ZA200401569B - A method and a system for transfering AMR signaling frames on halfrate channels. - Google Patents

A method and a system for transfering AMR signaling frames on halfrate channels. Download PDF

Info

Publication number
ZA200401569B
ZA200401569B ZA200401569A ZA200401569A ZA200401569B ZA 200401569 B ZA200401569 B ZA 200401569B ZA 200401569 A ZA200401569 A ZA 200401569A ZA 200401569 A ZA200401569 A ZA 200401569A ZA 200401569 B ZA200401569 B ZA 200401569B
Authority
ZA
South Africa
Prior art keywords
bit stream
bits
block
bit
coding
Prior art date
Application number
ZA200401569A
Inventor
Thierry Bellier
Benoist Sebire
Original Assignee
Nokia Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nokia Corp filed Critical Nokia Corp
Priority to ZA200401569A priority Critical patent/ZA200401569B/en
Publication of ZA200401569B publication Critical patent/ZA200401569B/en

Links

Landscapes

  • Detection And Prevention Of Errors In Transmission (AREA)

Description

oo | 1
A method and a system for transfering AMR signaling frames on halfrate channels
Field of the invention ’ The present invention relates generally to the transmission of sig- naling information. More specifically, the present invention relates to the transference of signaling frames on GERAN (GSM/EDGE Radio Access
Network) half rate channels utilizing a modulation scheme.
Background of the invention
Recent development in mobile communications technologies has been towards increased bandwidths and faster data rates. The GSM (Global
System for Mobile Communications) has been one of the most successful communications technologies ever. However, as the relatively slow transmis- sion speed of GSM has been a bottleneck for creating better services to the consumer market, a lot of effort has been put into developing new faster technologies for mobile communications. One such example is EDGE (En- hanced Data rates GSM Evolution). The standardization of EDGE was com- pleted at the European Telecommunications Standards Institute in 1997.
For GSM/EDGE Radic Access Network (GERAN), there have been several new challenges to overcome. Higher data rates are achieved in part by changing channel coding. The transmission in a TDMA (Time Division
Multiple Access) system takes place in time frames. Each frame can be shared among users by dividing the frame into time slots. A TDMA time frame thus comprises physical channels used to physically transfer informa- tion from one place to another. The contents of the physical channels form logical channels, which can be divided into traffic and control channels. The control channels can be further divided into dedicated and common chan- network and the Mobile Stations (MS), whereas the common channels are used for broadcasting diferent information to the MS and for setting up sig- naling channels between the Mobile Swilching Center / Visitor Location Reg- ister (MSC/VLR) and the MSs. Over the radio path, different types of signal- ing channels are used to facilitate the discussions between the MSs and the
Base Transceiver Stations (BT1Ss), Base Station Controllers (BSCs), and the
MSC/VLR. The logical channels are mapped onto physical channels as de-
) . scribed in the technical specification 3GPP TS 45.002 (GERAN Muttipiexing and multipie access on the radio path). in the GSM system, the modulation method used 1s a phase modulation known as Gaussian Minimum Shift Keying (GMSK). in GMSK, : the phase of a true bit is shifted 90°, whereas the phase of a false bit is not shifted. With the increasing data rates of EDGE, new 8 Phase Shift Keying (8-PSK) with 8 possible shift values has been introduced (3GPP TS 45.004).
Each of the shift values corresponds to a certain symbol consisting of 3 bits.
The two types of speech traffic channels used in the GSM are the
Full Rate GMSK Traffic Channel (TCH/F) and the Half Rate GMSK Traffic
Channel (TCH/H). For the TCH/F channels, the voice codecs normally used are the Full Rate (FR) and the Enhanced Full Rate (EFR) codecs. The EFR . speech coder provides the best quality of voice. For the TCH/H channels a
Half Rate (HR) coder is normally used, which consumes less bandwidth as compared to the FR codecs. The HR coder can therefore be used to serve a double number of subscribers on a half rate speech traffic channel as com- pared to an FR coder on a full-rate speech traffic channel.
In order to achieve a better voice quality, a new Adaptive Multi-
Rate (AMR) coder has been introduced (Release 1998). Further, the intro- duction of AMR on TCH/H channels utilizing 8-PSK (O-TCH/H channels) has been considered. However, there has not been any channel coding for the
AMR signaling frames defined for such traffic channels (O-TCH/H). The AMR signaling frames are listed in Table 1.
Table 1. The different AMR signaling frames used on half rate channels.
AMR signaling Purpose frame
SID_FIRST_INH inhibits the second part of a SID_FIRST_P1 if a ) speech onset occurs after DTX inhibits the second part of a SID_UPDATE frame if . SID_UPDATE_INH a speech onset occurs
RATSCCH DATA [conveys the actual RATSSCHmessage
It is not possible to use GMSK for the AMR signaling frames and 8-PSK for the traffic at the same time, because some of the signaling frames, such as for the ONSET signaling messages, share the same bursts as the speech. .
As the same AMR signaling frames are needed for the new half rate channels using 8-PSK modulation (O-TCH/H), a new channel coding for these frames has to be introduced. A straightforward solution to this problem is now discussed with reference to FIG. 1. The numbers under the data flow arrows in the figure denote the number of bits included in a bit block used in the system. The reader is kindly referred to the document 3GPP TS 45.003
V5.1.0 (GSM/EDGE Radio Access Network; Channel Coding) and to the references therein about the different messages referenced below.
A crucial part of the system is the Channel Coder 100 in FIG. 1A.
Usually a block to be transmitted includes Inband Data 101 consisting of two bits. These bits are coded in a coding block 102 using predefined code words, which must be 48 bits in length in order to correspond to the reserved block length. Occasionally, the bits to be transmitted further include
Identification Marker Sequences 103 consisting of 9- or 11-bit sequence.
Eleven bits are used for the RATSSCH_MARKER, for which a repetition of 58 times in the repetition block 104 is required to get the comect total block length of 636 bits. For the other AMR signaling frames a repetition of 71 times is required in the repetition block 104. For the AMR signaling frames
SID_UPDATE, which convey Comfort Noise parameters during a Discontinu- ous Transmission (DTX) period, and for the RATSCCH_DATA AMR signaling ) frame the Comfort Noise parameters 185 also need to be coded. A Cyclic
Redundancy Check (CRC) is performed in the check block 108 to protect the
Comfort Noise against transmission emors. This checksum (14 bits) is added to the Comiort Noise parameters (49 bits total), and the result is fed through a convolutional encoder block 107, which increases the block length to 636 bits.
All signals coming from the Channel Coder 100 are muitiplexed ir a multiplexing block 108. The total number of bits to be sent in a block Is either 684 bits or 1368, depending on the AMR signaling frame. The AMR signaling frames are mapped in the mapping block 109 to 8-PSK symbols, which modifies the block size to 228 or 456 symbols. The symbols resulting from the signaling frames are then interleaved in the interleaving block 110 together with blocks from other frames, which may be speech frames, for example. After the interleaving a burst will be formatted in the burst formation block 111. Then the burst is modulated in a modulation block 112 and di- rected to the transmission block 113.
In FIG. 1B, after receiving a signal in a receiving block 129, the signal must be demodulated in the demodulation block 130. The content of the original burst has to be recovered in the recovery block 131. Because the burst consists of interleaved symbols, they must first be fed through the de- interleaving block 132 and then converted back to bits in the converting block 133. Before the messages can be passed to the Channel Decoder 120, the signaling must be de-multiplexed in the de-multiplexing block 134 so that the
Inband Data part 136 is decoded in the codeword decoder block 135, and the
Identification Marker Sequence 138 in the Identification Marker decoder block 137. if the AMR signaling frame includes Comfort Noise parameters, they are then decoded in the corresponding decoding block 139, and the CRC bits are verified in the verification block 140. Only after this are the Comfort Noise parameters 141 obtained.
The drawback of the solution described above is that both high rate convolutional codes and high rate block codes are required. The convo- lutional coder 107 encodes the 49 bits sequence into 636 bits, and the block coder 102 encodes the 2 bits into 48 bits. In the reverse direction, the convo- lutional decoder 139 decodes the 636 bits into the 49 bits, and the code word decoder 135 decodes the 48 bits into the 2 bits. These so-called fast-rate conversions are not desirable, as they increase the development cost and demand larger coding tables in the network elements and terminals. They are also computationally heavy and consume a lot of memory.
The rate of the convolutional encoder 107 for the Comfort Noise is 1/12, which is more demanding than the % for the GMSK. The constraint length can also be increased from k=5 to k=7, as is done for speech, and the existing polynomials G4-G7 can be used. The interested reader may find descriptions of the polynomials and the constraint lengths in the document 3GPP TS 45.003 V5.1.0 (GERAN Channel Coding). As explained above, the shorter 9 bit sequence of the identification marker has to be repeated 71 times. 5 The purpose of the present invention is to address the problem discussed above. This can be achieved using a method and system for proc- essing AMR signaling frames as described in the independent claims.
Summary of the invention
In order to be able to utilize the existing channel coding developed for chammels with different modulations and to avoid high rate convolutional and block codes, the channel coding has to be performed in a novel way for channels that use a modulation scheme carrying several bits in one symbol.
The present invention provides such a novel mechanism, particularly for traffic channels utilizing the 8-PSK modulation method referred to above.
Using a modulation scheme carrying several bits in one symbol, n bits can be represented with one symbol if there are 2" possible modulation states.
The idea of the invention is that by using bit repetition the current channel encoders and interleavers defined for the use of corresponding
GMSK channels can be utilized when transmitting AMR signaling frames on the GERAN half rate channels utilizing 8-PSK.
Thus, a coding procedure identical to the one already standard- ized for the GMSK channels can be used for the AMR signaling frames. In the method of the invention, in each block of m bits coded by means of said standardized coding process, each bit c(i) is first repeated 3 times {c'(3(i- 1)+1), c'(3(1)+2), c'(3(i-1)+3)}. These 3 bits are then converted into one symbol C(i). After converting all m signaling bits, the resulting symbols C(1), .... C{m) are interleaved. After interleaving, the symbols are modulated and then transmitted. in the receiver, the signal has 10 be demodulated. The received symbols {C(1), C(2), C(3)} are then de-interieaved, after which the symbol
C(i) is converted into 3 bits {C'(3(i-1)+1), c'(3(i-1)+2)), c'(3(i-1)+3)}. The 3- multiple bit sequence has 10 be converted back to the original bit c(i). The sol value of c(i) can be computed by combining the soft values of the 3- miltiple bit sequence received.
Brief description of the drawings
FIG. 1A and 1B show a straightforward solution for transfenng AMR signaling frames on a GERAN half rate channel,
FIG. 2A and 2B illustrate the method of the invention, and
FIG. 3A and 3B illustrate the system of the invention.
Detailed description of the invention
FIG. 2A illustrates the transmission side of the system of the in- vention. Each bit c(i) transmitted from the existing channel encoder block 21, which is defined for the corresponding GMSK channel, is repeated in a repe- tition block 22 n times to correspond to the number of bits carried by one symbol. The value of n depends on the keying algorithm used; for the 8-PSK n=3.The n bits {¢'((i-1)n+1), ..., ¢'((i-1)n+n)} generated on the basis of c(i) are then converted to a symbol C(i) in a conversion block 23. The symbols are then interleaved using an interleaver 24, which is also already defined for the corresponding GMSK channel. It is important not to map one bit directly onto one symbol but to repeat them first, as, depending on the modulation, every time a bit is marked on a symbol there may be a rotation on the constellation.
For example, the 8-PSK symbols are continuously rotated with 32/8 radians per symbol before pulse shaping.
The receiving side of the system, as seen from FIG. 2B, is the re- verse of the sending side. The received symbols C(i) have to be de- interleaved using a known de-interieaver 25 defined for the corresponding
GMSK channel. After this, the n bit sequence {c'((i-1)n+1), ..., ¢'((i-1)n+n)} is obtained using the normal symbol to bit conversion in the conversion block 26. The n bit sequence is combined in a combination block 27 to form a bit value corresponding to the original bit value c(i). Preferably, this is done by combining the soft values of the received n-multiple bit sequence. In this way the bit sequence is reduced to one bit, which is ready to be fed into the chan- nel decoder 28. The channel decoder 28 is substantially the same as the one defined for the corresponding GMSK channel.
In the following, the application of the invention on the AMR signal- ing frames of an 8-PSK modulated HR AMR Speech Traffic Channel (O-
TCH/AHS) is described. On the transmission side (FIG. 3A), the Inband Data
301 to be transmitted is encoded in the coder block 302 of the channel coder part 300. The Identification Marker sequence 303 is repeated in the repeating block 304. Also for the Comfort Noise 305, if included in the signaling frame, a CRC is calculated in the calculation block 306 and then the resulting bit sequence is convolutionally encoded in the corresponding coding block 307.
Blocks 304, 306, and 307 are identical to the respective blocks already stan- dardized for the corresponding GMSK channel (see 3GPP TS 45.003 GE-
RAN Channel Coding).
Thanks to the solution according to the invention, the code words block 302 can be selected to correspond to the already existing TCH/AHS 16 bit code words, for example. Consequently, the tables in the Chamnel Coder 300 need not include longer high-rate codes, thus saving memory. Another benefit of the invention is that if already existing code words are selected, there are only minor changes required in the network elements and in the terminal equipment. The 9-bit sequence of the Identification Marker 303 need not be repeated 71 times but 24 times, and for the AMR signaling frame
RATSCCH the 11-bit sequence of the Identification Marker 303 needs to be repeated only 20 times. After the repetition in the repetition block 304, the
Identification Marker block consists of 212 bits. The Comfort Noise parame- ters 305 can be convolutionally encoded in the coder block 307 with a shorter code, which results in 212 bits. This is computationally cheaper, and, also saves memory. As the convolutional encoder block 307 and the encoder block 302 are identical to the ones already standardized for the correspond- ing GMSK channels, the rate of the convolutional encoder will be %. The constraint length value k=5 and, similarly, the already existing polynomials
G1, G2, and G3 can be used.
The multiplexing block 308 functions similarly to the solution de- scribed above. In the repetition part 309, when the invention is applied to O-
TCHM, the available bandwidth for the AMR signaling frames is threefold (stretched from 228 bits to 684 bits, or from 456 bits to 1368 bits). Each bit is therefore repeated thwee times: a bit c(i) will be mapped to a bit triplet {c'(3(- 1)+1), c(A-1)+2), c'(3(i-1)+3)}. The repeating ensures a correct block length for the AMR signaling frames. When all bits have been repeated, the bit triplets are mapped in a mapping block 310 © 8-PSK symbols C(i) according to Table 1 in 3GPP TS 45.004 (GERAN Modulation). :
The symbols are then interleaved in the interleaving block 311 to- gether with other symbois. and then a burst is formattec in the formation block 312. The bursts are transmitted by the transmission block 314 after modulation in the modulation block 313.
In the receiving part (FIG. 3B) the received signal 320 is first de- modulated in a demodulating block 321. The content of the original burst is recovered in the recovery block 322 and de-interleaved in the de-interleaving block 323. At this stage the signals are symbols C(i), which need to be con- verted back to bit triplets {c'(3(i-1)+1), c'(3(i-1)+2), ¢'(3(i-1)+3)}. The conver- sion is performed using the ordinary 8-PSK conversion table. The triplet has to be combined in the combining block 325 to correspond to one bit. The probable soft value of c(i) can be decided, for example, by combining the soft values of the received bit triplet, for instance c(i)=0.4*c'(3(i-1)+1)+0.4*c’(3(i- 1)+2)+0.2*c’(3(i-1)+3).
The combined soft values c(i) of the bits are then de-multiplexed in the de-muitiplexing block 326, before the channel decoder 340 processes them. The Inband Data 328 is then decoded in the decoding block 327 using the code word table, whereas the Identification Marker sequence 330 is de- coded in the Identification Marker decoding part 212. If present, the Comfort
Noise 333 is obtained after convolutionally decoding the bit sequence coming from the demultiplexer 326 in the decoding block 331 and CRC checking them in the checking block 332. Again, the decoder 340 and de-interleaver block 323 are the same as those used for the TCH/AHS traffic channels.
Although the invention was described above with reference to the examples shown in the appended drawings, it is obvious that the invention is not limited to these, but may be modified by those skilled in the art without departing from the scope and spirit of the invention. For example, another suitable modulation scheme, such as 16-PSK, can be selected to implement the invention.

Claims (5)

‘ Claims
1. A system for processing an AMR signaling frame to be trans- mitted on a GERAN half rate channel utilizing a modulation scheme carrying n bits (n 2 2) in one symbol, the system comprising: - coding means (300), compatible with coding means defined for a GSM half rate channel utilizing a GMSK modulation scheme, said coding means being adapted to output a coded bit stream, - repeating means (309), responsive to the coding means, for re- peating each bit of the coded bit stream n times. whereby a repeated bit stream Is obtained. and - symboi-forming means (310). responsive 1c saic repeating means. for forming symbols to be transmitted from said repeated bit stream.
2. A system for processing an AMR signaling frame received on a GERAN half rate channel utilizing a modulation scheme carrying n bits (n22) in one symbol, the system comprising: - conversion means (324), for converting a received symbol stream to a first bit stream, - processing means (325), for segmenting the first bit stream into successive blocks, each block consisting of n bits. and for converting each block into one bit, whereby a second bit stream is obtained, and - decoding means (340), compatible with decoding means defined for a GSM half rate channel utilizing a GMSK modulation scheme, said de- coding means being adapted to decode said second bit stream.
3. A method for processing an AMR signaling frame to be transmitted on a GERAN half rate channel utilizing a modulation scheme carrying n bits (n 2 2) in one symbol, the method comprising the steps of: - coding the AMR signaling frame using a coding procedure de- fined for a GSM half rate channel utilizing a GMSK modulation scheme, the coding procedure outputting a coded output stream, - in response to said coding, repeating each bit of the coded out- put stream n times, whereby a repeated bit stream is obtained, and ‘ - forming symbols to be transmitted from the repeated bit stream.
4. A method for processing an AMR signaling frame received on a GERAN half rate channel utilizing a modulation scheme carrying n bits (n 2 2) in one symbol, the method comprising the steps of:
- converting a received symbol stream into a first bit stream. - segmenting said first bit stream into successive blocks. each block consisting of n bits, - converting each block into one bit, whereby a second bit stream is obtained, and - decoding said second bit stream using a coding procedure de- fined for a GSM half rate channel utilizing a GMSK modulation scheme.
5. A system for transferring an AMR signaling frame on a GERAN half rate channel utilizing a modulation scheme carrying n bits (n = 2) in one symbol, the system comprising - at a transmitting end: -coding means (300), compatible with coding means defined for a GSM half rate channel utilizing a GMSK modulation scheme, said cod- ing means being adapted to output a coded bit stream, - repeating means (309), responsive to the coding means, for repeating each bit of the coded bit stream n times, whereby a repeated bit stream is obtained, and - symbol forming means (310), responsive to said repeating means, for forming symbols to be transmitted from said repeated bit stream, and - at a receiving end: - conversion means (324), for converting a received symbol stream to a first bit stream, - processing means (325), for segmenting the first bit stream into successive blocks, each block consisting of n bits, and for converting each block into one bit, whereby a second bit stream is obtained, and - decoding means (340), compatible with decoding means defined for a GSM half rate channel utilizing the GMSK modulation scheme, said decoding means being adapted to decode said second bit stream.
ZA200401569A 2004-02-27 2004-02-27 A method and a system for transfering AMR signaling frames on halfrate channels. ZA200401569B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
ZA200401569A ZA200401569B (en) 2004-02-27 2004-02-27 A method and a system for transfering AMR signaling frames on halfrate channels.

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
ZA200401569A ZA200401569B (en) 2004-02-27 2004-02-27 A method and a system for transfering AMR signaling frames on halfrate channels.

Publications (1)

Publication Number Publication Date
ZA200401569B true ZA200401569B (en) 2004-12-29

Family

ID=34574975

Family Applications (1)

Application Number Title Priority Date Filing Date
ZA200401569A ZA200401569B (en) 2004-02-27 2004-02-27 A method and a system for transfering AMR signaling frames on halfrate channels.

Country Status (1)

Country Link
ZA (1) ZA200401569B (en)

Similar Documents

Publication Publication Date Title
US7817679B2 (en) Method and a system for transferring AMR signaling frames on halfrate channels
US6608828B1 (en) Methods and systems for decoding headers that are repeatedly transmitted and received along with data on a radio channel
FI109251B (en) Communication method, radio system, radio transmitter and radio receiver
US8730999B2 (en) Method and system for reduction of decoding complexity in a communication system
US7188300B2 (en) Flexible layer one for radio interface to PLMN
US20030135797A1 (en) Method and apparatus for enhancing the transmission of error in the IEEE 802.11e systems
JP2001508608A (en) Apparatus and associated method for transmitting and receiving multistage encoded and interleaved digital communication signals
AU757352B2 (en) A method, communication system, mobile station and network element for transmitting background noise information in data transmission in data frames
US7342956B2 (en) System and method to extract uplink status flag bits in a cellular wireless network
JP4276261B2 (en) Bit-swapping method, bit-swapping apparatus and computer program
US20030093747A1 (en) Method for transporting real-time data frames comprising at least two bit portions having different relevance, corresponding transmitter and receiver
CA2762915C (en) A method and a system for transfering amr signaling frames on halfrate channels
ZA200401569B (en) A method and a system for transfering AMR signaling frames on halfrate channels.
WO2004047355A1 (en) Method and apparatus for channel coding and rate matching
MXPA01002701A (en) Method and system for alternating transmission of codec mode information