media stream transmitted within unit time relative to the total number of data packets contained in the media stream, etc. The data attribute analysis device can com-pare the voice recognition result with the preset recog-nition success threshold while determining the data at-tribute of the media stream currently transmitted over the voice channel in accordance with the voice recognition result and the preset recognition success threshold; and if a comparison result is that the voice recognition result is below or equal to the preset recognition success threshold, then the data attribute of the media stream currently transmitted over the voice channel is deter-mined as non-voice data, otherwise the data attribute of the media stream currently transmitted over the voice channel is determined as voice data.
[0044] For example, a preset first recognition success
threshold assumed to be 70% is compared with a voice recognition result for successful recognition, for example, the percentage of the number of successfully recogniz-able data packets contained in the media stream trans-mitted within unit time relative to the total number of data packets contained in the media stream, and if the per-centage exceeds (or is above) 70%, then the data at-tribute of the media stream transmitted within unit time is determined as voice data.
[0045] If 70% is not exceeded, then a further
determi-nation is made to compare the next index, i.e., the number of successfully recognizable voice words contained in the media stream transmitted within unit time, with a cor-responding preset second recognition success thresh-old; and if the number exceeds (or be above) the preset second recognition success threshold, then the data at-tribute of the media stream transmitted within unit time is determined as voice data, otherwise the data attribute of the media stream transmitted within unit time is deter-mined as non-voice data.
[0046] In the step S107, the communication process
of transmitting the media stream with the data attribute of non-voice data is controlled in accordance with pre-configured policy information, that is, firstly a control mode for the communication process of the currently transmitted media stream with the data attribute of non-voice data is determined, and the control mode may be an interruption control mode in which the current com-munication process is interrupted, a disturbance control mode in which the current communication process is dis-turbed, for example, by adding trashy data such as noise or the like in the currently transmitted media stream to make it difficult for a receiver to recover the original initial data, an alarm control mode, etc.
[0047] If the determined control mode is the
interrup-tion control mode, then the data attribute analysis device instructs the voice switch to perform an interruption con-trol on the current communication process according to
the interruption control mode, and the voice switch inter-rupts the media communication connection between both communication parties according to the instruction.
[0048] If the determined control mode is the
distur-bance control mode, then the data attribute analysis de-vice instructs the voice switch to perform a disturbance control on the current communication process according to the disturbance control mode, and the voice switch transmits a corresponding instruction to the media server according to the instruction. The media server adds trashy data such as noise or the like in the currently trans-mitted media stream and transmits the media stream to the receiver according to the instruction from the voice switch.
[0049] If the determined control mode is the alarm
con-trol mode, the data attribute analysis device may notify the voice switch according to the alarm control mode, and the voice switch may generate an alarm signal to notify a network administrator according to the notifica-tion from the data attribute analysis device. Alternatively, the data attribute analysis device may perform corre-sponding processing by itself according to the deter-mined control mode, that is, directly generate an alarm signal to notify the network administrator.
[0050] The above descriptions of the second
embodi-ment have been given taking voice recognition of a media stream in the media server by the recognition server as an example. Additionally, the recognition server may also recognize a media stream by a currently existing human voice detection method. For example, a spectrum model method is employed, a spectrum profile of a media stream is compared with a general human voice spec-trum profile; and if an error of the specspec-trum profile is within an allowable range, the data attribute of the media stream may be determined as voice data, and if the error of the spectrum profile is beyond the range, the data attribute of the media stream may be determined as non-voice data. Obviously, a finally obtained human voice detection result contains information about whether the detected media stream is voice or non-voice data, and the recog-nition server transmits the human voice detection result as a recognition result to the data attribute analysis de-vice. Thus, the data attribute analysis device can deter-mine whether the data attribute of the media stream cur-rently transmitted over the voice channel is voice or non-voice data directly in accordance with the human non-voice detection result.
[0051] The third embodiment of the invention provides
a data attribute analysis device as structured in Fig.3, which includes an obtaining unit and a detection unit, the detection unit can further include a comparison sub-unit and a detection sub-unit. For further control a current communication, the data attribute analysis device can further include a control mode determination unit and a control unit. The control unit can further include a first control sub-unit and/or a second control sub-unit.
[0052] The obtaining unit is adapted to obtain a
5 10 15 20 25 30 35 40 45 50 55 channel. Particularly, the obtaining unit can obtain a voice recognition result from a recognition server.
[0053] The detection unit is adapted to determine the
data attribute of the media stream transmitted over the voice channel in accordance with the recognition result. Particularly, the detection unit includes:
[0054] The comparison sub-unit compares the voice
recognition result obtained by the obtaining unit with a preset recognition success threshold which can be pre-configured in the data attribute analysis device or in con-figuration information of a system where the data attribute analysis device is located, and then transmits a compar-ison result to the detection sub-unit.
[0055] The detection sub-unit determines that the data
attribute of the media stream transmitted over the voice channel is voice data when the comparison result is that the voice recognition result exceeds (or is above) the preset recognition success threshold and determines that the data attribute of the media stream transmitted over the voice channel is non-voice data when the com-parison result is that the voice recognition result is below or equal to the preset recognition success threshold.
[0056] The control mode determination unit is adapted
to determine a control mode for a communication process of transmitting the media stream with the data attribute of non-voice data in accordance with policy information preconfigured for different communication processes, and the determined control mode may be an interruption control mode in which a current communication process is interrupted, a disturbance control mode in which a cur-rent communication process is disturbed, for example, by adding trashy data such as noise or the like in the currently transmitted media stream to make it difficult for a receiver to recover the original initial data, or an alarm control mode, etc.
[0057] The control unit is adapted to control the
com-munication process of transmitting non-voice data ac-cording to the control mode determined by the control mode determination unit specifically as follows.
[0058] The first control sub-unit is adapted to instruct
voice switches which both communication parties are subordinate to control the communication process of transmitting non-voice data. Processing details thereof are analogous to those in the related descriptions of the first embodiment and will not be repeated here.
[0059] The second control sub-unit is adapted to
per-form corresponding control by itself on the communica-tion process of transmitting non-voice data according to the control mode determined by the control mode deter-mination unit. For example, an alarm signal is generated for the current communication process of transmitting non-voice data over the voice channel to notify a network administrator. The rest details are substantially analo-gous to those of the second embodiment and will not be repeated here.
[0060] The fourth embodiment of the invention further
provides another data attribute analysis device including an obtaining unit and a detection unit, and the detection
unit can include a data attribute determination sub-unit. For further control on a current communication, the data attribute analysis device can further include a control mode determination unit and a control unit. The control unit can further include a first control sub-unit and/or a second control sub-unit.
[0061] Particularly, the control mode determination
unit, the control unit and the first control sub-unit and/or the second control sub-unit in the control unit have func-tions analogous to the respective units in the third em-bodiment and will not be described here. Functions of only the obtaining unit and the detection unit will be de-scribed below.
[0062] The obtaining unit is adapted to obtain a
recog-nition result of a media stream transmitted over a voice channel, and it can obtain a human voice detection result from a recognition server.
[0063] The detection unit is adapted to determine the
data attribute of the media stream transmitted over the voice channel in accordance with the recognition result obtained by the obtaining unit. Particularly, the detection unit includes:
[0064] A data attribute determination sub-unit
deter-mines whether the data attribute of the media stream transmitted over the voice channel is voice or non-voice data in accordance with the human voice detection result obtained by the obtaining unit.
[0065] As apparent from the specific implementation
solutions provided according to the embodiments of the invention, they can detect whether the media stream transmitted over the voice channel is non-voice data by using the method for determining the data attribute of the media stream transmitted over the voice channel in ac-cordance with the recognition result, so that particularly the non-voice data can be prevented from being trans-mitted to the outside over the voice channel, thereby im-proving the security of important data information.
[0066] Evidently, those skilled in the art can modify and
vary the invention without departing from scope of the invention. Accordingly, the invention is intended to en-compass these modifications and modifications provided that they come into the scope of the claims appended to the invention.
Claims
1. A data attribute detection method, comprising:
obtaining (105a-105b) a media stream transmit-ted over a voice channel established through a voice switch;
recognizing (106a) the obtained media stream to obtain a recognition result;
and
determining (106b) whether the recognition re-sult is above a preset recognition success threshold; if the recognition result is above a
pre-5 10 15 20 25 30 35 40 45 50 55 set recognition success threshold, determining that the data attribute of the media stream trans-mitted over the voice channel is voice data; oth-erwise, determining that the data attribute of the media stream transmitted over the voice chan-nel is non-voice data;
if the data attribute of the media stream is non-voice data, determining(107) a corresponding control mode for a communication process of transmitting the media stream with the data at-tribute of non-voice data in accordance with pre-configured policy information;
if the determined control mode is a disturbance control mode, instructing the voice switch to per-form a disturbance control on the current com-munication process according to the distur-bance control mode, so that the voice switch transmits a corresponding instruction to a media server to instruct the media server to add trashy data, such as noise, in the currently transmitted media stream and transmit the media stream with the trashy data to a receiver, wherein all media streams exchanged between both com-munication parties are forwarded via the media server.
2. The method according to claim 1, wherein when the
recognition result is a human voice detection result, the data attribute of the media stream transmitted over the voice channel is determined in accordance with the human voice detection result.
3. The method according to claim 2, wherein if the
hu-man voice detection result is determined by a spec-trum profile, the determining of the data attribute of the media stream transmitted over the voice channel in accordance with the human voice detection result comprises:
comparing a spectrum profile of the media stream with a preset human voice spectrum pro-file; if a comparison error is within a preset range, determining the data attribute of the media stream as voice data; otherwise determining the data attribute of the media stream as non-voice data.
4. The method according to any one of claims 1 to 3,
further comprising:
if the determined control mode is an interruption control mode, instructing the voice switch to per-form an interruption control on the current com-munication process according to the interruption control mode, so that the voice switch interrupts the media communication connection between both communication parties according to the in-struction.
5. A data attribute detection system, comprising: a
rec-ognition server, a data attribute analysis device and a voice switch;
the recognition server is adapted to obtain a media stream transmitted over a voice channel established through the voice switch, to recognize the obtained media stream to obtain a recognition result which is then transmitted to a data attribute analysis device; and
the data attribute analysis device is adapted to de-termine whether the recognition result is above a pre-set recognition success threshold; if the recognition result is above a preset recognition success thresh-old, determining that the data attribute of the media stream transmitted over the voice channel is voice data; otherwise, determining that the data attribute of the media stream transmitted over the voice chan-nel is non-voice data; if the data attribute of the media stream is non-voice data, is further adapted to de-termine a corresponding control mode for a commu-nication process of transmitting the media stream with the data attribute of non-voice data in accord-ance with preconfigured policy information, and if the determined control mode is a disturbance control mode, is further adapted to instruct the voice switch to perform a disturbance control on the current com-munication process according to the disturbance control mode;
the voice switch is adapted to establish the voice channel, and is further
adapted to transmit a corresponding instruction to a media server according to instruction of the data at-tribute analysis device to instruct the media server to add
data, such as noise, in the currently transmitted me-dia stream and transmit the meme-dia stream with the trashy data to a receiver, wherein all media streams exchanged between both communication parties are forwarded via the media server.
6. The system according to claim 5, further comprising:
a media server adapted to obtain a certain number of media streams transmitted over the voice channel and to provide the recognition server with the media streams;
wherein the recognition server is specifically adapted to receive the media streams from the media server and recognize the media streams to obtain the recognition result.
7. The system according to claim 6, wherein the voice
switch is adapted to establish the voice channel, wherein
the voice switch is further adapted to establish a con-nection channel between both communication par-ties and the media server.
5 10 15 20 25 30 35 40 45 50 55
8. The system according to any one of claims 5 to 7,
wherein, if the determined control mode is an inter-ruption control mode, the data attribute analysis de-vice is further adapted to instruct the voice switch to perform an interruption control on the current com-munication process;
the voice switch is further adapted to interrupt media communication connection between both communi-cation parties according to instruction of the data at-tribute analysis device.
9. A data attribute analysis device, comprising:
an obtaining unit adapted to obtain a recognition result of a media stream transmitted over a voice channel established through a voice switch; and a detection unit comprising a comparison sub-unit and a detection sub-sub-unit, wherein, the com-parison sub-unit is adapted to compare the rec-ognition result with a preset recrec-ognition success threshold; and the detection sub-unit adapted to determine that the data attribute of the media stream transmitted over the voice channel is voice data when a comparison result is that the recognition result is above the preset recogni-tion success threshold, or to determine that the data attribute of the media stream currently transmitted over the voice channel is non-voice data when a comparison result is that the rec-ognition result is below or equal to the preset recognition success threshold;
a control mode determination unit adapted to determine a corresponding control mode for a communication process of transmitting the me-dia stream with the data attribute of non-voice data in accordance with preconfigured policy in-formation if the data attribute of the media stream is non-voice data; and, if the determined control mode is a disturbance control mode, in-struct the voice switch to perform a disturbance control on the current communication process according to the disturbance control mode, so that the voice switch transmits a corresponding instruction to a media server to instruct the me-dia server to add trashy data, such as noise, in the currently transmitted media stream and transmit the media stream with the trashy data to a receiver, wherein the media server is used to forward media streams exchanged between both communication parties.
10. The data attribute analysis device according to claim
9, wherein, if the determined control mode is an in-terruption control mode, the control mode determi-nation unit is further adapted to instruct the voice switch to perform an interruption control on the cur-rent communication process, so that the voice switch interrupts media communication connection
be-tween both communication parties according to in-struction of the control mode determination unit.
11. The data attribute analysis device according to claim
9, wherein the detection unit comprises:
a data attribute determination sub-unit adapted to determine the data attribute of the media stream transmitted over the voice channel in ac-cordance with a human
voice detection result when the recognition re-sult is the human voice detection rere-sult.
12. The data attribute analysis device according to any
one of claims 9 to 11, wherein the data attribute anal-ysis device is integrated into a voice switch.
Patentansprüche
1. Datenattribut-Detektionsverfahren, umfassend:
Erhalten (105a-105b) eines Medienstroms, der über einen durch eine Sprachvermittlung herge-stellten Sprachkanal übertragen wird;
Erkennen (106a) des erhaltenen Medienstroms, um ein Erkennungsergebnis zu erhalten; und Bestimmen (106b), ob das Erkennungsergebnis über einer voreingestellten Erkennungserfolgs-schwelle liegt; wenn das Erkennungsergebnis über einer voreingestellten Erkennungserfolgs-schwelle liegt, Bestimmen, dass das Datenattri-but des über den Sprachkanal übertragenen Medienstroms Sprachdaten ist; andernfalls Be-stimmen, dass das Datenattribut des über den Sprachkanal übertragenen Medienstroms Nicht-Sprachdaten ist;
wenn das Datenattribut des Medienstroms Nicht-Sprachdaten ist, Bestimmen (107) eines entsprechenden Steuermodus für einen Kom-munikationsprozess zum Übertragen des Me-dienstroms mit dem Datenattribut von Nicht-Sprachdaten gemäß vorkonfigurierten Richtlini-eninformationen;
wenn der bestimmte Steuermodus ein Stö-rungssteuermodus ist, Anweisen der Sprach-vermittlung, gemäß dem Störungssteuermodus eine Störungssteuerung an dem aktuellen Kom-munikationsprozess auszuführen, so dass die Sprachvermittlung eine entsprechende Anwei-sung zu einem Medienserver überträgt, um den Medienserver anzuweisen, Schund-Daten, wie etwa Rauschen, in dem aktuell übertragenen Medienstrom hinzuzufügen und den Medienst-rom mit den Schund-Daten zu einem Empfänger zu übertragen, wobei alle zwischen beiden Kommunikationsteilnehmern ausgetauschten Medienströme über den Medienserver
weiter-5 10 15 20 25 30 35 40 45 50 55 geleitet werden.
2. Verfahren nach Anspruch 1, wobei, wenn das
Er-kennungsergebnis ein Menschliche-Stimme-Detek-tionsergebnis ist, das Datenattribut des über den Sprachkanal übertragenen Medienstroms gemäß dem Menschliche-Stimme-Detektionsergebnis be-stimmt wird.
3. Verfahren nach Anspruch 2, wobei, wenn das
Menschliche-Stimme-Detektionsergebnis durch ein Spektrumprofil bestimmt wird, das Bestimmen des Datenattributs des über den Sprachkanal übertrage-nen Medienstroms gemäß dem Menschliche-Stim-me-Detektionsergebnis Folgendes umfasst:
Vergleichen eines Spektrumprofils des Me-dienstroms mit einem voreingestellten Mensch-liche-Stimme-Spektrumprofil; wenn ein Ver-gleichsfehler innerhalb eines voreingestellten Bereichs liegt, Bestimmen des Datenattributs des Medienstroms als Sprachdaten; andernfalls Bestimmen des Datenattributs des Medienst-roms als Nicht-Sprachdaten.
4. Verfahren nach einem der Ansprüche 1 bis 3, ferner
umfassend:
wenn der bestimmte Steuermodus ein Unterbre-chungssteuermodus ist, Anweisen der Sprach-vermittlung, gemäß dem Unterbrechungssteu-ermodus eine Unterbrechungssteuerung an dem aktuellen Kommunikationsprozess auszu-führen, so dass die Sprachvermittlung die Me-dienkommunikationsverbindung zwischen bei-den Kommunikationsteilnehmern gemäß der Anweisung unterbricht.
5. Datenattribut-Detektionssystem, umfassend: einen
Erkennungsserver, eine Datenattribut-Analyseein-richtung und eine Sprachvermittlung;
wobei der Erkennungsserver ausgelegt ist zum Er-halten eines Medienstroms, der über einen durch die Sprachvermittlung hergestellten Sprachkanal übtragen wird, um den erhaltenen Medienstrom zu er-kennen, um ein Erkennungsergebnis zu erhalten, das dann zu einer Datenattribut-Analyseeinrichtung übertragen wird; und die Datenattribut-Analyseein-richtung ausgelegt ist zum Bestimmen, ob das kennungsergebnis über einer voreingestellten Er-kennungserfolgsschwelle liegt; wenn das nungsergebnis über einer voreingestellten Erken-nungserfolgsschwelle liegt, Bestimmen, dass das Datenattribut des über den Sprachkanal übertrage-nen Medienstroms Sprachdaten ist; andernfalls Be-stimmen, dass das Datenattribut des über den Sprachkanal übertragenen Medienstroms Nicht-Sprachdaten ist; wenn das Datenattribut des
Me-dienstroms Nicht-Sprachdaten ist, sie ferner ausge-legt ist zum Bestimmen eines entsprechenden Steu-ermodus für einen Kommunikationsprozess zum Übertragen des Medienstroms mit dem Datenattri-but von Nicht-Sprachdaten gemäß vorkonfigurierten Richtlinieninformationen; und wenn der bestimmte Steuermodus ein Störungssteuermodus ist, sie fer-ner ausgelegt ist, zum Anweisen der Sprachvermitt-lung, gemäß dem Störungssteuermodus eine Stö-rungssteuerung an dem aktuellen Kommunikations-prozess auszuführen,
wobei die Sprachvermittlung ausgelegt ist zum Her-stellen des Sprachkanals und ferner ausgelegt ist zum Übertragen einer entsprechenden Anweisung zu einem Medienserver gemäß der Anweisung der Datenattribut-Analyseeinrichtung, um den Medien-server anzuweisen, Schund-Daten, wie etwa Rau-schen, in dem aktuell übertragenen Medienstrom hinzuzufügen und den Medienstrom mit den Schund-Daten zu einem Empfänger zu übertragen, wobei alle zwischen beiden Kommunikationsteilneh-mern ausgetauschten Medienströme über den Me-dienserver weitergeleitet werden.
6. System nach Anspruch 5, ferner umfassend:
einen Medienserver, ausgelegt zum Erhalten ei-ner bestimmten Anzahl von über den Sprachka-nal übertragenen Medienströmen und zum Be-reitstellen der Medienströme für den Erken-nungsserver;
wobei der Erkennungsserver spezifisch ausge-legt ist zum Empfangen der Medienströme von dem Medienserver und Erkennen der Medien-ströme, um das Erkennungsergebnis zu erhal-ten.
7. System nach Anspruch 6, wobei die
Sprachvermitt-lung ausgelegt ist zum Herstellen des Sprachkanals; wobei die Sprachvermittlung ausgelegt ist zum Her-stellen eines Verbindungskanals zwischen beiden Kommunikationsteilnehmern und dem Medienser-ver.
8. System nach einem der Ansprüche 5 bis 7, wobei,
wenn der bestimmte Steuermodus ein Unterbre-chungssteuermodus ist, die Datenattribut-Analyse-einrichtung ferner ausgelegt ist zum Anweisen der Sprachvermittlung, eine Unterbrechungssteuerung an dem aktuellen Kommunikationsprozess auszu-führen; und
die Sprachvermittlung ferner ausgelegt ist zum Un-terbrechen der Medienkommunikationsverbindung zwischen beiden Kommunikationsteilnehmern ge-mäß der Anweisung der Datenattribut-Analyseein-richtung.
5 10 15 20 25 30 35 40 45 50 55 eine Erhalteeinheit, ausgelegt zum Erhalten ei-nes Erkennungsergebnisses eiei-nes Medienst-roms, der über einen durch eine Sprachvermitt-lung hergestellten Sprachkanal übertragen wird; und
eine Detektionseinheit, die eine Vergleichssu-beinheit und eine DetektionssuVergleichssu-beinheit um-fasst, wobei die Vergleichssubeinheit ausgelegt ist zum Vergleichen des Erkennungsergebnis-ses mit einer voreingestellten Erkennungser-folgsschwelle; und
die Detektionssubeinheit ausgelegt ist zum Be-stimmen, dass das Datenattribut des über den Sprachkanal übertragenen Medienstroms Sprachdaten ist, wenn ein Vergleichsergebnis darin besteht, dass das Erkennungsergebnis über der voreingestellten Erkennungserfolgs-schwelle liegt, oder zum Bestimmen, dass das Datenattribut des gerade über den Sprachkanal übertragenen Medienstroms Nicht-Sprachda-ten ist, wenn ein Vergleichsergebnis darin be-steht, dass das Erkennungsergebnis unter oder gleich der voreingestellten Erkennungserfolgs-schwelle ist;
eine Steuermodus-Bestimmungseinheit, aus-gelegt zum Bestimmen eines entsprechenden Steuermodus für einen Kommunikationspro-zess zum Übertragen des Medienstroms mit dem Datenattribut von Nicht-Sprachdaten ge-mäß vorkonfigurierten Richtlinieninformatio-nen, wenn das Datenattribut des Medienstroms Nicht-Sprachdaten ist; und, wenn der bestimmte Steuermodus ein Störungssteuermodus ist, An-weisen der Sprachvermittlung, eine Störungs-steuerung gemäß dem Störungssteuermodus an dem aktuellen Kommunikationsprozess aus-zuführen, so dass die Sprachvermittlung eine entsprechende Anweisung zu einem Medien-server überträgt, um den MedienMedien-server anzu-weisen, Schund-Daten, wie etwa Rauschen, in dem aktuell übertragenen Medienstrom hinzu-zufügen und den Medienstrom mit den Schund-Daten zu einem Empfänger zu übertragen, wo-bei der Medienserver zum Weiterleiten von zwi-schen beiden Kommunikationsteilnehmern aus-getauschten Medienströmen verwendet wird.
10. Datenattribut-Analyseeinrichtung nach Anspruch 9,
wobei, wenn der bestimmte Steuermodus ein Unter-brechungs-Steuermodus ist, die Steuermodus-Be-stimmungseinheit ferner ausgelegt ist zum Anwei-sen der Sprachvermittlung, eine Unterbrechungs-steuerung an dem aktuellen Kommunikationspro-zess auszuführen, so dass die Sprachvermittlung die Medienkommunikationsverbindung zwischen beiden Kommunikationsteilnehmern gemäß der An-weisung der Steuermodus-Bestimmungseinheit un-terbricht.
11. Datenattribut-Analyseeinrichtung nach Anspruch 9,
wobei die Detektionseinheit Folgendes umfasst: eine Datenattribut-Bestimmungssubeinheit, ausgelegt zum Bestimmen des Datenattributs des über den Sprachkanal übertragenen Me-dienstroms gemäß einem Menschliche-Stim-me-Detektionsergebnis, wenn das Erkennungs-ergebnis das Menschliche-Stimme-Detektions-ergebnis ist.
12. Datenattribut-Analyseeinrichtung nach einem der
Ansprüche 9 bis 11, wobei die Datenattribut-Analy-seeinrichtung in eine Sprachvermittlung integriert ist.
Revendications
1. Procédé de détection d’attribut de données,
com-prenant les étapes suivantes :
obtenir (105a-105b) un flux de média transmis sur un canal vocal établi par l’intermédiaire d’un commutateur vocal ;
reconnaître (106a) le flux de média obtenu pour obtenir un résultat de reconnaissance ; et déterminer (106b) si le résultat de reconnais-sance est supérieur à un seuil de succès de re-connaissance prédéfini ; si le résultat de recon-naissance se situe au-dessus d’un seuil de suc-cès de reconnaissance prédéfini, déterminer que l’attribut de données du flux de média trans-mis sur le canal vocal correspond à des données vocales ; sinon,
déterminer que l’attribut de données du flux de média transmis sur le canal vocal correspond à des données non vocales ;
si l’attribut de données du flux de média corres-pond à des données non vocales,
déterminer (107) un mode de contrôle corres-pondant pour un processus de communication de transmission du flux de média avec l’attribut de données de données non vocales conformé-ment à des informations de politique préconfigurées ;
si le mode de contrôle déterminé est un mode de contrôle des perturbations, donner pour ins-truction au commutateur vocal d’effectuer un contrôle des perturbations sur le processus de communication en cours conformément au mo-de mo-de contrôle mo-des perturbations, mo-de manière à ce que le commutateur vocal transmette une ins-truction correspondante à un serveur de média pour donner pour instruction au serveur de mé-dia d’ajouter des données inutiles, telles que du bruit, dans le flux de média en cours de trans-mission et transmettre le flux de média avec les données inutiles à un récepteur, où tous les flux
5 10 15 20 25 30 35 40 45 50 55 de média échangés entre les deux parties en communication sont retransmis par l’intermé-diaire du serveur de média.
2. Procédé selon la revendication 1, dans lequel,
lors-que le résultat de reconnaissance est un résultat de détection de voix humaine, l’attribut de données du flux de média transmis sur le canal vocal est déter-miné conformément au résultat de détection de voix humaine.
3. Procédé selon la revendication 2, dans lequel, si le
résultat de détection de voix humaine est déterminé par un profil spectral, la détermination de l’attribut de données du flux de média transmis sur le canal vocal conformément au résultat de détection de voix humaine comprend les étapes suivantes :
comparer un profil spectral du flux de média à un profil spectral de voix humaine prédéfini ; si une erreur de comparaison se situe dans une plage prédéfinie, déterminer l’attribut de don-nées du flux de média comme correspondant à des données vocales ; sinon, déterminer l’attri-but de données du flux de média comme cor-respondant à des données non vocales.
4. Procédé selon l’une quelconque des revendications
1 à 3, comprenant en outre l’étape suivante : si le mode de contrôle déterminé est un mode de contrôle d’interruption, donner pour instruc-tion au commutateur vocal d’effectuer un con-trôle d’interruption sur le processus de commu-nication en cours conformément au mode de contrôle d’interruption, de manière à ce que le commutateur vocal interrompe la connexion de communication multimédia entre les deux par-ties en communication conformément à l’ins-truction.
5. Système de détection d’attribut de données,
comprenant : un serveur de reconnaissance, un dis-positif d’analyse d’attribut de données et un commu-tateur vocal ;
le serveur de reconnaissance est conçu pour obtenir un flux de média transmis sur un canal vocal établi par l’intermédiaire du commutateur vocal, pour re-connaître le flux de média obtenu pour obtenir un résultat de reconnaissance qui est ensuite transmis à un dispositif d’analyse d’attribut de données ; et le dispositif d’analyse d’attribut de données est con-çu pour déterminer si le résultat de reconnaissance est supérieur à un seuil de succès de reconnaissan-ce prédéfini ; si le résultat de reconnaissanreconnaissan-ce se si-tue au-dessus d’un seuil de succès de reconnais-sance prédéfini, déterminer que l’attribut de données du flux de média transmis sur le canal vocal
corres-pond à des données vocales ; sinon, déterminer que l’attribut de données du flux de média transmis sur le canal vocal correspond à des données non vocales ; si l’attribut de données du flux de média correspond à des données non vocales, il est en outre conçu pour déterminer un mode de contrôle correspondant pour un processus de communica-tion de transmission du flux de média avec l’attribut de données de données non vocales conformément à des informations de politique préconfigurées, et si le mode de contrôle déterminé est un mode de con-trôle des perturbations, il est en outre conçu pour donner pour instruction au commutateur vocal d’ef-fectuer un contrôle des perturbations sur le proces-sus de communication en cours conformément au mode de contrôle des perturbations ;
le commutateur vocal est conçu pour établir le canal vocal, et est en outre conçu pour transmettre une instruction correspondante à un serveur de média conformément à une instruction du dispositif d’ana-lyse d’attribut de données pour donner pour instruc-tion au serveur de média d’ajouter des données inu-tiles, telles que du bruit, dans le flux de média en cours de transmission et transmettre le flux de média avec les données inutiles à un récepteur, où tous les flux de média échangés entre les deux parties en communication sont retransmis par l’intermédiaire du serveur de média.
6. Système selon la revendication 5, comprenant en
outre :
un serveur de média conçu pour obtenir un cer-tain nombre de flux de média transmis sur le canal vocal et pour fournir au serveur de recon-naissance les flux de média ;
où le serveur de reconnaissance est spécifique-ment conçu pour recevoir les flux de média du serveur de média et pour reconnaître les flux de média pour obtenir le résultat de reconnaissan-ce.
7. Système selon la revendication 6, dans lequel le
commutateur vocal est conçu pour établir le canal vocal ;
où le commutateur vocal est conçu pour établir un canal de connexion entre les deux parties en com-munication et le serveur de média.
8. Système selon l’une quelconque des revendications
5 à 7, dans lequel, si le mode de contrôle déterminé est un mode de contrôle d’interruption, le dispositif d’analyse d’attribut de données est en outre conçu pour donner pour instruction au commutateur vocal d’effectuer un contrôle d’interruption sur le proces-sus de communication en cours ;
le commutateur vocal est en outre conçu pour inter-rompre la connexion de communication multimédia
5 10 15 20 25 30 35 40 45 50 55 entre les deux parties en communication conformé-ment à une instruction du dispositif d’analyse d’attri-but de données.
9. Dispositif d’analyse d’attribut de données,
comprenant :
une unité d’obtention conçue pour obtenir un ré-sultat de reconnaissance d’un flux de média transmis sur un canal vocal établi par l’intermé-diaire d’un commutateur vocal ;
et
une unité de détection comprenant une sous-unité de comparaison et une sous-sous-unité de dé-tection, où la sous-unité de comparaison est conçue pour comparer le résultat de reconnais-sance à un seuil de succès de reconnaisreconnais-sance prédéfini ; et la sous-unité de détection est con-çue pour déterminer que l’attribut de données du flux de média transmis sur le canal vocal cor-respond à des données vocales lorsqu’un résul-tat de comparaison indique que le résulrésul-tat de reconnaissance se situe au-dessus du seuil de succès de reconnaissance prédéfini, ou pour déterminer que l’attribut de données du flux de média actuellement transmis sur le canal vocal correspond à des données non vocales lors-qu’un résultat de comparaison indique que le résultat de reconnaissance est inférieur ou égal au seuil de succès de reconnaissance prédéfini ;
une unité de détermination de mode de contrôle conçue pour déterminer un mode de contrôle correspondant pour un processus de communi-cation de transmission du flux de média avec l’attribut de données de données non vocales conformément à des informations de politique préconfigurées si l’attribut de données du flux de média correspond à des données non vocales ; et, si le mode de contrôle déterminé est un mode de contrôle des perturbations, don-ner pour instruction au commutateur vocal d’ef-fectuer un contrôle des perturbations sur le pro-cessus de communication en cours conformé-ment au mode de contrôle des perturbations, de manière à ce que le commutateur vocal trans-mette une instruction correspondante à un ser-veur de média pour donner pour instruction au serveur de média d’ajouter des données inuti-les,
telles que du bruit, dans le flux de média en cours de transmission et transmettre le flux de média avec les données inutiles à un récepteur, où le serveur de média est utilisé pour retransmettre des flux de média échangés entre les deux par-ties en communication.
10. Dispositif d’analyse d’attribut de données selon la
revendication 9, dans lequel, si le mode de contrôle déterminé est un mode de contrôle d’interruption, l’unité de détermination de mode de contrôle est en outre conçue pour donner pour instruction au com-mutateur vocal d’effectuer un contrôle d’interruption sur le processus de communication en cours, de ma-nière à ce que le commutateur vocal interrompe la connexion de communication multimédia entre les deux parties en communication conformément à l’instruction de l’unité de détermination de mode de contrôle.
11. Dispositif d’analyse d’attribut de données selon la
revendication 9, dans lequel l’unité de détection comprend :
une sous-unité de détermination d’attribut de données conçue pour déterminer l’attribut de données du flux de média transmis sur le canal vocal conformément à un résultat de détection de voix humaine lorsque le résultat de recon-naissance est le résultat de détection de voix humaine.
12. Dispositif d’analyse d’attribut de données selon l’une
quelconque des revendications 9 à 11, dans lequel le dispositif d’analyse d’attribut de données est inté-gré dans un commutateur vocal.
REFERENCES CITED IN THE DESCRIPTION
This list of references cited by the applicant is for the reader’s convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.
Patent documents cited in the description • US 2004260560 A1 [0005]
Non-patent literature cited in the description • RUNSHENG WANG et al. VoIP Development in
Chi-na. COMPUTER IEEE SERVICE CENTER, 01 Sep-tember 2004, vol. 37, 30-37 [0005]
