JP7048931B2 - Wireless relay device and wireless relay method - Google Patents

Description

The present invention relates to a wireless relay device and a wireless relay method.

The traffic of wireless communication is rapidly increasing due to the expansion of the use of wireless LAN and the increase of devices with communication functions such as wireless surveillance cameras. To improve the frequency utilization efficiency and accommodate more traffic with limited wireless resources. Is required.

That is, in recent years, wireless local area networks (WLANs) have been densely deployed, and due to their demand for IoT and M2M communications, the shortage of frequency resources is rapidly progressing. In general, a plurality of system-shared frequency bands, such as an ISM (Industry-Science-Medical) band, which are unlicensed bands, are often used as accommodation destinations for such wireless communication traffic.

Further, in such a wireless communication system, a convolutional code may be used as the coding method and Viterbi decoding may be used as the decoding method for the purpose of error correction.

Viterbi decoding is practically important as a decoding method that can easily realize maximum likelihood decoding for convolutional codes, and since soft determination decoding can be easily applied, excellent error performance can be obtained. It has been known.

For such bitabi decoding, for example, in Patent Document 1, an addition comparison selection operation (Add, Compare, Select operation, hereinafter referred to as an ACS operation) is performed, and the input data for which the ACS calculation has been performed reaches the “cutoff length”. Disclosed is a configuration in which the traceback processing is performed by the traceback calculator and the decoding result is acquired. Patent Document 1 discloses a configuration in which an ACS arithmetic unit executes an addition comparison selection operation on received data again when there is an error in the decoded data for the purpose of sufficiently improving the error correction capability. There is.

Further, Patent Document 2 discloses the following configuration for the purpose of reducing power consumption. That is, the receiving device disclosed in Patent Document 2 includes a Viterbi decoder that decodes convolutional coded data and a traceback length of parameters used in the Viterbi decoder from the data described in the header of the received data frame. It is composed of a traceback length control means for controlling the data and a header analysis means for reading the information in the header using the header of the decoded data. The receiving device receives the convolutional coded data, and then uses the information in the header of the received data when the most probable decoding is performed by a Viterbi decoder capable of parallel processing with multiple memories. It controls the traceback length, sets whether to use or not use multiple parallel memories, and if there are parallel memories that are not used, quiesce them and put them in standby mode.

However, these conventional techniques for Viterbi decoding are aimed at improving the error correction capability of each receiver and reducing the power consumption when the configuration of the Viterbi decoder becomes complicated. ..

By the way, as a method of expanding the coverage area without increasing the maximum transmission power of each transmitter, there is a relay method (relay method), which is standardized by, for example, IEEE 802.11s.

There are mainly the following two types of conventional relay methods (see, for example, Patent Document 3).

That is, there are a relay method called Decode-and-Forward (DF) and a method called Amplify-and-Forward (AF).

JP-A-2009-159482 JP-A-2011-35568 Japanese Patent Application Laid-Open No. 2013-175817

In the DF method, the frame to be relayed is decoded, it is confirmed that there is no error in the received frame, and then the bit string after the error correction is re-encoded / modulated to perform relay transmission. In this DF method, error propagation to the transfer destination can be prevented by performing error correction in the relay device, but relay transmission cannot be started until the decoding process is completed, so that a large delay occurs. In particular, in the case of a good communication condition in which error correction becomes unnecessary, the delay time until the decoding is completed becomes a large loss.

On the other hand, in the AF method, the signal is received without demodulation / decoding and the transfer is started immediately. However, since the noise contained in the received signal is amplified and transmitted as it is, the received signal power is high. The S / N ratio may deteriorate. In addition, the relay is performed without confirming whether the received frame needs to be transferred. Therefore, unnecessary frame transmission may occur, resulting in waste of resources and increased interference.

That is, the conventional relay method has a problem that it is difficult to suppress the error rate and shorten the delay time at the same time under a variable communication condition.

In other words, it was not always clear how to control the decoding process, coding and modulation processing in the repeater to reduce the delay time under the communication situation where the S / N ratio can fluctuate.

The present invention has been made to solve the above problems, and an object of the present invention is to provide a wireless relay device and a wireless relay method capable of suppressing a waste of resources and an increase in interference.

Another object of the present invention is to provide a wireless relay device and a wireless relay method capable of reducing the delay time while suppressing the error rate.

According to one aspect of the present invention, a radio for communicating by a wireless communication method capable of full-duplex communication, relaying a first frame signal from a source node and transferring it to a destination node as a second frame signal. A relay device, a receiving means for receiving a signal from a source node, and a demodulation / decoding means for demodulating the first frame signal received by the receiving means and decoding at least the header information of the first frame signal. , A relay control means for controlling the relay process according to the determination that it is necessary to relay the first frame signal based on the information in the header, and a first depending on the control of the relay control means. Coding / modulation means for encoding and modulating the data from the demodulation / decoding means to generate a second frame signal, and coding / modulation during a period that at least partially overlaps with the reception period of the frame signal of. A transmission means for transmitting the output of the means to the destination node is provided, and the demodulation / decoding means is a demodulation means for performing a predetermined demodulation process to generate a bit string, and decoding for a trellis code based on the bit string. A decoding means for executing processing and an error detecting means for performing error detection on the output of the decoding means are included, and the coding / modulation means is a demodulation means or a demodulation means or a demodulation means according to the control of the relay control means. Data output from any of the decoding means is selectively received, encoded and modulated to generate a second frame signal .

Preferably, the coding / modulation means obtains the second frame signal by a coding / modulation method controlled independently of the demodulation / decoding method in the demodulation / decoding means according to the control of the relay control means. Generate.

Preferably, the wireless communication method capable of full-duplex communication is a method of wireless communication using each wireless channel of a plurality of frequency bands separated from each other.

Preferably, the decoding means can set the cutoff length on the trellis for the decoding process for the trellis code, and the relay control means is between the source node and the wireless relay device based on the first frame signal. A communication quality estimating means for estimating the communication quality of the above, and a relay condition setting means for setting the cutoff length so as to reduce the expected value of the time required for transferring the second frame signal based on the estimated communication quality. including.

Preferably, the communication quality estimating means includes an error rate estimating means that acquires a propagation path estimated value and a noise power estimated value based on the first frame signal and estimates the frame error rate as the communication quality.

Preferably, the relay condition setting means calculates the expected value of the time required for transmitting the first frame signal, including retransmission, based on the communication quality, and encodes / modulates the means so that the expected value is minimized. The coding and modulation method in the above is selected from the predetermined methods.

Preferably, the relay condition setting means calculates the relay time required for the relay transmission, and if the relay time is longer than the frame period of the first frame, the relay transmission is started after the header is received.

Preferably, the relay condition setting means calculates the relay time required for relay transmission, and when the relay time is shorter than the frame period of the first frame, the first is until the timing at which the transfer data is not insufficient. After the frame data is buffered, relay transmission is started.

According to another aspect of the present invention, a radio for communicating in a wireless communication method capable of full-duplex communication, relaying a first frame signal from a source node and transferring it to a destination node as a second frame signal. It is a relay method, that is, a step of receiving a signal from a source node, a demodulation / decoding step of demodulating the received first frame signal and decoding at least header information of the first frame signal, and header information. Based on the above, the relay control step that controls the relay processing according to the determination that it is necessary to relay the first frame signal, and the reception period of the first frame signal according to the control of the relay processing. A coding / modulation step that receives data before the completion of processing in the demodulation / decoding step and encodes and modulates it to generate a second frame signal, and a second post-modulation period, at least in a partially overlapping period. The demodulation / decoding step includes a step of transmitting a frame signal to the destination node, a demographic step of executing a predetermined demodulation process to generate a bit string, and a decoding step of performing a decoding process for a trellis code based on the bit string. And a step of performing error detection on the result of the decoding step , the coding / modulation step selectively selects data of either the demodulation step or the result of the decoding step according to the control in the relay control step. In response to, it is encoded and modulated to generate a second frame signal.

According to the present invention, it is possible to suppress the waste of resources and the increase of interference in the wireless relay device.

Further, it is possible to reduce the delay time in the wireless relay device.

Further, according to the present invention, it is possible to reduce the delay time while suppressing the error rate.

It is a figure for demonstrating the concept of the relay operation of the wireless relay apparatus 1000 of this embodiment. It is a functional block diagram for demonstrating the configuration of the wireless relay apparatus 1000 of this embodiment. It is a conceptual diagram for demonstrating the composition of the frame signal which a wireless relay apparatus 1000 receives from a source node. It is a block diagram for demonstrating the structure of the demodulation / decoding unit 110, the relay control unit 112, and the coding / modulation unit 120. It is a conceptual diagram explaining the process executed by a relay node in the conventional DF method. It is a conceptual diagram explaining the process executed by a relay node in a TDF system. It is a figure which shows the time lapse of transmission in a relay system. It is a flowchart for demonstrating the retransmission control. It is a figure which shows the simulation parameter. It is a figure which shows the modulation and coding method (MCS) and the simulation condition. It is a graph which shows the delay time of each relay transmission system under the condition 5 of FIG. 10 (b).

Hereinafter, the configuration of the wireless communication system and the wireless relay device according to the embodiment of the present invention will be described. In the following embodiments, the components and processing steps having the same reference numerals are the same or equivalent, and the description thereof will not be repeated if they are not necessary.

In the following, the wireless relay device of the present invention will be described as operating in a wireless communication system capable of "full-duplex communication". In that case, as an example to explain the wireless relay device of the present invention, a plurality of existing unlicensed bands (for example, 920 MHz band used for IoT etc., 2.4 GHz band and 5 GHz used for wireless LAN) which are largely separated from each other. In (band), an example is a wireless relay device that uses multiple frequency bands at the same time and simultaneously executes transmission from the source mode to the repeater and transmission from the repeater to the destination node in different frequency bands. The embodiment to be performed will be described.

However, the method for realizing full-duplex communication is not necessarily limited to the method using such a plurality of frequency bands, and for example, even in the same frequency band, full-duplex communication is performed by so-called spatial division. May be a possible method.

The following documents disclose as a method for realizing a method for realizing full-duplex communication in such the same frequency band.

Literature: Jung Il Choiy, Mayank Jainy, Kannan Srinivasany, Philip Levis, Sachin Katti, "Achieving Single Channel, Full Duplex Wireless Communication", pp.1-12, Proc. Of the 16th Annual International Conference on Mobile Computing and Networking (Mobicom) 2010).
In the following description, as described above, it is assumed that different frequency bands are used at the same time.
[Embodiment]
FIG. 1 is a diagram for explaining the concept of relay operation of the wireless relay device 1000 of the present embodiment.

In the following, for the sake of simplicity, it is assumed that one wireless relay device 1000 is interposed between the source node and the destination node. However, the configuration and operation of the wireless relay device 1000 is not necessarily limited to such a case, and can be applied to a case of multi-hop in which a plurality of relay devices intervene.

As shown in FIG. 1, the wireless relay device 1000 receives a signal from the source node 1010 and relays the signal to the destination node 1020.

Here, as will be described in detail later, the wireless relay device 1000 performs a low-delay relay by starting relay transmission of a signal whose decoding is partially completed in a frequency band different from reception during reception processing. Realize.

Further, the wireless relay device 1000 adaptively selects the decoding processing method of the received signal according to the error rate prediction result of the received frame. For example, by selecting the cutoff length of the decoding process for the convolutional code including no cutoff, the delay in the decoding process is minimized to enable a low-delay relay.

In the following, as an example of the modulation method in each frequency band of a plurality of frequency bands as described above, the orthogonal frequency division multiplexing (OFDM) method is taken as an example, and the communication path coding for error correction is performed. As a method, a convolutional code is used, and as a decoding method, a method using bitabi decoding will be described as an example. However, as will be described later, the present invention is not limited to such an example.

FIG. 2 is a functional block diagram for explaining the configuration of the wireless relay device 1000 according to the present embodiment.

With reference to FIG. 2, the radio relay device 1000 is lower than the receiving antenna 100 for receiving the frame signal from the source node transmitted in the first frequency band and the frame signal from the receiving antenna. The RF unit 102 for executing high-frequency processing such as noise amplification, the synchronization processing unit 104 that executes synchronization processing using the preamble signal of the signal received by the RF unit 102, and the synchronization signal from the synchronization processing unit 104. Based on the OFDM demodulation unit 106 for executing the OFDM demodulation processing for the signal from the RF unit 102 and the pilot signal from the OFDM demodulation unit 106, the transmission path estimation value and the noise power estimation value are synchronized with. It is provided with a propagation path estimation unit 108 for calculating.

In the wireless relay device 1000, as will be described later, the demodulation / decoding unit 110 receives a frame signal (header and frame body data signal) from the OFDM demodulation unit 106 and receives at least a header portion of the received frame signal. Decrypt the information. The relay control unit 112 determines whether the received frame signal requires relay processing based on the decoded header information, and if it requires relay processing, estimates the transmission path from the propagation path estimation unit 108. The error rate of the received frame is predicted based on the value and the noise power estimate, and the decoding processing method for the received frame signal is adaptively selected according to the prediction result.

Although not particularly limited, for example, the relay control unit 112 shall switch the decoding method for the data of the frame body as follows according to the rank preset for the prediction result of the error rate of the received frame. ..

i) No censoring in Viterbi decoding ii) Execution of decoding with a censored length L in Viterbi decoding, and transfer the execution result.

In this case, a plurality of values may be preset for the cutoff length L according to the plurality of ranks of the prediction result.

iii) Transfer the demodulation result without executing the Viterbi decoding process.

In this case, in the Viterbi decoding, not performing the censoring process corresponds to the highest error correction capability in the wireless relay device 1000, so that the prediction of the error rate of the received frame has the highest error rate. Selected when it corresponds to a high rank. However, in this case, since the processing time required for decoding and error correction is the longest, the delay in the wireless relay device 1000 is the largest.

Since the longer the cutoff length L is, the higher the error correction capability is. Therefore, in the prediction of the error rate of the received frame, the lower the error rate is, the shorter the cutoff length L can be set. In this case, the shorter the cutoff length L, the smaller the delay in the wireless relay device 1000.

Further, transferring the demodulation result without executing the Viterbi decoding process is selected when the prediction of the error rate of the received frame corresponds to the rank with the lowest error rate. In this case, since the error rate is low, it is expected that the error rate at the destination node can be suppressed to a low level even if the wireless relay device 1000 does not perform error correction. Then, in this case, the delay in the wireless relay device 1000 is minimized.

Returning to FIG. 2, in the wireless relay device 1000, when the relay control unit 112 requires relay processing, the reception frame predicted based on the transmission line estimated value and the noise power estimated value from the propagation path estimation unit 108. It is assumed that the modulation method / coding rate for the frame signal to be transferred independently can be changed with respect to the modulation method / coding rate at the time of reception according to the error rate of.

Here, the modulation method and the coding rate can be indexed in advance as an MCS (Modulation and Coding Scheme).

By changing the modulation method and the coding rate in this way, as will be described later, from the wireless relay device 1000 with respect to the frame period of the frame signal transmitted from the source node to the wireless relay device 1000. The frame period of the frame signal to the destination node can be changed, and the relay control unit 112 can adjust the transmission timing of the relay signal.

The coding / modulation unit 120 performs coding and modulation processing on the data received from the demodulation / decoding unit 110 under the control of the relay control unit 112 with the selected modulation method and coding rate. ..

The OFDM modulation unit 122 executes OFDM modulation processing on the signal from the coding / modulation unit 120, and the RF unit 124 is a frame relayed in a second frequency band different from the first frequency band. A relay wave is transmitted from the transmitting antenna 130 by executing harmonic processing such as power amplification for transmitting the signal.

In the above description, the coding method for channel coding is "convolutional coding", and the decoding process is "Vitterbi decoding". However, in the wireless relay device 1000 of the present embodiment, it is more generally possible to use a "trellis code" as a coding method for channel coding.

Here, the "trellis code" is a tree code that can be defined on the trellis diagram, and the trellis diagram is the process in which the encoder, which is a finite state machine, changes the state according to the input bit sequence. It represents the code string generated by. In other words, in a trellis diagram, the path from the start point to the end point of the labeled graph corresponds to the codeword sequence. Therefore, in the trellis code, not only the information sequence at the time when the code sequence is output from the encoder but also the information sequence before that is determined.

Then, in the trellis code, when the number of states of the encoder is not so large, maximum likelihood decoding by the Viterbi decoding method utilizing the characteristics of the trellis is possible. However, the trellis code decoding method is not limited to the Viterbi decoding method.

Here, the "convolutional code" is a "linear code" among the "trellis codes" that is determined depending not only on the information series at the time when the code series is output from the encoder but also on the information series before that. To say.

In addition, in the Viterbi decoding, the "censored length" was explained. Normally, when Viterbi decoding is performed, decoding is terminated by the cleavage length, and traceback is started from the state having the most likely state metric at the time of termination on the trellis. Then, as a result of performing traceback, the obtained data becomes the decoding result. A state having a high likelihood state metric is a state having the smallest state metric value among the states existing at the time on the trellis diagram at which the execution of the Viterbi algorithm is discontinued.

However, as a decoding method that can consider the "censored length" in the decoding process, there are, for example, SW-BCJR algorithm and SOVA (soft output Viterbi algorithm) in addition to Viterbi decoding. The SW-BCJR algorithm is disclosed in the following documents, for example.

Documents: Japanese Unexamined Patent Publication No. 2010-73264 Also, SOVA is disclosed in the following documents.

Document: Japanese Patent Application Laid-Open No. 2002-217748 FIG. 3 is a conceptual diagram for explaining the configuration of a frame signal received from a source node by the wireless relay device 1000.

As shown in FIG. 3A, the frame signal includes a preamble, a header, and a frame body, and is received in this order.

Here, the preamble is not convolutionally coded, and the signal in this portion executes the synchronization process as described above.

The header and frame body data are convolutional coded. As described above, the demodulation / decoding unit 110 first performs Viterbi decoding of the header, and the relay control unit 112 analyzes the contents of this header to determine whether or not the received frame is the frame to be relayed. do.

FIG. 3B is a diagram illustrating a configuration of a MAC header in the case of conforming to the current IEEE 802.11.

From the information of Address1 to Address3 in the MAC header, it is possible to determine whether relay transmission is necessary or not. That is, since the address of the destination / source to be relayed is instructed in advance, it is possible to determine whether or not the relay is necessary by checking the address.

In this case, in order to confirm the address, it is necessary to decode the PSDU (Physical Layer convergence protocol Service Data Unit) portion existing in the data body in the data frame structure of the physical layer. The advantage of selecting "demodulation only" like this is diminished.

FIG. 3C is a diagram illustrating a configuration of a physical header in the case of conforming to the current IEEE 802.11a.

Therefore, it is also possible to specify the necessity of relay by using the information of the physical header instead of the MAC header. That is, as shown in FIG. 3C, one reserve bit rsv is assigned to the physical header. Therefore, it is possible to use this reserve bit to indicate the necessity / unnecessity of the relay. When an instruction requiring relay is indicated, the configuration may be such that relay transmission is performed to a predetermined destination.

In this case, decoding of the physical header is necessary, but decoding up to the PSDU part is not essential, and if only demodulation is performed, the processing delay will be until the physical header is read, and relaying with a smaller delay may be performed. It will be possible.

Since IEEE 802.11n, IEEE 802.11ac, etc. also have a reserve bit in the physical header, it is possible to instruct the necessity of relay processing by using this reserve bit in the same way.

FIG. 4 is a block diagram for explaining the configuration of the demodulation / decoding unit 110, the relay control unit 112, and the coding / modulation unit 120 described with reference to FIG.

With reference to FIG. 4, as described above, the relay control unit 112 predicts the frame error rate for the received frame based on the transmission path estimation value and the noise power estimation value from the propagation path estimation unit 108, and the frame error rate prediction unit 1122. When the header analysis unit 1124 that determines the necessity of relay processing and the header analysis unit 1124 determine that relay processing is necessary based on the header information from the demodulation / decoding unit 110, the frame error rate It includes a relay condition setting unit 1126 that sets conditions for decoding processing as described above and conditions for a modulation method and a coding rate according to the prediction result.

The demodulation / decoding unit 110 has, for example, a demodulation processing unit 1101 that executes demodulation processing for multi-value modulation on the signal from the OFDM demodulation unit 106, and a deinterleave processing that executes deinterleave processing on the demodulated bit string. It includes a processing unit 1102 and a depuncturing unit 1103 for executing depunching processing on the deinterleaved signal. The depuncturing unit 1103 includes an internal memory 1103.2 for executing the depuncturing process and a depuncturing processing unit 1103.1 for executing the depuncturing process.

Further, the demodulation / decoding unit 110 includes a Viterbi decoding unit 1104 for performing Viterbi decoding on the depunctured signal. The Viterbi decoding unit 1104 includes an internal memory 1104.2 for executing the Viterbi decoding process and a Viterbi decoding processing unit 1104.1 for executing the Viterbi decoding process. The Viterbi decoding processing unit 1104.1 executes, for example, the ACS calculation and the traceback processing as described above. At this time, the cutoff length is set by the setting of the relay condition setting unit 1126.

Further, the demodulation / decoding unit 110 includes an error detection unit 1105 for performing error detection with an error detection code such as CRC (Cyclic Redundancy Check) for the Viterbi-decoded signal.

The output of the error detection unit 1105 is temporarily stored in the memory 111.

Next, the coding / modulation unit 120 includes an error detection code addition unit 1201 that adds an error detection code such as CRC to the data stored in the memory 111, and a convolution for executing convolution coding. Includes a coding unit 1202.

When the data stored in the memory 111 is sent to the error detection code addition unit 1201, it corresponds to the above-mentioned "no censoring in Viterbi decoding".

As will be described later, depending on the setting of the relay condition setting unit 1126, the data stored in the internal memory 1104.2 in the Viterbi decoding unit 1104 may be directly input to the convolutional coding unit 1202. This corresponds to the above-mentioned "execution of decoding with a cutoff length L in Viterbi decoding and transfer of execution results".

Further, the coding / modulation unit 120 includes a puncturing unit 1203 that executes a puncturing process on the convolutional coded signal. As will be described later, depending on the setting of the relay condition setting unit 1126, the data stored in the internal memory 1103.2 in the depuncturing unit 1103 may be directly input to the puncturing unit 1203. This corresponds to the above-mentioned "transfer the demodulation result without executing the Viterbi decoding process".

Further, the coding / modulation unit 120 includes an interleaving processing unit 1204 that executes interleaving processing on the punctured signal, and a modulation processing unit 1205 that executes modulation processing on the interleaved signal. including.
(Relay transmission method of this embodiment)
Hereinafter, the conventional relay transmission method (DF method) and the relay transmission method of the present embodiment will be described in comparison with each other.

To simplify the representation, we define the terms "S-R frame" and "R-D frame". The former represents a frame between the source node (S) and the relay node (R), and the latter represents a frame between the relay node (R) and the destination node (D).

In the wireless relay device 1000 of the present embodiment, the relay node receives the data frame in one frequency band, while transferring the data frame on another frequency band. [Conventional relay transmission method (DF method)]
FIG. 5 is a conceptual diagram illustrating the processing executed by the relay node in the conventional DF method.

FIG. 7 is a diagram showing each relay method.

As shown in FIG. 5, it is assumed that a convolutional code is used, and the relay node by the relay transmission method of the DF method is a frame using demodulation processing, deinterleaving processing, depuncturing processing, Viterbi decoding, and CRC code. Perform error checking.

If no frame error is detected in the received frame, the decoded bit string is transmitted after the error detection code is added and then after convolutional coding processing, puncturing processing, interleaving processing, and modulation processing.

Here, T _SR and T _RD are assumed to be frame transmission periods of SR frames and RD frames, respectively.

FIG. 7A shows the transmission time of the DF method.

From FIG. 7A, it can be seen that in the DF method, the relay transmission time (relay delay time) T _DF is the total of T _SR and TR _D. [Relay transmission method (TDF method) of the present embodiment]
Next, the relay transmission method of the present embodiment for reducing the delay time will be described below.

In the relay method of the present embodiment, the decoded bit string is started to be transferred before the completion of frame reception by interrupting the decoding process in the middle.

Here, such a relay method of this embodiment will be referred to as a "censored DF (TDF: Truncated-Decode-and-Forward) method" or simply a "TDF method".

FIG. 6 is a conceptual diagram illustrating the processing executed by the relay node in the TDF method.

As shown in FIG. 6A, in the case of "transferring the demodulation result without executing the Viterbi decoding process" (corresponding to L = 0), the result of the demodulation processing performed by the demodulation / decoding unit 110 is , Is sent to the coding / modulation unit 120, and is then transmitted after interleaving and modulation processing.

On the other hand, as shown in FIG. 6B, in the case of "execution of decoding with the cutoff length L in Viterbi decoding and transfer of the execution result", the demodulation / decoding unit 110 performs demodulation processing, deinterleave processing, and depunk. Charling processing and Viterbi decoding processing with path truncation are executed. The bit string decoded by pass censoring is transmitted in the coding / modulation unit 120 after convolutional coding, puncturing processing, interleaving processing, and modulation processing.

Therefore, although the TDF method in this case performs censored decoding, it does not check for S-R frame errors by CRC before starting relay transmission.

Even in the relay method of the present embodiment, in the case of "no censoring in Viterbi decoding", the relay is executed by the same process as the conventional DF method of FIG.

With reference to FIG. 7, in the TDF method, the relay node receives the header of the frame on one frequency band, and when relay transmission is required, the relay node starts relay transmission on another frequency band. , The relay node is receiving a frame as shown in FIG. 7 (b).

In the TDF method, the period from the completion of reception of the orthogonal frequency division multiplexing (OFDM) symbol to the start of its relay transmission is short, so the period of "processing delay" cannot be ignored. Here, the "processing delay" is considered to be a period during which a bit string is received in order to obtain L decoded bits. Further, L is the length of the censored path in the Viterbi decoding.

The delay time of the TDF method is expressed as follows.

ここで、Tpreは、プリアンブルの期間であり、Tsymは、ガードインターバル(GI)のあるＯＦＤＭシンボルの期間であり、NbをOFDMシンボルごとの情報ビットの数であるとするとき、Ｎは、以下の式で表される。

Where Tpre is the preamble period, Tsym is the period of the OFDM symbol with a guard interval (GI), and Nb is the number of information bits per OFDM symbol, then N is: It is expressed by an expression.

ここで、以下の関数は、一般に天井関数と呼ばれ、引数ｘに対して、ｘ以上の最小の整数を表す。

Here, the following function is generally called a ceiling function, and represents the smallest integer greater than or equal to x with respect to the argument x.

Ｔ_SR≦Ｔ_RDである場合、リレー制御部１１２は、S-Rフレームのヘッダの受信および復号の完了後、所定の処理遅延の時間後に、R-Dフレームの中継伝送が始められるように制御する。この場合、処理遅延の時間は、打ち切り長によって異なり、事前に設定されているものとする。

When T _SR ≤ TR, the relay control unit 112 controls so that the relay transmission of the _RD frame is started after a predetermined processing delay time after the reception and decoding of the header of the SR frame are completed. In this case, the processing delay time differs depending on the censoring length and is set in advance.

When T _SR > T _RD , the relay control unit 112 controls the relay transmission so that the relay transmission can be started after the buffering of the received data for generating the transfer frame is completed. By doing so, it is possible to avoid running out of data to be transferred during the transfer.

In this case, it is considered that the end of the transmission of the R-D frame is 2 Tsym after the end of the S-R frame in consideration of the processing delay as shown in the equation (1).

However, when L = 0, the term N · T _sym representing the processing delay in the equation (1) becomes zero. [Adaptive relay transmission selection (ARTS)] As described above, the relay transmission method of the present embodiment can reduce the delay time. However, since the error correction capability in decoding is reduced by the termination of decoding, the frame error rate (FER) tends to deteriorate.

This can lead to retransmissions and can result in higher delay times.

Therefore, in the relay method of the present embodiment, relay transmission is selected according to a channel condition such as a signal-to-noise power ratio (SNR) between a source node and a relay node.

In order to select an appropriate relay transmission method, the frame error rate predictor 1122 uses, for example, mutual information based effective SNR mapping (MIESM) technology to frame the S-R frame. Estimate the error rate (FER).

Here, MIESM is a technique for evaluating FER using mutual information obtained from SNR, and is disclosed in the following documents, for example.

References: L. Wan, S. Tsai, and M. Almgren, "A Fading-Insensitive Performance Metric for a Unified Link Quality Model," Proc. IEEE WCNC2006, Apl. 2006.
Hereinafter, the prediction of FER will be briefly described according to the above-mentioned literature.

For M value modulation (M = 2 ^B ), symbol information for the kth subcarrier whose SNR is γ _k is calculated.

The symbol information is defined by the following equation (2).

ここで、xは、送信されたシンボルであり、yは、SNRがγ_kであるチャネルを通過した受信シンボルであり、P(x)はxの事前確率であり、P(y｜x，γ_k)は、条件付き確率密度関数である。

Where x is the transmitted symbol, y is the received symbol that has passed through the channel where SNR is γ _k , P (x) is the prior probability of x, and P (y | x, γ). _k ) is a conditional probability density function.

Considering that each transmitted symbol is generated with the same probability, P (x) = 1 / M.

Here, the equation (2) can be rewritten as follows.

ここで、zは、分散０．５γ_kで平均０の複素ガウス確率変数であり、Ｘは変調シンボルの集合である。

Here, z is a complex Gaussian random variable with a variance of 0.5γ _k and an average of 0, and X is a set of modulation symbols.

For the SNR of each subcarrier, SI (γ _k , M) is calculated by Eq. (3).

Since the calculation of the equation (3) is generally a large amount of calculation, a look-up table containing the calculation result of the equation (3) is generated in advance for each modulation method, and the equation (3) is calculated each time. It shall be used instead of the calculation.

Next, the received bit information rate (RBIR) is calculated by averaging SI (γ _k , M) for all subcarriers.

Therefore, RBIR is given by the following equation (4).

ここで、Ｋはサブキャリアの数である。

Here, K is the number of subcarriers.

The frame error rate prediction unit 1122 estimates the FER by storing the FER curve in the memory in advance according to the RBIR value and the coding rate and referring to the curve.

It is assumed that the FER curves for each available code rate have been pre-measured under the Additive White Gaussian Noise (AWGN) channel and are also stored in the reference table.

The estimated delay time is calculated using the estimated FER.

(Retransmission control)
FIG. 8 is a flowchart for explaining the retransmission control.

FIG. 8A shows retransmission control by the DF method.

In the DF method, the relay node first determines whether the number of retransmissions has reached the maximum number of transmissions (S100), and if the number of retransmissions has not reached the maximum number (N in S100), the relay node receives the S-R frame and receives the S-R frame. Check for S-R frame errors (S104).

When an error occurs in the S-R frame (N in S104), a retransmission request is transmitted to the source node, the S-R frame is retransmitted from the source node, and the relay node receives it (S102).

Retransmissions are repeated until the number of retransmissions reaches a predetermined upper limit.

After the S-R frame transmission is successful (Y in S104), the relay node transmits the R-D frame (S106), and determines whether the normal reception is returned from the destination node or the maximum number of transmissions is reached (S108). ). If the reception is not normal and the maximum number of transmissions has not been reached (N in S108), the R-D frame is retransmitted. If not (Y in S108), the transmission is terminated.

FIG. 8B shows retransmission control by the TDF method.

In the TDF method, the relay node does not check the S-R frame error, and when the relay node determines that the number of retransmissions is not the maximum number of transmissions (N in S200), the relay node receives the S-R frame (N). S202), the R-D frame is transmitted to the destination node as it is (S204), and the destination node checks for an error in the R-D frame.

If the reply from the destination node is not normally received and the maximum number of transmissions has not been reached (N in S206), the S-R frame is retransmitted from the source node, and the R-D frame is retransmitted from the relay node. If not (Y in S206), the transmission is terminated.

Therefore, the waiting time predicted by the DF method and the delay time predicted by the TDF method are expressed by the following equations (5) and (6), respectively.

ここで、Ｔ_TDF ^currentは、ＴＤＦ方式を使用する現在の伝送の遅延時間であり、Ｔ_TDF ^nextは、ＴＤＦ方式を使用してフレームエラーによって引き起こされる次の再送信の遅延時間であり、これらは、それぞれ式（１）で与えられる。

Where T _TDF ^current is the delay time of the current transmission using the TDF method and T _TDF ^next is the delay time of the next retransmission caused by a frame error using the TDF method. , Respectively given by equation (1).

The relay condition setting unit 1126 selects the conditions for the decoding process so that the relay transmission method is such that the minimum delay time is predicted.

Therefore, in the TDF method, if the relay condition setting unit 1126 has several options for the path cutoff length, the expected delay time is calculated for all the options, and the relay node also has the path cutoff length. , Select to have the minimum delay time.
(Performance evaluation)
In the following, the evaluation results of the performance of the TDF method and the ARTS method will be described by computer simulation.

FIG. 9 is a diagram showing simulation parameters.

FIG. 10 is a diagram showing modulation and coding schemes (MCS) and simulation conditions.

The coding rate and modulation scheme as shown in FIG. 10 (a) are based on IEEE 802.11a standards.

FIG. 10B is a list of conditions in the simulation.

FIG. 11 is a graph showing the delay time of each relay transmission method under the condition 5 of FIG. 10 (b).
The initial MCS used for the SR frame is shown in FIG. The MCS used between RDs is BPSK R = 1/2 and SNR between RDs = 5 dB.

In FIG. 11, it can be seen that the DF / TDF achieves the minimum or almost the minimum delay time in the adaptively controlled ARTS method, although the superiority and inferiority of the DF / TDF may change depending on the situation.

Therefore, it can be seen that the ARTS method can select an appropriate relay transmission method and reduce the delay time.

As described above, as described as the specific configuration and operation of the wireless relay device 1000 as the embodiment, the TDF method transfers the bit string obtained by censoring the decoding before the completion of reception. In the TDF method, the relay node receives the header of the frame of a certain frequency band. When relay transmission is required by the header information, the relay node starts relay transmission on one frequency band, while the relay node receives a frame on another frequency band.

The TDF method reduces the delay time of relay transmission, but its FER characteristics tend to deteriorate as compared with the DF method that performs decoding without censoring. Therefore, the relay transmission method is selected according to the channel state so that the total delay time until the transmission from the source node to the destination node is completed does not increase. In this case, a relay transmission method that achieves the minimum delay time is selected.

That is, as described above, according to the wireless relay device 1000 of the present embodiment and the wireless relay method executed by the wireless relay device 1000, unnecessary relay transmission can be suppressed by confirming the header and then performing the transfer.

In addition, when the communication condition is good, relay transmission can be performed with low delay by using a relay method with less delay, and when the communication condition is bad, a relay method with excellent error rate characteristics can be used for retransmission. Delays can be avoided and minimized.

The embodiments disclosed this time are examples of configurations for concretely implementing the present invention, and do not limit the technical scope of the present invention. The technical scope of the present invention is shown by the scope of claims, not the description of embodiments, and includes modifications within the wording scope of the claims and the scope of equal meaning. Is intended.

100 Receiving antenna 102, 124 RF unit, 104 synchronization processing unit, 106 OFDM demodulation unit, 108 propagation path estimation unit, 110 demodulation / decoding unit, 111 memory, 112 relay control unit, 120 coding / modulation unit, 122 OFDM modulation unit , 130 transmit antenna, 1000 wireless repeater, 1010 source node, 1020 destination node, 1101 demodulation processing unit, 1102 orthogonal leaving processing unit, 1103 depuncturing unit, 1104 bitabi decoding unit, 1105 error detection unit, 1122 frame error rate. Prediction unit, 1124 header analysis unit, 1126 relay condition setting unit, 1201 error detection code addition unit, 1202 convolution coding processing unit, 1203 puncturing unit, 1204 interleaving processing unit, 1205 modulation processing unit.

Claims (9)

A wireless relay device for communicating by a wireless communication method capable of full-duplex communication, relaying a first frame signal from a source node, and transferring it to a destination node as a second frame signal.
A receiving means for receiving a signal from the source node, and
A demodulation / decoding means that demodulates the first frame signal received by the receiving means and decodes at least the header information of the first frame signal.
A relay control means for controlling the relay process according to the determination that it is necessary to relay the first frame signal based on the information in the header.
According to the control of the relay control means, the data from the demodulation / decoding means is encoded and modulated to obtain the second frame signal during a period that at least partially overlaps with the reception period of the first frame signal. Coding / modulation means for generation,
A transmission means for transmitting the output of the coding / modulation means toward the destination node is provided.
The demodulation / decoding means is
A demodulation means for executing a predetermined demodulation process to generate a bit string,
Decoding means for executing the decoding process for the trellis code based on the bit string,
Includes an error detecting means for performing error detection on the output of the decoding means.
The coding / modulation means selectively receives the data output from either the demodulation means or the decoding means according to the control of the relay control means, encodes and modulates the second data. A wireless relay device that generates a frame signal of . The coding / modulation means is a coding / modulation method controlled independently of the demodulation / decoding method of the demodulation / decoding means according to the control of the relay control means, and the second frame signal is used. The wireless relay device according to claim 1 . The wireless relay device according to claim 1 or 2, wherein the wireless communication method capable of full-duplex communication is a method of wireless communication using each wireless channel of a plurality of frequency bands separated from each other. The decoding means can set the cutoff length on the trellis for the decoding process for the trellis code.
The relay control means includes a communication quality estimating means for estimating the communication quality between the source node and the wireless relay device based on the first frame signal.
The first aspect of the present invention includes a relay condition setting means for setting the cutoff length so as to reduce the expected value of the time required for transferring the second frame signal based on the estimated communication quality. Wireless relay device. 4. The communication quality estimating means includes an error rate estimating means that acquires a propagation path estimated value and a noise power estimated value based on the first frame signal and estimates a frame error rate as the communication quality. The described wireless repeater. The relay condition setting means calculates an expected value of the time required for transmission of the first frame signal, including retransmission, based on the communication quality, and encodes the expected value so as to minimize the expected value. The wireless relay device according to claim 4 , wherein the coding and modulation method in the modulation means is selected from predetermined methods. The relay condition setting means calculates the relay time required for the relay transmission, and if the relay time is longer than the frame period of the first frame signal , the relay transmission is started after the header is received, according to claim 6 . Wireless relay device. The relay condition setting means calculates the relay time required for relay transmission, and if the relay time is shorter than the frame period of the first frame signal , the first is until the timing at which the transfer data is not insufficient. The wireless relay device according to claim 6 , wherein relay transmission is started after the data of the frame signal of 1 is buffered. It is a wireless relay method for communicating by a wireless communication method capable of full-duplex communication, relaying a first frame signal from a source node, and transferring it to a destination node as a second frame signal.
The step of receiving the signal from the source node and
A demodulation / decoding step of demodulating the received first frame signal and decoding at least header information of the first frame signal.
A relay control step that controls relay processing according to the determination that it is necessary to relay the first frame signal based on the information in the header.
In response to the control of the relay processing, the data before the completion of the processing in the demodulation / decoding step is received, encoded and modulated in a period at least partially overlapping with the reception period of the first frame signal. The coding / modulation step to generate the second frame signal,
A step of transmitting the modulated second frame signal toward the destination node is provided.
The demodulation / decoding step
A demodulation step that performs a predetermined demodulation process to generate a bit string,
A decoding step that executes a decoding process for the trellis code based on the bit string,
Including a step of executing error detection for the result of the decoding step .
The coding / modulation step selectively receives data of either the demodulation step or the result of the decoding step according to the control in the relay control step, encodes and modulates the second frame. A wireless relay method that generates a signal.

Images