KR100723018B1 - Method for adaptive modulation using residual power and apparatus thereof - Google Patents

Abstract

An adaptive modulation method and apparatus thereof using residual power are disclosed.

According to an embodiment of the present invention, a plurality of streams are arranged in order of eigenvalues, and a predetermined power is assigned to each of the streams. Determining an energy level of a symbol to be transmitted at, if there is residual power in the priority stream, further assigning the residual power to a next stream having an eigenvalue less than the priority stream, and the code rate of the next stream and the Determining an energy level of a symbol to be transmitted in the next stream using feedback information.

According to the present invention, the energy level of a symbol is determined in consideration of different code rates for each stream, and the remaining power is additionally allocated to a stream having a low eigenvalue, thereby preventing waste of residual power and having a constraint of a target bit error rate. Even if the maximum size of the constellation is limited, the performance of the system can be improved, and no additional computation is required for the implementation.

Description

{Method for adaptive modulation using residual power and Apparatus}

1A and 1B illustrate a MIMO BIC-OFDM system according to an embodiment of the present invention.

2 is a block diagram of the present invention.

3 is a flowchart of the present invention.

4 is a detailed flowchart of an energy level determination process of FIG. 3.

FIG. 5 illustrates an example of a coding offset table used in the energy level determination process of FIG. 3.

6 is a graph showing a conventional modulation method and the performance according to the present invention.

The present invention relates to orthogonal frequency multiple division modulation, and more particularly, to an adaptive modulation method and apparatus using residual power.

Recently, in the field of wireless communication, a system that supports high-speed packet data transmission is pursued. As a technique for supporting a high data rate, a multiple input multiple output (MIMO) system is attracting attention, and the application of an orthogonal frequency multiple division (OFDM) modulation scheme can effectively overcome the frequency selective nature of the channel. Here

The combination of Bit Interleaved Coded Modulation (BICM) to achieve good diversity gain in the frequency domain has been widely applied to wireless standard technology in recent years. Examples are technologies such as IEEE 802.11a wireless LAN (WLAN) and IEEE 802.16e wireless metropolitan area network (WMAN).

When channel state information (CSI) is known to the transmitter, the transmission rate may be maximized through an adaptive modulation and coding (AMC) scheme. Recently, a bit allocation scheme using the Levin-Campello (LC) algorithm has been proposed in the SISO BIC-OFDM system, which uses an approximate coding gain (asymptotic coding gain) to apply the LC algorithm. The constraints of the bit error rate (BER) are relaxed.

Also, if the transmitter knows the channel state information, the multiple input multiple output (MIMO) channel corresponding to each subcarrier frequency can be divided into independent parallel channels by applying a single value decomposition (SVD) technique. Can be. System performance can be maximized by using an optimized power allocation scheme through water-filling (WF) in the spatial domain and a bit allocation scheme in each subchannel. However, according to the conventional modulation method as described above, in the high SNR range, there is residual power even after allocating the maximum possible number of bits.

Therefore, in the conventional modulation method, when the maximum size of the constellation is limited, there is a problem of limiting performance improvement and wasting residual power due to the inherent property of the single-value decomposition technique.

Accordingly, the first technical problem to be achieved by the present invention is to provide an adaptive modulation method using residual power that can prevent the waste of residual power and improve performance even when the maximum size of the constellation is limited.

A second technical problem to be achieved by the present invention is to provide an adaptive modulation device using the residual power to which the adaptive modulation method using the residual power is applied.

In order to solve the first technical problem described above, the present invention comprises the steps of arranging a plurality of streams in order of eigenvalues, allocating a predetermined power for each of the streams, Determining an energy level of a symbol to be transmitted in the priority stream by using feedback information of a receiver; if residual power is present in the priority stream, additionally assigning the residual power to a next stream having a lower eigenvalue than the priority stream And determining an energy level of a symbol to be transmitted in the next stream using the code rate of the next stream and the feedback information.

In order to solve the above second technical problem, the present invention provides an initialization unit for arranging a plurality of streams in order of eigenvalues and allocating a predetermined power for each of the streams. Adaptive symbol generation unit for determining the energy level of a symbol to be transmitted in the streams using the code rate of the streams and the feedback information of the receiving end, after determining the energy level of the symbol in any of the streams to the arbitrary stream If there is residual power, there is provided an adaptive modulation apparatus that uses the residual power including a power reassignment for further allocating the residual power to the next stream having an eigenvalue less than the arbitrary stream.

The present invention is characterized by an adaptive modulation scheme that provides improved performance in a MIMO BIC-OFDM system. In particular, unlike the conventional scheme using a coding gain, the bit allocation scheme is considered in consideration of the code rate used for each stream. It is further optimized and increases the power efficiency by delivering the remaining power allocated to each stream to the lower stream.

Hereinafter, with reference to the drawings will be described a preferred embodiment of the present invention.

1A and 1B illustrate a MIMO BIC-OFDM system according to an embodiment of the present invention. 1A and 1B assume an adaptive modulated MIMO BIC-OFDM system with the number of transmit antennas N_t = 4 and the number of receive antennas N_r = 4.

1A is a block diagram illustrating a transmitting end of an adaptive modulated MIMO BIC-OFDM system.

Referring to FIG. 1A, the user information (Data of user 1, 2, 3, ..., K) transmitted from the base station forms a total of N_t pieces, that is, four streams. Here, the i-th stream means user information transmitted through the i-th antenna.

The scheduler 100 outputs a plurality of user information in a limited stream (1, 2, 3, 4).

Then, the feedback information is used to determine the current channel status and adjust the data rate and power level based on the current channel status. Once the rate and power are determined, the data modulated according to the BICM scheme for each stream is transmitted through an orthogonal frequency multiplexing (OFDM) modulation technique having N_c subcarriers.

Feedback channels that provide channel information usually use good performance channel coding and low modulation levels to ensure reliable transmission. Therefore, the present invention assumes an ideal feedback channel without error.

The AMC block 110 allocates and outputs bits to be transmitted to each stream.

The plurality of BICMs 120 expands the size of data according to the code rate of each stream.

The plurality of IFFTs 130 performs an inverse fast Fourier transform on each stream.

Here, by using the BICM technique, it is possible to form a diversity gain in the frequency domain through channel coding in a frequency selective channel. OFDM is implemented by combining a binary convolutional coder and a symbol mapper through a bitwise interleaver. BIC-OFDM is implemented through the inverse fast Fourier transform (IFFT) and fast Fourier transform (FFT) modulation schemes in the BICM technique.

1B is a block diagram illustrating a receiving end of an adaptive modulated MIMO BIC-OFDM system.

The plurality of FFTs 140 performs fast Fourier transform on the signal received by each receive antenna. The weight applying unit 150 applies a weight to each stream, the demultiplexer 160 combines each stream according to the allocation information, and the demapper 170, the deinterleaver 180, and the decoder 190 are original bits. Restore it.

2 is a block diagram of the present invention when the number of streams is N. FIG.

The initialization unit 200 arranges the plurality of streams in order of eigenvalues, and allocates a predetermined power to each stream. In addition, the initialization unit 200 may calculate the eigenvalue for each stream from the feedback information of the receiver. In addition, the initialization unit 200 may detect a state of a channel through which the symbol generated by the adaptive symbol generator 210 is transmitted from feedback information of the receiver, and determine a predetermined power according to the state of the detected channel.

The adaptive symbol generator 210 determines the energy level of a symbol to be transmitted in the streams using the code rate of the streams and the feedback information of the receiver in the order of high eigenvalues for each stream. In this case, the symbol is information of a bit unit consisting of a plurality of bits included in the user information.

In addition, the adaptive symbol generator 210 may compensate an energy level of a symbol to be transmitted in the streams by using a coding offset according to the code rate of the streams. Preferably, the coding offset may be determined with reference to a coding offset table which previously stored an offset corresponding to the code rate of the streams based on a predetermined bit error rate. At this time, the output streams 1 to N are streams which have completed the energy level determination process and the additional allocation process of residual power.

The power reallocator 220 further allocates the remaining power to the next stream having a lower eigenvalue than the above stream if there is residual power in the above stream after determining the energy level of the symbol in any of the streams.

Preferably, the adaptive modulation device of FIG. 2 may be implemented within the AMC block 110 of FIG. 1A.

3 is a flowchart of the present invention.

First, a plurality of streams are arranged in order of eigenvalues, and a predetermined power is allocated to each of the streams (step 300). This process 300 may include detecting a state of a channel through which a symbol is transmitted from feedback information of a receiver and determining a predetermined power according to the state of the channel. In addition, this process (process 300) may include a process of calculating the eigenvalue for each stream from the feedback information of the receiving end.

Next, the streams are selected in order of increasing eigenvalues among the streams, and the energy level of the symbol to be transmitted in the selected stream is determined using the code rate of the selected stream and the feedback information of the receiver (step 310). This process 310 may include generating a symbol to be transmitted from the user information. At this time, the step 310 of determining the energy level of the symbol consumes at least a portion of power initially allocated to each stream.

When the process of determining the energy level of the symbol in the selected stream is completed, it is determined whether residual power exists in the selected stream (step 320).

If there is residual power in the selected stream, the remaining power is additionally allocated to the next stream having a smaller eigenvalue than the selected stream (step 330). On the other hand, if there is no residual power in the selected stream, the process proceeds to the next step (340).

Next, it is determined whether the above processes (steps 310-330) are completed for all streams (step 340). At this time, if the above process (steps 310-330) is not completed for all streams, the process proceeds to the next stream (350) and repeats the above steps (310-330) (step 340).

On the other hand, if the above steps (steps 310-330) have been completed for all streams, all steps are terminated (step 340).

Here, a stream having the largest eigenvalue may be defined as a priority stream, a stream having a unique value smaller than the priority stream, a next stream, and a stream having a unique value less than the next stream as a subordinate stream.

4 is a detailed flowchart of the energy level determination process 310 of FIG. 3.

First, the state of the channel through which the symbol for each stream is transmitted is determined from the feedback information (step 411).

Next, a specific position of the coding offset table is referred to according to the code rate and the channel state for each stream (step 412).

Next, the coding offset according to the code rate for each stream is determined (step 413).

Finally, the energy level of the symbol transmitted in the stream is compensated with the determined coding offset (step 414).

FIG. 5 illustrates an example of a coding offset table used in the energy level determination process 310 of FIG. 3.

FIG. 5 is a table illustrating coding offsets corresponding to various channel codes and modulation schemes used in a MIMO BIC-OFDM system according to an embodiment of the present invention. Here, the leftmost column represents the code rate of the stream, and the topmost column represents various modulation schemes. The coding offset described here is an experimental value obtained based on the bit error rate (BER) 10 ^ 4 and is expressed in decibels (dB).

The present invention is described as follows.

The received signal in the nth subcarrier is represented by Equation 1 below.

In Equation 1, H_ (ij) ^ n indicates a channel frequency response between the i th transmit antenna and the j th receive antenna in the n th subcarrier. z_j ^ n is a complex Gaussian noise of independent coincidence distribution following the variance of σ_z ^ 2, and x_i ^ n denotes a symbol transmitted to the i th transmit antenna having a variance of σ_x ^ 2. It is assumed that the total power used in the transmission stage is P = N_t * σ_x ^ 2.

The available channel codes and the magnitude of the maximum constellation are assumed to be V and 2 ^ (m_max), respectively. In each transmission, the i-th stream undergoes encoding by selecting one of V different channel codes, and the n-th subcarrier of the i-th stream is b_n ^ i ∈ {m_1, m_2,... , m_max} (n = 1, 2, ..., N_c) bits are transmitted. [b_1 ^ i, b_2 ^ i,... , B_ (N_c) ^ i) ] If the bit vector b _i la of the i-th stream, i spectral efficiency (spectral efficiency of the second stream) is expressed as the following equation (2).

In Equation 2, R_c (C_v *) denotes a coding rate selected by the channel encoder C_v *. The channel encoder C_v * and the number of bits b_n ^ i allocated to the nth subcarrier are determined by the AMC scheme.

In more detail, the n-th subchannel H_n can be expressed by Equation 3 through the SVD method.

In Equation 3, V_n and U_n are unitary matrices, and Σ_n is a diagonal matrix having eigenvalues of H_n as components. Knowing the SVD information, the MIMO channel H_n can be diagonalized through filtering using U_n and V_n ^ *. The equivalent model of the signal received at the receiver can be expressed as Equation 4 below.

As in Equation 4 above, x _ n indicates a symbol transmitted in the nth subcarrier. The most common method of controlling x _ n is a power allocation scheme through eigenmodes. The optimized power allocation technique is obtained through the well-known water-filling technique in the spatial domain. However, this requires a lot of complexity.

On the other hand, the bit allocation scheme is a method of allocating more bits to a mode having a larger eigenvalue. The bit rate is maximized in the MIMO BIC-OFDM system according to another embodiment of the present invention. This is optimized in terms of channel capacity through spatio-temporal vector coding (STVC). However, this method is very complicated because SVD must be performed in N_rN_c × N_tN_c dimension. In addition, it is known that the algorithm performing N_c times the SVD of N_r × N_t dimension almost follows the optimized performance as N_c increases. Thus, the present invention takes the latter method to reduce complexity. This type of rate maximization method can be formulated as in Equation 5 below.

역시

보다 작은 값을 갖는다. 따라서

만큼의 잔여 파워가 생겨난다.As can be seen from Equation 5, the conventional Levin-Campello (LC) algorithm is optimized bit allocation vector b _i through equal power distribution of each subcarrier when the maximum constellation size is not limited. And the code index v_i ^ *. However, in practical situations where the maximum supported feature size is limited, the above-mentioned residual power problem arises, resulting in poor performance. SVDs performed on each subcarrier are sorted in descending order of eigenvalues. At high SNR, this concentrates on streams of indices where the number of bits of m_max is low. In this case, the energy required to transmit the m_max bits is less than Ε_x,

Also

Has a smaller value. therefore

As much residual power is generated.

Conversely, high index streams with small eigenvalues have room to carry more information when allocating additional power. Therefore, delivering extra power from a low index stream to another stream can increase the overall throughput of the system. This may be arranged as an equation as shown in Equation 6 below.

As shown in Equation 6 above, when applied to the LC Efficientizing (EF) algorithm and the E-Tightening (ET) algorithm, a bit rate is allocated to a stream having a small eigenvalue, thereby increasing the transmission rate.

 Meanwhile, the discrete bit allocation algorithm (MIMO) (BLCM / RPD (E_x)) will be described below.

In Equation 7, the value of c) can be directly obtained through the ET algorithm without any additional calculation amount.

In addition, the discrete bit allocation algorithm (DBLC / CS (H, σ_x)) in the encoding system will be described with Equation (8).

As shown in Equation 8, the generalized LC algorithm (GLC) of the present invention will be described. The conventional LC algorithm is one of the best known optimized bit allocation algorithms, and performs the algorithm using an approximate coding gain through a fixed channel gap (Γ). The energy function obtained through gap approximation in M-QAM is expressed by Equation 9 below.

As shown in Equation 9, in the uncoded system, the channel gap is assumed to be constant, but in the encoding system, a value corresponding to the gap is changed according to the characteristics of the applied channel code. The gap of the system to which the code rate R_c is applied is expressed by Equation 10 below.

(M, R_c) is an SNR value required to satisfy the bit error rate (BER) 10 ^-4 when the code rate and the constellation size are applied as R_c and 2 ^m , respectively. When channel coding is applied to each stream, the number of bits carried in each subcarrier no longer means the number of information bits. In this case, the channel gap should be replaced with a coding offset defined as in Equation 11 below.

 The energy function considering this coding offset as a new parameter is given by Equation 12 below.

6 is a graph showing a conventional modulation method and the performance according to the present invention.

In this case, we consider an OFDM system consisting of N_c = 64 samples and CP = 16 samples, and used a channel with exponentially decaying power consisting of five taps. A 64-state RCPC code was used and the target frame error rate (FER) was set to 1%. The corresponding bit error rate is 10 ^-4 .

6 is a graph showing the performance when a single user of N_t = N_r = 4 is used. The conventional SVD, distinguished by the rightmost asterisk, represents the throughput of a conventional water-filling bit allocation system. In other words, the system of the conventional SVD uses a fixed channel gap in the conventional LC algorithm and does not consider the remaining power distribution.

Performance Curve of System for Reallocating Power According to the Invention (RPD with GLCRPD) According to the present invention, the performance curve (RPD) of the system which compensates for the energy of the symbol in consideration of the code rate is higher than the conventional SVD of the conventional system. It can be seen that the throughput is high, that is, the performance is improved by about 4 dB based on the spectral efficiency of 18 bps / Hz.

Since most practical systems all have a limit on the maximum feature size that can be supported, the problem of residual power in the high SNR range is very common. The present invention can obtain a large performance gain especially in the high SNR range by proposing a bit allocation scheme that compensates for this.

Preferably, a program for executing the adaptive modulation method using the residual power of the present invention on a computer can be recorded on a computer-readable recording medium.

The invention can be implemented via software. When implemented in software, the constituent means of the present invention are code segments that perform the necessary work. The program or code segments may be stored on a processor readable medium or transmitted by a computer data signal coupled with a carrier on a transmission medium or network.

Although the present invention has been described with reference to one embodiment shown in the drawings, this is merely exemplary and will be understood by those of ordinary skill in the art that various modifications and variations can be made therefrom. However, such modifications should be considered to be within the technical protection scope of the present invention. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

As described above, according to the present invention, the energy level of the symbol is determined in consideration of different code rates for each stream, and the remaining power is additionally allocated to the stream having a low eigenvalue, thereby preventing waste of residual power and reducing the target bit error rate. Even when there are constraints or when the maximum size of constellations is limited, the performance of the system can be improved and the implementation does not require additional computation.

Claims (15)

Arranging a plurality of streams in order of eigenvalues, and allocating a predetermined power for each of the streams; Determining an energy level of a symbol to be transmitted in the priority stream by using a code rate of a priority stream having a high eigen value among the streams and feedback information of a receiver; If there is residual power in the seniority stream, further assigning the residual power to the next stream having an eigenvalue less than the seniority stream; And Determining an energy level of a symbol to transmit in the next stream using the code rate of the next stream and the feedback information. The method of claim 1, Allocating a predetermined power for each stream Detecting a state of a channel through which the symbol is transmitted from feedback information of the receiving end; And And determining the predetermined power in accordance with the state of the channel. The method of claim 1, Allocating a predetermined power for each stream And calculating a eigenvalue for each stream from the feedback information of the receiving end. The method of claim 1, Determining an energy level of a symbol to be transmitted in the priority stream Compensating for an energy level of a symbol to be transmitted in the priority stream using a coding offset according to the code rate of the priority stream. The method of claim 4, wherein The coding offset is And a coding offset table that stores an offset corresponding to a code rate of the priority stream based on a predetermined bit error rate. The method of claim 1, Determining the energy level of the symbol to be transmitted in the next stream Compensating for an energy level of a symbol to be transmitted in the next stream using a coding offset according to the code rate of the next stream. The method of claim 6, The coding offset is Adaptive modulation method using the residual power, characterized in that determined using a coding offset table that previously stored an offset corresponding to the code rate of the next stream on the basis of a predetermined bit error rate. The method of claim 1, If there is residual power in the next stream, further assigning the residual power to a subordinate stream having an eigenvalue less than the next stream; And Determining an energy level of a symbol to transmit in the subordinate stream using the code rate of the subordinate stream and the feedback information; And further allocating the residual power and determining the energy level of a symbol to be transmitted in the subordinate stream for all streams. The method of claim 8, Determining the energy level of the symbol to be transmitted in the subordinate stream Compensating for an energy level of a symbol to be transmitted in the subordinate stream using a coding offset according to the code rate of the subordinate stream; And the coding offset is determined by using a coding offset table which previously stores an offset corresponding to a code rate of the subordinate stream based on a predetermined bit error rate. A computer-readable recording medium having recorded thereon a program for executing the method of claim 1 on a computer. An initialization unit for arranging a plurality of streams in order of eigenvalues and allocating a predetermined power for each of the streams; An adaptive symbol generator configured to determine an energy level of a symbol to be transmitted in the streams by using the code rate of the streams and feedback information of a receiver in the order of increasing eigenvalues for each of the streams; If there is residual power in the arbitrary stream after determining the energy level of the symbol in any of the streams, a power reallocation section for further allocating the residual power to the next stream having an eigenvalue less than the arbitrary stream Adaptive modulation device using the residual power comprising. The method of claim 11, wherein The initialization unit And detecting the state of a channel through which the symbol is transmitted from the feedback information of the receiving end, and determining the predetermined power according to the state of the channel. The method of claim 11, wherein The initialization unit Adaptive modulation apparatus using the residual power, characterized in that for calculating the eigen value for each stream from the feedback information of the receiving end. The method of claim 11, wherein The adaptive symbol generation unit And an energy level of a symbol to be transmitted in the streams by using a coding offset according to the code rate of the streams. The method of claim 14, The coding offset is And an offset corresponding to a code rate of the streams based on a predetermined bit error rate.

Images