TWI478514B - The method of using precode to distinguish the equivalent channel and the relay module thereof - Google Patents

Description

Method for separating equivalent channel by precoding and relay device thereof

The present invention relates to a method for separating an equivalent channel and a relay device thereof, and more particularly to a method for separating a channel response of an equivalent channel by precoding and a relay device therefor.

In the communication technology, the Two Way Relay Network (TWRN) architecture is a Cooperative Communications system, which has a relay system for auxiliary transmission of data, and Following the system, data can be exchanged between the two data transceivers, thereby improving transmission efficiency and communication quality.

Referring to FIG. 1 , a conventional bidirectional transmission relay network architecture 9 is provided with a plurality of relay devices 93 between a first source end 91 and a second source end 92 to enable the first source end. 91 and the second source 92 can exchange data at the plurality of relay devices 93. When the second source 92 receives the data transmitted from the plurality of relay devices 93, a plurality of input and output (Multiple Input) can be formed between the plurality of relay devices 93 and the second source terminal 92. The Single Output (MISO) system enables the second source 92 to obtain the advantages of spatial diversity, thereby improving communication quality.

Further, when data is transmitted in different channels, it will be affected by the channel response of different channels. Therefore, if channel compensation is performed on the data, if the different channels cannot be effectively estimated, it will affect the whole. Communication quality. In the study of the conventional separation channel response, Space-Time Coding (STC) or space-frequency coding is introduced in the relay device. Space-Frequency Coding (SFC) technology, which separates the channels formed by different relay devices. However, the research is mainly directed to a One Way Relay Network (OWRN) architecture, and in the one-way transmission relay network architecture, each of the relay devices forms a simple channel, including only The channels formed by the transmitting end, the relay device and the receiving end are sequentially, and the above coding is mostly limited by the condition that the channel environment is a single path.

In the two-way transmission trunking network architecture 9, since the first source end 91 and the second source end 92 can exchange data, not only the transmitting end and the receiving end, respectively, for the second source end 92, The data received by the second source end 92 has a first equivalent channel formed by the first source terminal 91, the relay device 93, and the second source terminal 92, in addition to the first equivalent channel formed by the first source terminal 91, the relay device 93, and the second source terminal 92. There is a second equivalent channel formed by the second source end 92 to the relay device 93 and then returned by the relay device 93 to the second source end 92. If the first equivalent channel and the second equivalent channel cannot be effectively separated, the channel estimation result will be affected, thereby causing the overall communication quality to decrease.

In view of this, for a relay device of a two-way transmission relay network, it is necessary to have a design capable of separating channel responses of different channels, thereby improving communication quality while maintaining communication efficiency, and having lower operation. the complexity.

The main object of the present invention is to provide a method for separating an equivalent channel by using precoding and a relay device thereof, and the method and the relay device thereof can improve communication quality while maintaining spectrum use efficiency.

Another object of the present invention is to provide a method for separating an equivalent channel using precoding and a relay device thereof, the method and the relay device having lower computational complexity.

In order to achieve the foregoing object, the method for separating an equivalent channel by using a precoding method comprises: precoding a first processing data by using a precoding matrix by using a precoding module to obtain a first pre- Encoding data; using the precoding module to perform a cyclic shift on the precoding matrix to obtain a displacement precoding matrix, and precoding the second processing data by using the displacement precoding matrix to obtain a second pre- Encoding data; and superimposing the first pre-encoded data and the second pre-encoded data by using a superposition operation module to obtain an overlay data.

The method for separating equivalent channels by using precoding according to the present invention, wherein the precoding matrix is a cyclic matrix, and the column vector of the cyclic matrix can be expressed as: P _i = [0 _{(i-1) × 1} A _i 0 _(Ni)×1 ] ^T

The precoding matrix can be expressed as:

Wherein, A _i is an arbitrary number other than 0, and N represents the length of the column vector, and the dimension of the precoding matrix is N×N.

The method for separating equivalent channels by using precoding according to the present invention, wherein the cyclic shift amount between the precoding matrix and the displacement precoding matrix is satisfied: 2L-1≦ cyclic displacement ≦N-L-1

Where L represents the maximum channel length from any of the transmitting terminals to the relay device, N represents the column vector length of the precoding matrix, and N is greater than L.

The method for separating the equivalent channel by using the precoding, wherein the method further comprises: receiving, by the receiving processing module, the first data from the first transceiver end and the second data of the second transceiver end, and The data and the second data are serially converted and the cyclic prefix is removed to obtain a first processed data and a second processed data, respectively.

The method for separating an equivalent channel by using a precoding method, wherein the method further comprises adding a cyclic prefix and a parallel string conversion to the superimposed data by using the transfer processing module to obtain an output data, and transmitting the output data to the The first transceiver end or the second transceiver end.

The relay device using the precoding separation equivalent channel of the present invention comprises: a receiving processing module, which receives a first data transmitted by a first transceiver and a second data transmitted by a second transceiver; And serially converting and removing the cyclic prefix from the first data and the second data to obtain a first processing data and a second processing data respectively; and a precoding module receiving the first processing data and a second processing data, and having a precoding matrix, and performing a cyclic shift on the precoding matrix to obtain a displacement precoding matrix, and then using the precoding matrix and the displacement precoding matrix, respectively, respectively, the first processing The data and the second processing data are pre-coded to obtain a first pre-encoded data and a second pre-encoded data respectively; a superimposed computing module receives the first pre-encoded data and a second pre-encoded data, and Superimposing the first pre-encoded data and the second pre-encoded data to obtain a superimposed data; and a transfer processing module, receiving the superimposed data, and Adding a cyclic prefix and a parallel conversion to the superimposed data to obtain an output data.

The relay device of the present invention utilizes precoding to separate an equivalent channel, wherein the column vector of the cyclic matrix can be expressed as: P _i = [0 _{(i-1) × 1} A _i 0 _{(N- i) × 1} ] ^T

The precoding matrix can be expressed as:

Wherein, A _i is an arbitrary number other than 0, and N represents the length of the column vector, and the dimension of the precoding matrix is N×N.

The relay device using the precoding separation equivalent channel of the invention, wherein the cyclic displacement between the precoding matrix and the displacement precoding matrix is satisfied: 2L-1≦ cyclic displacement ≦N-L-1

Where L represents the maximum channel length between any transceiver and the relay device, N represents the column vector length of the precoding matrix, and N is greater than L.

The above and other objects, features and advantages of the present invention will become more <RTIgt; (Circulant Matrix) means that a matrix H consists of a column vector h = [h ₀ , h ₁ , ... , h _L-1 ] ^T , and the representation of the matrix H is as follows:

The "cyclic displacement" in the present invention means that in an N×N matrix, each element of each column vector is displaced by a block unit in the same direction, and the last element of each column vector before the displacement The first element of each column vector is formed after the displacement.

The "equivalent channel" as used in the present invention refers to a channel path through which a data source starts from any transceiver end and finally reaches any transceiver end.

Please refer to FIG. 2, which is a preferred implementation apparatus and environment for the method for separating equivalent channels by using precoding according to the present invention. The relay device 1 is disposed in a bidirectional transmission relay network, and the bidirectional transmission is performed. The network further has a first transceiver terminal 2 and a second transceiver terminal 3, wherein the first transceiver terminal 2 and the second transceiver terminal 3 are both available for outputting or receiving data, and the first transceiver terminal 2 transmits one. The first data, the second transceiver 3 transmits a second data, and the first data and the second data can be processed at the relay device 1 to achieve data exchange.

As shown in FIG. 3 , the relay device 1 includes a receiving processing module 11 , a precoding module 12 , a superimposing computing module 13 , and a transfer processing module 14 .

The method for separating equivalent channels by using precoding according to the present invention, firstly The device 1 receives the first data from the first transceiver 2 and the second data from the second transceiver 3 by using the receiving processing module 11, and performs serial-to-parallel conversion on the first data and the second data. The Cyclic Prefix (CP) is removed to obtain a first processing data and a second processing data, respectively.

In this embodiment, the receiving processing module 11 may be a related structure such as serial-to-parallel conversion and removal of a cyclic prefix in a conventional communication device, and is not limited herein.

In more detail, after the receiving processing module 11 removes the cyclic prefix from the first data to obtain the first processing data, the first processing data may be expressed as follows:

among them Representing the first processing data, Representing the transmission power of the first transceiver 2, s ₁ represents the first data, and n _{R, 1} represents Additive White Gaussian Noise (AWGN), Representing a channel loop matrix between the first transceiver end 2 and the relay device 1, and the channel loop matrix is composed of a channel response between the first transceiver end 2 and the relay device 1, wherein the channel The response column vector can be expressed as , L represents the channel length of the channel response.

After the receiving processing module 11 removes the cyclic prefix from the second data to obtain the second processing data, the second processing data may be expressed as follows:

among them Representing the second processing data, Representing the transmission power of the second transceiver 3, s ₂ represents the second data, and n _{R, 2} represents Additive White Gaussian Noise (AWGN), Representing a channel loop matrix between the second transceiver end 3 and the relay device 1, and the channel loop matrix is composed of a channel response between the second transceiver end 3 and the relay device 1, wherein the channel The response column vector can be expressed as , L represents the channel length of the channel response.

After obtaining the first processing data, the pre-coding module 12 is used to pre-code the first processing data in a precoding matrix to obtain a first pre-encoding data.

The precoding matrix is a cyclic matrix, and the column vector of the cyclic matrix can be expressed as: P _i = [0 _{(i-1) × 1} A _i 0 _{(Ni) × 1} ] ^T (3)

Where A _i is an arbitrary number other than 0, and N represents the length of the column vector.

When the dimension of the precoding matrix is N×N, the precoding matrix can be expressed as follows:

Moreover, the precoding matrix can perform N cyclic shifts when the dimension is N×N, and can have N types of precoding matrices. Taking N=8 as an example, as shown in Fig. 4, there may be eight types of precoding matrices. Each of the squares represents a block unit of the precoding matrix, and the precoding matrix When the block unit is 0, it is represented by a space, and the rest are elements having the same value A and not 0.

The first pre-encoded data can be expressed as follows:

among them, Representing the first pre-encoded data, Str ₁ represents a precoding matrix for precoding the first processed data.

After obtaining the second processing data, the precoding matrix is cyclically shifted by the precoding module 12 to obtain a displacement precoding matrix, and the second processing data is pre-processed by the displacement precoding matrix. Encoding to obtain a second pre-encoded material.

In more detail, the precoding module 12 receives the first processing data and the second processing data, and has the precoding matrix, and can perform the cyclic shift on the precoding matrix to additionally obtain the displacement precoding. And using the precoding matrix and the displacement precoding matrix to precode the first processing data and the second processing data, respectively, to obtain the first precoding data and the second precoding data respectively.

The second pre-encoded data can be expressed as follows:

among them, Representing the second pre-encoded data, Str ₂ represents a displacement precoding matrix for precoding the second processed data.

The cyclic displacement between the precoding matrix and the displacement precoding matrix preferably satisfies the following condition: 2L-1≦ cyclic displacement ≦N-L-1

Where L represents the maximum channel length between any transceiver and the relay device 1, N represents the column vector length of the precoding matrix, and N is greater than L. The overlapping of the channel response of the first processed data and the second processed data is avoided.

Since the cyclic matrix only needs a simple cyclic shift, it can present different types of precoding matrices and precode different data, so it does not need too complicated operations and has low computational complexity.

After obtaining the first pre-encoded data and the second pre-encoded data, the superimposed operation module 13 performs superposition operation on the first pre-encoded data and the second pre-encoded data to obtain an overlay data.

The superposition operation module 13 receives the first encoded data and a second encoded data, and performs a superposition operation to obtain the superimposed data, and the superimposed data can be expressed as follows:

Where y _R represents the superimposed data, and α represents a signal amplification factor of the relay device 3, and the signal amplification factor is expressed as follows:

Where P _r represents the transmission power of the relay device and σ represents the average variation amount.

In more detail, after the first pre-encoded data and the second pre-encoded data are superimposed to obtain the superimposed data, since the superimposed data can simultaneously include the first pre-encoded data and the second pre-encoded data, the overall The spectrum use efficiency of the communication system.

However, since the first pre-encoded data and the second pre-encoded data are respectively sent through the first transceiver end 2 and the second transceiver end 3, the first pre-encoded data and the second pre-encoded data have different channels. Path, and each channel has a different channel response to form a channel loop matrix and . In order to effectively separate the channel response carried by the first pre-encoded data and the second pre-encoded data, to improve the accuracy of the channel estimation, the first processed data and the second processed data are pre-coded. It is necessary to use precoding matrices with different cyclic shifts to avoid the loop matrix of the channel. and Overlap, which improves communication quality.

In addition, since the precoding matrix only needs to perform cyclic shift to achieve the purpose of separating the loop matrix of the channel, it does not need too complicated operation and has low computational complexity.

After the superimposed data is obtained, the superimposed data is received by the transfer processing module 14, and a cyclic prefix and a parallel conversion are added to the superimposed data to obtain an output data, and the output data is transmitted to the first Transceiver 2 or second transceiver 3. In this embodiment, the transmission processing module 14 may be a related structure such as a cyclic prefix and a parallel conversion in a conventional communication device, and is not limited herein.

In order to clarify the channel separation method of the present invention, the following illustrative examples only list the operational relationship between the channel response and the precoding matrix to explain that the present invention can indeed achieve separation of equivalent channels. In addition, since the first transceiver terminal 2 and the second transceiver terminal 3 have similar manners in signal processing, only the first transceiver terminal 2 is exemplified.

Referring to FIG. 2 again, the first transceiver terminal 2 has an equivalent channel included in the output data received, and is transmitted to the relay device 1 via the first transceiver terminal 2 And then passed back to the relay device 1 to a first equivalent channel formed by the first transceiver end 2; and a second equivalent channel formed by the second transceiver terminal 3, the relay device 1 and the first transceiver terminal 2 in sequence.

More specifically, before the output data is superimposed, the operational relationship between the first equivalent channel and the precoding matrix is expressed as follows:

among them, Representing the first equivalent channel, Representing a channel loop matrix formed by the channel response of the relay device 1 to the first transceiver terminal 2, Representing the channel loop matrix formed by the channel response of the first transceiver 2 to the relay device 1.

The operational relationship between the second equivalent channel and the displacement precoding matrix is expressed as follows:

among them, Representing the second equivalent channel, Representing the channel loop matrix formed by the channel response of the second transceiver 3 to the relay device 1.

Please refer to FIG. 5a, which is a calculation process of precoding the first equivalent channel and the precoding matrix, wherein the channel circulant matrix and The channel lengths of the channel responses h ₁ and h ₂ are both 2, and the dimension of the precoding matrix is 8×8. After the encoding operation of the precoding matrix, a matrix as shown on the right side of the equal sign is obtained, wherein the matrix indicates the block unit at the shadow, that is, the block unit representing the channel response of the first equivalent channel.

In order to effectively separate the first equivalent channel and the second equivalent channel, the precoding matrix performs cyclic shift before precoding the second processing data from the second transceiver terminal 3, in order to satisfy the cyclic displacement. Quantity must wait In the present embodiment, for example, if the length of each channel is assumed to be 2, and the length of the column vector of the precoding matrix is 8, it is selected here. The cyclic displacement is 4, and the calculation process is as shown in Fig. 5b. Wherein, the matrix on the right side of the equal sign indicates the block unit at the shadow, that is, the block unit representing the channel response of the second equivalent channel.

When the first pre-encoded data and the second pre-encoded data are superimposed, the operational relationship between the first equivalent channel, the second equivalent channel, and the precoding matrix is expressed as follows:

Among them, H _Mix represents a mixed equivalent channel.

Please refer to FIG. 5c, which is a schematic diagram of a mixed equivalent channel formed by superposing the first equivalent channel and the second equivalent channel. It can be seen from FIG. 5c that the channel responses of the first equivalent channel and the second equivalent channel do not overlap, and the first equivalent channel and the second equivalent channel can be effectively separated.

Please refer to the figures 6a and 6b. Further, when the mixed equivalent channel is to be de-precoded in the future, only the conjugate phase of the mixed equivalent channel and the precoding matrix and the displacement precoding matrix need to be converted. By performing a (Hermitian) matrix operation, the channel response of the first equivalent channel and the second equivalent channel can be respectively moved to the block unit of the matrix diagonal to obtain the channel response of the correct equivalent channel. In the process of de-precoding, only the conjugate transposed matrix of the pre-coding matrix and the displacement pre-coding matrix can be used, and the complicated operation is not needed, which helps to reduce the overall computational complexity.

In more detail, the first pre-encoded data and the second pre-encoded data are performed. After superimposing to obtain the superimposed data, since the superimposed data can simultaneously include the first pre-encoded data and the second pre-encoded data, the spectrum use efficiency of the overall communication system can be maintained.

In addition, since the first pre-encoded data and the second pre-encoded data are pre-coded by using the pre-coding matrix of different cyclic shifts, respectively, the first processed data and the second processed data are used, so that the first pre-encoding can be avoided. The channel response of the equivalent channel of each of the second pre-encoded data overlaps, and the first equivalent channel and the second equivalent channel are effectively separated, which helps to improve communication quality.

The method for precoding separation equivalent channel and the relay device thereof of the invention only use cyclic shift to change the precoding matrix, and only need to use the conjugate transposed matrix relative to the precoding matrix when deprecoding Completed with lower computational complexity.

While the invention has been described in connection with the preferred embodiments described above, it is not intended to limit the scope of the invention. The technical scope of the invention is protected, and therefore the scope of protection of the present invention is subject to the scope defined in the appended patent application.

〔this invention〕

1‧‧‧Relay device

11‧‧‧Receive processing module

12‧‧‧ Precoding module

13‧‧‧Superimposed computing module

14‧‧‧Transfer Processing Module

2‧‧‧First transceiver

3‧‧‧Second transceiver

[study]

9‧‧‧Two-way transmission relay network

91‧‧‧First source

92‧‧‧second source

93‧‧‧Relay device

Figure 1: A schematic diagram of a conventional two-way transmission relay network.

Figure 2: Schematic diagram of the present invention applied to a two-way transmission relay network.

Fig. 3 is a block diagram of a relay device using the precoding separation equivalent channel of the present invention.

Fig. 4 is a view showing the pattern of cyclic displacement of the precoding matrix of the present invention.

Figure 5a is a schematic diagram of the operation of the precoding matrix and the first equivalent channel of the present invention.

Figure 5b is a schematic diagram of the operation of the precoding matrix and the second equivalent channel of the present invention.

Figure 5c: Schematic diagram of the superposition operation of the hybrid channel of the present invention.

Figure 6a: The first situation diagram of the pre-precoding of the hybrid channel of the present invention.

Figure 6b is a second diagram of the pre-precoding of the hybrid channel of the present invention.

1‧‧‧Relay device

11‧‧‧Receive processing module

12‧‧‧ Precoding module

13‧‧‧Superimposed computing module

14‧‧‧Transfer Processing Module

Claims (8)

A method for separating an equivalent channel by using a precoding method comprises: precoding a first processing data by a precoding matrix by using a precoding module to obtain a first precoding data; using the precoding mode Performing a cyclic shift on the precoding matrix to obtain a displacement precoding matrix, and precoding the second processing data by using the displacement precoding matrix to obtain a second precoding data; and using a superposition operation The module performs a superposition operation on the first pre-encoded data and the second pre-encoded data to obtain an overlay data. The method for separating an equivalent channel by using a precoding according to claim 1, wherein the precoding matrix is a cyclic matrix, and a column vector of the cyclic matrix can be expressed as: P _i = [0 _{(i- 1) × 1} A _i 0 _{(Ni) × 1} ] ^T The precoding matrix can be expressed as: Wherein, A _i is an arbitrary number other than 0, and N represents the length of the column vector, and the dimension of the precoding matrix is N×N. For example, the method for separating the equivalent channel by using the precoding according to Item 1 of the patent application, wherein the cyclic displacement between the precoding matrix and the displacement precoding matrix is satisfied: 2L-1≦ cyclic displacement ≦N-L-1 Where L represents the maximum channel length between any transceiver and the relay device, N represents the column vector length of the precoding matrix, and N is greater than L. The method for separating an equivalent channel by using a precoding according to the first aspect of the patent application, wherein the method further comprises: receiving, by the receiving processing module, a first data from the first transceiver end and a second transceiver end Second data, and serially converting and removing the cyclic prefix from the first data and the second data to obtain a first processing data and a second processing data respectively. The method for separating an equivalent channel by using a precoding according to the first aspect of the patent application, wherein the method further comprises: using the transmission processing module to add a cyclic prefix and a parallel conversion to the superimposed data to obtain an output data, And transmitting the output data to the first transceiver or the second transceiver. A relay device for separating an equivalent channel by precoding includes: a receiving processing module, which receives a first data transmitted by a first transceiver and a second data transmitted by a second transceiver, and The first data and the second data are serially converted and removed from the cyclic prefix to obtain a first processed data and a second processed data respectively; and a precoding module receives the first processed data and the second Processing the data, and having a precoding matrix, and performing a cyclic shift on the precoding matrix to obtain a displacement precoding matrix, and then using the precoding matrix and the displacement precoding matrix, respectively, respectively, the first processing data and The second processing data is pre-coded to obtain a first pre-encoded data and a second pre-encoded data respectively; a superimposed computing module receives the first encoded data and a second encoded Data, and superimposing the first pre-encoded data and the second pre-encoded data to obtain a superimposed data; and a transfer processing module, receiving the superimposed data, and adding a cyclic prefix to the superimposed data and String conversion to get an output data. A relay apparatus using a precoding separation equivalent channel as described in claim 6 wherein the column vector of the cyclic matrix is expressed as: P _i = [0 _{(i-1) × 1} A _i 0 _{( Ni) × 1} ] ^T The precoding matrix can be expressed as: Wherein, A _i is an arbitrary number other than 0, and N represents the length of the column vector, and the dimension of the precoding matrix is N×N. The relay device using the precoding separation equivalent channel as described in claim 6 wherein the cyclic displacement between the precoding matrix and the displacement precoding matrix is satisfied: 2L-1≦ cyclic displacement ≦NL- 1 where L represents the maximum channel length between any transceiver and the relay device, N represents the column vector length of the precoding matrix, and N is greater than L.