JP2012004865A - Fading frequency estimation apparatus, fading frequency estimation method and base station device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a fading frequency estimation apparatus capable of estimating a fading frequency without narrowing down a fading frequency estimation enabled range and without reducing throughput.SOLUTION: A fading frequency estimation apparatus 1 has: a channel estimation section 11 for determining a channel estimate between a communication terminal and a receiving apparatus in which the fading frequency estimation apparatus 1 is packaged, over a plurality of frequency bands; a frequency correlation value calculation section 12 for determining a correlation value between channel estimates of two frequency bands to calculate a frequency correlation value of a channel estimate; a time/frequency correlation value calculation section 13 for correlating the channel estimates of the two frequency bands while deviating them just by a predetermined time interval to calculate a time/frequency correlation value of the channel estimate; a time correlation value calculation section 14 for calculating a ratio of the time/frequency correlation value with respect to the frequency correlation value; and a frequency estimation section 15 for estimating a fading frequency corresponding to that ratio as a fading frequency of a radio signal.

Description

  The present invention relates to, for example, a fading frequency estimation device that estimates a fading frequency of a radio signal received from a communication terminal, a fading frequency estimation method, and a base station device that controls communication with a mobile station device using the estimated fading frequency. .

ここでc(m/sec)は光速であり、v(m/sec)は基地局装置に対する移動局装置の相対的な移動速度を表す。またf_c(Hz)は、移動局装置から送信された無線信号のキャリア周波数を表す。（１）式から明らかなように、基地局装置に対する移動局装置の相対的な移動速度が速いほど、フェージング周波数も大きくなり、その結果、受信信号の電力レベルの変化量及び位相の変化量も大きくなる。そのため、基地局装置と移動局装置間の通信を適切に制御するために、基地局装置においてフェージング周波数を正確に推定できる技術が求められている。 In a mobile radio communication system, communication may be performed between a mobile station apparatus and a base station apparatus while a user carrying a mobile station apparatus such as a mobile terminal moves. When the mobile station apparatus moves in an environment where there are a large number of scattered waves from the surroundings, the base station apparatus is moved from the mobile station apparatus by the Doppler effect according to the relative moving speed of the mobile station apparatus with respect to the base station apparatus. The frequency of the received signal that is the received radio signal varies. Such frequency variation due to the Doppler effect is called a fading frequency. This fading frequency f _D (Hz) is expressed by the following equation.

Here, c (m / sec) is the speed of light, and v (m / sec) represents the relative moving speed of the mobile station apparatus with respect to the base station apparatus. F _c (Hz) represents the carrier frequency of the radio signal transmitted from the mobile station apparatus. As is clear from the equation (1), the faster the relative moving speed of the mobile station apparatus with respect to the base station apparatus, the larger the fading frequency becomes. As a result, the received signal power level change amount and phase change amount also increase. growing. Therefore, in order to appropriately control communication between the base station apparatus and the mobile station apparatus, a technique capable of accurately estimating the fading frequency in the base station apparatus is required.

ここでr(t)は、時刻tにおける既知のシンボルについての受信信号から得られたチャネル推定値を表し、r^*(t)は、r(t)の複素共役を表す。またΔTは時間間隔を表す。そして関数E_t[x(t)]は、時間によって変動する変数x(t)の時間平均を求める関数を表す。
チャネル推定値r(t)が理想的なレイリーフェージングに従って変動し、かつ雑音を含まない場合、時間相関値ρ_t(ΔT)は次式により表される。

ここで関数J₀(x)は、第1種0次ベッセル関数である。従って、基地局装置は、（２）式に基づいて算出された時間相関値ρ_t(ΔT)を（３）式に代入することにより、フェージング周波数を求めることができる。
ただし、第1種0次ベッセル関数J₀(x)の出力値は、入力値xが増加するにつれて振動するので、出力値に対応する入力値が一意に定められる出力値の範囲は限られている。例えば、時間相関値ρ_t(ΔT)が0.2であれば、対応するf_DΔTの値は0.2、0.62、0.79の何れかとなる。フェージング周波数f_Dを一意に決定することが可能な時間相関値ρ_t(ΔT)は、0.3よりも大きく、その時間相関値ρ_t(ΔT)に対応するf_DΔTの値は0.3未満である。例えば、時間間隔ΔTが10msecであれば、推定可能なフェージング周波数f_Dは30Hz未満となる。 Therefore, based on the time correlation value of the channel estimation value representing the communication state between the mobile station apparatus and the base station apparatus, which is obtained from a known symbol, such as a pilot signal, included in the received signal, the fading frequency A technique for estimating the above has been developed. For example, the theoretical value ρ _t (ΔT) of the time correlation value is expressed by the following equation.

Here, r (t) represents a channel estimation value obtained from a received signal for a known symbol at time t, and r ^* (t) represents a complex conjugate of r (t). ΔT represents a time interval. The function E _t [x (t)] represents a function for obtaining a time average of a variable x (t) that varies with time.
When the channel estimation value r (t) varies according to ideal Rayleigh fading and does not include noise, the time correlation value ρ _t (ΔT) is expressed by the following equation.

Here, the function J ₀ (x) is a first kind 0th-order Bessel function. Accordingly, the base station apparatus can obtain the fading frequency by substituting the time correlation value ρ _t (ΔT) calculated based on the equation (2) into the equation (3).
However, since the output value of the first type 0th-order Bessel function J ₀ (x) oscillates as the input value x increases, the range of output values in which the input value corresponding to the output value is uniquely determined is limited. Yes. For example, if the time correlation value ρ _t (ΔT) is 0.2, the corresponding value of f _D ΔT is 0.2, 0.62, or 0.79. The time correlation value ρ _t (ΔT) capable of uniquely determining the fading frequency f _D is greater than 0.3, and the value of f _D ΔT corresponding to the time correlation value ρ _t (ΔT) is less than 0.3. . For example, if the time interval ΔT is 10 msec, the fading frequency f _D that can be estimated is less than 30 Hz.

ただし、Sは信号成分s(t)の平均信号電力E_t[|s(t)|²]を表し、Nは雑音成分n(t)の平均雑音電力E_t[|n(t)|²]を表す。このように、時間相関値ρ_t(ΔT)は受信信号の信号対雑音比に応じて変動する。そのため、チャネル推定値に含まれる雑音成分が大きいほど、フェージング周波数f_Dの推定値の誤差も大きくなる。
このような問題を解決するために、互いに時間間隔の異なる二つの時間相関値の比に基づいてフェージング周波数を推定する技術が提案されている（例えば、特許文献１を参照）。例えば、特許文献１に開示されたフェージング周波数推定回路は、時間間隔ΔTに対するチャネル推定値の時間相関値とその2倍の時間間隔に対するチャネル推定値の時間相関値との比を算出し、その比に基づいてフェージング周波数を推定する。 In practice, the channel estimation value r (t) includes a noise component. That is, the channel estimation value r (t) is represented by the sum of the signal component s (t) and the noise component n (t). Therefore, the time correlation value ρ _t (ΔT) is obtained by the following equation by averaging the correlation value of the channel estimation value r (t) over a sufficiently long time that the time variation of the noise component is canceled: expressed.

Where S represents the average signal power E _t [| s (t) | ² ] of the signal component s (t), and N represents the average noise power E _t [| n (t) | ² of the noise component n (t). ]. Thus, the time correlation value ρ _t (ΔT) varies according to the signal-to-noise ratio of the received signal. Therefore, as noise components included in the channel estimation value is large, the error of the estimate of the fading frequency f _D also increases.
In order to solve such a problem, a technique for estimating a fading frequency based on a ratio of two time correlation values having different time intervals has been proposed (see, for example, Patent Document 1). For example, the fading frequency estimation circuit disclosed in Patent Document 1 calculates the ratio between the time correlation value of the channel estimation value with respect to the time interval ΔT and the time correlation value of the channel estimation value with respect to the twice time interval. The fading frequency is estimated based on

JP 2001-358621 A

By calculating the ratio of two time correlation values having different time intervals, the noise component included in the channel estimation value is canceled. Therefore, the fading frequency estimation circuit disclosed in Patent Document 1 can estimate the fading frequency regardless of the signal-to-noise ratio of the channel estimation value.
This fading frequency estimation circuit uses a time correlation value of a time interval that is twice the time interval when the fading frequency is estimated based on the equation (3). The (3) as shown in the formula, the range of estimable fading frequency f _D is determined by the product of the fading frequency f _D and the time interval. Therefore, if the time interval used for calculating the time correlation value is doubled, the range of the fading frequency f _D that can be estimated is halved.
On the other hand, by shortening the transmission interval of known symbols, the time interval for calculating the time correlation value is shortened, so that the range of fading frequency f _{D that} can be estimated is widened. However, as the known symbol transmission interval becomes shorter, the number of subframes or time slots that can be used for data transmission decreases, so the throughput of the entire communication system decreases.

  Accordingly, it is an object of the present specification to provide a fading frequency estimation device, a fading frequency estimation method, and a base station device capable of estimating a fading frequency without reducing the estimable range of the fading frequency and without reducing the throughput. And

  According to one embodiment, there is provided a fading frequency estimation device that estimates a fading frequency of a radio signal that is mounted on a reception device and transmitted from a communication terminal that is movable relative to the reception device. The fading frequency estimation device includes a channel estimation unit that obtains a channel estimation value that represents a transmission characteristic of a radio signal between a communication terminal and a reception device over a plurality of frequency bands, and a second frequency difference that has a predetermined frequency difference among the plurality of frequency bands. A frequency correlation value calculation unit for calculating a frequency correlation value of channel estimation values by obtaining a correlation value between channel estimation values of two frequency bands and a correlation between the channel estimation values of the two frequency bands shifted by a predetermined time interval A time frequency correlation value calculation unit for calculating a time frequency correlation value of the channel estimation value, a time correlation value calculation unit for calculating a ratio of the time frequency correlation value to the frequency correlation value, and a fading frequency corresponding to the ratio A frequency estimation unit that estimates the fading frequency of the signal.

  According to another embodiment, there is provided a fading frequency estimation method for estimating a fading frequency of a radio signal transmitted from a communication terminal that can move relative to a receiving apparatus. In this fading frequency estimation method, a channel estimation value representing a transmission characteristic of a radio signal between a communication terminal and a receiving apparatus is obtained over a plurality of frequency bands, and two frequency bands having a predetermined frequency difference among the plurality of frequency bands are obtained. Calculate the frequency correlation value of the channel estimation value by obtaining the correlation value between the channel estimation values, and shift the channel estimation values of the two frequency bands by a predetermined time interval to correlate them. Calculating a value, calculating a ratio of a temporal frequency correlation value to a frequency correlation value, and estimating a fading frequency corresponding to the ratio as a fading frequency of the radio signal.

  According to still another embodiment, a base station apparatus is provided. The base station apparatus includes a channel estimation unit that obtains a channel estimation value representing a transmission characteristic of a radio signal transmitted from the mobile station apparatus between the mobile station apparatus and the base station apparatus over a plurality of frequency bands, and a plurality of frequency bands. A frequency correlation value calculation unit that calculates a correlation value between channel estimation values of two frequency bands having a predetermined frequency difference among them, and a channel estimation value of the two frequency bands is calculated. A time-frequency correlation value calculating unit that calculates a time-frequency correlation value of the channel estimation value by correlating by shifting by a predetermined time interval; a time-correlation value calculating unit that calculates a ratio of the time-frequency correlation value to the frequency correlation value; A frequency estimator that estimates the fading frequency corresponding to the ratio as the fading frequency of the radio signal, and the fading frequency And a communication control unit which controls the communication between the mobile station apparatus and the base station apparatus Te.

The objects and advantages of the invention will be realized and attained by means of the elements and combinations particularly pointed out in the appended claims.
It should be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are not restrictive of the invention as claimed.

  The fading frequency estimation device, the fading frequency estimation method, and the base station device disclosed in this specification can estimate the fading frequency without reducing the fading frequency estimable range and without reducing the throughput.

It is a schematic block diagram of the fading frequency estimation apparatus by one embodiment. It is a schematic block diagram of a channel estimation part. It is a schematic block diagram of a frequency correlation part. It is a schematic block diagram of a time frequency correlation part. It is a conceptual diagram which shows the relationship between a time frequency correlation value and a frequency correlation value. It is a graph which shows the relationship between the product of a fading frequency and a time interval, and a time correlation value. It is an operation | movement flowchart of a fading frequency estimation process. It is a schematic block diagram of the base station apparatus incorporating the fading frequency estimation apparatus by one Embodiment.

Hereinafter, a fading frequency estimation apparatus according to an embodiment will be described with reference to the drawings.
This fading frequency estimation device is mounted on, for example, a reception device. The receiving apparatus is, for example, a base station apparatus, and receives a radio signal transmitted from a communication terminal such as a mobile station apparatus that can move relative to the receiving apparatus. This fading frequency estimation device obtains a channel estimation value representing a communication state between a communication terminal and a reception device over a plurality of frequency bands based on a known symbol included in a radio signal transmitted from the communication terminal. The fading frequency estimation apparatus obtains a frequency correlation value that is a correlation value between channel estimation values of two frequency bands having different frequencies. In addition, this fading frequency estimation apparatus obtains a time-frequency correlation value that is a correlation value obtained by correlating channel estimation values of two frequency bands having different frequencies with each other by a predetermined time interval. The fading frequency estimation device calculates a ratio of the time frequency correlation value to the frequency correlation value to obtain a time correlation value that is not affected by the noise component included in the channel estimation value, and the fading frequency based on the time correlation value. Is estimated.

  In this embodiment, the signal transmitted from the communication terminal is an uplink modulation scheme stipulated in Long Term Evolution (LTE), that is, a single-carrier frequency-division multiple access (SC-FDMA) scheme. Is modulated according to However, the signal transmitted from the communication terminal may be modulated by other modulation schemes.

  In LTE, Sounding Reference Signal (SRS) is used as a known symbol for measuring uplink channel quality. This SRS is repeatedly transmitted at a constant period in the physical channel. Therefore, in the present embodiment, the fading frequency estimation apparatus obtains a channel estimation value based on the SRS, and estimates the fading frequency of the radio signal transmitted from the communication terminal based on the channel estimation value.

SRSの系列長、すなわち、通信端末から送信される信号が搬送されるサブキャリアの数がMである場合、系列長NをM以下の最大の素数として、（５）式における変数nに(m MOD N)を代入して得られた系列にサイクリックシフト演算を行った系列が求められる。この系列y_M,q(m)は次式で表される。

なお、(m MOD N)は、mをNで除したときの剰余である。またαはサイクリックシフト量である。同じ周波数帯域を用いる通信端末が複数存在する場合、サイクリックシフト量αは、通信端末ごとに異なる値に設定される。そしてサイクリックシフト量αが通信端末ごとに異なることにより、各通信端末から送信された信号間の直交性が保たれる。
SRSは、系列y_M,q(m)を直交周波数分割多重方式にて変調することにより生成される。すなわち、SRSは、次式で表されるように、系列y_M,q(m)を互いに直交する周波数を用いて逆高速フーリエ変換することにより生成される。

ここでZ_M,q(t)はSRSである。またf_mは、m番目のサンプル点に対応するサブキャリアの周波数を表す。より具体的には、周波数f_mは、通信端末ごとに設定される、互いに直交する所定数の連続する周波数の何れかである。
通信端末は、SRSを、所定の時間長を持つ一つのフレームを時間方向に複数に分割したサブフレームの何れかにSRSを割り当てる。そして通信端末は、SRSを含む送信信号が複数の経路を通って受信装置に達しても、受信装置が受信信号からSRSを分離できるように、SRSを含むフレームにCyclic Prefix(CP)を付加する。そして通信端末は、CPが付加されたSRSを含むフレームを無線周波数を持つ搬送波に重畳して送信する。 The SRS is generated, for example, by orthogonal frequency division multiplexing of a signal obtained by cyclic shifting a Zadoff-Chu (ZC) sequence that is one of Constant Amplitude Zero Auto-Correlation (CAZAC) sequences. The details of SRS are as follows.
A ZC sequence x _{N, q} (n) having a sequence length of N and a sequence number q is expressed by the following equation. However, N is an odd number.

When the SRS sequence length, that is, the number of subcarriers carrying a signal transmitted from a communication terminal is M, the sequence length N is set to the maximum prime number equal to or less than M, and the variable n in equation (5) is expressed as (m A sequence obtained by performing a cyclic shift operation on a sequence obtained by substituting MOD N) is obtained. This sequence y _{M, q} (m) is expressed by the following equation.

Note that (m MOD N) is a remainder when m is divided by N. Α is a cyclic shift amount. When there are a plurality of communication terminals using the same frequency band, the cyclic shift amount α is set to a different value for each communication terminal. Since the cyclic shift amount α is different for each communication terminal, the orthogonality between signals transmitted from each communication terminal is maintained.
The SRS is generated by modulating the sequence y _{M, q} (m) by the orthogonal frequency division multiplexing method. That is, the SRS is generated by performing inverse fast Fourier transform on the sequence y _{M, q} (m) using mutually orthogonal frequencies, as represented by the following equation.

Here, Z _{M, q} (t) is SRS. F _m represents the frequency of the subcarrier corresponding to the mth sample point. More specifically, the frequency f _m is set for each communication terminal is one of a predetermined number of consecutive frequencies that are orthogonal to each other.
The communication terminal assigns the SRS to one of the subframes obtained by dividing one frame having a predetermined time length into a plurality of frames in the time direction. Then, the communication terminal adds a cyclic prefix (CP) to the frame including the SRS so that the receiving apparatus can separate the SRS from the received signal even if the transmission signal including the SRS reaches the receiving apparatus through a plurality of paths. . Then, the communication terminal transmits a frame including the SRS to which the CP is added, superimposed on a carrier having a radio frequency.

FIG. 1 is a schematic configuration diagram of a fading frequency estimation apparatus 1 according to one embodiment. The fading frequency estimation device 1 is mounted on a receiving device such as a base station device that communicates with a mobile station device, for example. The fading frequency estimation apparatus 1 includes a channel estimation unit 11, a frequency correlation value calculation unit 12, a time frequency correlation value calculation unit 13, a time correlation value calculation unit 14, and a fading frequency estimation unit 15.
Each of these units included in the fading frequency estimation apparatus 1 is formed as a separate circuit. Alternatively, these units included in the fading frequency estimation device 1 may be mounted on the fading frequency estimation device 1 as one integrated circuit in which circuits corresponding to the respective units are integrated.

ここで、r_8m+k(n)は、サブフレームnのサブキャリア(8m+k)について、SRSの一部に対応する復調された受信信号を表す。またαは、サイクリックシフト量である。N_RB ^SRSは、SRSに割り当てられたリソースブロック(Resource Block、RB)数、すなわち、周波数帯域数を表す。そして1RBあたりのサブキャリア数は6である。N_RB ^SRSは、4の倍数であり、例えば、4、8または20に設定される。
チャネル推定部１１は、サブフレームnの各周波数帯域のチャネル推定値h_m(n)を、周波数相関値算出部１２及び時間周波数相関値算出部１３へ出力する。 The fading frequency estimation device 1 converts the frequency of a radio signal received from a mobile station device into a baseband frequency, and receives a reception signal obtained by analog / digital conversion from the reception device. Then, fading frequency estimation apparatus 1 passes the received signal to channel estimation unit 11.
The channel estimation unit 11 obtains a channel estimation value over a plurality of frequency bands based on the SRS included in the received signal.
FIG. 2 is a schematic configuration diagram of the channel estimation unit 11. The channel estimation unit 11 includes a CP removal unit 111, a Fourier transform unit 112, a subcarrier demapping unit 113, a cyclic shift removal unit 114, and a subcarrier average unit 115.
CP removing section 111 removes CP from a frame including SRS in the received signal. Then, CP removing section 111 outputs a received signal including SRS from which CP has been removed to Fourier transform section 112.
The Fourier transform unit 112 demodulates the received signal by performing a fast Fourier transform on the received signal including the SRS from which the CP has been removed. Then, Fourier transform section 112 outputs the demodulated received signal to subcarrier demapping section 113.
The subcarrier demapping unit 113 selects a received signal in a frequency band to which SRS is allocated from among the demodulated received signals. Then, the subcarrier demapping unit 113 passes the received signal in the selected frequency band to the cyclic shift removing unit 114.
The demodulated received signal is cyclically shifted using a cyclic shift amount α that differs for each communication terminal, as shown in the equation (6). For example, in LTE, signals from mobile station apparatuses that are a maximum of eight communication terminals are multiplexed using this cyclic shift. Therefore, the cyclic shift removal unit 114 and the subcarrier averaging unit 115 remove the cyclic shift that differs between the communication terminals for the eight consecutive subcarriers used by one communication terminal from the demodulated reception signal. A channel estimation value for each communication terminal is calculated by averaging.
For this purpose, the cyclic shift removal unit 114 multiplies the demodulated reception signal corresponding to a part of the SRS by the phase shift amount e- ^jkα corresponding to the corresponding cyclic shift amount α. Then, cyclic shift removal section 114 ^passes the product of the received signal and the phase shift amount e ^−jkα to subcarrier averaging section 115.
The subcarrier averaging unit 115 calculates the channel estimation value by averaging the product of the received signal and the phase shift amount e ^−jkα .
The channel estimation value h _m (n) for the subframe n of the mth frequency band assigned to one communication terminal is expressed by the following equation.

Here, r _{8m + k} (n) represents a demodulated received signal corresponding to a part of the SRS for the subcarrier (8m + k) of subframe n. Α is a cyclic shift amount. N _RB ^SRS represents the number of resource blocks (Resource Blocks, RBs) allocated to the SRS, that is, the number of frequency bands. The number of subcarriers per RB is six. N _RB ^SRS is a multiple of 4, and is set to 4, 8, or 20, for example.
The channel estimation unit 11 outputs the channel estimation value h _m (n) of each frequency band of the subframe n to the frequency correlation value calculation unit 12 and the time frequency correlation value calculation unit 13.

ここで、h^* _m+1(n)は、チャネル推定値h_m+1(n)の複素共役である。またN_cは、チャネル推定値が算出された周波数帯域の総数である。上記のように、1RBあたりのサブキャリア数は6であり、一つの周波数帯域のチャネル推定値は8個のサブキャリアの復調信号の平均値であるため、N_cは6N_RB ^SRS/8である。また上記のように、N_RB ^SRSは4の倍数であるため、N_cは3以上の整数となる。周波数差Δfは、チャネル推定値h_m(n)に対応するm番目の周波数帯域とh_m+1(n)に対応する(m+1)番目の周波数帯域間の周波数の差である。括弧内に示されるように、h_m(n)、h_m+1(n)を用いて表された項は、より一般的なチャネル推定値の関数h(t,f)を用いて表すことができる。関数h(t,f)は、サブフレームnに相当する時刻tにおける周波数fのチャネル推定値を表し、関数h^*(t,f+Δf)は、サブフレームnに相当する時刻tにおける周波数(f+Δf)のチャネル推定値の複素共役を表す。また関数E_f[x]は、変数xの周波数平均を表す。
なお、相関演算回路１２１は、隣接しない二つの周波数帯域のチャネル推定値間の相関演算を行うことにより、瞬時周波数相関値ρ_f(n,Δf)を算出してもよい。例えば、相関演算回路１２１がp個の帯域だけ離れたチャネル推定値間の相関演算を行う場合、（９）式において、h_m(n)h_m+1(n)の代わりにh_m(n)h_m+p(n)が計算される。ただし、SRSについて求められる、周波数帯域ごとのチャネル推定値の総数はN_cであるため、サブフレームnについての瞬時周波数相関値ρ_f(n,Δf)は、(N_c-p)個の相関値h_m(n)h_m+p(n)の平均値となる。また周波数差Δfは、チャネル推定値h_m(n)に対応するm番目の周波数帯域とh_m+p(n)に対応する(m+p)番目の周波数帯域間の周波数の差となる。 The frequency correlation value calculation unit 12 calculates a frequency correlation value ρ _f (Δf) by performing a correlation operation between channel estimation values in two frequency bands having different frequencies.
FIG. 3 is a schematic configuration diagram of the frequency correlation value calculation unit 12. The frequency correlation value calculation unit 12 includes a correlation calculation circuit 121, an averaging circuit 122, and a buffer memory 123.
Each time the correlation calculation circuit 121 receives the channel estimation value h _m (n) of each frequency band of the subframe n from the channel estimation unit 11, the correlation calculation circuit 121 calculates the instantaneous frequency correlation value ρ _f (n, Δf) of the subframe n. Calculate according to the formula.

Here, h ^* _{m + 1} (n) is a complex conjugate of the channel estimation value h _{m + 1} (n). N _c is the total number of frequency bands for which channel estimation values are calculated. As described above, since the number of subcarriers per RB is 6, and the channel estimation value of one frequency band is the average value of the demodulated signals of 8 subcarriers, N _c is 6N _RB ^SRS / 8 . As described above, since N _RB ^SRS is a multiple of 4, N _c is an integer of 3 or more. The frequency difference Δf is a frequency difference between the mth frequency band corresponding to the channel estimation value h _m (n) and the (m + 1) th frequency band corresponding to h _{m + 1} (n). As shown in parentheses, terms expressed using h _m (n), h _{m + 1} (n) should be expressed using the more general channel estimate function h (t, f). Can do. The function h (t, f) represents the channel estimation value of the frequency f at the time t corresponding to the subframe n, and the function h ^* (t, f + Δf) is the frequency at the time t corresponding to the subframe n ( f + Δf) represents the complex conjugate of the channel estimate. The function E _f [x] represents the frequency average of the variable x.
Note that the correlation calculation circuit 121 may calculate the instantaneous frequency correlation value ρ _f (n, Δf) by performing a correlation calculation between channel estimation values of two frequency bands that are not adjacent to each other. For example, when the correlation calculation circuit 121 performs correlation calculation between channel estimation values separated by p bands, in the formula (9), h _m (n) instead of h _m (n) h _{m + 1} (n) ) h _{m + p} (n) is calculated. However, since the total number of channel estimates for each frequency band obtained for SRS is N _c , the instantaneous frequency correlation value ρ _f (n, Δf) for subframe n is (N _c -p) correlations The average value of the values h _m (n) h _{m + p} (n). The frequency difference Δf is a frequency difference between the m-th frequency band corresponding to the channel estimation value h _m (n) and the (m + p) -th frequency band corresponding to h _{m + p} (n).

なお、関数h(t,f)は、サブフレームnに相当する時刻tにおける周波数fのチャネル推定値を表し、関数h^*(t,f+Δf)は、サブフレームnに相当する時刻tにおける周波数(f+Δf)のチャネル推定値の複素共役を表す。また関数E_f[x]は、変数xの周波数平均を表し、関数E_t[x]は、変数xの時間平均を表す。
忘却係数βが大きいほど、周波数相関値ρ_f(Δf)は長期間にわたる瞬時周波数相関値ρ_f(n,Δf)の時間平均値となるので、雑音成分による変動が十分に除去される。一方、忘却係数βが小さいほど、周波数相関値ρ_f(Δf)は短期間の瞬時周波数相関値ρ_f(n,Δf)の時間平均値となるので、周波数相関値ρ_f(Δf)はチャネル推定値の時間変動に追従し易くなる。本実施形態では、一例として、忘却係数βは0.95に設定される。
周波数相関値算出部１２は、周波数相関値ρ_f(Δf)を時間相関値算出部１４へ出力する。また周波数相関値算出部１２は、SRSが含まれる次のサブフレームについての瞬時周波数相関値ρ_f(n+1,Δf)が得られたときに、（１０）式に従って周波数相関値ρ_f(Δf)を更新するために、周波数相関値ρ_f(Δf)を、バッファメモリ１２３に記憶する。 The averaging circuit 122 cancels the temporal fluctuation of the instantaneous frequency correlation value due to the noise component included in the received signal, and the time average of the instantaneous frequency correlation value ρ _f (n, Δf) calculated for each subframe including the SRS. Thus, the frequency correlation value ρ _f (Δf) is calculated. For example, the averaging circuit 122 receives the instantaneous frequency correlation value ρ _f (n, Δf) from the correlation calculation circuit 121. In addition, the averaging circuit 122 reads the frequency correlation value ρ _f (Δf) obtained for the subframe including the SRS immediately before the subframe n from the buffer memory 123. Then, the averaging circuit 122 calculates the frequency correlation value ρ _f (Δf) using the forgetting factor β as in the following equation.

The function h (t, f) represents the channel estimation value of the frequency f at the time t corresponding to the subframe n, and the function h ^* (t, f + Δf) is the time t corresponding to the subframe n. Represents the complex conjugate of the channel estimate at frequency (f + Δf). The function E _f [x] represents the frequency average of the variable x, and the function E _t [x] represents the time average of the variable x.
As the forgetting factor β increases, the frequency correlation value ρ _f (Δf) becomes a time average value of the instantaneous frequency correlation value ρ _f (n, Δf) over a long period of time, and thus fluctuations due to noise components are sufficiently removed. On the other hand, as the forgetting factor β is small, the frequency correlation value ρ _f (Δf) is the time average value of the short term instantaneous frequency correlation value ρ _f (n, Δf), frequency correlation value ρ _f (Δf) is the channel It becomes easy to follow the time fluctuation of the estimated value. In the present embodiment, as an example, the forgetting factor β is set to 0.95.
The frequency correlation value calculation unit 12 outputs the frequency correlation value ρ _f (Δf) to the time correlation value calculation unit 14. Further, when the instantaneous frequency correlation value ρ _f (n + 1, Δf) for the next subframe including the SRS is obtained, the frequency correlation value calculation unit 12 obtains the frequency correlation value ρ _f ( In order to update (Δf), the frequency correlation value ρ _f (Δf) is stored in the buffer memory 123.

The time-frequency correlation value calculation unit 13 calculates a time-frequency correlation value ρ _tf (ΔT, Δf) by performing a correlation operation by shifting the channel estimation values of two frequency bands having different frequencies by a predetermined time interval. To do.
FIG. 4 is a schematic configuration diagram of the time-frequency correlation value calculation unit 13. The time-frequency correlation value calculation unit 13 includes a delay circuit 131, a correlation calculation circuit 132, an averaging circuit 133, and a buffer memory 134.

Delay circuit 131, each time it receives a channel estimation value of each frequency band of the sub-frame n h _m (n) from the channel estimation unit 11, delayed by a period of the respective channel estimate h _m (n) corresponding to the delta _SRS . Δ _SRS is the number of subframes corresponding to an integral multiple of the SRS transmission period. Higher delta _SRS is short, since the range of the estimated possible fading frequency is widened, delta _SRS is preferably short. Therefore, Δ _SRS is set to, for example, the number of subframes corresponding to one or two times the transmission period of SRS.

In order to suppress radio wave interference, the communication terminal may change the transmission frequency of a radio signal including SRS, for example, at a predetermined time period for each subframe. In such a case, Δ _SRS is equal to the number of subframes corresponding to the period in which the transmission frequencies of the subframes including SRS are the same frequency, in particular, the number of subframes corresponding to the shortest period of such periods. It is preferably set.
The delay circuit 131 outputs the delayed channel estimation value h _m (n−Δ _SRS ) to the correlation calculation circuit 132.

ここで、h^* _m+1(n-Δ_SRS)は、チャネル推定値h_m+1(n-Δ_SRS)の複素共役である。またNcは、チャネル推定値が算出された周波数帯域の総数である。また周波数差Δfは、チャネル推定値h_m(n)に対応するm番目の周波数帯域とh_m+1(n)に対応する(m+1)番目の周波数帯域間の周波数の差である。また時間間隔ΔTはΔ_SRSに相当する期間であり、一つのサブフレームの時間長がt_subであれば、ΔT=Δ_SRSt_subである。また、括弧内に示されるように、h_m(n)、h_m+1(n)を用いて表された項は、より一般的なチャネル推定値の関数h(t,f)を用いて表すことができる。関数h(t,f)は、サブフレームnに相当する時刻tにおける周波数fのチャネル推定値を表す。関数h^*(t-ΔT,f+Δf)は、サブフレーム(n-Δ_SRS)に相当する時刻(t-ΔT)における周波数(f+Δf)のチャネル推定値の複素共役を表す。そして関数E_f[x]は、変数xの周波数平均を表す。
なお、相関演算回路１３２は、周波数相関値算出部１２の相関演算回路１２１と同様に、隣接しない二つの周波数帯域のチャネル推定値間の相関演算を行うことにより、瞬時時間周波数相関値ρ_tf(n,ΔT,Δf)を算出してもよい。ただし、瞬時時間周波数相関値ρ_tf(n,ΔT,Δf)の算出のために設定される周波数差Δfは、瞬時周波数相関値ρ_f(n,Δf)の算出のために設定される周波数差Δfと同一となるように設定される。
また、相関演算回路１３２は、予め設定された複数の時間間隔のうち、何れかの時間間隔を用いて瞬時時間周波数相関値ρ_tf(n,ΔT,Δf)を算出してもよい。 The correlation calculation circuit 132 is based on the channel estimation value h _m (n) of each frequency band of the subframe n received from the channel estimation unit 11 and the channel estimation value h _m (n−Δ _SRS ) received from the delay circuit. The instantaneous time-frequency correlation value ρ _tf (n, ΔT, Δf) of the subframe n is calculated according to the following equation.

Here, h ^* _{m + 1} (n− _ΔSRS ) is a complex conjugate of the channel estimation value h _{m + 1} (n− _ΔSRS ). Nc is the total number of frequency bands for which channel estimation values are calculated. The frequency difference Δf is a frequency difference between the mth frequency band corresponding to the channel estimation value h _m (n) and the (m + 1) th frequency band corresponding to h _{m + 1} (n). The time interval [Delta] T is a period corresponding to delta _SRS, if the time length t _sub of one subframe, a [Delta] T = delta _SRS t _sub. Also, as shown in parentheses, the terms expressed using h _m (n), h _{m + 1} (n) are expressed using the more general channel estimate function h (t, f). Can be represented. The function h (t, f) represents the channel estimation value of the frequency f at the time t corresponding to the subframe n. Function ^{h * (t-ΔT, f} + Δf) represents the complex conjugate of the channel estimation value of a frequency (f + Δf) at a time corresponding to the sub-frame _{(n-Δ SRS) (t} -ΔT). The function E _f [x] represents the frequency average of the variable x.
Similar to the correlation calculation circuit 121 of the frequency correlation value calculation unit 12, the correlation calculation circuit 132 performs a correlation calculation between channel estimation values of two frequency bands that are not adjacent to each other, whereby an instantaneous time frequency correlation value ρ _tf ( n, ΔT, Δf) may be calculated. However, the frequency difference Δf set for calculating the instantaneous time frequency correlation value ρ _tf (n, ΔT, Δf) is the frequency difference set for calculating the instantaneous frequency correlation value ρ _f (n, Δf). It is set to be the same as Δf.
The correlation calculation circuit 132 may calculate the instantaneous time-frequency correlation value ρ _tf (n, ΔT, Δf) using any one of a plurality of preset time intervals.

本実施形態では、一例として、忘却係数βは0.95に設定される。なお、関数h(t,f)は、サブフレームnに相当する時刻tにおける周波数fのチャネル推定値を表し、関数h^*(t-ΔT,f+Δf)は、サブフレーム(n-Δ_SRS)に相当する時刻(t-ΔT)における周波数(f+Δf)のチャネル推定値の複素共役を表す。また関数E_f[x]は変数xの周波数平均を表し、関数E_t[x]は変数xの時間平均を表す。 The averaging circuit 133 time-averages the instantaneous time-frequency correlation value ρ _tf (n, ΔT, Δf) calculated for each subframe including the SRS in order to cancel the deviation of the instantaneous time-frequency correlation value due to the noise component. Thus, the time-frequency correlation value ρ _tf (ΔT, Δf) is calculated. For example, the averaging circuit 133 receives the instantaneous time-frequency correlation value ρ _tf (n, ΔT, Δf) from the correlation calculation circuit 132. In addition, the averaging circuit 133 reads the time-frequency correlation value ρ _tf (ΔT, Δf) obtained for the subframe including the SRS immediately before the subframe n from the buffer memory 134. Then, the averaging circuit 133 calculates the time-frequency correlation value ρ _tf (ΔT, Δf) using the forgetting factor β as in the following equation.

In the present embodiment, as an example, the forgetting factor β is set to 0.95. Incidentally, the function h (t, f) represents a channel estimation value of the frequency f at time t corresponding to the sub-frame n, the function ^{h * (t-ΔT, f} + Δf) , the sub-frame (n-delta _SRS ) Represents the complex conjugate of the channel estimation value of frequency (f + Δf) at time (t−ΔT) corresponding to. The function E _f [x] represents the frequency average of the variable x, and the function E _t [x] represents the time average of the variable x.

The time frequency correlation value calculation unit 13 outputs the time frequency correlation value ρ _tf (ΔT, Δf) to the time correlation value calculation unit 14. Further, the frequency correlation value calculation unit 12 stores the time-frequency correlation value ρ _tf (ΔT, Δf) in the buffer memory 134. As described above, the time-frequency correlation value ρ _tf (ΔT, Δf) stored in the buffer memory 134 is obtained when the instantaneous time-frequency correlation value for the next subframe including the SRS is obtained. Used to update the value ρ _tf (ΔT, Δf).

FIG. 5 is a conceptual diagram showing the relationship between the instantaneous time frequency correlation value ρ _tf (n, ΔT, Δf) and the instantaneous frequency correlation value ρ _f (n, Δf). In FIG. 5, the horizontal axis represents frequency. The upper arrows 501a to 501f represent channel estimation values of the respective frequency bands calculated based on the SRS included in the nth subframe. The lower arrows 502a to 502f represent channel estimation values of the respective frequency bands calculated based on the SRS included in the (n−Δ _SRS ) th subframe. Furthermore, the bidirectional arrows 511a to 511e represent correlation operations calculated to obtain the instantaneous frequency correlation value ρ _f (n, Δf), and the bidirectional arrows 521a to 521e represent the instantaneous time frequency correlation value ρ _tf (n , ΔT, Δf) represents the correlation operation calculated.
The instantaneous frequency correlation value ρ _f (n, Δf) is, as represented by the double arrows 511a to 511e, out of the channel estimation values 501a to 501f of each frequency band in the same subframe. Calculated by averaging the correlation values between channel estimates. On the other hand, with respect to the instantaneous time-frequency correlation value ρ _tf (n, ΔT, Δf), as represented by the bidirectional arrow 521a, the channel band estimated value 501a of the frequency band in the subframe n and the subframe (n− _ΔSRS) The correlation value with the channel estimation value 502b of the adjacent frequency band in () is calculated. Similarly, correlation values between channel estimation values 501b to 501e of each frequency band in subframe n and channel estimation values 502c to 502f of adjacent frequency bands in subframe (n−Δ _SRS ) are respectively calculated. . The instantaneous time-frequency correlation value ρ _tf (n, ΔT, Δf) is calculated by averaging the correlation values.

The time correlation value calculation unit 14 calculates a time correlation value ρ _t (ΔT) of the channel estimation value based on the time frequency correlation value ρ _tf (ΔT, Δf) and the frequency correlation value ρ _f (Δf).
Here, the time-frequency correlation value ρ _tf (ΔT, Δf) can be expressed by the product of the time correlation value ρ _t (ΔT) and the frequency correlation value ρ _f (Δf). The reason is as follows.

したがって、周波数間隔Δfだけずらした周波数応答H(t,f)同士の相関値である周波数相関値R_fは以下のように表される。

ただし、関数<>は、周波数f及び時間tの両方についての平均を表す。ここで、各パスは独立であるため、異なるタイミングで受信装置に到達する信号同士の相関は無いとすると、<h^*(t,τ)h(t,τ')>は以下のようになる。

したがって、（１４）式及び（１５）式に基づき、周波数相関値R_fは以下のように表される。

ここでp(τ)は遅延プロファイルを表し、p(τ)=<h^*(t,τ)h(t,τ)>である。 A general multipath is assumed as a path through which a radio signal transmitted from a communication terminal and received by a receiving device passes. Each path is independent from each other, and fading due to the Doppler transition of this radio signal follows Rayleigh fading. The impulse response of the channel at time t is represented by h (t, τ). Note that τ represents an elapsed time from when the radio signal is transmitted until it is received.
In this case, the frequency response H (t, f) at time t is obtained by performing Fourier transform on the impulse response h (t, τ) of the channel with respect to τ = 0 and thereafter. This frequency response H (t, f) is expressed by the following equation.

Therefore, a frequency correlation value R _f that is a correlation value between frequency responses H (t, f) shifted by the frequency interval Δf is expressed as follows.

Where function <> represents the average for both frequency f and time t. Here, since each path is independent, if there is no correlation between signals reaching the receiving device at different timings, <h ^* (t, τ) h (t, τ ′)> is as follows: .

Therefore, based on the equations (14) and (15), the frequency correlation value R _f is expressed as follows.

Here, p (τ) represents a delay profile, and p (τ) = <h ^* (t, τ) h (t, τ)>.

ここで、どのパスを通った無線信号についても、通信端末と受信装置間の相対的な移動速度は同一であるため、フェージング周波数は、無線信号が通るパスに依存せず、全てのパスについて同一であると仮定できる。そのため、各パスについての周波数応答の時間相関値も同一であると仮定できる。したがって、<h^*(t,τ)h(t+Δt,τ')>は以下のようになる。

ただし、ρ_t(Δt)は、時間間隔Δtだけずらした周波数応答H(t,f)同士の相関値である時間相関値を表す。（１７）式及び（１８）式に基づき、時間周波数相関値R_tfは以下のように表される。

したがって、（１６）式及び（１９）式に基づいて、時間周波数相関値R_tfは次式で表される。

このように、時間周波数相関値R_tfは時間相関値ρ_t(Δt)と周波数相関値R_fの積となる。 On the other hand, a time-frequency correlation value R _tf that is a correlation value between frequency responses H (t, f) shifted by a frequency interval Δf and a time interval Δt is expressed as follows.

Here, since the relative moving speed between the communication terminal and the receiving device is the same for any radio signal passing through any path, the fading frequency does not depend on the path through which the radio signal passes, and is the same for all paths. Can be assumed. Therefore, it can be assumed that the time correlation value of the frequency response for each path is the same. Therefore, <h ^* (t, τ) h (t + Δt, τ ′)> is as follows.

However, ρ _t (Δt) represents a time correlation value that is a correlation value between the frequency responses H (t, f) shifted by the time interval Δt. Based on the equations (17) and (18), the time-frequency correlation value R _tf is expressed as follows.

Therefore, based on the equations (16) and (19), the time-frequency correlation value R _tf is expressed by the following equation.

As described above, the time-frequency correlation value R _tf is a product of the time correlation value ρ _t (Δt) and the frequency correlation value R _f .

あるいは、時間相関値算出部１４は、次式に従って、周波数相関値ρ_f(Δf)の絶対値の２乗に対する時間周波数相関値ρ_tf(ΔT,Δf)の絶対値の２乗の比を時間相関値ρ_t(ΔT)の絶対値の２乗として算出してもよい。

このように、チャネル推定値の時間相関値は、チャネル推定値の周波数相関値に対するチャネル推定値の時間周波数相関値の比として算出される。また、受信信号に含まれる雑音成分は、時間軸及び周波数軸に関して相関性を持たないので、異なる時間及び異なる周波数帯域におけるチャネル推定値同士の相関値の時間平均である周波数相関値及び時間周波数相関値における雑音成分に関する値は平均する相関値の数を増やすにつれて0に近づく。したがって、平均する相関値の数を十分に多く設定すれば、算出された時間相関値は、受信信号に含まれる雑音成分に影響されない。
時間相関値算出部１４は、時間相関値ρ_t(ΔT)またはその実数部をフェージング周波数推定部１５へ出力する。 Therefore, for example, time correlation value calculating unit 14 according to the following equation, the time-frequency correlation value of the real component of the frequency correlation value _{ρ f (Δf) ρ tf (} ΔT, Δf) time correlation value ratio of the real component of the [rho _t Calculated as the real component of (ΔT).

Alternatively, the time correlation value calculation unit 14 calculates the ratio of the square of the absolute value of the time frequency correlation value ρ _tf (ΔT, Δf) to the square of the absolute value of the frequency correlation value ρ _f (Δf) according to the following equation: The correlation value ρ _t (ΔT) may be calculated as the square of the absolute value.

Thus, the time correlation value of the channel estimation value is calculated as the ratio of the time frequency correlation value of the channel estimation value to the frequency correlation value of the channel estimation value. In addition, since the noise component included in the received signal has no correlation with respect to the time axis and the frequency axis, the frequency correlation value and the time frequency correlation, which are time averages of correlation values between channel estimation values in different times and different frequency bands, The value for the noise component in the value approaches zero as the number of averaged correlation values increases. Therefore, if a sufficiently large number of correlation values to be averaged is set, the calculated time correlation value is not affected by the noise component included in the received signal.
The time correlation value calculation unit 14 outputs the time correlation value ρ _t (ΔT) or a real part thereof to the fading frequency estimation unit 15.

The fading frequency estimation unit 15 estimates the fading frequency based on the time correlation value ρ _t (ΔT) or the value of its real number component. As described above, the relationship between the fading frequency f _D and the time interval [Delta] T product and time correlation value [rho _t ([Delta] T) is represented by the equation (3).
Figure 6 is a graph showing the relationship between the fading frequency f _D and the time interval [Delta] T of the product and time correlation value ρ _t (ΔT). In FIG. 6, the horizontal axis represents the product f _D ΔT of the fading frequency f _D and the time interval ΔT, and the vertical axis represents the time correlation value ρ _t (ΔT) represented by the first type 0th-order Bessel function. A graph 601 represents the relationship between f _D ΔT and the time correlation value ρ _t (ΔT). As shown in the graph 601, the value of f _D ΔT is uniquely determined only when the time correlation value ρ _t (ΔT) is greater than 0.3. The time interval ΔT is a preset value. Therefore, when the time correlation value ρ _t (ΔT) received from the time correlation value calculation unit 14 is larger than 0.3, the fading frequency f _D is uniquely obtained.

上記の参照テーブルにおいて、左端の列は、時間相関値の各サンプリング点に対するインデックスn(n=0,1,2,...,7)を表す。また中央の列は、インデックスnに対応する時間相関値のサンプリング点R_n(n=0,1,2,...,7)の値を表す。さらに右端の列は、時間相関値のサンプリング点R_nに対応するフェージング周波数f_nを表す。この参照テーブルを参照すると、例えば、インデックスn=1に対応する時間相関値のサンプリング点R₁の値は0.9であり、そのサンプリング点R₁に対応するフェージング周波数は10.2Hzであることが分かる。
上記のように、各サンプリング点に対するフェージング周波数は、時間間隔ΔT、すなわち、Δ_SRSに応じて異なる。そこで、時間相関値を算出するために複数のΔ_SRSが用いられる可能性がある場合、フェージング周波数推定部１５は、Δ_SRSごとに別個に作成された参照テーブルを記憶しておくことが好ましい。そしてフェージング周波数推定部１５は、選択されたΔ_SRSに応じた参照テーブルを参照して、フェージング周波数を推定する。 Therefore, a reference table representing the relationship between the fading frequency f _D and the time correlation value ρ _t (ΔT) or its real number component is stored in advance in a nonvolatile semiconductor memory included in the fading frequency estimation unit 15.
Table 1 is an example of such a reference table when the time interval ΔT is 10 msec.

In the above lookup table, the leftmost column represents the index n (n = 0, 1, 2,..., 7) for each sampling point of the time correlation value. The center column represents the value of the sampling point R _n (n = 0, 1, 2,..., 7) of the time correlation value corresponding to the index n. Further, the rightmost column represents the fading frequency f _n corresponding to the sampling point R _n of the time correlation value. With reference to this reference table, for example, it can be seen that the value of the sampling point R ₁ of the time correlation value corresponding to the index n = 1 is 0.9, and the fading frequency corresponding to the sampling point R ₁ is 10.2 Hz.
As described above, the fading frequency for each sampling point, the time interval [Delta] T, that is, varies depending on the delta _SRS. Therefore, when there is a possibility that a plurality of _ΔSRSs are used to calculate the time correlation value, it is preferable that the fading frequency estimation unit 15 stores a reference table created separately for each _ΔSRS . The fading frequency estimating unit 15 refers to the reference table in accordance with the delta _SRS chosen to estimate the fading frequency.

サンプリング点R_nは、参照テーブルに示された時間相関値ρ_t(ΔT)よりも大きいサンプリング点のうちの最小値である。またnは、その最小値に対応するインデックスであり、f_nは、サンプリング点R_nに対応するフェージング周波数である。例えば、表１を参照すると、時間相関値ρ_t(ΔT)が0.75であれば、サンプリング点R_n、R_n+1はそれぞれ0.8、0.7であり、フェージング周波数f_n、f_n+1はそれぞれ14.6(Hz)、18.2(Hz)である。したがって、フェージング周波数の推定値f_DESTは16.4(Hz)となる。 The fading frequency estimation unit 15 refers to the reference table and identifies the two sampling points R _n and R _{n + 1} that are closest to the time correlation value ρ _t (ΔT). Then, the fading frequency estimation unit 15 performs linear interpolation using the two sampling points R _n and R _{n + 1} and the fading frequencies f _n and f _{n + 1} corresponding to the sampling points R _n and R _{n + 1,} thereby obtaining a time correlation value ρ _t. An estimated value f _DEST of the fading frequency corresponding to (ΔT) is obtained. This estimated value f _DEST is calculated by the following equation.

The sampling point R _n is the minimum value among sampling points larger than the time correlation value ρ _t (ΔT) shown in the reference table. N is an index corresponding to the minimum value, and f _n is a fading frequency corresponding to the sampling point R _n . For example, referring to Table 1, if the time correlation value ρ _t (ΔT) is 0.75, the sampling points R _n and R _{n + 1} are 0.8 and 0.7, respectively, and the fading frequencies f _n and f _{n + 1} are respectively 14.6 (Hz) and 18.2 (Hz). Therefore, the estimated value f _DEST of the fading frequency is 16.4 (Hz).

Note that the fading frequency estimation unit 15 uses the two sampling points R _n and R _{n + 1} closest to the time correlation value ρ _t (ΔT) and the fading frequencies f _n and f _{n + 1} corresponding to the sampling points. A three-dimensional spline function may be obtained. The fading frequency estimation unit 15 may obtain an estimated value f _DEST of the fading frequency corresponding to the time correlation value ρ _t (ΔT) based on the three-dimensional spline function. In this case, in order to reduce the amount of calculation for calculating the three-dimensional spline function, it is preferable that the reference table also stores the primary differential value and the secondary differential value of the fading frequency at each sampling point.
Or the fading frequency estimation part 15 may obtain | require the estimated value _fDEST of a fading frequency by performing a numerical analysis based on (3) Formula.
The fading frequency estimation unit 15 outputs the fading frequency estimation value f _DEST to another circuit that uses the estimation value.

FIG. 7 is an operation flowchart of fading frequency estimation processing executed by the fading frequency estimation apparatus 1.
The channel estimation unit 11 calculates channel estimation values over a plurality of frequency bands based on known symbols included in a signal received from a movable communication terminal (step S101). Then, the channel estimation unit 11 outputs the channel estimation value of each frequency band to the frequency correlation value calculation unit 12 and the time frequency correlation value calculation unit 13.

The frequency correlation value calculation unit 12 calculates an instantaneous frequency correlation value ρ _f (n, Δf) by performing a correlation operation between channel estimation values of two frequency bands having different frequencies (step S102). The frequency correlation value calculation unit 12, the instantaneous frequency correlation value ρ _f (n, Δf) to calculate a frequency correlation value [rho _f (Delta] f) by averaging time (step S103). The frequency correlation value calculation unit 12 outputs the frequency correlation value ρ _f (Δf) to the time correlation value calculation unit 14.

On the other hand, the time-frequency correlation value calculation unit 13 performs the correlation operation by shifting the channel estimation values of two frequency bands having different frequencies by a predetermined time interval ΔT, so that the instantaneous time-frequency correlation value ρ _tf (n, ΔT , Δf) is calculated (step S104). The time-frequency correlation value calculator 13, the instantaneous time-frequency correlation value _{ρ tf (n, ΔT, Δf} ) time-frequency correlation value by averaging the time ρ _tf (ΔT, Δf) is calculated (step S105). The time frequency correlation value calculation unit 13 outputs the time frequency correlation value ρ _tf (ΔT, Δf) to the time correlation value calculation unit 14.

Time correlation value calculating unit 14, time-frequency correlation value of the frequency correlation value _{ρ f (Δf) ρ tf (} ΔT, Δf) the ratio of _{(ρ tf (ΔT, Δf)} / ρ f (Δf)) by calculating the Then, a time correlation value ρ _t (ΔT) is obtained (step S106). Then, the time correlation value calculation unit 14 outputs the time correlation value ρ _t (ΔT) to the fading frequency estimation unit 15.
Fading frequency estimation unit 15, by specifying the fading frequency f _D, corresponding to the time correlation value ρ _t (ΔT) based on the reference table to estimate the f _D (step S107).
And the fading frequency estimation apparatus 1 complete | finishes a fading frequency estimation process.

  As described above, this fading frequency estimation apparatus calculates the time correlation value of the channel estimation value as the ratio of the time frequency correlation value of the channel estimation value to the frequency correlation value of the channel estimation value. Therefore, since this fading frequency estimation apparatus can reduce the noise component contained in the channel estimation value, it can accurately estimate the fading frequency. Furthermore, since this fading frequency estimation apparatus can estimate the fading frequency using only one time interval, the fading frequency can be estimated without reducing the fading frequency estimable range and without reducing the throughput. Further, the specific function representing the frequency correlation value depends on the multipath model representing the path between the communication terminal and the receiving apparatus, but the frequency correlation value itself is canceled when the time correlation value is calculated. Therefore, this fading frequency estimation device can estimate the fading frequency without considering a specific function of the frequency correlation value. Therefore, this fading frequency estimation device can estimate the fading frequency regardless of the communication environment that affects the path between the communication terminal and the reception device.

In addition, this invention is not limited to said embodiment. For example, instead of SRS, the channel estimation unit obtains a channel estimation value based on other symbols that are included in a radio signal transmitted from a communication terminal and that have components over a plurality of frequency bands and are known to the receiving device. May be. For example, the channel estimator uses a reference signal included in a control channel such as a reference signal for data demodulation or a physical uplink control channel (Physical Uplink Control Channel, PUCCH) specified by LTE. May be calculated.
In addition, the receiving device on which the fading frequency device is mounted may be not only a base station device that supports LTE, but also a communication device that supports other communication standards, for example, Worldwide Interoperability for Microwave Access (WiMAX). Alternatively, the receiving device on which the fading frequency device is mounted may be a vehicle-mounted device or a roadside device compatible with Dedicated Short Range Communication (DSRC) used in an intelligent road traffic system.

FIG. 8 is a schematic configuration diagram of a base station apparatus in which the fading frequency estimation apparatus according to the above-described embodiment is incorporated. The base station apparatus 100 includes an interface unit 101, a baseband processing unit 102, a transmission unit 103, a transmission amplifier 104, a duplexer 105, an antenna 106, a reception amplifier 107, and a reception unit 108.
In addition, the baseband processing unit 102, the transmission unit 103, and the reception unit 108 may be separate circuits, or each of these units may be a single integrated circuit in which these circuits are integrated. .

  The interface unit 101 has a communication interface for connecting the base station apparatus 100 to the core network. The interface unit 101 receives a downlink signal transmitted to the mobile station apparatus 200 from the core network, and outputs the downlink signal to the baseband processing unit 102. On the other hand, the interface unit 101 receives the uplink signal received from the mobile station apparatus 200 from the baseband processing unit 102, and outputs the uplink signal to the core network.

The baseband processing unit 102 performs transmission processing such as error correction coding processing such as convolution coding or turbo coding on the downlink signal. Further, the baseband processing unit 102 performs orthogonal modulation processing such as orthogonal frequency division multiplexing (OFDMA) on the encoded downlink signal, and multiplexes the downlink signal. The baseband processing unit 102 outputs the orthogonally modulated downlink signal to the transmission unit 103.
The baseband processing unit 102 demodulates the uplink signal having the intermediate frequency received from the receiving unit 108. Then, the baseband processing unit 102 performs reception processing such as error correction decoding processing on the demodulated uplink signal. Then, the baseband processing unit 102 outputs the decoded uplink signal to the interface unit 101.

Further, the baseband processing unit 102 includes the fading frequency estimation device 1021 according to the above embodiment and a communication control unit 1022. Then, fading frequency estimation apparatus 1021 estimates the fading frequency for the uplink signal received from mobile station apparatus 200.
The communication control unit 1022 controls communication between the mobile station device 200 and the base station device 100 based on the estimated fading frequency. For example, when the fading frequency is equal to or lower than a predetermined value, the communication control unit 1022 can be expected to improve channel estimation accuracy by averaging channel estimation values for a long time. On the other hand, when the fading frequency is equal to or higher than a predetermined value, the communication control unit 1022 can improve the follow-up property to the fluctuation by setting the channel estimation value to an instantaneous value (not averaged in the time direction).
The communication control unit 1022 generates a control signal including control information such as a modulation scheme and coding efficiency, and includes the control signal in the downlink signal.

The transmission unit 103 performs digital / analog conversion on the orthogonally-converted downlink signal, and then superimposes the downlink signal on a carrier wave having a radio frequency. Transmitting section 103 then outputs the downlink signal superimposed on the carrier wave to transmitting amplifier 104.
The transmission amplifier 104 has a high power amplifier. Transmitting amplifier 104 amplifies the strength of the downlink signal superimposed on the carrier wave to a desired level, and transmits the signal to antenna 106 via duplexer 105. The antenna 106 radiates the downlink signal transmitted from the transmission amplifier 104.

The antenna 106 receives an uplink signal transmitted from the mobile station apparatus 200 and transmits the uplink signal to the reception amplifier 107 via the duplexer 105.
The receiving amplifier 107 has a low noise amplifier. Then, the reception amplifier 107 amplifies the received uplink signal and outputs the amplified uplink signal to the reception unit 108.

  The receiving unit 108 converts the frequency of the uplink signal from a radio frequency to an intermediate frequency by superimposing a periodic signal having a local transmission frequency on the uplink signal. Then, the receiving unit 108 performs analog / digital conversion on the uplink signal having the intermediate frequency, and then passes it to the baseband processing unit 102.

  All examples and specific terms listed herein are intended for instructional purposes to help the reader understand the concepts contributed by the inventor to the present invention and the promotion of the technology. It should be construed that it is not limited to the construction of any example herein, such specific examples and conditions, with respect to showing the superiority and inferiority of the present invention. Although embodiments of the present invention have been described in detail, it should be understood that various changes, substitutions and modifications can be made thereto without departing from the spirit and scope of the present invention.

The following supplementary notes are further disclosed regarding the embodiment described above and its modifications.
(Appendix 1)
A fading frequency estimation device for estimating a fading frequency of a radio signal transmitted from a communication terminal mounted on a reception device and movable relative to the reception device,
A channel estimation unit for obtaining a channel estimation value representing a transmission characteristic of the radio signal between the communication terminal and the receiving device over a plurality of frequency bands;
A frequency correlation value calculation unit that calculates a frequency correlation value of the channel estimation value by obtaining a correlation value between channel estimation values of two frequency bands having a predetermined frequency difference among the plurality of frequency bands;
A time-frequency correlation value calculating unit that calculates a time-frequency correlation value of the channel estimation value by correlating the channel estimation values of the two frequency bands by a predetermined time interval; and
A time correlation value calculating unit for calculating a ratio of the time frequency correlation value to the frequency correlation value;
A frequency estimation unit that estimates a fading frequency corresponding to the ratio as a fading frequency of the radio signal;
A fading frequency estimation device comprising:
(Appendix 2)
The channel estimation unit obtains the channel estimation value based on symbols known by the reception device included in the radio signal,
The fading frequency estimation apparatus according to attachment 1, wherein the predetermined time interval is a transmission period of the known symbol.
(Appendix 3)
The channel estimation unit obtains the channel estimation value based on symbols known by the reception device included in the radio signal,
The fading frequency estimation apparatus according to appendix 1, wherein when the frequency of the radio signal fluctuates at a predetermined period, the predetermined time interval is a minimum period at which the known symbol has the same frequency.
(Appendix 4)
The time-frequency correlation value calculating unit calculates the time-frequency correlation value using any one of a plurality of preset time intervals as the predetermined time interval,
The frequency estimation unit stores a plurality of reference tables that are set for each of the plurality of time intervals and represents a relationship between the ratio and a fading frequency, and the predetermined time intervals of the plurality of reference tables are stored. The fading frequency estimation apparatus according to any one of appendices 1 to 3, wherein a fading frequency of the radio signal is estimated using a reference table corresponding to.
(Appendix 5)
A fading frequency estimation method for estimating a fading frequency of a radio signal transmitted from a communication terminal movable relative to a receiving device,
Obtaining a channel estimation value representing transmission characteristics of the radio signal between the communication terminal and the receiving device over a plurality of frequency bands;
Calculating a frequency correlation value of the channel estimation value by obtaining a correlation value between channel estimation values of two frequency bands having a predetermined frequency difference among the plurality of frequency bands;
Calculating a time-frequency correlation value of the channel estimation value by correlating the channel estimation values of the two frequency bands by shifting by a predetermined time interval;
Calculating a ratio of the time-frequency correlation value to the frequency correlation value;
Estimating a fading frequency corresponding to the ratio as a fading frequency of the radio signal;
A fading frequency estimation method.
(Appendix 6)
A base station device,
A channel estimation unit for obtaining a channel estimation value representing a transmission characteristic of a radio signal transmitted from the mobile station device over a plurality of frequency bands between the mobile station device and the base station device;
A frequency correlation value calculation unit that calculates a frequency correlation value of the channel estimation value by obtaining a correlation value between channel estimation values of two frequency bands having a predetermined frequency difference among the plurality of frequency bands;
A time-frequency correlation value calculating unit that calculates a time-frequency correlation value of the channel estimation value by correlating the channel estimation values of the two frequency bands by a predetermined time interval; and
A time correlation value calculating unit for calculating a ratio of the time frequency correlation value to the frequency correlation value;
A frequency estimation unit that estimates a fading frequency corresponding to the ratio as a fading frequency of the radio signal;
A communication control unit that controls communication between the mobile station apparatus and the base station apparatus according to the fading frequency;
A base station apparatus.

DESCRIPTION OF SYMBOLS 1 Fading frequency estimation apparatus 11 Channel estimation part 12 Frequency correlation value calculation part 13 Time frequency correlation value calculation part 14 Time correlation value calculation part 15 Fading frequency estimation part 100 Base station apparatus 101 Interface part 102 Baseband process part 1021 Fading frequency estimation apparatus 1022 Communication Control Unit 103 Transmitting Unit 104 Transmitting Amplifier 105 Duplexer 106 Antenna 107 Receiving Amplifier 108 Receiving Unit 200 Mobile Station Device

Claims (4)

A fading frequency estimation device for estimating a fading frequency of a radio signal transmitted from a communication terminal mounted on a reception device and movable relative to the reception device,
A channel estimation unit for obtaining a channel estimation value representing a transmission characteristic of the radio signal between the communication terminal and the receiving device over a plurality of frequency bands;
A frequency correlation value calculation unit that calculates a frequency correlation value of the channel estimation value by obtaining a correlation value between channel estimation values of two frequency bands having a predetermined frequency difference among the plurality of frequency bands;
A time-frequency correlation value calculating unit that calculates a time-frequency correlation value of the channel estimation value by correlating the channel estimation values of the two frequency bands by a predetermined time interval; and
A time correlation value calculating unit for calculating a ratio of the time frequency correlation value to the frequency correlation value;
A frequency estimation unit that estimates a fading frequency corresponding to the ratio as a fading frequency of the radio signal;
A fading frequency estimation device comprising: The channel estimation unit obtains the channel estimation value based on symbols known by the reception device included in the radio signal,
2. The fading frequency estimation apparatus according to claim 1, wherein when the frequency of the radio signal fluctuates at a predetermined period, the predetermined time interval is a minimum period at which the known symbol has the same frequency. A fading frequency estimation method for estimating a fading frequency of a radio signal transmitted from a communication terminal movable relative to a receiving device,
Obtaining a channel estimation value representing transmission characteristics of the radio signal between the communication terminal and the receiving device over a plurality of frequency bands;
Calculating a frequency correlation value of the channel estimation value by obtaining a correlation value between channel estimation values of two frequency bands having a predetermined frequency difference among the plurality of frequency bands;
Calculating a time-frequency correlation value of the channel estimation value by correlating the channel estimation values of the two frequency bands by shifting by a predetermined time interval;
Calculating a ratio of the time-frequency correlation value to the frequency correlation value;
Estimating a fading frequency corresponding to the ratio as a fading frequency of the radio signal;
A fading frequency estimation method. A base station device,
A channel estimation unit for obtaining a channel estimation value representing a transmission characteristic of a radio signal transmitted from the mobile station device over a plurality of frequency bands between the mobile station device and the base station device;
A frequency correlation value calculation unit that calculates a frequency correlation value of the channel estimation value by obtaining a correlation value between channel estimation values of two frequency bands having a predetermined frequency difference among the plurality of frequency bands;
A time-frequency correlation value calculating unit that calculates a time-frequency correlation value of the channel estimation value by correlating the channel estimation values of the two frequency bands by a predetermined time interval; and
A time correlation value calculating unit for calculating a ratio of the time frequency correlation value to the frequency correlation value;
A frequency estimation unit that estimates a fading frequency corresponding to the ratio as a fading frequency of the radio signal;
A communication control unit that controls communication between the mobile station apparatus and the base station apparatus according to the fading frequency;
A base station apparatus.

Images