WO2008063108A1

WO2008063108A1 - Method for multiplexing/modulating layered data in broadcast transmissions

Info

Publication number: WO2008063108A1
Application number: PCT/SE2006/050491
Authority: WO
Inventors: Stefan Felter; Yi-Pin Eric Wang; Carmela Cozzo; Jörg Huschke
Original assignee: Telefonaktiebolaget Lm Ericsson (Publ)
Priority date: 2006-11-20
Filing date: 2006-11-20
Publication date: 2008-05-29

Abstract

The present invention relates to broadcasting in a cellular communications system. An object of the present invention is to broadcast user data according to a hierarchical data layer model, whereby a basic data layer has superior coverage in the cell and provides for a basic QoS. Enhanced data layers are available in addition to the basic data layer, albeit with less wide coverage in the cell, and provide for a higher QoS than if the basic layer only was detected. According to the invention the hierarchical model is provided by channel coding the different data layers at different coding rates. The coded data layers are multiplexed on a physical channel separately from the modulation and broadcasted in the cell. An advantage with the solution is that robust modulation methods may be used, and flexibility in the allocation of resources to the different layers.

Description

Method for Multiplexing/Modulating Layered Data in Broadcast Transmissions

TECHNICAL FIELD OF THE INVENTION

5 The present invention relates to the field of cellular tele- and data- communication and in particular to a radio base station method for broadcasting multimedia, and to a radio base station adapted for performing the method.

DESCRIPTION OF RELATED ART

Multimedia Broadcast and Multicast Service, abbreviated to MBMS, is a concept within 3GPP WCDMA standard for transmitting the same information such as news, video clips and movies to more than one user simultaneously.

In wireless transmission the different UEs (User equipment) in a cell typically receive a MBMS service from a RBS (Radio Base Station) at different channel quality. The MBMS solution today implies that the transmit power of a broadcast channel is determined based on users with worst channel conditions, typically the user at the cell border. This is necessary in order to guarantee, e.g., 95% coverage. This means e.g. that users with better channel conditions will have "too high" SINR. Another way to see it is that much of the base station's transmission power will be used for satisfying the worst user(s) .

DVB-H is a standard for television broadcasting that uses two hierarchical layers of data at transmission. The basic idea is that receivers with good radio conditions will be able to detect both levels of data, which results in high quality reproduction of the data sent, whereas terminals in geographical positions with less good radio conditions will at least be able to detect the basic data layer and be able to reproduce the source code with less good quality. Thereby

SUBSTITUTE SHEET fπULF 26) the basic layer data will have superior coverage than the enhancement data layer. The variation in robustness against disturbances on the radio link for the two data layers is accomplished by layered modulation. In layered modulation a symbol transmitted, is a superposition of two other information symbols .

Also a system called MediaFLO employs layered modulation for broadcasting data in two hierarchical levels. MediaFLO is a product of Qualcomm Inc and disclosed in a document retrievable from website http: //www. qualcomm.com/mediaflo/news/pdf/flo_datasht.pdf .

In both the above cases the hierarchical data layers provide a layered QoS for terminals, were the quality depends on the radio conditions in the location of the terminal and the performance of the UE receiver.

SUMMARY OF THE INVENTION

The object of the present invention is to provide a feasible implementation for layered QoS versus coverage, and to do this in an alternative way than that stated in the prior art.

The present invention relates to a method for a radio base station, in which a stream of user data divided into a basic layer and at least one enhancement layer is received over separate transport channels. The different layers are channel coded at different coding rates, time multiplexed on a physical channel separately from the modulation and then broadcasted.

The present invention also relates to a radio base station adapted for performing the method.

An advantage by the present invention is that more robust modulation methods may be used, than is possible to use in layered modulation. Thereby for example GMSK or QPSK are alternatives not possible for layered modulation and that provides robustness that is favourable in cellular telecommunications system because superior coverage can be provided, than what is possible with modulation methods having a higher symbol alphabet .

A further advantage is that it is easy to provide more than two hierarchical levels of coverage in a cell .

A further advantage is that the stream of user data may comprise two or more groups of data, for example relating to audio and video, and they may each be divided into two or more layers independently of the number of layers in the other group.

A further advantage is the increased flexibility in how the physical channel may be shared among different data layers, groups of data and plural streams of user data. For example plural streams of TV-programs or other streams of user data may be multiplexed on the same physical channel, which is an advantage if the physical channel is spread by a spreading code and there is a lack of such codes. Should instead it be found more important to reduce mobile power to save battery, the different user data streams can be mapped to corresponding physical channel each. In the same system some data stream/s may be mapped to a separate physical channel/s, while other physical channels are shared between two or more data streams.

DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates a geographical coverage of the transmission of different data layers from a RBS.

Figure 2a is a hierarchical data layer structure. Figure 2b is block diagram of some nodes in a cellular telecommunications system and a source coder connected to the cellular telecommunications network.

Figure 3 is a simplified functional block diagram of a RBS. The block functions correspond to the steps of a method in the RBS.

Figure 4 is a flowchart that in more detail discloses the coding and multiplexing functions disclosed in figure 3.

Figure 5 is a functional block diagram illustrating the steps from source coding to the broadcasting of plural applications .

Figure 6 is a functional diagram similar to fig 5, disclosing one application comprising two groups of data.

Figure 7 is a vector diagram in complex plane coordinate system.

Figure 8 is an alternative vector diagram in complex plane coordinate system

DESCRIPTION OF PREFERRED EMBODIMENTS

The purpose of the present invention is to broadcast a multimedia service according to a hierarchical data layer structure that provides for a differentiated QoS (Quality of Service) , that means a basic data layer with a basic QoS level have superior coverage in a cell than any enhancement data layer. Thereby the better QoS is available for UEs (User Equipment) in a less wide area within the cell.

An implication to the description above is the QoS depends on the performance of the UE as well as on the radio conditions were the UE is located. Therefore a UE at the cell border with a high performance receiver might have as high QoS as a UE with a receiver of lower performance but better channel conditions. For simplicity in the further description we refer channel conditions to include both the channel and the receiver of the UE.

Figure 1 illustrates a physical area C, also referred to as a cell, that is serviced by a RBS, (Radio Base Station) . A number of UEs are located within the cell, and depending on geographical conditions and distance to the RBS, it may or may not be possible for a UE to detect a signal broadcasted by the RBS. In the example a video service will be broadcasted, with the data structured into a hierarchical layer model comprising a basic layer, a first and a second enhancement layer. The hierarchical structure is disclosed in figure 2a. The basic layer is to be received by UEs in almost the total cell C area, and that are UEl, UE2 & UE3 in figure 1. UE2 Sc UE3 in figure 1 are located in spots were the signal strength is enough for detecting also the first enhancement layer. This result in a higher quality of the video is possible to reproduce in UE2 and UE3. Yet better quality of the video is possible in UE3 , because it can also detect the 3^rd layer data in addition to the first two data layers .

The invention is based on the assumption that a source coder encodes the service data with at least two classes of source bits, in the example it is class I, class II and class III bits. These 3 types of bits correspond to a basic layer, and the two enhancement layers of data. When the three classes of bits are used to reconstruct the source information, the distortion between the source and reconstructed information is the lowest, as is the case for UE3. If only classes I and II bits are used to reconstruct the source information, the distortion between the source and reconstructed information is medium. If only the class I bits are used, then the distortion is the highest, as is the case for UEl. However, for UEl, the receiver still recovers part of the source information, at a lower quality of course. Using this scheme, the channel coder can then use a low-rate FEC

(Forward Error Correcting) code to protect class I, enabling class I bits to be received by almost all users. Class II bits can be protected with a medium coding-rate FEC code, facilitating users with reasonable channel conditions to receive it. Lastly, class III bits can be protected with a high coding-rate FEC code, making it only possible for users with good channel conditions to receive, detect and decode it correctly. With this scheme though all three classes of bits are transmitted, the quality of services can be differentiated at the terminals, depending on their location in the cell. Terminals at the cell edge will be able to receive and reproduce a TV program, let be at less good quality than the terminals closer the base station.

H.264 is a standard for source coding a video stream into differentiated classes of bits, and that may well be used in combination with the present invention.

The invention is further based on the assumption that the streams of the different layers, corresponding to the different classes of bits are transmitted from the source to a RBS over corresponding transport channels . Figure 2b is a block diagram of nodes essential for the present invention. The source coder is connected to a RNC (Radio Network Controller) in a cellular network. The RNC is connected with an RBS and that broadcast a service within a cell to a UE. For simplicity only one RNC, one RBS and one UE is illustrated whereas in practise a cellular network comprises plural RNCs, plural RBS connected to each RNC and in each cell serviced by a RBS there are plural UEs . The main focus of the present invention is in the RBS and to some extent the signalling from the RNC to the RBS for controlling the broadcasting.

The 3GPP WCDMA terminology is used in the description, albeit not strictly, because 3GPP is intended as the primary implementation of the present invention. The correct WCDMA term for RBS (radio base station) is NodeB, while the more generic term RBS is used in the description. It should be noted the invention may well be implemented also in other cellular data- and tele-communications networks than WCDMA. The generic term for UE (user Equipment) is radio terminal, and it should be understood that the invention can be implemented with use of a generic radio terminal.

Differentiated channel coding rates

Figure 3, illustrate the functional blocks of a RBS, essential for the present invention. The functions of these blocks also correspond to the method steps of the present invention. A stream of data, which may be the video service, is received by the RBS over a number N of FACHs (Forward Access Channels) and that are a transport type of channel. In the RBS, the data streams of respective FACH are channel coded with a respective FEC code rate, see 21. The differently channel coded data layers are then multiplexed, 22, on one common S-CCPCH (Secondary Common Control Channel) and that is a physical channel. After the multiplexing, and separate from it is the stream of channel coded and multiplexed data modulated, by modulator 23. Before S-CCPCH is broadcasted it is amplified, by amplifier 24, to a predefined transmit power level.

Multiplexing in detail

Figure 4 is a flowchart of the steps taken for coding and multiplexing the transport channels FACH_0 to FACH_N on to the same physical channel S-CCPCH. The last steps of modulating 23 and amplifying 24 are though not included. All steps of figure 4 are disclosed in the 3GPP WCDMA standard prior to this invention. The differentiated coding of different data layers as is disclosed in figure 3, influence the third step, i.e. the Channel coding step. In figure 4, a separate channel coding rate is applied to each FACH , such that if Channel coding_0 is allocated to FACH_O, in figure 4, that implies a unique FEC indexed 0 is used, and FACH_N is applied to Channel coding_N which implies a corresponding unique FEC indexed N.

In more detail the following steps are performed separately for each FACH; CRC (Cyclic Redundancy Check) bits are attached, 410, to transport blocks containing the data for enabling error detection in a receiver. Next, 402, all transport blocks in a TTI are serially concatenated. If the number of bits in a TTI is larger than the maximum size of a code block in question, then code block segmentation is performed after the concatenation of the transport blocks. In the following step, 403, channel coding is performed with a separate coding rate for each FACH according to the invention. The coding may consist of Convolutional coding, and/or turbo coding. Rate matching, by puncturing or padding of dummy bits is performed, 404. In case of DTX (Discontinuous Transmission) bits indicating DTX is added, 405. A first interleaving is performed, 406, and last radio frame segmentation is made 407.

In a following step, 408, radio frame segments from all the FACH relating to one application are multiplexed on the same CCTrCH (Coded Composite Transport Channel) . A second indication of DTX is added 409, if relevant. Next, physical channel segmentation is performed 410, a second interleaving 411, follows. Finally, the radio frames are mapped 412, to the physical radio channel. This result in radio frame segments of the hierarchical data layers are multiplexed on the same physical channel.

A very last optional step 413, may follow in which STTD (Space Time block) Antenna diversity is applied for the transmission. For deeper information of the various multiplexing steps, see 3GPP TS 25.212 V.6.7.0. Considerations relating to protocol layers above the physical layer

Assume MBMS will include transmission of e.g. several of TV programs in one cell simultaneously, and each TV program will correspond to one MBMS application, it is an advantage that all data layers of an application can be mapped on the same S-CCPCH, and that even several applications can be mapped on the same S-CCPCH because each S-CCPCH requires a unique spreading code, and the number of spreading codes having good properties is often short compared to the need.

Figure 5 illustrates the mapping from an application layer, to physical layer of a number of M applications or MBMS services, such as TV programs. Here application and physical layers refers to layers of a protocol stack. A number of boxes indicate the handling of a respective application 1 to M. In each box, the application is source coded and demultiplexed 51. The data layers of one application are provided over separate transport channels FACH_0 - FACH_N to the RBS.

Some applications may involve two or more groups of information. An example is TV broadcasting were the information consists of both video and audio. Excluding the time synchronization the video and audio should be coded independently. Thus, there will be one base layer for the video and one base layer for the audio. In other words the data layers are divided into sub-group layers . Figure 6 is almost the same as figure 5, with the exception of only one service being disclosed and with the additions of the data layers being divided into the sub-groups. In the example illustrated in figure 6, the audio consists of two layers and the video consists of three layers. In figure 6, the base layer for the audio corresponds to FACH_0 and the base layer for the video corresponds to FACH_2. An MBMS service is received by the RNC using the internet protocol (IP) . The RNC maps each combination of information group and layer of the MBMS service to a separate MTCH (MBMS traffic channel) . The MTCHs are mapped by the RNC to one or a plural of FACH. When a MBMS application is set up, [moved to the end of section] The RNC will do the mapping of the different data layers to the respective FACH. Assume there are two MBMS applications, both TV-programs. The RNC may as an example allocate a number of FACH according to:

TV-program 1:

Audio: layer (1) -> (FACH (1) ,MTCH (1) ), layer(2)-> (FACH(2) ,MTCH(I) )

Video: layer (1) -> (FACH (3) ,MTCH(I) ), layer(2)-> (FACH (4) ,MTCH(I) ), layer (3 )-> (FACH ( 5 ), MTCH ( 1 ))

TV-program 2 :

Audio: layer (1) -> (FACH (6) ,MTCH(I) ), layer(2)-> (FACH(7) ,MTCH(I) )

Video: layer (1) -> (FACH (8) ,MTCH(I) ), layer(9)-> (FACH(IO) ,MTCH(I) ) , layer (3 )-> (FACH ( 11 ) ,MTCH(I) )

By this mapping the RBS will handle each MBMS service separately and map it to a corresponding physical channel as is illustrated in figures 5 and 6. If, alternatively, the two or more MBMS services are to be broadcasted on the same physical channel, the RNC may map the two or more MBMS applications to the same FACH. The basic layers of the two applications should then be mapped to the same FACH, while the first enhancement layer of the two applications, and subgroup if applicable, should be mapped to another FACH. For example the mapping for TV-program X could be:

Audio: layer (1) -> (FACH(I) ,MTCH(X) ), layer(2)-> (FACH(2) ,MTCH(X) )

Video: layer (l)-> (FACH (3) ,MTCH(X) ), layer(2)-> (FACH (4) ,MTCH(X) ) , layer (3 )-> (FACH (5 ) ,MTCH(X) ) where the MTCH counting starts at 1 for each FACH Applied to two TV programs the mapping would be:

TV-program 1 : Audio: layer (l)-> (FACH(I) ,MTCH(I) ), layer(2)-> (FACH (2) ,MTCH(I) )

Video: layer (l)-> (FACHO) ,MTCH(I)) , layer (2) -> (FACH ( 4 ) , MTCH (I)), layer ( 3 ) -> (FACH ( 5 ) , MTCH (I))

TV-program 2 :

Audio: layer (l)-> (FACH(I) ,MTCH (2) ), layer(2)-> (FACH(2) ,MTCH(2) ) Video: layer (1) -> (FACH (3 ) ,MTCH (2) ), layer(2)-> (FACH(4) ,MTCH(2) ) , layer (3 ) -> (FACH ( 5 ) ,MTCH(2) )

The RNC signals the establishment of the FACHs to the RBS. In the establishment messages, the RNC defines the transmit parameters, i.e. the spreading factor, channel coding and transmit power. All these parameters are configured via O&M in the RNC.

The data layers that are mapped to the same FACH will be coded at the same rate.

Data layer dependency and sub-group layers

There is an incremental dependency between the data layers of one MBMS service; to be able to benefit from an enhancement layer the basic layer and any lower enhancement must be correctly decoded. For example, assume that a picture has been source coded into two layers, to be carried by respective FACHs, the basic layer data corresponds to a medium resolution picture. If both layers are decoded correctly it will result in a high-resolution picture.

The information bits of the basic and enhancement layers are multiplexed into one CCTrCH per TTI (Transmission Time Interval) . Even though the channel and fading conditions might change over the TTI , the layers , due to the multiplexing and interleaving, experience the same effective channel. This way the layered hierarchy is preserved, i.e. the probability of correctly decoding an enhancement layer without being able to correctly decode the basic layer is minimized. The multiplexer, see 408, needs information on the synchronization between the different FACHs within one MBMS application to be able to schedule in the same TTI, sequences of data of the hierarchical layers that will be reproduced simultaneously.. This can be performed by monitoring of sequence numbers included on the headers of the FACHs data frames received. The data received in the frames should further be in its sequential order of reproduction.

Modulation method comparison

According to the present invention, the different coverage within a cell by the different data layers is accomplished by different channel coding rates. The channel coded bits of the two layers are time multiplexed on the same physical channel and after that modulated with a traditional modulation method.

In opposite to the present invention, in the prior art multiplexing of the two data layers on a physical channel is made by layered modulation, which result in a superposition in time of two layers .

Figure 7 shows the plane of complex numbers illustrating the principle of layered modulation. At a time instant, two symbols, representing data from two different layers, are represented by a respective vector Vl, V2. The length of the vectors corresponds to the power used for respective symbols. A resultant vector R is formed by the addition of the two vectors Vl and V2. The vector R, and its end point at a constellation point in the complex diagram, correspond to the two super positioned symbols transmitted. The constellation points of figure 7 correspond to the modulation alphabet of 16 QAM. The modulator has two input streams, one for each of the two data layers. This is necessary for superposing the symbols for the two layers and applying the different power levels.

Figure 8 is a complex plane illustrating constellation points in the modulation alphabet of QPSK, which is one of several possible modulation methods applicable for the present invention. The vector V in figure 8 represents one symbol of the modulation alphabet only. The QPSK modulator has a single input stream consisting of time multiplexed data from the two or more layers. Distinct pairs of consecutive bits in the data stream will be modulated into one of the 4 alternative symbols of the alphabet.

GMSK is an alternative modulation method for the present invention, with its binary modulation alphabet being more robust than QPSK. Robustness is desired for cellular telecommunications system because it provides for superior coverage. Coverage and capacity is the most important factors in the design of cellular systems.

Modulation methods with higher alphabets such as 16 QAM is nevertheless possible to use with the present invention.

Alternatives to the embodiments specifically disclosed

In the embodiments three different layers of data has been disclosed. There is however nothing that prevents any number of layers from two or more to be used.

All the above embodiments have been described as implemented in 3GPP WCDMA, it is however possible to implement the present invention in a system that uses OFDMA access technology, or CDMA based on TDD instead of FDD as in the examples above. In any system the physical radio channel, exampled with the S-CCPCH channel could be: • A time slot in a TDMA system

• A carrier frequency in any form of multi-carrier system, including OFDM

• A code in a CDMA system

• A combination of the above three in any system employing a combination of the above three techniques .

A TV program has been the example of MBMS service broadcasted in the above description. The inventions is however not restricted to a TV-program, it can be used for any user data streams that shall be broadcasted. The most common examples of such broadcasted data streams are expected to include video or audio or both. The data stream may alternatively include text and or other data, for example it could be information on traffic disturbances including a map of roads and routes alternative to those disturbed. Downlink user data stream is the generic term used for the data stream of one service such as a MBMS. It will mainly comprise data intended for being reproduced, probably as video or audio, while it may also include control information for the reproduction.

The invention can be applied with a generic radio terminal, in the description above the generic radio terminal is exampled with a UE.

Above, the purpose of the basic data layer has been described as providing as wide coverage in the cell as possible. This is probably the most common purpose of the basic layer, however, it may also be applied for providing coverage in a smaller part of the cell, for example in an antenna beam covering the location of some event . The purpose of the layered data model is in both cases that the basic layer shall have superior coverage than any enhancement layer.

Claims

1. A method for a base station in a cellular data- and tele- communications system, comprising the steps of: a) receiving a stream of downlink user data that has been source coded into a hierarchical data layer structure comprising a basic layer and at least one enhancement layer and wherein said layers are received over separate transport channels; b) channel coding the basic layer data at a first coding rate and channel coding the enhancement layer/s data at a respective coding rate that is higher than the coding rate used for any lower layer data; c) time multiplexing the channel coded data layers on one physical channel; d) modulating the data on the physical channel; wherein the modulation step is made after and separate from the multiplexing step, and, e) broadcasting the physical channel.

2. The method of claim 1 wherein data of the basic layer and the one or more enhancement layer/s are time multiplexed into the same Transmission Time Interval.

3. The method of claim 1, wherein the stream of user data is divided into two or more subgroups, and at least one of them being source coded into two or more of said data layers .

4. The method of claim 1, wherein the stream of downlink user data is an audio strean.

5. The method of claim 1, wherein the stream of downlink user data is a video stream.

6. The method of claim 1, wherein the stream of downlink user data includes a video and an audio stream.

7. The method of claim 1, wherein the stream of downlink user data includes at least one of the two options text and data.

8. The method of claim 6 wherein in the stream of downlink user data is a TV program.

9. The method of claim 1, wherein the radio base station receives two or more streams of downlink user data, and for each stream steps a) - e) are performed, and wherein each downlink user data stream is multiplexed on a separate physical channel.

10. The method of claim 1, wherein more than one stream of user data is multiplexed on the same physical channel.

11. The method of claim 1 or 9, wherein the physical channel is a S-CCPCH channel.

12. A radio base station adapted for performing the method in claim 1 or in any of the claims dependent on claim 1.