Data processing: speech signal processing – linguistics – language – Speech signal processing – For storage or transmission
Reexamination Certificate
1998-09-10
2001-01-30
Zele, Krista (Department: 2748)
Data processing: speech signal processing, linguistics, language
Speech signal processing
For storage or transmission
Reexamination Certificate
active
06182032
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of The Invention
The present invention relates to a communication system comprising a network, and a plurality of terminals, the network and the terminals comprising multi-rate speech encoding and decoding means. Such a communication system can be a cellular or cordless telephony system, a wired system such as a public switched telecommunications network (PSTN) or an Internet, a mixed wired/wireless system, or any other suitable system with terminals having speech encoding and decoding means.
The present invention further relates to a terminal for use in such a communication system.
2. Description of the Related Art
A communication system of the above kind is known from the article “A Multi-rate Transcoder”, A. Lovrich et al., IEEE Transactions on Consumer Electronics, Vol. 35, No. 4, November 1989, pp.
716-722
. In this article, a multi-rate speech transcoder is described for speech encoding of 64 kbit/s PCM (Pulse Code Modulation) data or decoding encoded speech to 64kbit/s PCM data. The described multi-rate transcoder is a DSP (Digital Signal Processor) programmed to support three coding rates, 64 kbit/s PCM, 32 kbit/s ADPCM (Adaptive Differential Pulse Code Modulation), and 16 kbit/s SBC (Subband Coding). On page 721, various telecommunication applications of this transcoder are described, such as mobile radio telephony and voice mail systems in which the recipient can be another PC (Personal Computer) on a LAN (Local Area Network). For mobile radio, the said article advises to use a speech transcoder with a bit rate of 16 kbit/s, and for voice mail it is advised to use a 32 kbit/s ADPCM or a 16 kbit/s SBC transcoder. In a communication system such as a mobile radio system, the usual service is a real time bi-directional communication voice connection, also called a full duplex voice communication link. To such a voice communication link, a typical voice transmission problem is related to an echo caused by several network elements, but pre-dominantly a terminal at the other end of the communication link, the so-called B-end-terminal. The said B-end is coupling back the voice signal arriving at its receiver via two mechanisms, namely intentional electronically implemented coupling, and via acoustic connection from its earpiece to the microphone of the transmitter. This voice coupling becomes then transported back to the point of origin of the voice, in case the so-called A-end-terminal. This coupling, called as loop-back is to give the impression to the subscriber at the A-end of having a “live voice connection” with the B-end. Furthermore, there are local acoustic loop backs at the A-end and the B-end to give the A-end and B-end subscribers the impression of a “live connection” if the subscriber at the other end is silent.
In digital mobile radio systems like GSM (Global System for Mobile Communications), DTX (Discontinuous Transmission) is applied when the B-end is silent. Before actually disconnecting the transmission at the B-end, first an estimated background noise at the B-end is transmitted to the A-end. Then, at the A-end so-called comfort noise is generated to give the A-end subscriber the impression of a still “live link”. Because of the above acoustic coupling paths, in a digital mobile radio such as GSM having a relatively long round trip signal delay between the A-end and B-end terminal, caused by encoding and decoding delays and transmission delays, without echo cancelling, the voice signal of the A-end subscriber would be strongly repeated in the earpiece of the A-end terminal. In a GSM system, the round trip signal delay would typically be 200 msec. Without echo cancelling, such a delayed echo of own voice would be experienced as extremely annoying. For this reason, in such systems echo cancellers are applied. If the round trip signal delay would become too large, it would be virtually impossible to design an echo canceller within the boundaries of a given specification. Particularly, a too long encoding and decoding delay would be problematic. At present, particularly mobile radio systems offer a still increasing number of services, including voice mail. Such a service is a one-way speech or voice service.
Furthermore, the GSM system explained above has the capability of deploying the radio resources in a flexible way, in particular when using its slow frequency hopping mode. Therein the data burst of one user is transmitted as a sequence of shorter bursts spread over a random sequence of frequences. If one or even several of these random frequencies would be simultaneously used by other users as interferers, despite of system procedures to minimise such an interference, the channel coding is able to recover the data of interfered bursts. Accordingly, even short instantaneous pauses in any radio transmission direction, change-over from full frames to transmitting shorter half frames, or the switch-off of an entire transmission direction will reduce, especially in high load conditions, the radio interference between users and, accordingly, increase the number and/or quality of communications which the system can support within a radio environment.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a communication system of the above kind, particularly a mobile radio system, offering various kinds of speech services, in which an optimal use is made of system resources.
To this end the communication system according to the present invention is characterised in that at least one of the terminals comprises switching means for switching over from acoustic voice paths to at least one non-acoustic path, the system being arranged for establishing whether a connection of a terminal to another terminal is a one-way voice communication link, and for causing the switching means to switch over to the at least one non-acoustic path and the multi-rate speech encoding and decoding means to adopt a lower voice coding rate than used for a two-way voice communication link, if it is established that the connection is a one-way voice communication link.
The present is based upon the insight that there is no acoustic or electronic noise or voice loop back if the B-end terminal is a non-acoustically coupled device and thus neither would need a loop back for the voice signal, nor any echo cancelling. It is realised that under such circumstances a longer one-way signal delay can be accepted. It is further realised possible to modify the encoding and decoding delays of the speech codec as the contribution to the one-way signal delay, and that encoding and decoding delay relates to the bit rate of the speech codec. Generally, for a typical conversation voice signal and a particular codec type, a longer encoding and decoding delay enables coding into and from a lower bit rate.
The advantages of the present advantages are thus a less demand for radio resources and immediate radio transmission, resulting in a better and more efficient use of the available system resources such as radio resources such as radio transmission capacity. Accordingly and herewith, more voice links can be accommodated to a given frequency band, resulting in a reduction of operational costs for the system operator and/or subscriber. Modern communication systems, such as GSM, are equipped to use even those radio resources that remain otherwise unused in one direction or instantaneously remain unused in one direction. For a cheaper resource, the operator could charge lower costs to the subscriber. In addition to voice mail according to the present invention, other services such as voice messaging, and voice recognition based services can benefit from this invention.
Embodiments are claimed in the dependent claims. If the systems knows a priori that the communication link is a one-way communication link, then the switching between coding rates and adjustment of side tones can be controlled via signalling. Herewith, a very robust system is obtained. A terminal could be programmed to inform the system it desires a one-way voice communication, and the sys
Opsasnick Michael N.
U.S. Philips Corporation
Zele Krista
LandOfFree
Terminal switching to a lower speech codec rate when in a... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Terminal switching to a lower speech codec rate when in a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Terminal switching to a lower speech codec rate when in a... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2474377