MXPA00001395A - Method for the transmission of speech inactivity with reduced power in a tdma system - Google Patents

Abstract

A method is disclosed for reducing the power during discontinuous transmission in a digital cellular system. In order to reduce battery usage and decrease interference to other users, a method is disclosed which employs the transmission of a silence descriptor frame and the following reduction of power once the frame is properly received in the receiver. While the power is reduced there is still the ability to receive and decode inactivity frames.

Description

METHOD FOR THE TRANSMISSION OF VOICE INACTIVITY WITH REDUCED POWER IN A TDMA SYSTEM. BACKGROUND The present invention relates generally to radio communication systems and, more particularly, to transmission during periods of voice inactivity. Many communication systems apply discontinuous transmission techniques (DTX) to turn off mobile station transmitters (MSs) during periods of silence, for example: conversation breaks. This saves power and reduces the interference caused to other users who are transmitting concurrently on the air interface. More specifically, however, systems that use DTX techniques do not interrupt transmission completely during periods of inactivity. In order to allow realistic reproduction in the receiver of background noise in the transmitter, silence descriptor frames (SID) are transmitted from the transmitter to the receiver during periods of silence. In conventional Full-Speed Enhanced (EFR) GSM systems, the SID frames are synchronous and aligned with the Associated Slow Control Channel (SACCH) frame structure. These systems are described, for example, in the following standards: "ETSI TC-SMG;" digital cellular telecommunications system (2D Phase); Coded channels ", GSM 05.03, version 5.2.1, EISL, November 1996, ETSI TC-SMG:" Digital cellular telecommunication system; substitution and silence of lost frames for voice traffic channels of Total Speed Increased (EFR) ", GSM 06.61, final version 5.1.1, ETSI, November 1996 and ETSI TC-SMG;" Digital telecommunications systems: Discontinuous transmission (DTX) ) for channels of Increased Total Speed (EFR) ", GSM 06.81, final version 5.1., ETSI, November 1996. In these conventional DTX systems there is a SID frame transmitted every 24 TDMA frames (this being a SID frame every 480 ms) During the remaining period, the MS transmitter is turned off, unless speech activity is resumed.For the transmission of SID frames, the same channel code as for the regular speech frame is used. three bits in each SID frame for noise comfort parameters which describe the spectral shape and gain of the signal to be reproduced in the receiver, ninety-five bits are used for a special SID bit pattern to identify r the box as a SID box and to distinguish it from regular voice boxes. He DTX basic principle as described in the GSM Enhanced Full Range standard is indicated in Figure 1. In this, the transmission of active voice frames ends in IT with the transmission of the first SID box which indicates the beginning of the period of silence and transports the noise comfort parameters. When the silence period continues, update SID boxes are transmitted every 24 frames, for example, in T2 and T3. The period of silence ends in T4 with the resumption of the transmission of the voice box. There are several disadvantages associated with these discontinuous transmission systems. First, the transmitted SID frames are transmitted with high relative power and produce strong interference for other users who are transmitting at the same time. Second, except for the transmitted SID frames, the transmission on this channel is otherwise completely stopped. This last characteristic of conventional DTX systems is particularly problematic for recent system innovations of communication systems, such as Adaptive Multiple Rate (AMR) for the GSM system. Here it is necessary to monitor the current channel state to have a more or less continuous internal band control channel, which require more or less continuous reception of transmitted frames. By Consequently, it would be desirable to find a solution with the benefits of battery conservation of conventional discontinuous transmissions in one of the bands and, on the other hand, to provide continuous (or more frequent) frame transmissions to satisfy the requirements of the innovations of new communication systems. Additionally, said solution must also eliminate the aforementioned interference pulses associated with the transmission of conventional DTX SID frames. COMPENDIUM OF THE INVENTION The disadvantages and limitations of conventional communication systems that implement DTX techniques to reduce battery discharge in mobile units can be overcome in accordance with the exemplary embodiments of the present invention which provide continuous and discontinuous transmission of information. SID during periods of inactivity to allow channel monitoring and transmission of control data in the band to continue during those periods. At the same time, however, SID information is transmitted at a reduced power level in relation to voice and data transmissions, which is made possible by the use of more powerful error correction techniques for SID information. Since the SID information uses less bits per frame to transmit than voice and data information, it is possible to use additional bits and therefore reduce the level of transmit power. In addition, the interference associated with powerful SID pulses used in conventional DTX techniques is avoided. The exemplary embodiments of the present invention transmit SID information continuously during the period of inactivity at a reduced power level. However, alternative exemplary embodiments of the present invention may transmit less than continuously, that is, for periods of time less than or equal to the individual SID pulses used in conventional DTX techniques. BRIEF DESCRIPTION OF THE DRAWINGS These and other objects, features and advantages of the present invention will be more readily apparent to those skilled in the art with reading the following detailed description in conjunction with the drawings in which: Figure 1 shows a schematic of Conventional DTX transmissions for a system that operates in accordance with the GSM Standard. Figures 2a, 2b and 2c show decoding schemes of SID frame channels according to modalities of example of the present invention. Figure 3 is a flow chart of an exemplary frame algorithm received in accordance with exemplary embodiments of the present invention. Figure 4 presents a traffic transmission scheme according to example modalities of the present invention. Figure 5 shows a frame processing scheme according to the present invention applied to a GSM system. DETAILED DESCRIPTION While the present invention will be described in connection with various exemplary embodiments, it will be understood by those skilled in the art that the invention is not intended to be limited by these particular examples. On the contrary, the present invention is intended to cover all alternatives, modifications and equivalent arrangements within the spirit and scope of the invention. For example: while the example modalities described below are based on the GSM standard, the present invention can be applied to any other TDMA-based system, for example: DAMPS, PDC or NMTS. In addition, the present invention is not limited to access methodology which include TDMA, but can also be applied to multiple division frequency access (FDMA), code division multiple access (CDMA) or hybrids thereof. As mentioned above, an objective of the present invention is to provide continuous transmission (or more frequent or at equal speed) but at reduced power during periods without voice exploiting the fact that the number of bits of information per frame to be transmitted for SID information It is much smaller than what is associated with voice charts. This allows additional redundancy (this being, more bits of code) that can be used which provides a large amount of protection against error. The resulting gain in protection against error allows transmission at a reduced power level satisfying the requirements concerning the robustness of the channel error. The present invention can be implemented in different ways. A first example embodiment uses the full available bit rate to transmit encoded SID information to provide continuous transmission but at reduced power during periods without speech. A second example mode, as described below, it also reduces the transmission power of frames without voice but does not use all the raw bit rate to transmit SID information. For example modalities employing continuous transmission, the following example parameters are used as a basis to describe these example modalities. First, noise comfort parameters are encoded with n bits, for example: with n = 43 bits in systems that operate according to GSM that have discontinuous transmission systems. Second, the noise comfort parameters are updated at a speed of one SID frame per p transmission frames. This means that, except for the first SID period after the voice and the last SID period before the voice, each SID period has a duration of p frames. An appropriate selection in the example solution is p = 12 frames. Finally, each channel box consists of g bits, example: g = 456 bits in the case of a GSM frame. In these example modalities, different channel coding schemes are used depending on the type of frame (voice or SID) being transmitted. Voice frames are conventionally channels encoded using existing methods, example: according to the standard GSM. However, the SID information is more protected and, therefore, uses a different channel coding scheme. Therefore, the channel coding of the SID information involves diagramming the SID information bits belonging to a SID period in raw bits available in such a way that the SID information is more adequately protected against channel errors. For example: the n information bits (example: 43) of a SID period can be encoded with up to p channel boxes (example: 12) of g bits each (example: 456, being this, with a maximum of p times g (example: 5472) channel bits The implementation of such a highly protective channel encoder can, however, be extremely complex and unnecessary to achieve the desired power level reduction.Therefore, another example embodiment involves encoding n bits of SID information using any suitable conventional channel coding method, example: convolutional encoding, so that they fit into a single channel box This single channel box is sent repeatedly during the duration of the SID period. Only one can be used for convolutionally encoded SID information and bits Remaining bits of the channel box can be used for a special SID bit pattern which marks the box as a SID box. Optionally, these remaining bits can also be used for other purposes, for example: for the inner band channel. The task of the channel decoder is to identify the received frame as a speech frame or a SID frame and, depending on that result, to apply the appropriate method of channel decoding. First, examples of decoding will be described, then an example identification algorithm. Voice frames are decoded by channel by any suitable conventional method, eg: in accordance with the GSM standard. The SID frame channel decoder can be implemented using any suitable conventional channel decader which allows decoding of unique channel frames, eg: decoded Viterbi, but modified to take advantage of the repetition of transmission of SID frames. Example solutions are illustrated in Figures 2 (a). 2 C) . Specifically, as shown in Figure 2 (a), an average of soft information pertaining to the input channel frames to channel box decoding is a possible solution.

Second, as shown in Figure 2 (a), the individual decoding (soft output) of the input and decision channel frames of most of the decoded net SID bits can be carried out. The majority decision is made by averaging the net bit output of the channel frame decoder. Finally, as shown in Figure 2 ©, the individual decoding of the channel input boxes and the selection of the frame with fewer errors can be used to decode frames. The number of gross bit errors is an estimate for example: re-encode the decoded sequence and calculate the Hamming distance to the raw bits of the received channel frame. To identify an input box either as a speech box or as a SID box, a test decoding method can be used. When the input frame is severely corrupted by channel errors such that it can not be decoded either as a speech box or as a SID frame, the input frames can also be classified as bad, this being useless frames. The decision on the type of frame made in the decoder is based on a number of different frame characteristics determined whose characteristic can be derived in several stages of the receiver, being this, in the channel decoder or in the equalizer. Examples of suitable characteristics for channel frame decoders are the following: There are indicators of bad frames of cyclic redundancy check (CRC), example: according to the GSM standard. Second, estimates of the number of gross bit errors in the channel box, referred to as SFQ-SP, can also be used as frame characteristics. Said frame characteristics can be determined by re-coding the coded frame and calculating the Hamming distance to the raw bits of received channel frame. When used, the non-existence of the SID bit pattern is also indicative of the occurrence of voice frames. Examples of suitable characteristics for SID channel frame decoders are as follows: First, an estimate of the number of gross bit errors in the channel box can be carried out (when the channel box decoder is applied SID), referred to as SFQ-SID. Again, this characteristic is determined by re-encoding the decoded SID bits and calculating the Hamming distance to the raw bits of the received channel frame.

Second, determining the Hamming distance to the special SID bit pattern (if any) used to make SID frames. These characteristics are used in the identification scheme described below with respect to the flow chart of Figure 3. First, a table is considered to be a voice chart if CRC indicates a good speech box (step 10), and SFQ-SP is below the SFQ-SP-THR limit (step 20), and (optionally) the Hamming distance to the SID bit pattern HAMM_SID is on or equal to an SFQ_SID_THR limit (step 40). Otherwise it is considered to be a SID box if SFQ_SID is below the SFQ_SID_THR limit (step 50). Otherwise, the painting is considered a bad picture. The type of frame detected is indicated by the markers SPEECH, S0FT_SID or BAD_FRAME, respectively. Only one of the markers is active at any given time. The omission of the optional condition concerning the SFQ_SID measure to decide on a speech frame makes it possible to operate the SID frame decoder only if an update of the SID information is required. This can serve to reduce the complexity of computing. Note: because the SID frames are transmitted with less power in relation to frames of voice, alternative measures can also be used for identification of type of frame. Suitable frame characteristics such as transmission power estimates may for example be taken from the equalizer or used to identify a frame received as a speech frame or an SID frame. For each frame identified as a bad frame, the decoded net bit frame must contain the decoded sequence of the speech channel decoder. This can improve the operation of the hide error function in the speech decoder. This example embodiment also includes implementations where a discontinuous transmission handler is located, for example, near the voice coder and a radio subsystem transmission (RSS) is located near a channel coder and the radio transmitter. It is further assumed that, as in the GSM standard, the discontinuous transmission driver (TX DTX) continuously passes traffic frames (this being, uncoded information frames) to the transmission radio subsystem (TX RSS). These boxes marked with a marker (marker SP) which indicates the type of box being this, voice or SID.

This example mode handles the processing of frames in radio transmission subsystems (TX RSS) which are marked as SID boxes, since said processing is different from conventional SID techniques for the reasons indicated above. Voice frames, on the other hand, are conventionally processed, for example: in accordance with the GSM standard. For the transmission of SID frames, a periodic marker is defined as the Silence Update Marker (SUF). This marker is set once in the last frame of each SID period. The SUF is also available at the end of the receiver. To ensure this functionality, the SUF can, for example in a GSM system, be aligned with the SACCH frame structure or can, for example, be synchronized during call setup. This example SID frame process procedure includes the following steps. First, if the box is the first SID box after the last voice box, the box is identified as a "first SID box", example: box 62 as shown in Figure 4. The net SID bits of the first SID box they are channel coded as described above to provide greater error protection than in the previously transmitted speech frame.

The corresponding gross (coded) bit channel box is handled in the same way as the voice channel frames. Specifically, the first encoded SID frame is passed through subsequent processing units, such as a conveniently used interposer and modulator, and is transmitted with the normal power level which is used for voice frame transmissions. Additionally, the raw bit channel box is stored in the storage unit. Then, if the box is not a first SID box and if the box is the first box after an SUF set, the box is referred to as a SID update box, example: box 64 in Figure 4. The SID update box it is processed in the same way as the first SID frames, except that it is transmitted at a reduced transmission power level. In particular, it is also stored in the storage unit. Finally, the other SID boxes (neither the first SID nor the SID update) arriving from the transmission DTX handler are skipped, eg: box 66. Instead, the SID box encoded from stored channel (first SID or SID update) is transmitted at the reduced TX power level as SID frames of upgrade. Note that only the first SID frames are sent with normal TX power. The other SID frames are transmitted at the reduced power level, as can be seen in the different power levels illustrated in the graph of the RZ radio subsystem of Figure 4. The full power transmission of the first SID frame ensures adequate detection at the beginning of a period without a voice. Due to the diagonal interleaving of coded channel bit frames over more than one TDMA frame, the situation can occur where a TDMA frame contains parts of one or more channel frames to be sent at a normal power level and parts of one or more channel boxes to be sent with reduced TX power. A possible solution for this situation is to transmit a TDMA box at normal power level if it contains a part of a channel box which is designated for transmission at normal power, for example: a first SID box. Analogous to the transmission side, this example mode also handles implementations that have a radio receiver (RSS) subsystem, located, for example, near a radio receiver, an equalizer and a channel decoder and a RX DTX handler, located, for example, near a voice decoder and an error concealment unit. It is further assumed that the RSS RX continuously passes decoded channel traffic frames (this being, net bit frames) to the RZ DTX handler which complies with the GSM standards of this example mode. In particular, the boxes are marked with markers that indicate the type of box detected, this being voice or SID, (ternary SID marker), and a bad box indicator (BFI). This example mode additionally handles the channel decoding control and frame type identification methods in the RX RSS and the decoding diagramming and the result of the identification to the input format of the RX DTX handler. As mentioned above, the Silence Update Marker (SUF) is available, as in the TX RSS. This procedure for receiving tables includes the following steps. First, channel decoding and frame type identification are carried out, example: as discussed above with respect to Figure 3. The outputs are the decoded net bit frame and, as a result of the identification, the markers SPEECH, SOFT_SID and BAD_FRAME. Second, the minimum selection units of the SID channel box decoder are re-set if the SPEECH or SUF marker is active, respectively. Third, the first SID box after voice is indicated. A marker, called FIRST_SID serves to provide this indication. FIRST_SOD becomes active: (1) if the S0FT_SID marker is active and the SPEECH marker was active in the preceding frame or (2) if the S0FT_SID marker is active and the BAD_FRAME marker was active during one or more consecutive frames after a frame with an active SPEECH marker. The marker S0FT_SID is re-set if the first marker FIRST_SID is active (fixed). Finally, the markers SPEECH, FIRST_SID, SOFT_SID, BAD_FRAME and SUF are diagrammed to the handles of the RX handler DTX, SID and BFI, according to Table 1, presented below. The general processing of frames on the receiver side is illustrated in Figure 5. RX RSS Markers Handler Markers RX DTX BFI SID Comment SPEECH 0 0 Good voice box FIRST-SID 0 2 Valid SID Frame SOFT_SID-SUF 0 2 Valid SID Frame SOFT-SID-SIF 1 0 Frame not usable BAD_FRAME 1 0 Frame not usable Table 1 - Diagramming of RX RSS markers to RX DTX handler markers The aforementioned example modes provide continuous transmission with reduced power during each SID period. However, it is also possible according to the present invention to apply the same concepts described above without transmitting continuously. The SID information may be coded to provide less redundancy (this being, lower error protection level) so that the SID information is transmitted on each gross bit frame n of the SID period. For n = 1, continuous transmission occurs, described above. For n greater than one, discontinuous transmission occurs. Although less error coding protection is provided, there is still considerable channel error protection than in conventional discontinuous transmission schemes, giving freedom to reduce transmission power. Additionally, although it is not transmitted continuously, transmissions are made more frequently than in conventional DTX techniques, therefore they provide systems, for example: Adaptive Multi-Range systems, a sufficient opportunity for channel measurement. An exemplary embodiment of the present invention which provides discontinuous transmissions will be described. On the transmitter side, the SID information bits belonging to a SID period are encoded in channel with a code rate that the resulting bits are accommodated within the number of raw bit frames to be used for transmission.

This number of bit frames can be determined based on, for example, the transmission time per SID period. Even if a single gross bit frame is transmitted per SID period, within the gross bit frame, logical subframe structures can be defined, and the reception of SID bit patterns described above can still be applied. On the receiver side, any known channel decoder which is suitable for decoding transmitted raw bits belonging to the SID period can be applied. In particular, if the above-described method with repeated gross or sub-frame transmissions is used, then the decoding methods described above with respect to the channel decoder and Figure 2 are applicable.

A modification of the above described method is possible considering the fact that, due to the diagonal interleaving of the raw bit frames on n channel frames, each raw bit frame to be transmitted is actually sent using n channel frames. Therefore, a method to transmit discontinuously (for example every m box where m is greater than or equal to n) which ignores this fact uses only a fraction, being this, 1 / n, of each channel box. The rest remain unused. Another alternative is to not use diagonal interleaving during periods of voice inactivity. If diagonal interleaving is not used, then all available raw bits of the channel boxes are used for the raw bit frames. In this case, the transmission power of the SID frame is reduced (keeping the frame transmission power SID constant). As an alternative, in addition to the conventional interleaving scheme, a second scheme can be employed which fills the channel frame parts with raw bits which are otherwise left unused. In this case, a higher gross bit rate is available to transmit SID information. Both transmissions, continuous and discontinuous, are also applicable is systems which operate with a period SID variable in time. Such situations occur for example in the case where the SID updates are only carried out if the noise comfort parameters have changed by more than a predetermined amount. Therefore, the transmission schemes according to the present invention create less interference than conventional DTX schemes since the interference pulses due to the transmission of SID frames are eliminated or reduced. further, such schemes can result in TX energy savings than in systems with conventional DTX. The system according to the present invention also supports, in contrast to the conventional DTX schemes, continuous internal band channels (this being, using the additional bandwidth not required by the SID information during continuous transmission or less than continuous for periods of silence) and continuous channel estimation. Also, this can lead to a more reliable channel quality estimate and produce its gains at higher SID update rates than, for example, systems according to GSM. This can result in a more accurate representation of background noise, possibly resulting in a better subjective perceived operation. Finally, such methods do not require, for example in a GSM application thereof, that the SID update frames are aligned either in phase or in frequency with the SACCH frame structure. Therefore, the DTX operation in MS to MS connections is simplified since no SID information should be delayed. It can also be appreciated by those skilled in the art that the present invention can be incorporated in other specific forms without departing from the spirit or characteristics thereof. The revealed modalities are considered, therefore, in all respects as illustrative and not restrictive. The scope of the invention as set forth in the appended Claims, and not the foregoing description, and all changes that fall within the meaning and equivalence range thereof are encompassed within the invention.

Claims (23)

1. CLAIMS n a TDMA radio communication system having a transmitting end and a receiving end, operating at a frame rate R, and having a voice traffic channel in which the source signal is transmitted at least two modes, of which one mode is applied during inactive speech periods, a method of transmitting coded source voice inactivity frames operating at a speech inactivity frame rate r being less than the speed R, which comprises : use a frame rate greater than r or equal to the transmission of inactive voice frames protected against error; and reducing the transmission power by an amount such that the transmission power is sufficiently high for the receiving end to decode the idle frames with a predetermined quality criterion.
2. The method of Claim 1, characterized by the following steps at the transmitting end of the link: for each of the encoded source voice inactivity frames that apply error protection of channel so that the resulting raw bits remain in a single raw bit frame; and sending the raw bit frame repeatedly to the maximum until the next coded source voice inactivity frame.
3. The method of Claim 2, characterized in that the receiving end of the link carries out the averaging of the raw bit frames received before the channel decoding.
4. The method of Claim 2, characterized in that the receiving end of the link carries out the individual channel decoding of each received raw bit frame and because it performs majority decision decoding to derive the encoded source voice inactivity frames .
5. The method of Claim 2, characterized in that the receiving end of the link carries out the individual channel decoding of each received raw bit frame and selects that decoded frame as the representative of the decoded source voice inactivity frame of which the frame of raw bits has the lowest number of gross bit errors estimated.
6. The method of Claim 1, characterized in that the logical subframe structure is defined within of gross bits of voice inactivity and that any of the claimed methods is applied to these sub-frames.
7. The method of Claim 6, characterized in that the speech inactivity frame transmission speed is equal to the voice inactivity range r.
8. The method of Claim 1, characterized in that the inactivity frame range is not constant.
9. The method of Claim 1, characterized in that the transmitting end of the link transmits at least the first frame which is coded as voice inactivity box, immediately after at least one frame which has not been coded as voice inactivity box , at a transmission level which is not reduced by the same amount as the power level used for other frames of voice inactivity.
10. The method of Claim 1, characterized in that the receiving end of the link detects the coding mode, which has been applied by the transmitting end, applying a decision algorithm to estimate the most likely coded mode of the raw bit frame, with base in measurements which are derived by at least one of the following steps: applying the channel decoder pertaining to at least one encoding mode and searching the net bits decoded by a bit pattern identification mode feature, or searching the raw bit frame for a characteristic of the bit pattern that identifies the mode of encoded.
11. The method of Claim 1, characterized in that the receiving end of the link detects the coding mode, which has been applied by the transmitting end., by means of a decoding test of the raw bit frames received which performs at least in part the decoders of at least two coding modes and in which a coding mode decision algorithm selects, based on measurements which are taken from different executed channel decoders, the estimated most likely encoding mode and the corresponding decoded net bit frame.
12. The method of Claim 10, characterized in that the receiving end of the link can classify the net bit array as not useful.
13. The method of Claim 10, characterized in that the coded mode decision algorithm At the receiving end of the link, the decision is based on at least one of the following measures: measurements which are taken from different channel encoders belonging to different encoding modes, measurements which are taken from the source decoder, measurements which are taken from the receiver components up to and including the equalizer.
14. The method of Claim 13, characterized in that the coded mode decision algorithm at the receiver end of the link bases the decision on the knowledge of different transmission power levels used for coded frames in different encoding modes.
15. The method of Claim 10, characterized in that the radio communication system is a radio communication system FDMA.
16. 16. The method of Claim 10, characterized in that the radio communication system is a CDMA radio communication system.
17. 17. A radio communication system comprising: means for transmitting voice information using a first type of error correction coding; and means to transmit silence information using a second type of error correction coding.
18. 18. The radio communication system of Claim 17, wherein said second type of error correction coding has greater redundancy than said first type of error correction coding.
19. 19. The radio communication system of Claim 17, wherein said means for transmitting voice information transmits at a higher power level than said means for transmitting silent information.
20. 20. The radio communication system of Claim 17, wherein said means for transmitting silent information continuously transmits during periods of silence.
21. The radio communication system of Claim 17, wherein said means for transmitting silence information transmits discontinuously during periods of silence.
22. The method of Claim 11, characterized in that the receiving end of the link can classify the decoded net bitrate as unusable.
23. 23. The method of Claim 11, characterized in that the coded mode decision algorithm at the receiving end of the link, the decision is based on at least one of the following measures: measurements which are taken from different channel encoders belonging to different coding modes, measurements which are taken from the receiver components up to and including the equalizer .