RU2395909C2 - Method and module of transfer for adaptive coding, modulation and transfer of information words in system of wireless communication - Google Patents

Abstract

FIELD: information technologies.

SUBSTANCE: as data is sent from module of transfer into module of reception along channel, there are own losses of throughput capacity observed for channel in system of wireless communication, with high variation of channel quality by its frequency range, for which purpose the following is suggested: to assess (301) quality of each subchannel in appropriate channel; to classify (302) subchannels into several quality groups based on assessed quality; and to select (303) frequency of code for a group of quality. Then specified selected frequency of code for group of quality is used, when information word, which is to be transferred, is coded (304), at the same time subchannels are grouped into a limited number of groups, for instance, from one to four groups, depending on variation of channel quality.

EFFECT: provision of high throughput capacity with minimum expenses for transfer of signals in the form of additional service information.

15 cl, 6 dwg

Description

FIELD OF THE INVENTION

The present invention relates to transmission methods and modules for encoding, modulating, and transmitting information words in a wireless communication system, and more particularly, relates to encoding, modulating, and transmitting information words to a receiver over a channel so that high throughput of the communication channel is achieved.

BACKGROUND

The data to be transmitted from the transmitter to the receiver in the communication system is basically encoded and modulated before being transmitted over the communication line between the transmitter and the receiver.

In wireless communication systems, the communication line between the transmitter and receiver is a wireless communication line. Hereinafter, this link is called a channel. The channel uses a certain frequency range for data transmission. A channel can be divided into subchannels, with each subchannel using a certain fraction of the frequency interval to transmit data. Data is encoded and modulated before being transmitted over the channel. In communication systems, basically, one modulation and coding scheme is used to encode and modulate all the data that must be transmitted over the communication line from the transmitter to the receiver within a certain time period, called, for example, the signal transit time interval (TTI), or time frame.

In wireless communication systems such as GSM (Global System for Mobile Communications), W-CDMA (Code Division Multiple Access Broadband), CDMA2000 (2000 Code Division Multiple Access), WiMAX, etc., channel quality over the frequency interval can vary much more than the quality of the communication line in fixed communication systems. As a result, the encoded information word will usually be transmitted over subchannels having different qualities. Channel quality can be estimated by measuring the signal-to-noise and noise (SINR) ratio over a frequency span. For broadband wireless systems, the channel spans a wide frequency range, i.e. the frequency span is wide. Therefore, the variation in channel quality over the entire frequency interval within the TTI interval is usually higher than in narrow-band systems.

For this reason, channel adaptation is widely used, as a result of which the modulation and coding scheme is dynamically selected for the channel in accordance with the state of the channel, for example, channel quality. This procedure is carried out, for example, in technologies of high-speed satellite packet access (HSDPA) or code-division multiple access only for evolutionary data (CDMA EV-DO). CDMA EV-DO technology is the first step in the evolution of CDMA2000 with some overhead channels only. In existing wireless communication systems, channel adaptation is traditionally calculated such that one modulation mode and one code frequency are used per channel TTI interval or time frame. That is, to encode and modulate data on all subchannels used for transmission on one TTI interval, the same modulation mode and the same code frequency are used.

When transmitting data over a channel, it is important to achieve high channel throughput for some level of power transmission.

The simulation performed by the authors shows that there are intrinsic losses in throughput for a channel experiencing a high variation in channel quality. According to the results of simulation, if a channel experiences a high variation in channel quality, then there is a need for a lower uniform code frequency than if a channel experiences a low variation in channel quality. Or if, in a different way, the greater the quality variation, the higher the transmit power is needed to achieve the same given block frequency with errors (BLER). The indicated simulation results are shown in the diagram of FIG. 1. In simulation, each subchannel of one channel can receive either the first or the second signal-to-noise ratio (SNR) state. Illustrative figures are given in the form of SNR ratios, because they are the results of a simulation of a communication line or the simulation results of a single cell area, where the SNR ratio is SINR. The curves in the diagram show the required average SNR ratio and the corresponding code frequency to achieve some BLER frequency. Each curve corresponds to a certain offset in dB between two states of the SNR relationship. As can be seen in FIG. 1, for some averaged SNR ratio, a lower coding rate is needed to achieve the same BLER frequency when the two SNR states are widely spaced compared to when they are closer to each other. For example, when two states have the same level, that is, a 0 dB offset (0 dB curve), the coding frequency should be 0.9 for the average SNR of the 8 dB ratio, but when the two states are offset by 30 dB (see 30 dB curves, that is, the curve is the farthest in the diagram), but the average SNR level of 8 dB is still used, then the coding frequency should be 0.43. Conversely, when two states are offset by 0 dB and a coding rate of 0.43 is used, then to achieve the same BLER frequency, the average SNR of the 8 dB ratio should be used, comparable to the average SNR of the 8 dB ratio for the case of a 30 dB offset the same coding rate of 0.43.

To further give some concepts of the effect of channel quality variation on channel capacity, the additional results of simulation are given in the four diagrams of Fig. 2. The diagrams of FIG. 2 show the normalized throughputs of various channel quality conditions. 2, it is assumed that the SNR relation has a log-normal distribution. In the upper left diagram, the standard deviation for the channel SNR is 5 dB, in the upper right diagram it is 10 dB, in the lower left diagram 15 dB, and in the lower right diagram it is 20 dB. The solid line in each diagram shows the throughput for perfect adaptation of the communication line, that is, each subchannel is assigned an individual modulation and coding scheme (MCS) based on its SNR ratio level (or state). The dashed line in each diagram shows the throughput for selecting one MCS scheme per frame, that is, when one modulation and coding scheme is selected for all subchannels regardless of the SNR ratio level. In the results obtained by simulation, for the same MCS circuit and for perfect adaptation of the communication line, the same code block size was used, so that the methods of one MCS circuit and perfect adaptation of the communication line can be compared, regardless of the block size. Also, the possible differences in the sizes of the necessary service information were not taken into account. In figure 2, the modulation was selected from the modulation schemes BPSK (binary phase shift keying), QPSK (quadrature phase shift keying), 16QAM (hexadecimal quadrature amplitude modulation), 64QAM (six-decimal quadrature amplitude shift modulation) for both the case of one circuit and the case of perfect adaptation of the communication line. Since it was necessary to use two bits per OFDM symbol to transmit a message to the receiver, the normalized channel capacity with several states increases from 0 to a maximum of 6 bits / symbol with an increase in the mathematical expectation of the SNR ratio in dB.

Figure 2 shows that when the assumption of the same block size and the same amount of overhead information is used, there is a difference in throughput between the case when the same MCS scheme is used for the entire channel and the case when an individual MCS scheme is selected for each subchannel . Figure 2 also shows that the difference increases with increasing channel variation. That is, the results of simulation showed that there is a clear opportunity to improve the adaptation of the communication line when the channel varies significantly during one TTI interval / frame. It can be noted that the difference in the quality of operation shown in FIG. 2 relates to practical codes and code block sizes. In the ideal case of the correct codes and infinitely large code blocks, we can expect that the difference in the quality of functioning will disappear when the same block size is used.

As shown above, the disadvantages of selecting one MCS scheme per frame or TTI interval include:

- loss of bandwidth in the channel with a significant variation in quality, since the imperative uniform code frequency and modulation mode limit the bandwidth for data sent over the subchannel with high quality (for good channel conditions);

- channel estimation is more error prone for low quality channel conditions. With poor channel ratings in the uniform adaptation of the link to the MCS circuit, the inclusion of low quality states effectively subtracts energy from the best quality states.

Figures 1 and 2 show that the throughput would be higher if the modulation and coding scheme could adapt to the channel quality level, or to the SNR level if the same block sizes are assumed. Then it can be assumed that the solution should use the method of perfect adaptation of the communication line, where MCS adaptation is performed for each channel, depending on its quality status. Although there are other disadvantages of the method of perfect adaptation of the communication line:

- The quality of operation or the actual throughput is limited, because the code blocks must be small, because each fraction of the information word has its own MCS circuit. In the simulation of FIG. 2, it was assumed that both methods had the same code block size so that it was possible to compare the throughput of the two methods, regardless of the size of the code blocks;

- too much overhead is needed to inform the receiver of the MCS scheme used in each code block.

In general, when the size of the code block and the size of the service information are turned on, the real throughput for the method of perfect adaptation of the communication line is not high.

There have also recently been offers of mixed modulation per TTI slot / frame for link adaptation. Such a proposal is described in the delivery of standardization software presented in 3GPP TSG RAN WG1 # 42 on LTE Rl-050942 in London August 29 - September 2, 2005 from NTT DoCoMo, NEC, SHARP called AMC and HARQ Using Frequency Domain Channel-dependent Scheduling in MIMO Channel Transmission. The specified document gives the adaptation of modulation for each piece of data or a subchannel depending on its status or quality level. Such a scheme causes a slight improvement in performance since higher order modulation is used for subchannels with a high SINR ratio and lower order modulation is used for subchannels with a low SINR ratio. However, the mixed modulation scheme requires significant additional signaling costs compared with the case of one MCS scheme, since for each fraction of the transmitted encoded and modulated information word, it is necessary to transmit information to the receiver which modulation scheme was used. If there are four different modulation schemes from which to choose, then each fraction of the encoded and modulated information word should include two bits formulating the modulation scheme used.

As shown above, there is a need to achieve high channel throughput, and it is possible to improve throughput for a channel experiencing high quality variation.

SUMMARY OF THE INVENTION

An object of the present invention is to achieve high throughput for data transmitted from a transmission module to a channel receiving module in a wireless communication system, the channel experiencing a large quality variation in its frequency range.

The aforementioned problem is solved by means of a certain method, transmission module and computer software product formulated in the characterizing part of the independent claims.

According to the invention, the throughput for a channel experiencing high quality variations is increased by evaluating the quality of each subchannel and classifying the subchannels into several quality groups, based on the estimated quality of each subchannel, and by choosing one code frequency for each quality group. When an information word is encoded, each fraction of the information word should be encoded in accordance with the selected code frequency, depending on which subchannel the fraction of the information word should be modulated. Channel quality should be assessed by measuring the signal-to-noise-to-noise ratio (SINR) or the signal-to-noise ratio (SNR) over the channel frequency range or by measuring the received bit rate on one subchannel in the receiving and transmitting module to the transmission module.

Simulation results have shown that near optimal bandwidth is achieved when the subchannels of a channel experiencing high quality variations are preferably classified into two, three or four groups if the channel is experiencing a very high quality variation, and into two groups if the channel is experiencing more moderate variation in quality. It may also be that all subchannels are classified into the same group if the channel has low quality variations. A good value for the spread of channel quality within one group is the spread of 0.3 received bit rate (RBIR).

An advantage of the present invention is that by classifying subchannels into groups depending on their quality level of the subchannel and by selecting one code frequency per group, the channel capacity for a certain transmit power is increased compared to using a constant code frequency.

Another advantage of the present invention is that a performance similar to that of using one code frequency per subchannel or one state, but with lower signaling costs, can be achieved.

Another advantage of the present invention is that if an embodiment of the present invention uses only two, three or four different quality groups, then only one or two bits are needed to transmit with the used frequency of the code to one code block, that is, high throughput is achieved with low signaling costs.

Another advantage of the present invention is that using any of the methods disclosed in the description to implement many code frequencies within a single coding unit, the quality of operation is further improved.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention is further explained in the description of specific options for its implementation with reference to the accompanying drawings, in which:

figure 1 depicts a diagram of the results of simulation performed by the inventors, the diagram shows the SNR ratio required for the decoder in the receiver (y axis) for various coding frequencies (x axis);

figure 2 shows various diagrams of the results of the simulation of channel throughput for different averaged values of the SNR of the ratio for four cases when the standard deviation of the SNR of the ratio is 5, 10, 15 or 20 dB;

Figure 3 depicts a procedural flowchart according to an embodiment of the method of the present invention;

4 is a flowchart according to another embodiment of the method of the present invention;

5 shows a schematic diagram of a transmission module according to an embodiment of the present invention;

6 shows various simulation results charts comparing various embodiments of the present invention with the prior art, the charts show channel throughput for various average SNR ratios when the standard deviation of the SNR ratios is 5, 10, 15, or 20 dB.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

The present invention will now be described in more detail with reference to the accompanying drawings, in which preferred embodiments are shown. However, the present invention can be implemented in many different forms, and should not be construed as limited by the embodiments presented here, rather, the above-mentioned embodiments are provided so that the description is thorough and complete, and will fully convey to the skilled person the scope of the present invention.

In wireless communication systems, especially for broadband systems, and even more specifically for broadband systems, which are systems based on orthogonal frequency division multiplexing (OFDM), a channel can have a large variation in channel quality over its frequency range, so that the encoded and modulated information word during transmission will experience channel conditions of very varying channel quality.

As can be seen from the results of the simulation performed by the inventors shown in FIGS. 1 and 2, there are intrinsic bandwidth losses for a channel experiencing a high variation in channel quality.

In order to be able to achieve high throughput of the transmitted data, even for a channel experiencing a high variation in channel quality over its frequency range, according to the method of the present invention, it is proposed:

evaluate the quality of each subchannel;

classify subchannels into several quality groups based on the estimated quality of each subchannel; and

Choose a certain frequency of code per quality group.

After that, the information word to be transmitted is encoded so that each fraction of the information word is encoded in accordance with the selected code frequency depending on which subchannel the fraction of the information word should be modulated. That is, subchannels are classified into several groups depending on their channel quality levels. For each group, the code frequency is selected to maximize or nearly maximize the bandwidth of the group. The information word is encoded so that the fractions of the information word that must be transmitted over a certain subchannel are encoded with the code frequency selected for the class into which this subchannel is classified.

According to one embodiment of the present invention, the code frequency for each group can be selected to maximize or nearly maximize the bandwidth of the group, as if the group was a permanent channel, that is, some channel without variation. For optimal adaptation of the constant channel link, the mapping of the SINR to capacity is a convex curve. Therefore, the code frequency could be selected for the group according to the following formula:

Medium (throughput (SINR_i)) ≥ throughput (average [SINR_i]).

Inequality claims that the average throughput as a function of the SINR ratio for each individual subchannel in a group can be higher than the throughput as a function of the average for all SINR relationships of the subchannels in a group. The frequency of the code per group can be selected according to the right side of the inequality. Since the ratio of SINR to throughput is a convex curve, in order to select the code frequency according to the right-hand side of the inequality, there must be an approximation to the optimal choice, which one way or another cannot be too optimistic. Therefore, there is no risk that the code frequency is too high (i.e., the encoding is too weak).

A similar formula can be used to display the ratio of RBIR to throughput. Other methods may also be used to select a code rate per group to maximize group throughput.

By classifying subchannels into quality groups and choosing a code frequency for all subchannels of the class, it is possible to choose a code frequency suitable for each subchannel, comparable to the case when a uniform MCS scheme is used, which can improve the quality of operation. At the same time, the size of code blocks can be kept large. Also, especially if the number of groups is kept low, then only a small number of bits are needed to send a message to the receiver about the selected code frequency. In general, the channel capacity will increase compared to using one MCS scheme for the entire channel. Also, since overhead information can be kept low, the throughput should be higher than when individual MCS schemes should be used for each subchannel. Simulation results showed that two, three, or four different quality groups are the best choice because, as can be seen in the simulation results shown in Fig. 6, the throughput must be high, and additional signaling overhead can be kept low. . For example, if only two groups are selected, then only one additional bit per code block is needed to send a message to the receiving module about the selected code frequency. The optimal number of groups should depend on the variation in channel quality: the higher the variation in channel quality, the higher the optimal number of quality groups. For a very high variation in channel quality, the optimum should be up to four groups, since the choice of the best code frequency should increase the bandwidth; more than an additional bit of overhead data should reduce the bandwidth. On the other hand, if the variation in channel quality is low, then only one quality group may be sufficient.

FIG. 3 depicts a method according to an embodiment of the present invention for transmitting information words from a transmission module to a channel receiving module in a wireless communication system. The method begins with an assessment of 301 quality of each subchannel of a channel. The subchannel spans the subharmonic interval, which is a fraction of the frequency interval used by the transmission channel. The subharmonic interval is covered by a subchannel depending on the evaluation method. According to one embodiment, the quality can be estimated based on recently obtained or instantaneous channel quality measurements, sampled for each subharmonic over a frequency interval, or on a statistical quality distribution. Then the subharmonic interval may depend on how many samples are taken over the channel frequency interval. After performing a quality assessment, the subchannels are classified 302 into several reception quality groups depending on the assessed quality for each subchannel. For each group, a 303 code frequency is selected that maximizes the bandwidth of the group. Then, the information word is encoded 304 with the selected code frequency so that depending on which subchannel the information word fraction should be modulated, each information word fraction is encoded according to the selected code frequency. After that, a modulation method 305 is selected. The modulation method is selected, for example, by choosing a uniform modulation method for all channels, or by choosing a modulation method for one quality group, or for one subchannel, or any combination thereof. After that, the encoded information word is modulated 306, and the encoded and modulated information word is transmitted 307 through the channel to the receiving module.

The following describes another method according to the present invention. The method uses, for example, data transmission from a transmission module to a reception module in a wireless communication system based on orthogonal frequency division multiplexing (OFDM). Figure 4 describes a procedural flowchart for adapting a communication link with many code frequencies according to the method. The method uses the concept of a data region to define a small area in a temporary and private domain with a small variation in channel quality. It is comparable with the subchannel of figure 3, for a short period of time for which the quality of the channel is evaluated. The method begins by assessing 401 the quality of the data region and sorting the data regions, respectively. Evaluation may be conducted by a transmission module or a reception module. The quality of the data section, for example, can be evaluated by the reception module, measuring the quality of the data section and transmitting it to the transmission module. Then the data sections are classified 402 preferably into two, three or four groups depending, for example, on the variation in channel quality. There are various ways to classify data sections. One example is based on normalized mutual information per encoded bit (MIPB), also called received bit rate (RBIR). RBIR is a measure of the information transfer efficiency of a data section to which a certain modulation scheme is applied, at a certain noise level, if the channel was ideal. Numerical studies have shown that it is an effective quality measure to use when setting the range of quality measures for a group. Classification is performed to limit the RBIR dynamic range to a certain range for each group. According to an advantageous embodiment of the present invention, the dynamic RBIR range is limited to about 0.3 for each group. Simulation results showed that if each group has a quality variation within RBIR 0.3 or approximately RBIR 0.3, then the channel can achieve a throughput that is very close to optimal. The corresponding simulation results are given in Fig.6. After that, for each group, the communication line (or channel) is adapted 403 by choosing a modulation mode or code frequency adapted for the channel state, before the information word is encoded and modulated with the selected code frequencies and modulation modes and, ultimately, transmitted to receiver. Also, when classifying data sections or subchannels into groups, other channel quality indicators can be used. Channel quality indicators may be SINR, RBIR, bandwidth or normalized bandwidth, or any other non-linear SINR function, or any other channel quality indicator. After evaluating the quality of the channel in order to obtain a channel quality indicator, the classification is based on the channel quality indicator.

There are various ways to apply different code frequencies to an information word. According to a first embodiment of the present invention, each group has an individual code block with a predetermined code frequency.

According to a second embodiment of the present invention, the groups share the same code block, but still have different code frequencies. This requires modifications for frequency matching algorithms and also, probably, for the interleaving configuration. In this embodiment, uneven error protection (UEP) is used to set different selected code frequencies for different groups. The latter requires some modifications of the frequency matching algorithms for the so-called turbo code specified by the body of the 3GPP standardization program and published as 3GPP TS 25.212 technical specification if this code is used, and preferably also for the interleaving configuration. Examples of possible methods for uneven error protection are as follows:

- Repeat in groups that require a low code frequency. In this case, the initial encoding is selected based on the code frequency for the best group, that is, the group that has the highest code frequency. For groups with lower selected code frequencies, the frequency of the code is reduced by repeating the encoded data in the private domain, so that the protection ability is increased. Such a repetition can be performed by other implementations such as, for example, scatter with some scatter configuration, etc .;

- Use an uneven interrupt frequency. In this case, the initial encoding is based on the group that has the lowest code frequency selected. For groups with higher selected code frequencies, some of the encoded bits are deleted to achieve a higher code frequency. An example of a convolutional code is a circuit called as non-uniform error protection, which is based on a frequency compatible flash convolutional code (RCPC). The RCPC code is described in the publication "Multirate convolutional codes", Frenger et al., In technical report 21 at the Communications Systems Group of Chalmers University of Technology in April 1998. A convolutional code is a code that decides which bit to delete. It is also based on some mother code, for example 1/3 of the frequency, with a given interrupt configuration in the RCPC code.

- To divide the shares of the information word into various parts, the first part of the share of the information word being transmitted to high quality groups, and the second part of the share of the information word is transmitted to low quality groups. For example, the best group with a high code rate contains only part of the 1st transmitted packet. The low code rate group contains a portion of the 1st transmitted packet and the 2nd transmitted packet. For turbo codes, a structure similar to the Hybrid Automatic Repeat Request - Incremental Redundancy (HARQ-IR) structure is possible. The order of the transmitted bit sequence may be based on the order of the HARQ-IR structure.

Combinations between the first and second embodiments of different code frequencies for the information word may also be used.

The modulation mode can be selected, for example, for each group. According to one embodiment of the present invention, each group has one modulation mode for all subchannels of the quality group. According to another embodiment of the present invention, the modulation mode is adapted to the state of each subchannel, for example, the modulation mode is selected per data section in an OFDM system. If the modulation mode is selected per data section or subchannel, more overhead transmission is required.

5 describes a transmission module 500 according to one embodiment of the present invention, configured to transmit information words to a reception module over a channel in a wireless communication system. The transmission module comprises a processor 504, which has means for evaluating 506 the quality of each subchannel of said channel. The quality of each subchannel can be estimated based on measurements showing the quality of each subchannel. Measurements can be taken in the processor 504 through a receiver 505 of some node in which a transmission module, such as a mobile station, can be located. The processor 504 also has means 507 for classifying subchannels into several groups based on the estimated quality of each subchannel. The processor also has a selection tool 508 arranged to select a code frequency for each group that maximizes the bandwidth of the group. A selector can also be arranged to select a modulation method for each group or each subchannel. The transmission module 500 comprises an encoder 501 for encoding a certain information word so that each fraction of the information word is encoded according to the code frequency selected by the processor, depending on which subchannel the fraction of the encoded information word should be modulated. For this reason, the processor is arranged to inform the encoder of the selected code frequency. The transmission module also comprises a modulator 502 arranged to receive the encoded information word from encoder 501 and to modulate the encoded information word. The processor 504 may also be arranged to inform the modulator 502 of the modulation method that the modulator should use to modulate the encoded information word. The transmission module further comprises a transmission unit 503 arranged to receive the encoded and modulated information word from the modulator and to transmit the encoded and modulated information word to the reception module. Evaluation tool 506, classification tool 507, and selection tool 508 may be separate modules in transmission module 500, and may also be located in a processor.

The transmission module according to the present invention is mainly described as embodied in hardware. Although the invention may also be embodied in software, or in a combination of software and hardware. In this case, to perform the steps according to the paragraphs of the method of the present invention, appropriate means with computer software in the transmission module can be used.

6 shows simulation results for an OFDM system. The results show a comparison between the various link adaptation solutions. All simulation results are listed below, where LA1 and LA2 are ideas proposed by the present invention. The illustrative figures are given in the form of a signal to noise ratio (SNR), since they are the results of a simulation of a communication line in which the SNR ratio is equal to the signal-to-noise-to-noise ratio (SINR).

LA0: Perfect link adaptation, that is, an adaptation of the modulation and code frequency for each channel state (or each subchannel, based on its channel state).

LA1: Numerous code frequencies with link adaptation, i.e. adaptation of modulation to a data section.

LA2: Multiple code frequencies with one modulation mode for each code frequency (adaptation of the MCS scheme in groups).

LA3: The only code frequency with modulation adaptation over data sections.

LA4: The only MCS circuit.

Simulation results include:

An OFDM system with 100 pieces of data per frame, with 128 characters per piece of data is assumed. The transmitted power is common to all data areas.

A channel with many states is modeled as 100 states with a normal SNR_dB distribution, with standard deviations of 5 dB in the upper left diagram, 10 dB in the upper right diagram, 15 dB in the lower left diagram, and 20 dB in the lower right diagram. A large variation (20 dB) is used to simulate a case with a very wide band, or when the user is close to the border of the cell area.

Estimated modulation modes are {BPSK, QPSK, 16QAM, 64QAM}. The receive module uses the BICM-logMAX algorithm.

To achieve a given BLER of 1%, the coding adaptation can choose any code frequency from 0 to 1.

In both, the transmission module and the reception module, knowledge of the perfect channel is assumed.

The classification of data sections is based on RBIR, by restricting the RBIR range of each group, designated D _RBIR , to within 0.3.

For all communication line adaptation decisions, the same block size is assumed. Also, the costs of signal transmission due to various sizes of service information are not considered.

In FIG. 6, it can be seen that when the channel varies significantly within one frame, LA1 and LA2 are clearly superior to LA4. Case LA1 is very close to perfect adaptation of the communication line, that is, LA0, LA2 and LA3 are quite close to each other in terms of quality of operation.

The following compares signal transmission costs for five different link adaptation solutions:

- LA4 only requires MCS circuit communication between the transmitter and the receiver, therefore the cost is the smallest overhead.

- LA2 is the second most effective case, which requires information from numerous MCS circuits and additional signaling to inform each section of the data about the group index. If only two groups are used, which is enough to achieve the results according to Fig.6, then to mark the group index, one bit is enough.

- LA3 implements adaptation of modulation over data sections with a common code frequency, therefore, its signal transmission costs are close to LA2. Although, if only two groups are used, one bit of code frequency information is enough, then two bits are needed to mark the modulation mode from among BPSK, QPSK, 16QAM and 64QAM for each data segment.

- LA1 requires some additional signaling compared to LA3 in order to communicate to each data region information of multiple code frequencies and group index for each data region.

- LA0 is the least efficient case, because it requires transmitting code frequency information and modulation mode information for each data section to the receiving module.

As shown above, the total bandwidth on a channel experiencing a high variation in channel quality over its frequency range will increase if the subchannels are classified into groups, and in which each group must use the selected code frequency. When using one code frequency per group, signal transmission costs should be low, especially when only a limited number of groups are used, for example, from one to four groups, depending on channel variation, but throughput should still be high regardless of costs to transmit signals.

The preferred embodiments of the invention have been disclosed in the drawings and description, and although specific terms are used, they are used only in a general and visual sense, and not for purposes of limitation, and the scope of the invention is defined in the claims below.

Claims (15)

1. A method of transmitting information words from a transmission module to a channel receiving module in a wireless communication system using adaptive channel coding, the channel being divided into several subchannels, the channel being allocated a corresponding frequency interval for transmission and each subchannel being allocated a corresponding subharmonic interval, each the subharmonic interval is a fraction of the frequency interval, the method comprising the steps of
encoding (304) the information word with code frequencies;
modulating the encoded information word into several subchannels, wherein one share of the encoded information word is modulated to one subchannel;
transmitting (307) the encoded and modulated information word over the channel to the receiver;
characterized in that it also contains steps
assessments (301) of the quality of each subchannel;
classifying (302) subchannels by the number of quality groups depending on the variation in channel quality based on the estimated quality of each subchannel; and
selection (303) of the frequency of the code per quality group, taking into account the expected frequency of errors in bits of the mentioned estimated quality;
wherein the encoding step of the information word comprises encoding the fractions of the information word that must be modulated into subchannels belonging to the same quality group with the code frequency selected for the quality group in an individual code block of the quality group or in a code block shared for a plurality of quality groups. 2. The method according to claim 1, wherein the step of modulating the encoded information word comprises selecting (305) one modulation method for each quality group and modulating (306) each fraction of the encoded information word according to the selected modulation method, depending on which subchannel should each fraction of the encoded information word is modulated. 3. The method according to claim 1, wherein the step of modulating the encoded information word comprises selecting (305) a modulation method for each subchannel and modulating (306) each fraction of the encoded information word according to the selected modulation method, depending on which subchannel should be modulated share of the encoded information word. 4. The method according to any one of claims 1 to 3, in which the step of evaluating (301) the quality of each subchannel is based on measuring the signal-to-noise ratio, SNR, or signal-to-noise and noise-to-noise ratio, SINR, on each subchannel. 5. The method according to any one of claims 1 to 3, in which the step of evaluating (301) the quality of each subchannel is based on measurements of the received bit rate, RBIR, on each subchannel. 6. The method according to claim 5, in which the subchannels are classified (302) so that the RBIR parameters for the subchannels of the same quality group have a dynamic range not exceeding 0.3. 7. The method according to claim 1, wherein the classification step (302) comprises the classification of subchannels into two, three or four different quality groups. 8. A transmission module (500) in a wireless communication system, adapted to transmit information words to a receiving module using adaptive channel coding over a communication channel in a wireless communication system, the channel being divided into several subchannels, the channel being allocated a corresponding frequency interval for transmission and each subchannel is allocated a corresponding subharmonic interval, each subharmonic interval being a fraction of the frequency interval containing an encoder (501) adapted for encoding information onnogo code word frequencies; a modulator (502) adapted to modulate the encoded information word into several subchannels, the modulator being further adapted to modulate the proportion of the information word to one subchannel; a transmitting unit (503) adapted for transmitting a coded and modulated information word over a channel to a receiver;
characterized in that it also contains means (506) for assessing the quality of each subchannel; means (507) for classifying subchannels by the number of quality groups, depending on the variation in channel quality, based on the estimated quality of each subchannel; and selection means (508) arranged to select a code frequency per quality group, taking into account the expected error rate of the bits of the estimated quality; wherein the encoder (501) is also arranged to encode fractions of the information word that should be modulated into subchannels belonging to the same quality group with a code frequency selected from a quality group in an individual code block of a quality group or in a code block shared for multiple quality groups . 9. The transmission module (500) according to claim 8, in which the selection means (508) are also arranged to select one modulation method for each quality group, and the modulator (502) is also arranged to modulate each fraction of the encoded information word according to the selected method modulation depending on which subchannel each fraction of the encoded information word should be modulated. 10. The transmission module (500) of claim 8, wherein the selection means (508) is also arranged to select a modulation method for each subchannel, and the modulator (502) is also arranged to modulate each fraction of the encoded information word according to the selected modulation method depending on for which subchannel each fraction of the encoded information word should be modulated. 11. The transmission module (500) according to any one of claims 8 to 10, wherein the estimator (506) is arranged to evaluate the quality of each subchannel based on measurements of the signal-to-noise ratio, SNR, or signal-to-noise and noise-to-noise ratio, SINR, on each subchannel. 12. The transmission module (500) according to any one of claims 8 to 10, wherein the estimator (506) is arranged to evaluate the quality of each subchannel based on measurements of the received bit rate, RBIR, on each subchannel. 13. The transmission module (500) of claim 12, wherein the classification tool (507) is arranged to classify the subchannels so that the RBIR parameters for the subchannels of the same quality group have a dynamic range not exceeding 0.3. 14. The transmission module (500) of claim 8, wherein the classification tool (507) is arranged to classify the subchannels into two, three or four different quality groups. 15. A digital computer device with memory located in the transmission module (500) and containing parts of the system program executed on the computer device and intended to perform the method according to any one of claims 1 to 7.

Images