Block encoding with a variable rate block code

Cross-reference to related applications This application claims priority over US Practical Patent Application No. 12 / 365,565 filed February 9, 2009 and US Provisional Patent Application No. 61 / 039,585 filed March 26, 2008. , All of which are incorporated herein by reference. In addition, the present application claims benefits from US Provisional Patent Application No. 61 / 038,383 filed March 20, 2008, and March 24, 2008, No. 61 / 038,942, which in their entirety. Incorporated herein by reference.

The present invention generally relates to channel quality indicators, more specifically LTE (Long Term Evolution) of 3GPP (3rd Generation Partnership Project) Radio Access Network Standards (formally known as E-UTRA) systems. Regarding such indicators in.

The 3rd Generation Partnership Project (3GPP) is in the process of defining the various communication protocols required for the evolving 3GPP cellular network standard, nicknamed LTE. As shown in Figure 1 referenced here, among other things, the mobile unit 14 has two operating channels defined for the uplink (ie, also referred to as "user equipment" or "UE" in the LTE standard. Its serving base station 10, ie the link to transmit to eNodeB in the LTE standard). One channel is the physical uplink channel 18 for data, called PUSCH, and the other channel is the physical uplink channel 16 for control, called PUCCH. Both data and control signals can pass over data channel 18, but control channel 16 is dedicated to control signals and is farthest from the farthest range of each cell 12 (ie, farthest possible from each base station 10). Designed to work (from a distance) and under very poor conditions.

Many different signals are defined. One is a CQI (Channel Quality Index) report in which each UE 14 reports to its serving base station 10 the quality of the downlink channel, that is, the propagation channel from base 10 to UE 14. Traditionally, CQI reports can be transmitted over the uplink control channel PUCCH.

It has recently been suggested that control reports (such as CQI reports) can be transmitted over data channels in the context of new LTE standards. However, because the two channels operate differently, the control report structure for the control channel cannot be easily copied to the data channel.

According to one embodiment, a user of a cellular communication system comprising a step of block coding a control signal with a variable rate block code whose coding rate generally matches the variable coding rate of the data channel on which the report is transmitted. Methods for device devices are provided.

Also, according to one embodiment, methods for user equipment devices of cellular communication systems are provided. The method is a target BLER (block error rate) that approximates the steps of transmitting and receiving a signal through both the data and the control channel and the target BLER used to transmit the control signal over the control channel. Includes the step of encoding the control signal using a coding system with a first coding rate to achieve. The control signal is encoded using a coding system with a second coding rate to achieve the target BLER for the data, which is different from the BLER for transmitting the control signal over the control channel. It shall be transmitted along with at least the data channel.

Further, according to one embodiment, the device is an LTE capable device.

Further, according to one embodiment, the code rate of the variable rate block code is proportional to the code rate of the data channel.

Further, according to one embodiment, the coding step is to provide a step of determining the output code length K as a function of the current code rate of the data channel and at least K columns in the block coding matrix. It includes an iterative step and a step of generating a coded control signal on the data channel from the input control signal using the first K column of the iterative block coding matrix.

Further, according to one embodiment, the coding step comprises a step of encoding with a cyclically repeated block code.

Further, according to one embodiment, the step of encoding includes the step of encoding a CQI (Channel Quality Index) report. In one embodiment, it is a short CQI report.

Further, according to one embodiment, the method also includes the step of drilling the block coded matrix if K is less than the number of columns in the block coded matrix.

Further, according to the preferred embodiment of the present invention, the repeating steps are The block coding produced by repeating the block code to match the convolutional code rate, further repeating the block code until the desired code rate is exceeded, and further repeating the convolutional code rate. Includes a matrix and a rate matching step.

According to one embodiment, a communication device is provided that includes a block encoder for encoding a control signal with a variable rate block code in which the variable code rate and the code rate of the data channel on which the report is transmitted generally match. To.

Also, according to one embodiment, a transceiver for transmitting and receiving signals over both data and control channels and a target BLER (block error rate) used to transmit control signals on the control channels are achieved. To do so, a communication device is also provided that includes a control channel encoder for encoding the control signal using a coding system having a first coding rate. The control signal is coded using a coding system with a second coding rate to achieve the target BLER for data different from the BLER used to transmit the control signal on the control channel. It is transmitted via a data channel together with the data to be converted.

Also provided according to one embodiment are control channel encoders, including a determinant, a table generator, and a block encoder. The determinant determines the current value of K for encoding a control signal of length M, where K is a function of the current code rate for the data channel. The table generator generates a variable-length block coded matrix from the fixed block coded matrix by iterating over the columns from the fixed block coded matrix so that it produces a variable length K iterative matrix, and the block coder , The M input control bit is encoded in the variable length block coding matrix as the K coding control bit.

Further, according to one embodiment, the device and encoder are implemented in an LTE-enabled communication system.

Further, according to one embodiment, the code rate of the variable rate block code is proportional to the code rate of the data channel.

Further, according to one embodiment, the encoder also includes a perforator for perforating the block coded matrix if K is less than the number of columns in the block code matrix.

Finally, according to one embodiment, the table generator first repeats the block code matrix an additional number of times to match the convolutional code rate and exceeds the desired code rate, and multiple convolutional code rates. Includes a rate matcher for matching with a repeated block coding matrix.

Embodiments of the present invention will be described with reference to the following drawings, where similar numbers indicate similar elements.

<figref num="1">It is a schematic diagram of two of a base station, two mobile units, and a channel between them.</figref>

<figref num="2A">It is a schematic diagram illustrating a variably repeating block code suitable for encoding an M input control signal that is constructed and operational according to an embodiment of the present invention.</figref><figref num="2B">It is a schematic diagram illustrating a variably repeating block code suitable for encoding an M input control signal that is constructed and operational according to an embodiment of the present invention.</figref>

<figref num="3">FIG. 2 is a block diagram of a control channel encoder useful for generating the block codes of FIGS. 2A and 2B.</figref>

<figref num="4">It is a block diagram of the alternative embodiment of the encoder of FIG.</figref>

Note that for the sake of simplification and clarification of the figure, the elements shown in the figure are not necessarily drawn to scale. For example, some dimensions of an element may be exaggerated compared to other elements for clarity. In addition, reference numbers may be repeated across multiple drawings to indicate the corresponding element or similar element, if deemed appropriate.

In embodiments for carrying out the following inventions, many specific details are provided to provide a complete understanding of the teaching principles of the described embodiments. However, it should be noted that the present invention may be practiced without some of these specific details.

FIG. 1 illustrates the problem of sending control reports over data channel 18. Base station 10 is located in cell 12. Mobile units such as UE14 can move across cell 12 and communicate with base station 10. As shown in the figure, UE14A is located close to the edge of cell 12, while UE14B is located close to base station 10. Various parameters, including the distance between the UE 14 and the base 10, as well as the nature of the terrain and the speed of each UE 14 relative to the base 10 affect the quality of the radio propagation channel (or simply the "radio channel").

Each UE 14 transmits and receives various types of information, including control and data, and the control information and data are transmitted on physical control channel 16 and physical data channel 18, respectively. Each control channel 16 is shown as a straight line because the control signal must be received accurately, quickly, and at a well-defined time (eg, during transmission) for the system to operate properly. , UE14A, etc., show that the control channel transmission system is defined so that the target block error rate (BLER) is as low as a few percent, even for UE14 with poor radio channel conditions. Note that UE14A may suffer poorer radio channel conditions than UE14B due to the greater distance of UE14A from base 10 as can be seen in FIG. Nevertheless, control signals from both UE14A and UE14B must be properly received despite the differences in their respective channel conditions.

However, the physical data channel 18 behaves differently. Therefore, the data channel 18 is indicated by a wavy line, indicating that the quality of the channel can vary over time. The data channel does not necessarily have to be received quickly and at a well-defined time. Instead, the data channel may employ techniques such as retransmission and incremental redundancy to optimize the transmission capacity of the data and ensure proper reception of the data. Therefore, the target maximum BLER for any single transmission on the data channel, higher than for the control channel rather, typically may be set to 10-30%.

In addition, according to one embodiment, the transmission parameters and error protection scheme of the physical data channel 18 can change dynamically as a function of the quality of the immediate radio propagation channel. This is called "link adaptation". In contrast, physical control channel 16 generally has no link adaptation. Therefore, control channel 16, typically configured to send less information than data channel 18, is designed to have significantly stronger error protection than data channel 18, eg, stronger forward error protection. ..

According to some mobile communication standards, control channel 16 and data channel 18 do not transmit at the same time. Selected subframes of the transmission cycle may be assigned to control channel 16, while other subframes of the cycle may be assigned to data channel 18. Base station 10 may request CQI (Channel Quality Indicator) reporting in various subframes and may assign either control or data channels to the reporting. Data channel 18 may send a variable number of bits ranging from a minimum of 240 to thousands. Control channel 16 may send a small number of bits that can be fixed for any given type of cell. For example, for smaller cells, the number of bits may be fixed at 20. Therefore, the CQI report transmitted on control channel 16 is relatively small and intended for subframes without data channel allocation.

According to the current 3GPP standard for LTE-enabled communications, the error protection scheme for CQI reporting on control channel 16 is the "algebraic block coding" type of general purpose Reed-Muller type coding, using the NxM generator matrix RM. To generate the vector O of the N output bits, the NxM matrix RM is binary multiplied by the input payload vector I of length M. Therefore, O = RM^*I. The resulting block code is relatively efficient for smaller CQI payloads having codeword length N and M being 10 bits or less. According to the current 3GPP standard for LTE-enabled communications, N is 20 and the code is called "RM20". In vector I, there can typically be a variable number of input (information) bits of 4 to 11 (which can be 4 to 14 for large cells), while the number of output (coded) bits. Is fixed at N.

In addition to the RM20 code, the modulation scheme for control channel 16 employs "spread" (ie, single) so that there is a gain of approximately 10 dB in the transmission of N-coded bits over similar transmissions in non-spread systems. Modulate multiple resource elements with bits to get signal gain). The link quality of the control channel 16 obtained by the combination of coding and spreading works well enough for UEs close to the cell edge, such as UE14A. Another aspect of spreading is that it allows up to 6 UEs of code multiplexing to be transmitted simultaneously and over the same frequency band without mutual interference.

According to one embodiment, the CQI report may be transmitted on the data channel 18. Base station 10 requests a larger (and more detailed) CQI report from the selected UE14 by allocating uplink transmissions for this purpose on the data channel 18 associated with the selected UE14. May be good. This may occur with or without an accompanying allocation for also transmitting data on the data channel 18. According to one embodiment, if the data is also transmitted, both the large CQI report and the data are first independently encoded with their own error protection code (the CQI target error rate is typically, Much smaller than the target error rate of the data, so the same code is not used for both), after which the reports and data are multiplexed and transmitted together on the PUSCH data channel 18.

In another scenario, UE 14 may periodically transmit a small CQI report on control channel 16. Base station 10 may dynamically or periodically provide data channel 18 to UE 14 for data transmission. Therefore, it may occasionally be necessary for both data channel 18 and control channel 16 to be transmitted at the same time. Since LTE does not allow such simultaneous transmission, CQI reports may be transmitted "piggyback" along with other data on data channel 18.

Another scenario is the requested small CQI reporting scenario on data channel 18, where this requested CQI report can have different content but is the same size as the piggybacked CQI. Has a range.

According to one embodiment, the error protection scheme for small CQI reports on data channel 18 may be treated in the same way regardless of whether it was specifically requested on data channel 18 or piggybacked from control channel 16. Good. In this embodiment, CQI reports and other control signals sent over data channel 18 are on control channel 16 despite the lower protection normally provided on data channel 18 (as shown in FIG. 1). It may have a level of error protection commensurate with the error protection available in. In addition, such reports when transmitted over data channel 18 have the same spread as control channel 16 to ensure that reports from UE 14 at the cell edge, such as UE 14B, are properly received. Gains may be provided.

In this embodiment, according to the current 3GPP standard for LTE-enabled communication, the control bits transmitted on the data channel 18 may follow the same link adaptation as the data. For example, one feature of the variable transmission parameter of data channel 18 is the variable coding rate. For block coding, the coding rate is given by the ratio M / N. The lower the code rate, the greater the protection. For example, if the number of input bits is M = 10 and the code length is N = 20, the coding rate is 1/2. If the code length is N = 32 for the same number of input bits M = 10, the coding rate is 10/32, or about 1/3, which is a stronger error than the 1/2 coding rate. Provide protection.

Finally, in this embodiment, the iteration of code information can provide a spreading operation, which can effectively provide power gain.

According to one embodiment, a short CQI report on data channel 18 may be configured to allow transmission of CQI reports with data having different code rates and at various link conditions as needed. It may have an error protection level. Other LTE coding and decoding schemes may be used and the number of encoded CQI bits on the data channel 18 may be configured to be less than or equal to the minimum required for proper reception. In this way, the efficiency of incidental data transmission on the data channel 18 can be generally optimized.

Here, with reference to FIGS. 2A and 2B, where the coded M input control signal, such as a short CQI report, is encoded so that the variable code rate on the data channel 18 and the code rate are approximately the same, K. A variably repeating block code suitable for generating output bits is shown. FIG. 2A shows a fixed block code matrix 20 and FIG. 2B shows a variable length block code matrix 22 that can be generated from the fixed block code matrix 20.

As can be seen from FIG. 2A, the payload 24 of length M may be encoded using the NxM matrix 20 by multiplying the payload by the NxM matrix. This may generate a coded payload 26 of length N. The data code rate for the coded payload 26 is M / N. For short CQI reports with M = 10 and N = 20 (typical size for short CQI reports on control channel 16), the data rate is 1/2. According to one embodiment, the fixed block code matrix 20 may be utilized at a data rate of 1/2 for reports of length M = 9-11.

According to one embodiment, the variable length block code matrix 22 may be utilized to generate a coded payload 32 for other data rates, as shown in FIG. 2B. The matrix 22 may be generated from the fixed matrix 20 by first iterating the fixed matrix 20 to generate the iterative matrix 30. FIG. 2B shows two iterations of block code 20, and additional iterations may be used if desired. Partial iterations may also be used, as the iteration matrix 30 may instead include, for example, 2.5 or 3.5 iterations of block code 20. The current variable length block code matrix 22 may then be generated by selecting the first K column of the iteration matrix 30, where K is the current data given the input vector length M. The value needed to generate the rate. Thus, K can be variable in size and can correspond to the number of columns in at least one or more integer matrix 20 and, more appropriately, fractional matrix 20. The current variable length block code matrix 22 may then be multiplied by the input payload 24 to produce a coded payload 32 of length K.

For example, if the current data rate is 1/3 for M = 10 as in the previous embodiment, K may be set to about 30 and the coded payload 32 is 30 bits long. There may be. If the current data rate is 1/4, K may be set to about 40 and the coded payload 32 may be 40 bits long. Note that K is not always a multiple of M. For example, for the current data rate of 0.40, M may be 10 and K may be set to 25. In such cases, the iteration matrix 30 may include 2.5 iterations of the fixed matrix 20, and some, but not all, columns of the partially fixed block code matrix 20 may also be included in the iteration matrix 30.

The fixed block code matrix 20 may be any suitable relatively short block code matrix. For example, the current 3GPP standard for LTE-enabled communications includes the definition of the RM20 block code used for control channel 16. The RM32 block code has also been proposed for use on data channel 18. Any of these may be suitable for the fixed block code matrix 20.

The current value for length K may be determined from the current data code rate according to the current 3GPP standard for LTE-enabled communications, as follows:

<img file="JP5440887B2_D0001.tif" />

In the equation, Δ is the offset, R is the current data code rate, and Q_mIs the current data modulation order, and Δ, R, and Qm are part of the transmission parameters for the current data channel 18. Therefore, the block code rate K / M can be proportional to the data code rate R. In equation 1, the proportions are Δ and Q_mIs a function of.

Here, with reference to FIG. 3, a control channel encoder 40 that is constructed and operational according to embodiments of the present invention and forms part of each UE 14 (FIG. 1) is shown. The encoder 40 may include a determinant 42, a table generator 44, and a block encoder 48, which are coupled together as shown.

The determinant 42 can determine the current value of K, which is currently required for the coded control signal of length M. As mentioned above, the number K may be a function of the code rate for the data channel 18 through which the UE 14 can carry data. At any given time, base station 10 (FIG. 1) may inform UE 14 of the current transmission format, which format includes the code rate for its data channel 18. As mentioned above, the code rate is typically a function of the channel condition. According to one embodiment, the control signal transmitted on the data channel 18 has a code that matches the code rate of the data bits simultaneously transmitted on the data channel 18, up to an offset (such as the offset Δ in Equation 1). Should have a rate. The determinant 22 may determine the number of output bits K according to equation 1.

The table generator 44 may generate the variable length block code matrix 22 from the fixed block code matrix 20 by iterating over the columns of the fixed matrix 20 to generate the iterative matrix 30. Note that, as mentioned above, the value of K can determine whether some of the columns of the partial block code matrix 20 can be included in the iterative matrix 30. The table generator 44 may then select the first K column of the iteration matrix 30 to generate the current type of variable length block code matrix 22.

As mentioned above, the number of output bits K can vary as indicated by the double-headed arrow 34 in FIG. 2B. As is known in the art, the ratio K / N is approximately equal to the coding gain achieved through operation with the fixed block code matrix 20 alone. Note that column iterations have about the same effect as the iterations brought about by diffusion. Therefore, if the K / N value is close to or equal to the following values, BCL_d/ BCL_c^*CSG_c Equation 2

(During the formula, BCL_dIs the data channel block code length, BCL_cIs the control channel block code length and CSG_cIs the control channel spread gain), the target BLER of the CQI bit on data channel 18 is close to or equal to the target BLER of control channel 16.

The block encoder 48 may utilize the variable length block code 22 to encode the M input control bit as a K coding control bit. The block encoder 48 controls whenever there are control bits transmitted over the data channel 18, or only when there are small control reports such as small CQI reports, or UE 14 is encoded with Reed-Muller codewords. It may operate at any suitable time during which the bits can be transmitted.

According to another embodiment, as can be seen from FIG. 4 referenced herein, the table generator 44 may "drill" the iterative table 30 using any suitable perforation system. In this embodiment, the table generator 44 may iterate over the fixed block code matrix 20 J times, and the block coder 48 may generate the encoded payload 32 from the iterative matrix 30. The resulting code rate (M / J ratio) may be lower than required, so the block coder 48 perforates the coded payload 32 as described in more detail below. Alternatively, J is reduced to J'until M / J'appears approximately to the desired code rate. It should be understood that both J and J'can be either integers or fractions, depending on whether the partially fixed block code matrix can be included in the iteration matrix 30.

In one embodiment, the fixed block code matrix 20 may be an RM32 block code matrix, where N is 32. If K&gt; 32, the table generator 44 may periodically iterate over the RM32 matrix to form the iterative matrix 30 as described above. However, for K &lt;32, the table generator 44 may send the fixed block code matrix 20 to the punch 50 to reduce the size of the matrix 20.

The perforator 50 may perform any suitable preceding action. For example, according to the current 3GPP standard for LTE-enabled communications, the RM32 matrix may be truncated until it has only the first K column in it.

In another embodiment, the perforator 50 may be designed by first looking for the optimal perforation pattern and optimizing the total payload size and the total nested perforation pattern (where "nesting" is once. Refers to the fact that column "j" is erased to produce a matrix of length N-1 and then a different column "K" is erased to obtain a matrix of length N-2).

In a further embodiment, the perforator 50 may perforate all of the surplus trailing bits of the iterative codeword block.

In a further embodiment, the table generator 44 may be designed to match the behavior of the convolutional coder used for large CQI reports on the data channel 18. The latter system can work for blocks that are an integral multiple of 3 in length (because the convolutional code has a 1/3 rate). Therefore, in this embodiment, the table generator 44 may first iterate over the fixed block code matrix 20 three times to generate the first iterative table 30. The table generator 44 may then iterate over the first iteration table 30 L times (L may be 1) until the desired code rate is exceeded, thereby regenerating the iteration table 30. Convolutional code rate matching may then operate on the output of the block coder 48 until the desired code rate is obtained.

An alternative method of the above iteration is to perform the above convolutional coding and rate matching on the output of the block coder 48.

Embodiments of the invention may include instruments for performing the operations herein. The instrument may be specially constructed for a desired purpose, such as an ASIC for use with LTE-enabled UEs, or may be selectively started or reconfigured by a computer program stored in the computer. It may be equipped with a computer. Such a computer program may be stored on a computer-readable storage medium, which may be any type of medium suitable for storing electronic instructions and connectable to a computer system bus. ..

Although certain features of the present disclosure have been exemplified and described herein, various modifications, alternatives, modifications, etc., without departing from the scope of the invention described in the claims below. And equivalents can be conceived by those skilled in the art.