WO2002019543A2

WO2002019543A2 - Method for improving the quality of voice signals in a mobile radio network

Info

Publication number: WO2002019543A2
Application number: PCT/DE2001/003317
Authority: WO
Inventors: Carsten Ball; Thomas Stark
Original assignee: Siemens Aktiengesellschaft
Priority date: 2000-08-29
Filing date: 2001-08-29
Publication date: 2002-03-07
Also published as: WO2002019543A3; DE10042352A1

Abstract

The aim of the invention is to improve the data rate of data packets with a fixed predetermined structure (bursts), which are used to transmit voice signals in time slots that are assigned to a transmission channel. To achieve this, the bursts are modulated using a phase shift-key modulation method, such as in particular the 8PSK, (8 phase shift keying) or QPSK, (quadrature phase shift keying) and/or several time slots (multislots) are simultaneously occupied for a voice communication.

Description

description

Method for improving the voice quality in a mobile radio network

Since in mobile radio systems such as GSM (Global System for Mobilcommunication) the bandwidth available per channel is limited for capacity reasons, voice signals of a voice connection or a channel have so far been transmitted with low data rates of less than 16 kbps, which results in a significantly poorer one Language quality than leads in the fixed network. In GSM, the gross data rate is 22.8 kbps per time slot or channel, of which about 13 kbps are available in a full rate channel for the transmission of voice signals. Nevertheless, various operators of mobile radio networks try to compete with or replace the fixed network by using a mobile radio terminal in the household of a subscriber instead of a fixed network connection (for example the GENION product from VIAG Intercom). However, due to the drastically reduced voice quality of mobile radio subscribers compared to a landline connection, it still has little resonance.

In the radio area, significant quality improvements in the transmission of voice signals have so far only been possible using so-called fixed wireless access systems, such as cordless phones according to the DECT (Digital Enhanced Cordless Telephone) standard or proprietary CDMA (Code Division Multiple Access) systems. Examples include the CDMAlink system and the DECTlink system from Siemens AG. With data rates from lβkbps (LDCELP: LD Code Excited Linear Prediction), 32kbps (ADPCM: Adaptive Delta pulse Code Modulation) up to 64kbps (ISDN BA: Integrated Services Digital Network Base Access), these systems offer a voice quality that is almost the same as in Fixed line corresponds. The invention is therefore based on the object of providing a method for improving the voice quality in a mobile radio network, which operates in particular according to the GSM standard.

This object is achieved by a method having the features of patent claim 1. Preferred embodiments and developments of the invention are the subject of the dependent claims.

The essence of the invention is the increase in the data rate of data packets with a predetermined structure (bursts), by means of which voice signals are transmitted in a transmission channel allocated time slots. This is done according to the invention in that the bursts are modulated with a higher-value modulation method than the GMSK method previously used in GSM, such as in particular phase shift keying modulation methods such as 8PSK (8 phase shift keying) or QPSK (quadrature phase shift keying) modulation, and / or several time slots (multislots) can be occupied simultaneously for a voice connection.

As a result, data rates are almost as high in a mobile radio system as in WLL systems (WLL: Wireless Local Loop) and the voice quality is essentially as good as that in the fixed network.

With the introduction of EDGE (Enhanced Data Rates for GSM Evolution) in the existing mobile radio systems with the ETSI Release 1999, the air interface of GSM and TDMA mobile radio systems according to IS-136 is expanded in the first phase (phase I) in such a way that bursts for Transmission of data signals can be modulated using 8PSK modulation. The introduction of QPSK modulation has already been discussed in the second phase (Phase II) (however, this is still the case not included in the previous ETSI specifications). However, the aforementioned improvements only serve to achieve higher data rates when transmitting data signals using EGPRS (Enhanced General Packet Radio Services) or ECSD (Enhanced Circuit Switched Data), i.e. generally to improve the data rates of (mostly packet-oriented) data services.

In comparison to the bursts modulated with GMSK (Gaussian Minimum Shift Keying) modulation, which are used in GSM systems for the transmission of voice signals, but also for data transmission using GPRS (General Packet Radio Service), the 8PSK modulation method can be used up to 69kbps and with the QPSK modulation up to 40kbps gross per time slot, which corresponds to a tripling of the data rate compared to the previously used GMSK with 8PSK modulation and a doubling with QPSK modulation.

Using suitable vocoders (speech encoders) for speech, it is therefore possible to not only record data but also speech depending on the propagation conditions, i.e. depending on the quality of the radio connection or the radio channel, to transmit with more than 16 kbps via the air interface without reducing the capacity (number of parallel connections) of the mobile radio system. Ideally (with small ranges, e.g. in the home or in offices), data rates up to 64 kbps of an ISDN basic channel per time slot can be achieved. For two ISDN basic channels and one ISDN D channel, up to three time slots must then be allocated in parallel or simultaneously by a mobile radio terminal. This method advantageously does not require more time slots than the previous GSM connections.

However, the 8PSK or QPSK modulation reduces the signal-to-noise ratio compared to GMSK, which means that the methods using 8PSK or QPSK modulation are only suitable for bridging small ranges. Usually, however, the distances to be bridged are small, particularly in the home (household, office, etc.), where the method is preferably used as a replacement for a landline connection.

With HSCSD (High Speed Circuit Switched Data), the parallel or simultaneous allocation of several time slots (multislots) is used to achieve high data rates during data transmission. This enables capacities of n x 14.4kbps per subscriber or connection to be achieved. According to the invention, this multislot method can also be used for the transmission of voice signals in order to achieve a high voice quality. Three time slots would then have to be occupied for a data rate of 32 kbps and up to five time slots per voice connection or channel for a data rate of 64 kbps, as in the case of an ISDN basic channel.

The advantage of this method is that the bursts can be modulated as in conventional GSM using GMSK modulation and therefore a high signal-to-noise ratio is achieved, which in turn requires greater robustness and range of the radio connection. The same ranges can be bridged as with the previous GSM radio connections for voice signal transmission. This does not affect the existing radio planning. However, several time slots or slots are occupied for a voice connection.

A combination of the two aforementioned methods is also possible, namely the transmission of bursts that are modulated with 8PSK or QPSK, with several time slots being occupied simultaneously for a voice connection. The very high data rates that can be achieved thereby enable a particularly high quality of the voice connection. However, only short ranges can be bridged with the radio connection due to the poorer signal-to-noise ratio by the 8PSK or QPSK modulation, and fewer voice or data connections per frequency are possible due to the multislots.

In order to be able to transmit even in the case of deteriorating radio connections, a so-called link adaptation function is preferably provided, which works between different speech coders with different data rates, e.g. 64 kbps, 50 kbps, 40 kbps, 32 kbps, 20 kbps, 16 kbps switches, • in order to always achieve the best call quality. A plurality of speech coders and corresponding speech decoders are therefore preferably provided in the mobile radio network in at least one base station and at least one mobile terminal. Depending on the quality of the radio connection between the mobile terminal and the base station, a speech encoder / decoder pair is selected which compresses the speech signals to be transmitted in accordance with the quality of the radio connection in such a way that the data rate is maximum. In other words, depending on the quality of the radio connection, the compression of the speech signals is selected by an encoder and corresponding decoder. In the case of a radio connection which deteriorates during a voice connection and is characterized by an increased bit error rate or decreasing reception level, it is possible to switch from a first encoder / decoder pair to a second encoder / decoder pair with a lower compression rate, in order to achieve a higher signal To get noise-to-noise ratio and not to let the connection break.

If the radio connection becomes so bad that none of the above methods are sufficient for a problem-free transmission, it is necessary to switch back to the robust GMSK modulation provided in the GSM. Therefore, in a preferred embodiment of the method modulates the bursts to be transmitted below a predetermined data rate by means of GMSK. This "fallback" on GMSK is necessary to avoid breaking the voice connection due to a too low signal-to-noise ratio.

However, the adaptive multirate method (AMR: adaptive multirate) previously specified in the GSM standard should preferably be used. This ensures that in the event of poorer radio connections, the fallback to the GSM call quality that was previously available is carried out automatically. For example, a connection or data rate adjustment to GMSK with 12.2 kbps, 10.2 kbps, 7.9 kbps, 7.4 kbps, 6.7 kbps, 5.9 kbps, 5.1 kbps or 4.7 kbps would be possible without interrupting a voice connection.

The switching of the modulation of the bursts is preferably carried out between the transmission of two successive bursts, since there is sufficient time between the bursts to be sent or received for a voice connection or a channel, the switching of the modulation in hardware or software in a mobile radio terminal and a base station (BTS : Base transceiver station).

A burst modulated with 8PSK transmits three times the data volume of a burst modulated with GMSK. Accordingly, the 69 kbps data rate per time slot or slot in GSM can be used for up to four full rates (full rate) - voice connections with 16 kbps (i.e. the previous GSM voice quality) or up to eight half rates (half rate) - voice connections on these Switch slot. This multiplexing of several conventional voice connections into one time slot allows the capacity of the mobile radio network in a radio cell to be increased without having to accept a loss in voice quality. If the number remains the same Voice connections per TDMA frame therefore leave time slots free, which are preferably used for monitoring neighboring radio cells or for transmitting signaling information.

It is also possible to transmit packet data, for example EGPRS packet data of a PDTCH (Packet data traffic Channel), together with voice signals that are modulated with 8PSK, in one slot. This would have the advantage that the transmission capacity of the radio link or the channel can be used as well as possible in the event of little packet data traffic. In a particularly expedient embodiment of the method, voice and data signals are therefore transmitted together in slots, the data rate of the voice signals being set as a function of the data signals to be transmitted in a slot.

In summary, the invention therefore uses the possibilities for data services, in particular already implemented in GSM, to improve the transmission of voice signals. On the one hand, this includes the use of higher modulation levels such as 8PSK or QPSK as well as the parallel assignment of several time slots (multislots). Likewise, the adaptive multirate method known from GSM can be used to advantage. Furthermore, the higher modulation levels allow the speech signals to be compressed while the call quality remains the same, as a result of which fewer bursts are sufficient for the transmission of the speech signals and therefore slots remain free which can be used, for example, for monitoring neighboring radio cells. This is important because, on the one hand, more and more functions are being integrated into the GSM system, such as GPRS, E-GPRS, etc., and there is also a convergence of GSM networks and UMTS networks. In addition, these free slots can also be used to increase the capacity of the GSM network by the bursts released can be used for additional calls.

The effort to implement the method according to the invention in a mobile radio network is very low. Modifications of the software are sufficient for BTS Base Transceiver Station) and BSC (Base Station Controller). Hardware and software may need to be modified in the TRAU (transcoder / rate adapter unit). Furthermore, the BSS internal interfaces such as the Abis interface between BTS and BSC and the A interface between BSC and MSC / TRAU can be expanded for the higher data rates per GSM time slot.

Some typical areas of application of the invention are briefly explained below:

In the household of a subscriber, the subscriber should be able to make phone calls with his mobile radio device like with a landline phone. In other words, the quality of voice connections over the mobile radio network and the fixed network should essentially not differ.

But even in the case of small radio cells in the inner city area, where the distance between a mobile radio subscriber or their mobile radio terminal and a base station is small, and in the rural area with so-called line-of-sight connections, the voice quality can be drastically improved.

The subscriber terminal measures the reception level of the BTS of the cell, for example on the broadcast channels (BCCH, CCCH) and thus concludes that the high voice quality is available due to the reciprocity of the radio channel. This can be indicated to the customer by the subscriber terminal by lighting up an icon on the display. The BTS is also intended to be able to receive channels with high speech quality via the Communicate broadcast channels to the subscriber terminals in the cell.

On the one hand, the customer can now make a call with either high voice quality or with the previously known low voice quality, e.g. by selection using menu items on the subscriber terminal. The subscriber terminal then shows the network its ability for the 8PSK modulated language during the dial-in process.

The operator can charge higher fees for using the premium service. The higher charge can be determined, for example, in the mobile radio network on the basis of measured parameters for the duration of the high 8PSK voice quality or, if link adaptation to GMSK has taken place, for the duration of the normal voice quality. Or the operator agrees a lump sum with the customer for this higher quality service, e.g. an increased basic fee.

In addition to the transmission of speech, suitable selection of the coding method could also transmit image information in the same channel. The voice transmission and the image transmission advantageously use an adaptive transmission method, both of which together use the full channel capacity. If, for example, only 16 kbps voice data are transmitted, then with a 64 kbps channel, 48 kbps image data could still be transmitted in parallel in the same channel. When there are pauses in the conversation or DTX (discontinuous transmission), full 64kbps are then available solely for the image information.

On the other hand, in radio cells with a high load, ie a high connection volume, it can make sense to make better use of the available capacities without deteriorating the voice quality. For this, several calls (with previous call quality) or several voice connections switched to a time slot or slot. The capacity of the mobile radio network is increased by the higher compression of the voice signals or the increase in the data rate by means of a higher modulation level such as 8PSK or QPSK. This is of particular interest for pico base stations, which are preferably used in buildings, airports, at train stations or, in general, in very busy locations with a high volume of mobile communications. Switching can also take place dynamically in the sense that the modulation is switched depending on the number of voice connections, that is to say the GMSK modulation is used for a few voice connections and the QPSK or 8PSK modulation is used for many voice connections. The higher voice quality that can theoretically be achieved by the higher modulation levels is given up here in favor of a higher subscriber density in a radio cell.

Depending on the interference situation and the capacity utilization in the cell, the network can, for example, switch off the 8PSK modulated service functions, which require more transmission power due to their increased signal-to-noise ratio and therefore generate more interference, and this via the broadcast channels to the participants in the Announce cell. Existing connections would then have to be reduced to normal speech quality using link adaptation.

The high-quality voice connections are preferably routed as line-switched services via the BSC (Base Station Controller) - TRAU (Transcoder / Rate Adapter Unit) - MSC (Mobile Station Controller) into the public fixed network PSTN (Public Switched Telephone Network). At this point it should only be mentioned briefly that the high-quality voice connections made possible by the invention can also be transmitted via packet data channels. The voice connections previously transmitted via the packet data channels using VoIP (Voice over Internet Protocol) only offer data rates of a few kbps, which can be improved noticeably by the invention. In this case, the voice signals run via the PCU (Packet Control Unit) - SGSN (Support GPRS Service Node) into the public data network PDN (Public Data Network).

Furthermore, quality gradations for transmission via packet-switched channels are possible by introducing a "Quality of Service" (QoS) function, which, for example, delays the signals to be transmitted by the radio channel, the throughput and the bit error rate BER (bit error rate). considered.

Finally, due to the free capacities due to the high data rates, a very fast power control FPC (Fast Power Control) for high quality voice connections by inband signaling is possible, which takes place every 20ms (instead of a slow power control by means of signaling about every 490 ms via the slow Associated Control Channel SACCH).

Abbreviations

8PSK Eight Phase Shift Keying

ADPCM Adaptive Differential Pulse Code Modulation

AMR Adaptive Multirate

BER bit error rate

BSC Base Station Controller

BTS Base Transceiver Station

CDMA Code Division Multiple Access CSD Circuit Switched Data

DECT Digital Enhanced Cordless Telephone

ECSD Enhanced CSD

EDGE Enhanced Data Rates for GSM Evolution

EGPRS Enhanced GPRS

GMSK Gaussian Minimum Shift Keying

GPRS General Packet Radio Service

GSM Global System for Mobile Communication

HSCSD High Speed CSD

ISDN BA Integrated Digital Services Network Base Access kbps kilobit per second

LDCELP LD code Excited Linear Predictive Coding

PCU packet control unit

PDN Public Data Network

PDTCH Packet Data Traffic Channel

PSK phase shift keying

PSTN Public Switched Telephone Network

QoS Quality of Service

QPSK Quadrature Phase Shift Keying

SGSN Serving GPRS Support Node

TDMA Time Division Multiple Access

TRAU transcoder / rate adapter unit

UMTS Universal Mobile Telecommunication System

VoIP Voice over Internet protocol

WLL Wireless Local Loop

Claims

claims

1. Method for improving the voice quality in a mobile radio network, in which in time slots (slots) which are assigned to a transmission channel, voice signals are transmitted at a predetermined data rate by means of data packets with a fixed structure (bursts), characterized in that for Increase in the data rate and the quality of the speech signals, the bursts are modulated with a higher modulation method or CDMA coding method and / or several time slots (multislots) are occupied simultaneously for one voice channel.

2nd Method according to claim 1, which also means that the mobile radio network is a GSM network and a phase shift keying modulation method such as 8PSK, QPSK or TD-SCDMA is used as the modulation method for the voice channels.

3. The method according to claim 1 or 2, characterized in that in the mobile radio network in at least one base station and at least one mobile device, a plurality of voice encoders and corresponding voice decoders are provided, with a voice encoder depending on the quality of the radio connection between the mobile device and the base station. Decoder pair is selected, which compresses the voice signals to be transmitted according to the quality of the radio link so that the data rate is maximum.

4. The method according to any one of the preceding claims, characterized in that that below a predetermined data rate, the bursts to be transmitted are modulated using GMSK.

5. The method according to any one of the preceding claims, characterized in that below a predetermined reception quality, the bursts to be transmitted are modulated by means of GMSK.

6. The method according to any one of the preceding claims, characterized in that the switching of the modulation of the bursts to be transmitted is carried out according to the adaptive multirate method of the GSM mobile radio system.

7. The method according to any one of the preceding claims, characterized in that the switching of the modulation of the bursts takes place between the transmission of two successive bursts.

8. The method according to any one of the preceding claims, characterized in that the bursts are modulated with the 8PSK modulation and four different full-rate or eight different half-rate voice channels are switched to each burst.

9. The method according to claim 8, characterized in that free slots are used for monitoring mobile radio cells of the mobile radio network, which are adjacent to the mobile radio cell in which the mobile device is located, and / or for the transmission of signaling information.

10. The method according to any one of claim 8 or 9, characterized in that voice and data signals are transmitted together in slots, the data rate of the voice signals depending on the data signals to be transmitted and the present transmission conditions being set in one slot.

11. The method according to any one of the preceding claims, characterized in that the network its ability for the higher

Modulation method and corresponding speech encoder in the

Radio cell transmits or when establishing the connection

Mobile terminal mi11ei11.

13. The method according to any one of the preceding claims, characterized in that the network measures and evaluates the interference situation in the cell and switches off the phase-modulated voice service when there is increased interference in the cell and modulates ongoing calls with GMSK.

14. The method according to any one of claims 3 to 13, characterized in that the speech coders support LDCELP, ADPCM or ISDN for the improved speech quality.

15. The method according to any one of the preceding claims, characterized in that in a connection to cellular subscribers

Cellular subscribers no recoding / 2. Coding the

Voice signal occurs in the network.

16. Method according to one of the preceding claims, characterized in that that the connection quality is set to the quality of the poorer connection in a connection between mobile subscribers to mobile subscribers.

17. The method according to any one of the preceding claims, characterized in that in addition to speech, image information is transmitted in parallel in the same channel.

18. The method according to claim 17, characterized in that the transmission of the image information uses free capacity in voice transmission, in particular in DTX (discontinuous transmission) and pauses in conversation.