WO1985005515A1 - Method for performing speech-detection in scpc systems - Google Patents

Method for performing speech-detection in scpc systems Download PDF

Info

Publication number
WO1985005515A1
WO1985005515A1 PCT/HU1985/000033 HU8500033W WO8505515A1 WO 1985005515 A1 WO1985005515 A1 WO 1985005515A1 HU 8500033 W HU8500033 W HU 8500033W WO 8505515 A1 WO8505515 A1 WO 8505515A1
Authority
WO
WIPO (PCT)
Prior art keywords
threshold
speech
noise
carrier
time
Prior art date
Application number
PCT/HU1985/000033
Other languages
French (fr)
Inventor
Erno^" BÁCS
Mihály GUBÁNYI
Lajos HANZÓ
László Hinsenkamp
László UNERECZKY
Original Assignee
Távközlési Kutató Intézet
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to GB08530834A priority Critical patent/GB2169776B/en
Application filed by Távközlési Kutató Intézet filed Critical Távközlési Kutató Intézet
Priority to GB08601284A priority patent/GB2173078B/en
Publication of WO1985005515A1 publication Critical patent/WO1985005515A1/en

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/14—Relay systems
    • H04B7/15—Active relay systems
    • H04B7/204—Multiple access
    • H04B7/208—Frequency-division multiple access [FDMA]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Definitions

  • the object of the patent is a method for performing speech-detection in SCPC /Single Channel Per Carrier/ satellite-communication systems.
  • the fundamental aim of the speech-detection is to disable the channel-carrier in speech-channels, where no speech is present.
  • speech-detectors operate basically on two different principles, namely either on analog or on digital coded speech-signals.
  • the basic problem of speech-detection in both cases is that a compromise has to be found between contradictory requirements for optimal speech-sensitivity and noise immunity:
  • the speech-detector has to detect speech, even if the speech-level is under the speech-threshold for a short time;
  • PCM- -coded speech-signals it is a known method to observe more, than one consecutive samples within a time interval and, wether they exceed the speech-threshold. This method has the advantage that in case of noise samples it is not likely that more, than one consecutive samples exceed the speech-threshold, and this way the false noise-triggering is decreased.
  • the known speech-detection techniques have the disadvantage that on channels with different, noise-levels the probability of false triggering is also different, as the speech-threshold is a fixed value. In order words, this means that a compromise between convenient speech- and noise-operation has to be found, as they represent conradictory requirements. For the sake of maximal speech- -sensitivity the speech-threshold must be as sensible, as possible, but in this case the probability of false triggering increases.
  • other disadvantage of the known-methods is that if the speech-detector is once activated by noise, then it is switched-off only, if the hangover-time elapsed, and this way valuable satellite-energy is wasted.
  • the aim of the invention is to propose a new speech- -detection method, which eliminates the above disadvantages, gives optical speech-sensitivity and noise-immunity at the same time and supplies a means to switch-off the detector as soon as possible, if once activated by noise, this way saving also a considerable amount of satellite-energy.
  • Our purpose with the invention is to provide a new, fast and reliable speech-detector.
  • the task to be solved by the patent is, as follows: to find a speech-detection method adjusting the speech- -threshold adoptively according to the actual level of the channel-noise. In order to achieve optimal speech- -sensitivity and noise-immunity, moreover to minimize the duration of false carrier-transmission.
  • the main point of the patent is that the received
  • PCM-noise-samples are statistically evaluated in that way that we investigate, which portion of them is greater, and which of them is smaller, than a given-noise-threshod.
  • the noise-threshold is adoptively adjusted in that sense that only a prescribed proportion of the noise-samples exceeds the noise-threshold.
  • the speech- -threshold is chosen to be a bit higher, than the noise- -threshold, this way resulting optimal noise-immunity and speech-sensitivity. Inspite of all these measures it is possible that the speech-detector is activated by noise. But for this case we propose also a security-means, which minimizes the duration of false carrier-transmission.
  • the hangover is set to be shorter, as its usual value.
  • the invented method for detecting speech signals in SCPC satillite communication systems in order to control the transmission of channel carrier samples of PCM coded channel signals are evaluated and the carrier is transmitted on consecutive exceeding of a first threshold during a first time interval till the end of a second time interval.
  • This second time interval is begun when the signal falls below said first threshold which is determined in dependence of the channel noise level.
  • the duration of said second time interval is controlled on hand of the duration of a third time interval, which is the time of consecutive exceeding of said first threshold.
  • Said third time interval is the length of a signal burst, i.e. the time difference between the beginning of said first and second time intervals.
  • Said second time interval is decreased when said third interval is shorter than a prescribed value characteristic tu human speech.
  • the channel noise level is used to determine said first threshold by adding a security difference to a second threshold which is regulated in a way, that the signal would exceed this second threshold at a predetermined range considering a statistically formed ratio of the exceedings while no carrier is transmitted. Said added security difference is decreased with the increase of said second threshold.
  • the invented method is preferably used with digitalized samples of channel signals, because in this case microprocessors can be used.
  • the speech- -detection is based on the evaluation, of PCM-samples in that way that the channel-carrier is enabled in the only case, if MA number of consecutive samples exceed the speech-threshold. If noise is detected, - i.e. the above condition is not met, and the hangover is elapsed - an in double sense adoptive noise- and speech-threshold adjustment process is initiated.
  • this in double sense adoptive process we adjust the noise- as well as the speech- threshold to achieve optimal noise-immunity and speech- -sensitivity, based on the statistical evaluation of the noise-samples.
  • the security-distance between the two thresholds is controled,so as to enhance noise-immunity on high-quality transmission-channels. If the duration of the activating burst is less, than the typical speech-burst-length, i.e. the detector is probably activated by a noise-spike, then the hangover is conveniently set to a lower, value, than in case of speech. This way the duration of false carrier-transmission is minimized.
  • Figure 1 is a flowchart, where according to the usual formalism in flowcharts, a rhombus represents a branching , and a rectangle an operation to be done.
  • the symbols, used in the flowchart are, as follows: INIT - Setting Initial conditions RS - Received signal
  • NBD Noise-burst-duration
  • NTC Counter for noise-samples, exceeding the noise- -threshold NT - Noise-threshold NSC - Noise-samples-counter MNSC - Maximum of NSC NTO - Noise-threshold for high-quality channels
  • the counter C1 is incremented by one, and a new sample is to be processed.
  • the noise-threshold is set so that out of MNSC number of noise- -samples at most MAX number of samples shall be above the noise-threshold NT, this way ensuring that the noise- -threshold is set approximately according to the level of the noise-spikes.
  • the speech - threshold ST is set with a security-distance SD above the noise-threshold NT. If the noise-threshold NT does not fulfil the above conditions, then the value of it is incremented or decremented, till it is not in the prescribed region, and so both thresholds are - in double sense adoptively - set according to the actual channel-noise. In the following, we are going to describe the sub-operations in this adoptive process.
  • the noise-threshold-counter is incremented by one, if the condition RS > NT is fulfiled, i.e. the sample, processed right now is greater, than the noise-threshold NT. If not, then the process goes on, as a function of the condition NSC ⁇ MNSC , without incrementing the noise-threshold-counter NTC. If the sample-counter NSC arrived at its maximum MNSC , or in other words, a noise-threshold-ajustment period elapsed, then in the block 0 ⁇ NSC the sample-counter is cleared, and a new averaging period is initiated.
  • the threshold-distance is set to a greater value, than SD1, as C>1, which means that on high-quality, noise-free channels it is not convenient to allow the speech-threshold ST to be quite low, as in this case the detector could easily be activated by high noise- -spikes, above the low-level average-noise.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Time-Division Multiplex Systems (AREA)
  • Radio Relay Systems (AREA)
  • Noise Elimination (AREA)

Abstract

A method to realize the speech-detection in SCPC (Single Channel Per Carrier) satellite-communication systems, where the speech-detection is based on the evaluation of PCM-coded samples that way that the channel-carrier is enabled to be transmitted, if MA number of consecutive samples exceed the speech-treshold. On detecting noise, i.e. if the above condition is not met, and the hangover-time of the detector elapsed, an in double sense adaptive noise- and speech-treshold adjustment process is initiated. In case the length of the activating sample-burst is less, than the - for speech typical minimal - segment-length, i.e. the detector was probably activated by noise, then the value of the detector-hangover-time is chosen conveniently to be shorter, than in case of speech. By the help of this idea the duration of false carrier-transmission is minimized, and a considerable quantity of valuable satellite-energy is saved.

Description

METHOD FOR PERFORMING SPEECH-DETECTION IN SCPC SYSTEMS
The object of the patent is a method for performing speech-detection in SCPC /Single Channel Per Carrier/ satellite-communication systems.
The fundamental aim of the speech-detection is to disable the channel-carrier in speech-channels, where no speech is present. According to the state of the art, it is known that speech-detectors operate basically on two different principles, namely either on analog or on digital coded speech-signals. The basic problem of speech-detection in both cases is that a compromise has to be found between contradictory requirements for optimal speech-sensitivity and noise immunity:
- the speech-detector has to detect speech, even if the speech-level is under the speech-threshold for a short time;
- on the other hand, it has to minimize the rate of false triggering caused by noise.
In other words, this means that the speech-detector has to detect speech even on noisy channels. It is conveniently realized in a way that a speech-threshold is chosen to be slightly higher, than the typical channel- -noise, but lower, than the weakest speech-level. False triggering on short noise-impulses can be prevented by the help of a so-called switching delay, while the final clipping of speech-bursts, caused by weak final consonants can be prevented by a so-called hangover. In case of PCM- -coded speech-signals it is a known method to observe more, than one consecutive samples within a time interval and, wether they exceed the speech-threshold. This method has the advantage that in case of noise samples it is not likely that more, than one consecutive samples exceed the speech-threshold, and this way the false noise-triggering is decreased.
The known speech-detection techniques have the disadvantage that on channels with different, noise-levels the probability of false triggering is also different, as the speech-threshold is a fixed value. In order words, this means that a compromise between convenient speech- and noise-operation has to be found, as they represent conradictory requirements. For the sake of maximal speech- -sensitivity the speech-threshold must be as sensible, as possible, but in this case the probability of false triggering increases. Ar. other disadvantage of the known-methods is that if the speech-detector is once activated by noise, then it is switched-off only, if the hangover-time elapsed, and this way valuable satellite-energy is wasted.
The aim of the invention is to propose a new speech- -detection method, which eliminates the above disadvantages, gives optical speech-sensitivity and noise-immunity at the same time and supplies a means to switch-off the detector as soon as possible, if once activated by noise, this way saving also a considerable amount of satellite-energy.
Our purpose with the invention is to provide a new, fast and reliable speech-detector.
The task to be solved by the patent is, as follows: to find a speech-detection method adjusting the speech- -threshold adoptively according to the actual level of the channel-noise. In order to achieve optimal speech- -sensitivity and noise-immunity, moreover to minimize the duration of false carrier-transmission. The main point of the patent is that the received
PCM-noise-samples are statistically evaluated in that way that we investigate, which portion of them is greater, and which of them is smaller, than a given-noise-threshod. After all this the noise-threshold is adoptively adjusted in that sense that only a prescribed proportion of the noise-samples exceeds the noise-threshold. Now the speech- -threshold is chosen to be a bit higher, than the noise- -threshold, this way resulting optimal noise-immunity and speech-sensitivity. Inspite of all these measures it is possible that the speech-detector is activated by noise. But for this case we propose also a security-means, which minimizes the duration of false carrier-transmission. If the number of samples, exceeding the speech-threshold - i.e. the duration of the noise-spike, activating the detector - does not exceed a limited length, which is surely shorter, than the typical speech-burst-length, then the hangover is set to be shorter, as its usual value.
According to the invented method for detecting speech signals in SCPC satillite communication systems in order to control the transmission of channel carrier samples of PCM coded channel signals are evaluated and the carrier is transmitted on consecutive exceeding of a first threshold during a first time interval till the end of a second time interval. This second time interval is begun when the signal falls below said first threshold which is determined in dependence of the channel noise level. The duration of said second time interval is controlled on hand of the duration of a third time interval, which is the time of consecutive exceeding of said first threshold. Said third time interval is the length of a signal burst, i.e. the time difference between the beginning of said first and second time intervals. Said second time interval is decreased when said third interval is shorter than a prescribed value characteristic tu human speech. The channel noise level is used to determine said first threshold by adding a security difference to a second threshold which is regulated in a way, that the signal would exceed this second threshold at a predetermined range considering a statistically formed ratio of the exceedings while no carrier is transmitted. Said added security difference is decreased with the increase of said second threshold. The invented method is preferably used with digitalized samples of channel signals, because in this case microprocessors can be used.
While implementing the invented method the speech- -detection is based on the evaluation, of PCM-samples in that way that the channel-carrier is enabled in the only case, if MA number of consecutive samples exceed the speech-threshold. If noise is detected, - i.e. the above condition is not met, and the hangover is elapsed - an in double sense adoptive noise- and speech-threshold adjustment process is initiated. By the help of this in double sense adoptive process we adjust the noise- as well as the speech- threshold to achieve optimal noise-immunity and speech- -sensitivity, based on the statistical evaluation of the noise-samples. On the other hand, according to the actual value of the noise-threshold, also the security-distance between the two thresholds is controled,so as to enhance noise-immunity on high-quality transmission-channels. If the duration of the activating burst is less, than the typical speech-burst-length, i.e. the detector is probably activated by a noise-spike, then the hangover is conveniently set to a lower, value, than in case of speech. This way the duration of false carrier-transmission is minimized.
The detailed description of carrying out the invented method is given on the drawing, wherein
Figure 1 is a flowchart, where according to the usual formalism in flowcharts, a rhombus represents a branching , and a rectangle an operation to be done. The symbols, used in the flowchart are, as follows: INIT - Setting Initial conditions RS - Received signal
ST - Speech-threshold /first threshold/ RSC - Received-samples-counter MA - Maximum of RSC /corresponds to first time interval/ HD - Hangover-duration /second time interval/ SHD - The small value of HD GHD - The great value of HD C1 - Counter
NBD - Noise-burst-duration NTC - Counter for noise-samples, exceeding the noise- -threshold NT - Noise-threshold NSC - Noise-samples-counter MNSC - Maximum of NSC NTO - Noise-threshold for high-quality channels
SD - Security-distance between the noise- and speech-
-threshold SD1 - Value of SD on average-quality channels C - Weighting-factor on high-quality channels MAX - Maximum of NTC MIN - Minimum of NTC
In the block INIT the initial conditons of the variables, used are set. If the condition RS > ST is met, - i.e. the received signal is greater, than the speech- -threshold - then the flow-chart is branched according to the condition RSC = MA, which means that the received- -signal-counter is at its maximum, i.e. we received MA number of consecutive samples, exceeding the speech-threshold, in which case the carrier is enabled, as speech is supposed to be present. If we did not receive up to now MA number of consecutive samples, exceeding the speech-threshold, then the counter RSC is incremented by one, and the two conditions mentioned before are investigated again. After enabling the carrier, we have to look, if the signal is "continuously" above the speech-threshold or only for a short time. If namely the number of consecutive samples, exceeding the speech-threshold ST - which is counted by the counter C1 - is less, than the noise-burt-duration NBD, then it is supposed that a false noise-triggering occured, i.e. false carrier-transmission was initiated. In this case the hangover HD is set in the block HD = SKD to its lower value out of the two possible values SHD and GHD, this way minimizing the duration of false-triggering. Now in the block C1 INCR the counter C1 is incremented by one, and a new sample is to be processed. But if the condition C1 > NBD is met, in the block HD = GHD the hangover is set to its higher value GHD, as it is supposed now that the detector was activated actually by speech. Now also a new sample is to be processed. The above operations describe the case when the condition RS > ST is fulfiled, i.e. the detector detects speech.
But if not so, i.e. no speech is detected, then in the block O → RSC the counter RSC is cleared, and in the block HD ELAPSED there must be checked, whether the hangover HD is elapsed. If so, and no new sample, exceeding the speech -threshold ST has been received, then noise is supposed to be present, and according to this in the block CARR . DISABLE the carrier is disabled, and the - in double sense adoptiv threshold-adjustment is initiated. In this process the noise-threshold is set so that out of MNSC number of noise- -samples at most MAX number of samples shall be above the noise-threshold NT, this way ensuring that the noise- -threshold is set approximately according to the level of the noise-spikes. Now the speech - threshold ST is set with a security-distance SD above the noise-threshold NT. If the noise-threshold NT does not fulfil the above conditions, then the value of it is incremented or decremented, till it is not in the prescribed region, and so both thresholds are - in double sense adoptively - set according to the actual channel-noise. In the following, we are going to describe the sub-operations in this adoptive process. After running through the block CARR. DISABLE the noise-threshold-counter is incremented by one, if the condition RS > NT is fulfiled, i.e. the sample, processed right now is greater, than the noise-threshold NT. If not, then the process goes on, as a function of the condition NSC < MNSC , without incrementing the noise-threshold-counter NTC. If the sample-counter NSC arrived at its maximum MNSC , or in other words, a noise-threshold-ajustment period elapsed, then in the block 0 → NSC the sample-counter is cleared, and a new averaging period is initiated. Now in the block NT < NTO is decided upon, if the noise-threshold NT is smaller, than the value NTO, which is characterizing high-quality channel-circuits, and if not, then the threshold-difference /security-distance/ SD - i.e. the distance between the noise-threshold NT and the speech-threshold ST - is set in the block SD = SD1 to the value SD1. On the other hand, if the noise-threshold NT is smaller, than the value NTO, then in the block SD = C.SD1 the threshold-distance is set to a greater value, than SD1, as C>1, which means that on high-quality, noise-free channels it is not convenient to allow the speech-threshold ST to be quite low, as in this case the detector could easily be activated by high noise- -spikes, above the low-level average-noise. After this, in the block NTC>MAX it is cheked, if the number of samples, exceeding the noise-threshold NT is greater, than MAX, and if so, then NT has to be incremented by one step in the block NT = NT + 1, and in the block ST = NT + SD the new speech-threshold is computed according to the actual threshold-difference SD. If the condition NTC > MAX is not fulfiled, we look in the block NTC < MIN , whether the number of samples, exceeding the noise-threshold NT is less, than MIN, and if the content of the noise-threshold-counter is less, than this minimum, then the value of the noise-threshold is decremented by one step, and in the block ST = NT + SD the new speech-threshold is computed. If the noise-threshold-counter content is in the prescribed interval MIN < NTC < MAX, then the noise-threshold does not have to be corrected, but after clearing the counter NTC in the block 0 > NTC, the whole process has to be reinitiated. Tracing the flowchart from the block NSC < MNSC along the "YES" branch, we have simply to increment the sample-counter NSC in the block NSC INCR. and also reinitiate the process. The fundamental advantage of the new method is that due to the - in double sense adoptive - threshold - adjustment the jointly optimized noise- and speech-thresholds are always set according to the actual channel-noise-level, so as to minimize the false triggering on noise, this way saving a considerable quantity of satellite-energy, and thanks to the idea of the adoptive threshold-difference, optimal speech-sensitivity and noise-immunity is achived. Furthermore, it is quite advantageous that in case the detector was - inspite of our efforts - activated by noise, and the noise-impulse duration does not exceed the typical talk-spurtlength, the speech and noise can be distinguished on this basis, and in this case the hangover is forced to a lower value, than the for speech typical value, this way further decreasing the duration of false carrier- -transmission. Also an important advantage of the proposed method is that due to the proper task-segmentation and flowchart-optimization, it is suitable for cost-effective, real-time microprocessor-implementation. It is namely plausible from the flowchart that most of the operations are to be done in that case, when no speech, but noise is present, and at this time the processor has no other task, and so it is able to accomplish the adoptive threshold adjustment. Beside all these advantages we must not forget to mention the conventiently short activation-time of the detector, which results in minimal initial clipping of speech-bursts.

Claims

C l a i m s
1. A method for detecting .speech signals in SCPC satellite communication systems, in order to control the transmission of channel carrier wherein the detection is carried out on evaluating PCM coded channel signals and the carrier is transmitted on continuous excession of a first threshold /ST/ during a first time interval /t1/ until the end of a second time interval /t2/ initiated at the time, when the signal falls below said first threshold /ST/, furtheron said first threshold /ST/ is determined as a function of the channel noise level in a way that during the time when no carrier is transmitted the excession ratio of a second threshold /NT/ is statistically defined and said threshold /NT/ is regulated so as to maintain this ratio in a predetermined range, furtheron said first threshold /ST/ is determined as a sum of said second threshold /NT/ and a security difference /SD/ the value of which is decreased with the increase of said second threshold /NT/, furtheron the duration of said second time interval /t2/ is controlled on hand of the duration of the third time inverval /t3 / when the signal consecutively exceeds said first threshold /ST/ wherein said second time interval /t2/ is decreased if said third interval / t3 / is shorther than a prescribed value.
2. A method as claimed in claim 1, wherein the speech-detection is based on the evaluation of PCM-coded signals of a PCM-channel in that way, that the channel- -carrier is transmitted, if a predetermined number /MA/ /t1/ of consecutive samples exceed the speech-threshold /ST/, and if noise is detected, i.e. if the condition of MA number of exceeding consecutive samples is not fulfiled, and the detector-hangover time /HD/ /t2/ elapsed, an in double sense adoptive noise- and speech-threshold adjustment process is initiated, by the help of which on one hand the noise-threshold and the - by a variable threshold-difference shifted - speech-threshold are set according to the statistical average-level, which is computed on hand of the accumulated noise-samples to achieve maximal noise- -immunity and at the same time optimal speech-sensitivity, on the other hand as a function of the actual noise-threshold /NT/ the threshold-difference /SD/ between the noise- -threshold /NT/ and the speech-threshold /ST/ is varied, so as to enhance the noise-immunity in case of channels; furthermore, if the length of the activating sample-burst is less, than the for speech typical burstlength, i.e. the detector was probably activated by noise, then the detector hangover time /HD/ is conveniently chosen to be shorter - HD = SHD -, than the for speech typical value - - HD = GHD -, this way minimizing the duration of false carrier-transmission.
PCT/HU1985/000033 1984-05-24 1985-05-24 Method for performing speech-detection in scpc systems WO1985005515A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
GB08530834A GB2169776B (en) 1984-05-24 1985-04-26 A method and circuit arrangement for implementing pcm channel units of scps satellite-communications systems
GB08601284A GB2173078B (en) 1984-05-24 1985-05-24 Method for performing speech-detection in scpc systems

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
HU2006/84 1984-05-24
HU842006A HU190304B (en) 1984-05-24 1984-05-24 Method for implementing speech detection of satellite communication systems of individual speech channel with fix or random marking-out of the carrier

Publications (1)

Publication Number Publication Date
WO1985005515A1 true WO1985005515A1 (en) 1985-12-05

Family

ID=10957259

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/HU1985/000033 WO1985005515A1 (en) 1984-05-24 1985-05-24 Method for performing speech-detection in scpc systems

Country Status (5)

Country Link
JP (1) JPS61502301A (en)
DE (1) DE3590220T1 (en)
GB (2) GB2169776B (en)
HU (1) HU190304B (en)
WO (1) WO1985005515A1 (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE1199805B (en) * 1963-12-09 1965-09-02 Siemens Ag Method and arrangement for secure data transmission via telephone systems
DE1803080A1 (en) * 1968-10-15 1970-06-04 Telefunken Patent Noise evaluation
SU522546A1 (en) * 1975-04-07 1976-07-25 Предприятие П/Я Р-6292 Device for extracting a signal affected by interference
SU640433A1 (en) * 1976-12-27 1978-12-30 Предприятие П/Я А-7306 Apparatus for transmitting and receiving pulse-modulated signals
SU756660A1 (en) * 1978-04-17 1980-08-15 Anatolij A Gladkikh Device for receiving information with noise suppression
US4455662A (en) * 1983-05-25 1984-06-19 Gutleber Frank S Noise code generator and pulse compression communications system
SU1115244A1 (en) * 1983-05-23 1984-09-23 Предприятие П/Я Г-4126 Device for compensating interference

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3334766A1 (en) * 1983-09-26 1985-04-11 Siemens AG, 1000 Berlin und 8000 München MULTIPROCESSOR CENTRAL CONTROL UNIT OF A SWITCHING SYSTEM, IN PARTICULAR TELEPHONE SWITCHING SYSTEM
SE439208B (en) * 1983-09-30 1985-06-03 Ericsson Telefon Ab L M PROGRAMMED CONTROL TELECOMMUNICATIONS

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE1199805B (en) * 1963-12-09 1965-09-02 Siemens Ag Method and arrangement for secure data transmission via telephone systems
DE1803080A1 (en) * 1968-10-15 1970-06-04 Telefunken Patent Noise evaluation
SU522546A1 (en) * 1975-04-07 1976-07-25 Предприятие П/Я Р-6292 Device for extracting a signal affected by interference
SU640433A1 (en) * 1976-12-27 1978-12-30 Предприятие П/Я А-7306 Apparatus for transmitting and receiving pulse-modulated signals
SU756660A1 (en) * 1978-04-17 1980-08-15 Anatolij A Gladkikh Device for receiving information with noise suppression
SU1115244A1 (en) * 1983-05-23 1984-09-23 Предприятие П/Я Г-4126 Device for compensating interference
US4455662A (en) * 1983-05-25 1984-06-19 Gutleber Frank S Noise code generator and pulse compression communications system

Also Published As

Publication number Publication date
JPS61502301A (en) 1986-10-09
GB2173078A (en) 1986-10-01
GB8601284D0 (en) 1986-02-26
GB2173078B (en) 1988-03-23
GB2169776A (en) 1986-07-16
GB8530834D0 (en) 1986-01-22
HU190304B (en) 1986-08-28
HUT37314A (en) 1985-11-28
DE3590220T1 (en) 1986-05-15
GB2169776B (en) 1988-09-01

Similar Documents

Publication Publication Date Title
US4052568A (en) Digital voice switch
HK1029701A1 (en) Method and apparatus for adjacent service area handoff in communication systems
US4028496A (en) Digital speech detector
EP0930719A3 (en) Impulse noise detector and noise reduction system
US4357491A (en) Method of and apparatus for detecting speech in a voice channel signal
US6735317B2 (en) Hearing aid, and a method and a signal processor for processing a hearing aid input signal
CA2209215A1 (en) Transmission rate judging unit
CA2035167A1 (en) Antenna selection diversity reception apparatus
EP0405839B1 (en) Speech detector with improved line-fault immunity
EP0654913A3 (en) Synchronization technique for pseudonoise signals
NO974393D0 (en) Procedure for detection of frame quality, as well as receiver
CA2341577A1 (en) Data rate detection device and method for a mobile communication system
US4975657A (en) Speech detector for automatic level control systems
WO1985005515A1 (en) Method for performing speech-detection in scpc systems
EP0831615A3 (en) Receiver allowing adjustment of error thresholds based upon message type and content
EP0692894A3 (en) Optimum signal decision apparatus for digital signal receiver
CA2422355A1 (en) Adaptive control of the detection threshold of a binary integrator
AU1233501A (en) Selective joint demodulation systems and methods for receiving a signal in the presence of noise and interference
US5771293A (en) Switching arrangement for mobile radio receivers
EP0574117B1 (en) Discriminating and suppressing incoming signal noise
WO2005027354B1 (en) Method and apparatus for transmit power control during a group call to a plurality of devices
CA1147071A (en) Method of and apparatus for detecting speech in a voice channel signal
CA2025370A1 (en) Transversal type automatic equalizer with tap coefficient protection
EP0148879B1 (en) Squelch circuit
US6516068B1 (en) Microphone expander

Legal Events

Date Code Title Description
AK Designated states

Designated state(s): DE GB JP SU

RET De translation (de og part 6b)

Ref document number: 3590220

Country of ref document: DE

Date of ref document: 19860515

WWE Wipo information: entry into national phase

Ref document number: 3590220

Country of ref document: DE