Patent Application
20180269938
This application is based upon and claims the benefit of priority of the prior Japanese Patent Application No. 2017-048165, filed on Mar. 14, 2017, the entire contents of which are incorporated herein by reference.
The embodiments discussed herein are related to a communication device, a communication system, and a communication method.
There are ongoing investigations into joint transmission multi-user multi-input multi-output (JT-MU-MIMO), which combines joint transmission (JT) with multi-user multi-input multi-output (MU-MIMO). Joint transmission is implemented using transmission points (referred to as TPs below) that have been arranged at different locations, and in MU-MIMO, spatial multiplexing is used to transmit downlink signals to plural communication terminals (referred to as user equipment (UE) below). In JT-MU-MIMO, in order to reduce downlink signal interference between UEs receiving spatially multiplexed downlink signals, at the TPs, transmission signals are multiplied by transmission weights acquired using zero-forcing (ZF) or an orthogonalization technique such as block-diagonalization.
Differences in propagation delays arise between TPs and UEs as a result of differing propagation distances between each TP and the UEs. For example, when a downlink signal is transmitted from a second TP to a UE while a downlink signal is being transmitted from a first TP to the UE, the following occurs. If the distance between the second TP and the UE is longer than the distance between the first TP and the UE, then the downlink signal transmitted from the second TP is received by the UE at a delay with respect to the downlink signal transmitted from the first TP.
In such cases, the received phase of the downlink signal transmitted from the second TP is rotated in the frequency domain with respect to the received phase of the downlink signal transmitted from the first TP. Namely, the downlink signals transmitted from the two TPs are received at the UE with a different phase difference at each frequency, and since the interference between the downlink signals from the two TPs is frequency dependent, the optimal transmission weights are frequency dependent.
In order to acquire optimal frequency-dependent transmission weights, in other technology, for example, the frequency band is divided into predetermined band segments and a transmission weight is computed for each segment. However, processing load increases as the number of segments increases.
There is also technology in which transmission weights are first computed without increasing the number of segments, namely, without narrowing the given bands in each segment division, and then frequency intervals are narrowed and transmission weights interpolated. However, in such technology, it is difficult to interpolate transmission weights with appropriate values approximating the optimal transmission weights when there are large differences in propagation delays.
According to an aspect of the embodiments, a communication device comprising: a memory; and a processor connected to the memory, the processor being configured to: compute reflected transmission weights based on timing correction coefficients to be applied to each of a plurality of antennas according to reception timings of uplink signals from a plurality of wireless terminals that receive multiplexed downlink signals transmitted from the plurality of antennas, and on respective channel estimation values between the plurality of antennas and the plurality of wireless terminals, the computed reflected transmission weights reflecting the timing correction coefficients and being computed at first frequency intervals; compute interpolated transmission weights by interpolating between the computed reflected transmission weights at second frequency intervals that are narrower than the first frequency intervals; and compute transmission weights to be applied to downlink signals by, based on the timing correction coefficients, correcting the interpolated transmission weights that have been interpolated such that the timing correction coefficients are reflected a second time.
The object and advantages of the invention will be realized and attained by means of the elements and combinations particularly pointed out in the claims.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are not restrictive of the invention.
Detailed description follows regarding an example first exemplary embodiment, with reference to the drawings.
A CBBU 20 is illustrated on the right side of the broken line D in
Each TP 30-k includes an uplink wireless processing section 31-k, a cyclic prefix (CP) removal section 32-k, a fast Fourier transform (FFT) section 33-k, a reception timing detection section 34-k, and a channel estimation section 35-k. Each TP 30-k also includes an inverse fast Fourier transform (IFFT) section 36-k, a CP appending section 37-k, and a downlink wireless processing section 38-k.
The uplink wireless processing section 31-k performs down conversion and analog-to-digital (A/D) conversion on a received signal, which is an uplink signal transmitted from a UE 40 that has been received by antenna 50-k. The CP removal section 32-k removes CP from the received signal. The FFT section 33-k performs fast Fourier transformation (FFT) on the received signal to convert the received signal into a signal in the frequency domain.
In each of the n TPs, the channel estimation section 35-k employs a known signal, such as a sounding reference signal (SRS), contained in the received signal to perform uplink channel estimation for each resource block (referred to as RB below) for the m UEs 40. The channel estimation section 35-k then converts an acquired uplink channel estimation value between a UE 40-i and the antenna 50-k into a downlink channel estimation value hi,k(r), as expressed by Equation (1). r is a number to distinguish between resource blocks, with r=1, . . . , N_RB. N_RB indicates the number of RBs.
h
i,k(r)=Ci,k(r)·{tilde over (h)}i,k(r) (1)
In Equation (1), the hi,k(r) with a tilde “˜” on top is referred to below as “tilde hi,k(r)”. Tilde hi,k(r) is an uplink channel estimation value for RB number r between UE 40-i and antenna 50-k. Ci,k(r) is a coefficient for converting an uplink channel estimation value to a downlink channel estimation value. Ci,k(r) is determined so as to adjust for a difference between an uplink signal transmission power and a downlink signal transmission power, and for a difference between an uplink circuit response and a downlink circuit response.
The reception timing detection section 34-k employs a known signal, such as an SRS, to detect a reception timing τi,k at the antenna 50-k of an uplink signal from a UE 40-i to the antenna 50-k. The reception timing τi,k corresponds to an amount of propagation delay between a UE 40-i and a TP 30-k. For example, when there is a long propagation distance between a UE 40-i and a TP 30-k, and there is a large amount of propagation delay, reception timing will be delayed. A reception timing τi,k at each of the n antennas is detected for each of the m UEs.
Based on the downlink channel estimation values hi,k(r) acquired by the channel estimation section 35-k, the scheduling section 21 selects UEs 40 for multiplexed signal transmission from out of the m UEs 40 being scheduled.
Based on the reception timing τi,k detected by the reception timing detection section 34-k and information relating to the UEs 40 selected by the scheduling section 21 for multiplexed signal transmission, the timing correction coefficient acquisition section 22-k determines a timing correction coefficient τ_ck for the antenna 50-k. As expressed by Equation (2), the timing correction coefficient τ_ck may, for example, be the reception timing τi,k for which the interval of time between a multiplexed signal transmission and a corresponding reception timing is smallest, namely the timing correction coefficient τ_ck may be the reception timing τi,k corresponding to the UE 40-i for which the reception timing at antenna 50-k is earliest. The reason for this is that received power p_rxi,k is strongest for UEs 40 having the earliest reception timings, and phase rotation suppression can be expected.
M is the collection of UEs 40 selected by the scheduling section 21 for multiplexed signal transmission.
The channel estimation value correction section 23-k employs the timing correction coefficient τ_ck, which was acquired by the timing correction coefficient acquisition section 22-k for the UEs 40 selected by the scheduling section 21, to correct the channel estimation values hi,k(r) acquired by the channel estimation section 35-k. Namely, as expressed by Equation (3), the timing correction coefficient τ_ck is applied so as to impart a phase rotation to the downlink channel estimation values hi,k(r).
h′
i,k(r)=hi,k(r)·exp(j·2π·Δ_RB·r·τ_ck) (3)
h′i,k(r) is the corrected channel estimation value, and Δ_RB is the frequency interval between RBs.
The transmission weight computation section 24 generates a concatenated channel matrix H′(r) that includes a channel estimation value h′i,k(r) for each RB of the antennas 50 with respect to a UE 40-i (i=1, . . . , |M|) acquired by the channel estimation value correction section 23, as expressed by of Equation (4).
The transmission weight computation section 24 employs the generated concatenated channel matrix H′(r) to compute transmission weights W′(r) for each RB. The transmission weights W′(r) for each RB may, for example, be computed using zero-forcing (ZF) as expressed by Equation (5).
W′(r)=H′H(r)·(H′(r)·H′(r)H)−1·P(r) (5)
In Equation (5), the superscripted H indicates a Hermitian transpose, and P(r) indicates a diagonal matrix that normalizes transmission weight power.
As expressed by Equation (6), the transmission weight interpolation section 25 interpolates between the transmission weights computed by the transmission weight computation section 24, namely, the transmission weights at each RB frequency interval, Δ_RB so as to give transmission weights at subcarrier (also referred to as SC below) frequency intervals.
The RB frequency interval Δ_RB is an example of a first frequency interval, the SC frequency interval is an example of a second frequency interval, and wk,i′(s) with “˜” on top, namely, tilde wk,i′(s), is transmission weight after interpolation. N_scrb is the number of SCs for each RB, where s represents an SC number, and w′k,i(r) is the element in the kth row and the ith column of W′(r). Moreover, “%” is an operator, whereby z1%z2 expresses a remainder obtained by dividing z1 by z2.
The transmission weight correction section 27-k employs the timing correction coefficient τ_ck acquired by the timing correction coefficient acquisition section 22-k to correct row number k of the transmission weights acquired by the transmission weight interpolation section 25, as expressed by Equation (7). Namely, the transmission weight correction section 27-k removes the phase rotation imparted to the channel estimation values in Equation (3).
w
k,i(s)={tilde over (w)}k,i′(s)·exp(j·2π·Δ_sc·s·τ_ck) (7)
wk,i(s) is the corrected transmission weight, and Δ_sc is the SC frequency interval.
The user data generation section 26 generates symbol data for transmission to a UE 40-i selected by the scheduling section 21. The transmission weight multiplication section 28-k multiplies each SC of the symbol data generated by the user data generation section 26 by the transmission weight wk,i(s) that was corrected by the transmission weight correction section 27-k.
The IFFT section 36-k performs inverse fast Fourier transform (IFFT) on the symbol data multiplied by the transmission weight wk,i(s) so as to convert the symbol data into a transmission signal in the time domain. The CP appending section 37-k appends CP to the transmission signal that was converted into the time domain. The downlink wireless processing section 38-k up converts the CP-appended transmission signal to a wireless frequency, and performs digital-to-analog (D/A) conversion on the up-converted transmission signal.
The CBBU 20 includes a processor 41, memory 42, and a network interface (NIF) 44. The NIF 44 is, for example, an interface section to connect to higher-level devices, such as a gateway device of a functional network.
The memory 42 holds a communication processing program. The processor 41 reads the communication processing program from the memory 42 and executes the communication processing program so as to respectively function as the scheduling section 21, the timing correction coefficient acquisition sections 22-k, and the channel estimation value correction sections 23-k in
Each TP 30-k includes a wireless processing section 39-k, a processor 56-k, and memory 57-k. The wireless processing section 39-k includes the uplink wireless processing section 31-k, the CP removal section 32-k, the FFT section 33-k, the IFFT section 36-k, the CP appending section 37-k, and the downlink wireless processing section 38-k illustrated in
The processor 41 and the processors 56-k may, for example, be configured by a central processing unit (CPU), a Field Programmable Gate Array (FPGA), or a Digital Signal Processor (DSP). The memory 42 and the memory 57-k may, for example, be configured by random access memory (RAM) or read only memory (ROM).
At step 101-k, the processor 56-k of TP 30-k employs a known signal, such as an SRS, contained in the received signal from a UE 40-i to detect a reception timing τi,k at the antenna 50-k of an uplink signal from the UE 40-i to the antenna 50-k. At step 102-k, the processor 56-k employs a known signal, such as an SRS, to acquire an uplink channel estimation value, and converts the acquired uplink channel estimation value to a downlink channel estimation value hi,k(r).
At step 103, based on the downlink channel estimation values hi,k(r), the processor 41 of the CBBU 20 selects UEs 40-i (i=1, . . . , |M|) for multiplexed signal transmission from out of the UEs 40 being scheduled. At step 104, based on the reception timings τi,k and information relating to the UEs 40-i selected at step 103 for multiplexed signal transmission, the processor 41 determines a timing correction coefficient τ_ck for each antenna 50-k.
At step 105, the processor 41 employs the timing correction coefficients τ_ck acquired at step 104 to correct the channel estimation values hi,k(r) acquired at step 102-k. At step 106, the processor 41 computes transmission weights W′(r) based on the channel estimation values corrected at step 105. At step 107, the processor 41 interpolates between the transmission weights at each RB frequency interval so as to give transmission weights at SC frequency intervals.
At step 108, the processor 41 employs the timing correction coefficients τ_ck acquired at step 104 to correct the transmission weights interpolated at step 107. At step 109, the processor 41 multiplies the symbol data in each SC transmitted to the UE 40-i by the corrected transmission weights.
Although an example has been described in which the timing correction coefficient τ_ck is the reception timing τi,k of the UE 40 having the earliest reception timing, the present exemplary embodiment is not limited thereto. For example, the received power p_rxi,k and reception timings τi,k of received signals from the selected UEs 40-i (i=1, . . . , |M|) may be employed to compute a weighted average of reception timings τi,k, which is then employed as the timing correction coefficient τ_ck, as expressed by Equation (8) below. The received power p_rxi,k at the antenna 50-k of the uplink signal from a UE 40-i to the antenna 50-k is detected by the reception timing detection section 34-k employing a known signal, such as an SRS.
However, in cases in which the UEs 40-i control uplink transmission power, pi,k is employed in place of received power p_rxi,k, as expressed by Equation (9).
p
i,k
=p_rxi,k−p_txf (9)
p_txi is the transmission power value of UE 40-i and may, for example, be estimated by employing the uplink power headroom (UPH) reported by the UE 40-i.
Moreover, the timing correction coefficient τ_ck may be computed as expressed by Equation (10). Equation (10) differs from Equation (8) in that p_rxi,k·τi,k is multiplied by a coefficient coe. The coefficient coe is a value determined based on the received power p_rxi,k, and may, for example, be set to 1 when the received power p_rxi,k exceeds a predetermined value, and set to 0 when not exceeding the predetermined value. Moreover, the coefficient coe may be a value that increases in steps as received power increases.
Note that although an example has been described in which the CBBU 20 and the TPs 30-k are separate bodies in
The processor 58 may, for example, be configured by a CPU, FPGA, or DSP. The memory 59 may, for example, be configured by RAM or ROM.
Note that although an example has been described in which the transmission weight multiplication sections 28-k and the transmission weight correction sections 27-k are included in the CBBU 20 in
Note that although an example has been described in which the TPs 30-k (or the wireless processing sections 39-k) each include a single antenna 50-k, the present exemplary embodiment is not limited thereto. Two or more of the antennas 50 may be connected to a single TP 30 (or to a single wireless processing section 39). In such cases the total number of the TPs 30 (or the wireless processing sections 39) is smaller than the total number n of the antennas.
Note that although an example has been described in which transmission weights computed at RB frequency intervals are interpolated at SC frequency intervals, the present exemplary embodiment is not limited thereto. For example, the transmission weights may be interpolated at sub-band frequency intervals covering a predetermined number of SCs. Note that although an example has been described in which the transmission weights are computed using zero-forcing, the present exemplary embodiment is not limited thereto. The transmission weights may, for example, be computed using diagonalized blocks.
The timing correction coefficient acquisition sections of the present exemplary embodiment acquire timing correction coefficients to apply to each of plural antennas based on the reception timings of uplink signals from plural wireless terminals that receive multiplexed downlink signals transmitted from the plural antennas. Based on the timing correction coefficients acquired by the timing correction coefficient acquisition sections, and on the respective channel estimation values between the plural antennas and the plural wireless terminals, the computation section computes reflected transmission weights at first frequency intervals that reflect the timing correction coefficients. The transmission weight interpolation section computes interpolated transmission weights by interpolating between the reflected transmission weights computed by the computation section at second frequency intervals that are narrower than the first frequency intervals. The transmission weight correction section computes transmission weights to be applied to downlink signals by correcting the interpolated transmission weights interpolated by the transmission weight interpolation section based on the timing correction coefficients, such that the timing correction coefficients are reflected a second time.
The computation section includes the channel estimation value correction section and the transmission weight computation section. The channel estimation value correction section corrects the channel estimation values based on the timing correction coefficients, and the transmission weight computation section computes the reflected transmission weights at the first frequency intervals based on the corrected channel estimation values that have been corrected by the channel estimation value correction section.
In the present exemplary embodiment, channel estimation values between each of plural antennas and plural wireless terminals are corrected for a set of plural wireless terminals using timing correction coefficients for each antenna. The timing correction coefficients suppress phase rotation, and are employed to impart phase rotation to the channel estimation values such that transmission weights are computed based on the channel estimation values in a state in which fluctuations in phase with respect to frequency due to timing offsets are suppressed. After the interpolation of transmission weights, the timing correction coefficients are employed to correct the transmission weights so as to remove the imparted phase rotation. The present exemplary embodiment thereby enables the acquisition of appropriate transmission weights to be applied to downlink signals from plural antennas to plural wireless terminals.
An example second exemplary embodiment will now be described. Description will be omitted for configuration and operation similar to that of the first exemplary embodiment. The second exemplary embodiment differs from the first exemplary embodiment in that channel estimation value correction is not performed, and that transmission weights are corrected prior to interpolating the transmission weights.
The transmission weight computation section 24′ concatenates the UEs 40-i (i=1, . . . , |M|) acquired by channel estimation sections 35-k and downlink channel estimation values hi,k(r) for antennas 50-k (k=1, . . . , n) to generate a concatenated channel matrix H(r), as expressed by Equation (11).
The transmission weight computation section 24′ employs the concatenated channel matrix H′(r) to compute transmission weights W(r). The transmission weights W(r) may, for example, be computed using zero-forcing as expressed by Equation (12).
W(r)=HH(r)·(H(r)·H(r)H)−1·P(r) (12)
The first transmission weight correction section 27A-k corrects the transmission weights W(r) computed by the transmission weight computation section 24′ using the timing correction coefficient τ_ck acquired by the timing correction coefficient acquisition section 22-k, as expressed by Equation (13). The transmission weight computation section 24′ and the first transmission weight correction sections 27A-k are an example of a computation section.
w
k,i(r)=wk,i(r)·exp(−j·2π·Δ_RB·r·τ_ck) (13)
wk,i′(r) is a corrected transmission weight, and wk,i(r) indicates the element in row k and column i of W(r) in Equation (12), namely, the transmission weight corresponding to antenna 50-k and UE 40-i.
The transmission weight interpolation section 25′ interpolates between transmission weights w′k,i(r) corrected by the first transmission weight correction sections 27A-k, namely transmission weights at each RB frequency interval, so as to give transmission weights at SC frequency intervals, as expressed by Equation (14).
Second transmission weight correction sections 27B-k employ the timing correction coefficients τ_ck acquired by the timing correction coefficient acquisition sections 22-k to correct the transmission weights interpolated by the transmission weight interpolation section 25′, as expressed by Equation (15). Namely, the second transmission weight correction sections 27B-k remove the phase rotation imparted to the transmission weights in Equation (13).
w
k,i(s)={tilde over (w)}i,k′(s)·exp(j·2π·Δ_sc·s·τ_ck) (15)
At step 104, based on the reception timings τi,k and information relating to the UEs 40-i selected at step 103 for multiplexed signal transmission, the processor 41 of the CBBU 20 determines a timing correction coefficient τ_ck for each antenna 50-k. At step 106, the processor 41 computes transmission weights W(r) based on the uncorrected downlink channel estimation values hi,k(r). At step 108A, the processor 41 employs the timing correction coefficients τ_ck acquired at step 104 to correct the transmission weights W(r) computed at step 106.
At step 107, the processor 41 interpolates between the transmission weights at each RB frequency interval so as to give transmission weights at SC frequency intervals. At step 108B, the processor 41 employs the timing correction coefficients τ_ck acquired at step 104 to correct the transmission weights interpolated at step 107.
Note that although an example has been described in which the second transmission weight correction sections 27B-k and the transmission weight multiplication sections 28-k are included in the CBBU 20, the present exemplary embodiment is not limited thereto. For example, the second transmission weight correction sections 27B-k and the transmission weight multiplication sections 28-k may be included in the TPs 30-k.
The timing correction coefficient acquisition sections of the present exemplary embodiment acquire timing correction coefficients to apply to each of plural antennas based on the reception timing of the uplink signals from plural wireless terminals that receive multiplexed downlink signals transmitted from the plural antennas. Based on the timing correction coefficients acquired by the timing correction coefficient acquisition sections, and on the respective channel estimation values between the plural antennas and the plural wireless terminals, the computation section computes reflected transmission weights at first frequency intervals that reflect the timing correction coefficients. The transmission weight interpolation section computes interpolated transmission weights by interpolating between the reflected transmission weights computed by the computation section at second frequency intervals that are narrower than the first frequency intervals. The transmission weight correction section computes transmission weights to be applied to downlink signals by correcting the interpolated transmission weights interpolated by the transmission weight interpolation section based on the timing correction coefficients, such that the timing correction coefficients are reflected a second time.
The computation section includes the transmission weight computation section and the transmission weight correction section. Based on the channel estimation values, the transmission weight computation section computes transmission weights prior to having the timing correction coefficient reflected therein at the first frequency intervals. Based on the timing correction coefficients, the transmission weight correction section corrects the transmission weights computed by the transmission weight computation section, thereby computing reflected transmission weights.
In the present exemplary embodiment, downlink-signal-multiplying transmission weights are corrected for a set of plural wireless terminals using timing correction coefficients for each antenna. The timing correction coefficients suppress phase rotation, and are employed such that transmission weights are interpolated in a state in which phase rotation has been imparted, namely, in which fluctuations in phase with respect to frequency due to timing offsets are suppressed. The timing correction coefficients are employed to correct the transmission weights so as to remove the imparted phase rotation from the transmission weights. The present exemplary embodiment thereby enables the acquisition of appropriate transmission weights to be applied to downlink signals from plural antennas to plural wireless terminals.
An example of related technology will now be described.
The communication system 110 differs from the communication system 10 of the first exemplary embodiment in that it does not include configuration elements corresponding to timing correction coefficient acquisition sections 22-k, channel estimation value correction sections 23-k, transmission weight correction sections 27-k, or reception timing detection sections 34. The communication system 110 also differs from the communication system 10 of the second exemplary embodiment in that it does not include timing correction coefficient acquisition sections 22-k, first transmission weight correction sections 27A-k, second transmission weight correction sections 27B-k, or reception timing detection sections 34-k.
The CBBU 120 includes a scheduling section 121, a transmission weight computation section 124, a transmission weight interpolation section 125, a user data generation section 126, and a transmission weight multiplication section 128-k. Each TP 130-k includes an uplink wireless processing section 131-k, a CP removal section 132-k, a FFT section 133-k, a channel estimation section 135-k, an IFFT section 136-k, a CP appending section 137-k, and a downlink wireless processing section 138-k.
The uplink wireless processing section 131-k, the CP removal section 132-k, and the FFT section 133-k are similar to the uplink wireless processing section 31-k, the CP removal section 32-k, and the FFT section 33-k in
In the related technology, the processors of the TPs 130-k acquire channel estimation values at step 202-k. At step 203, the processor of the CBBU 120 selects UEs 140-i ((i=1, . . . , |M|) to transmit a signal to using multiplexed transmission based on the channel estimation values.
At step 206, the processor of the CBBU 120 computes transmission weights based on the channel estimation values acquired at step 202, and at step 207, the processor of the CBBU 120 interpolates between the transmission weights at each RB frequency interval so as to give transmission weights at SC frequency intervals. At step 209, the processor of the CBBU 120 multiplies each SC of the symbol data to be transmitted to a UE 140-i by the respective interpolated transmission weights
In the present exemplary embodiment, phase rotation is imparted to the transmission weight at first frequency intervals so as to suppress fluctuations in phase with respect to frequency due to timing offsets. Then transmission weights are interpolated at second frequency intervals that are narrower than the first frequency intervals, and phase rotation that had been imparted is removed from the interpolated transmission weights. The present exemplary embodiment thereby enables the acquisition of transmission weights approximating optimal transmission weights.
There is also a related technology that obviates the interpolation of transmission weights by limiting the UEs receiving multiplexed downlink signals to the UEs for which a difference in propagation delays from the TPs is smallest, and adjusting the transmission timings from the TPs so as to make the reception timings at the UEs closer together. However, in this technology, the combinations of UEs are limited, and there is a chance that the number of UEs capable of receiving multiplexed downlink signals may be reduced. Moreover, for UEs with small propagation delay differences, it is highly likely that propagation losses from the TPs will converge, making orthogonalization of the downlink signals difficult.
On the other hand, the present disclosure enables the acquisition of appropriate transmission weights to be applied to downlink signals from plural antennae to plural wireless terminals.
All examples and conditional language provided herein are intended for the pedagogical purposes of aiding the reader in understanding the invention and the concepts contributed by the inventor to further the art, and are not to be construed as limitations to such specifically recited examples and conditions, nor does the organization of such examples in the specification relate to a showing of the superiority and inferiority of the invention. Although one or more embodiments of the present invention have been described in detail, it should be understood that the various changes, substitutions, and alterations could be made hereto without departing from the spirit and scope of the invention.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2017-048165 | Mar 2017 | JP | national |
