WO2011070614A1

WO2011070614A1 - Wireless apparatus and sir measurement method

Info

Publication number: WO2011070614A1
Application number: PCT/JP2009/006710
Authority: WO
Inventors: 末満大成
Original assignee: 三菱電機株式会社
Priority date: 2009-12-09
Filing date: 2009-12-09
Publication date: 2011-06-16
Also published as: JP4525867B1; JPWO2011070614A1

Abstract

A wireless apparatus and SIR measurement method which can accurately measure the signal-to-interference ratio, that is, SIR. An AGC control unit (40) controls the gain of a variable amplifier (60) to automatically control the gain of a received signal. A signal power/interference power calculator (7) calculates signal power and interference power based on a known predetermined received signal which had been subjected to automatic gain control. An SIR calculator (8) calculates an SIR based on the signal power and interference power that the signal power/interference power calculation unit (7) calculated, and outputs the SIR to the scheduler (10).

Description

Wireless device and SIR measurement method

The present invention relates to a radio communication technique, and more particularly, to a radio apparatus and a SIR measurement method in the mobile communication field.

In a wireless device, SIR (Signal-to-Interference-power Ratio) measurement is a function necessary for estimating the quality of a wireless transmission path. The base station, mobile station, base station host apparatus, and the like determine the quality of the wireless transmission path based on the measured SIR, and change the transmission rate, service, and the like according to the quality. Since SIR is the ratio of signal power to interference power, the higher the calculation accuracy of signal power and interference power, the higher the SIR measurement accuracy.

In a wireless communication system, a known symbol sequence (pilot signal) for detecting a phase in a physical channel is provided, for example, as in the W-CDMA system (for example, see Non-Patent Document 1). The pilot signal is a symbol sequence that is known in advance by both the base station and the mobile station, and signal power and interference power are calculated by measuring the power or amplitude of the pilot signal.

Conventional radio apparatuses generally use received signal power (hereinafter referred to as received power) as signal power, and calculate dispersion of received signals as interference power. The SIR is calculated from the signal power and the interference power calculated in this way. (For example, refer to Patent Document 1 and Patent Document 2).

JP 2006-67002 A JP 2006-80585 A Japanese Patent No. 4226641 Japanese Patent No. 3993556 Japanese Patent No. 3901026 Patent No. 4090969 Japanese Patent No. 4053419

Since the received signal includes not only the signal component but also the noise component, the signal power calculated by the conventional wireless device includes not only true signal power (power of only the signal component of the received signal) but also noise power. Therefore, the value is larger than the true signal power. For this reason, there is a problem that the measurement accuracy of SIR, which is the ratio of signal power to interference power, deteriorates and the quality of the wireless transmission path cannot be accurately grasped. In mobile communication, the amplitude of the received signal is not constant but varies. When the amplitude of the received signal is small, the influence of quantization noise becomes large, and when the amplitude of the received signal is large, overflow occurs, so that the signal power and the interference power cannot be calculated correctly. For this reason, the calculation accuracy of SIR deteriorates, and there is a problem that the quality of the wireless transmission path cannot be accurately grasped.

Further, as in Patent Document 3, SIR is calculated by using the average value of the result of despreading the pilot signal in the W-CDMA system as signal power, and as an improvement in SIR measurement accuracy, the NF included in the SIR when the interference component is low There is a method to remove (Noise Figure). However, the pilot signal is a sequence that takes positive / negative (or 0/1) (see Non-Patent Document 1), and if the despread results are averaged as they are, they are canceled out by positive and negative (or 0 and 1 are not equal). There is a problem that the signal power is not calculated correctly. Furthermore, in practice, a signal that was only the I component or only the Q component at the time of data transmission should be a signal having amplitudes in both the I component and the Q component due to distortion of the wireless transmission path. In Patent Document 3, there is a problem that this is not taken into consideration in the calculation formula of the signal power.

Patent Document 4 estimates the statistical data of noise more accurately to evaluate the reliability of the received signal and reconstructs the data based on it. Patent Document 5 estimates the ratio of signal power to interference noise power by increasing the accuracy of interference power and prevents deterioration of reception quality. Patent Document 6 mentions that observation information is more Gaussian than information source signals. However, none of the above-described patent documents removes the noise power included in the signal power and calculates the highly accurate signal power.

The present invention has been made to solve the above-described problems, and an object of the present invention is to obtain a radio apparatus that can calculate signal power with high accuracy and improve SIR measurement accuracy.

A radio apparatus according to the present invention includes a variable amplifier, an AGC control unit that automatically controls a gain of a variable amplifier by controlling a gain of the variable amplifier, and a signal power and an interference power using a predetermined reception signal that is automatically gain controlled. A signal power / interference power calculation unit to be calculated, and a SIR calculation unit to calculate SIR based on the signal power and interference power calculated by the signal power / interference power calculation unit.

In the present invention, signal power and interference power are calculated using a predetermined reception signal subjected to automatic gain control. Since the amplitude of the reception signal subjected to automatic gain control converges to a desired value, the influence of quantization noise in power calculation is small, and overflow does not occur, so that the calculation accuracy of signal power and interference power is improved. For this reason, since the measurement accuracy of SIR is improved and the quality of the wireless transmission path can be accurately grasped, the accuracy of scheduling processing according to the quality of the wireless transmission path is increased, and wireless communication can be performed at an appropriate transmission rate.

It is a block diagram which shows the structure of the radio | wireless apparatus in Embodiment 1 of this invention. It is a DPCCH frame configuration diagram. It is a table | surface which shows the slot format of DPCCH. It is a block diagram which shows the structure of the radio | wireless apparatus in Embodiment 1 of this invention. It is a graph which shows the relationship between SIR and an offset value. It is a graph which shows the difference of unbiased variance and sample variance. It is a graph which shows difference (DELTA) SIR [dB] between SIR calculated based on the prior art, and SIR calculated based on this Embodiment 2. FIG. 10 is a table showing ΔSIR [dB] when the number of pilot symbols = 5. 10 is a table showing ΔSIR [dB] when the number of pilot symbols = 6. 10 is a table showing ΔSIR [dB] when the number of pilot symbols = 7. 10 is a table showing ΔSIR [dB] when the number of pilot symbols = 8. It is a block diagram which shows the structure of the radio | wireless apparatus in Embodiment 3 of this invention. It is a block diagram which shows the structure of the input bit width adjustment part 20 in Embodiment 3 of this invention. It is explanatory drawing which shows operation | movement of the input bit width adjustment part 20 in Embodiment 3 of this invention. It is a graph which shows the error of signal power, interference power, and SIR which arise by frequency offset. It is a block diagram which shows the structure of the demodulation part in Embodiment 4 of this invention. It is a block diagram which shows the structure of the receiver in Embodiment 5 of this invention. It is a block diagram which shows the structure of the radio | wireless apparatus in Embodiment 5 of this invention.

DESCRIPTION OF SYMBOLS 1 Antenna 2 Receiver 3 A / D converter 4 Demodulation part 5 Timing synchronization part 6 Detection part 7 Signal power / interference power calculation part 8 SIR calculation part 9 Transmitter 10 Scheduler part 11 Modulation / coding part 12 D / A conversion 13 Quadrant conversion unit 14 Amplifier 15 Mixer 20 Input bit width adjustment unit 21 First symbol averaging unit 22 Bit position determination unit 23 Extraction unit 30 Frequency offset correction unit 40 AGC control unit 50 Desired wave / interference wave level calculation unit 60 Variable amplifier

Embodiments of a wireless device and an SIR measurement method according to the present invention will be described with reference to the drawings. In the following drawings, the same reference numerals indicate the same or corresponding configurations. In addition, this invention is not limited to each embodiment shown below.

Embodiment 1 FIG.
FIG. 1 is a block diagram showing a configuration of a radio apparatus according to Embodiment 1 of the present invention. The wireless device includes an antenna 1, a receiver 2, a transmitter 9, and a scheduler unit 10. The received signal received from the antenna 1 is converted from an analog signal to a digital signal by the A / D converter 3. The analog signal input to the A / D converter 3 is frequency-converted from a high frequency to a baseband frequency or an intermediate frequency by a mixer (not shown in FIG. 1). The digital signal output from the A / D converter 3 is input to the demodulator 4.

The demodulation unit 4 includes a timing synchronization unit 5, a detection unit 6, a signal power / interference power calculation unit 7, and an SIR calculation unit 8. The timing synchronization unit 5 synchronizes the timing of the input digital signal. The detector 6 detects the received signal and generates a complex signal composed of an I component and a Q component. In a state in which timing synchronization is established and the pilot signal can be extracted at an accurate timing, the signal power / interference power calculation unit 7 calculates the signal power based on the amplitude corresponding to the soft decision value for each symbol of the pilot signal. . Further, the variance of the amplitude of the pilot signal is calculated and used as interference power. The SIR calculation unit 8 calculates the SIR by dividing the signal power calculated by the signal power / interference power calculation unit 7 by the interference power.

The scheduler unit 10 determines the optimum transmission rate for the quality of the wireless transmission path based on the SIR calculated by the SIR calculation unit 8 and the communication partner station (if the station is a base station, the partner station is a mobile station, If the local station is a mobile station, the partner station instructs the base station) to transmit at that transmission rate. The modulation / encoding unit 11 encodes and modulates data including information instructed by the scheduler unit 10. The D / A converter 12 converts a digital signal into an analog signal.

Next, the operation of the wireless device in the first embodiment will be described. The demodulator 4 calculates the SIR using the signal power and interference power calculated for each predetermined data unit. Hereinafter, a method for calculating the signal power and the interference power for each slot of the DPCCH (Dedicated Physical Control Channel) defined in 3GPP TS25.211 will be described. FIG. 2 is a frame configuration diagram of the DPCCH. FIG. 3 is a table showing a slot format of DPCCH.

In the frame configuration shown in FIG. 2, the signal power / interference power calculation unit 7 calculates signal power and interference power using the pilot signal of each slot. The pilot signal is located at the head of each slot, and the number of symbols is N_pilot. The number of pilot signal symbols (hereinafter referred to as the number of pilot symbols) N_pilot differs according to the slot format as shown in FIG.

The pilot signal is known, and the radio apparatus holds in advance the pilot signal symbol sequence as a reference pilot signal P_I [isym] + j × P_Q [isym]. Isym is a symbol number. The detector 6 detects the input signal and extracts a complex signal represented by Rx_I [isym] + j × Rx_Q [isym]. Further, the detector 6 multiplies this complex signal by the complex conjugate of the reference pilot signal (level-converted from 0 → + 1, 1 → −1), as shown in the following equation, for each symbol in the pilot signal. I component amplitude sym_I [isym] and Q component amplitude sym_Q [isym] are obtained.

The signal power / interference power calculation unit 7 calculates the signal power and the interference power by using the signal after adjustment so that each symbol of the pilot signal is in the same quadrant. For example, in the W-CDMA system, the detection unit 6 multiplies the despread signal by the inverse characteristic of the pilot signal, as shown in Equation 1, so that all symbols are complex planes (IQ planes). The signal power / interference power calculation unit 7 calculates the signal power using the corrected data (sym_I [isym], sym_Q [isym]). The signal expressed as A + j × 0 or 0 + j × A at the time of transmission from the communication partner station becomes C + j × D due to transmission path distortion. Although the real component C becomes the I component and the imaginary component D becomes the Q component, the signal power / interference power calculation unit 7 calculates the signal power and the interference power in consideration of both the I component and the Q component. Therefore, the power calculation accuracy is high.

In the receiver 2 shown in FIG. 1, the detection unit 6 in the demodulation unit 4 performs quadrant conversion after detecting the I component and Q component, but the I component in the preceding stage of the A / D converter 3. Each of the signals separated into Q components may be A / D converted and input to the demodulator 4. FIG. 4 is a block diagram showing a configuration of the radio apparatus according to Embodiment 1 of the present invention. A signal received from the antenna 1 is amplified by an amplifier 14 and separated into an I component and a Q component by a mixer 15. The I component and Q component signals are converted from analog signals to digital signals by the A / D converter 3 and input to the demodulator 4. The I component and Q component signals input to the demodulator 4 are subjected to quadrant conversion represented by Equation 1 by the quadrant converter 13, and the corrected data transferred to the first quadrant is signal power / interference power calculator. 7 is input.

The signal power / interference power calculation unit 7 calculates signal power and interference power using sym_I [isym] and sym_Q [isym]. First, the received power for each slot (= signal power before correction, RXPOW _slot ) is expressed by Equation 2.

Here, N_pilot is the number of pilot symbols per slot as described above, in other words, the number of samples.

Next, the interference power is obtained from the variance of the amplitude of the received signal. Here, the sample variance is used as the variance. The interference power for each _slot (EPPPOW _slot ) is expressed by the following equation.

When calculating the signal power, if the number of samples of the received signal used for power calculation is small, sufficient averaging cannot be performed, and noise components are mixed in the calculated power. The noise component is removed. This indicates that the power of the noise component included in the received signal is expressed by (variance / number of samples). In order to calculate an accurate signal power, a value obtained by averaging the variance of the received signal by the number of symbols may be subtracted from the received power. The signal power (SIGPOW _slot ) for each slot calculated by the signal power / interference power calculation unit 7 is shown in the following equation.

The signal power correction described above can also be considered as follows. It is considered that the signal power RXPOW _slot before correction includes Gaussian noise generated inside the apparatus. Gaussian noise has a normal distribution. If all the received pilot signals are a population, the pilot signal for one slot is a sample obtained by randomly extracting data from the population, and the pilot symbol number N_pilot for one slot can be said to be the number of samples. Since population variance is unknown, for example, ERRPOW _slot is assumed to be population variance. According to the central limit theorem, when the population is sufficiently large, the average of the variance of the received signal is represented by (ERRPOW _slot ÷ N_pilot). The central limit theorem is that if the variable X follows the distribution of mean μ and standard deviation σ, the sample mean X based on a random sample of size n is the mean μ and standard deviation σ when n becomes infinitely large. It approaches the normal distribution of / √n. According to the central limit theorem, the average of the variances (ERRPOW _slot ÷ N_pilot) is normally distributed, so that it can be regarded as Gaussian noise included in the received signal. Therefore, the calculation accuracy of the signal power can be improved by correcting the signal power shown in Formula 6.

Further, in calculating the interference power, the pilot signals for one slot may not be used as a population, but pilot signals of as many slots as possible may be used as the population. For example, a variance is obtained by using pilot signals for 4 frames (ie, 60 slots) as a population, and is assumed to be a population variance σ _S60 . The ERRPOW _slot of the slot for which signal power is to be calculated can be calculated as follows. First, a pilot signal for one slot is extracted from the population and used as a sample. The number of samples of the sample is the number N_pilot of pilot symbols in one slot, and an average of received signal dispersion (σ _s60 ÷ N_pilot) in the sample may be calculated and set as an ERRPOW _slot . There is an effect that the accuracy of the signal power and the interference power is higher in the case where the variance of the pilot signals for four frames is used as the population variance than in the case where the variance of the pilot signal of 1 slot is the population variance.

Although the received signal includes a noise component, the power of the noise component is removed from the received power to obtain the signal power (SIGPOW _slot ), so that the calculation accuracy of the signal power is increased. Note that the denominator of (ERRPOW _slot / N_pilot) to be subtracted from the received power is not limited to the number of samples, and may be a changeable value for adjustment. This is because the noise component includes thermal noise and distortion noise in addition to the interference component due to interference from other users.

Based on the signal power (SIGPOW _slot ) and interference power (EPPPOW _slot ) calculated by the signal power / interference power calculation unit 7, the SIR calculation unit 8 calculates SIR (SIR _slot ) for each _slot . The SIR _slot is a value obtained by dividing the signal power (SIGPOW _slot ) for each slot by the interference power (EPPPOW _slot ) for each _slot , as shown in the following equation.

When the received power shown in Formula 2 is used as the signal power as in a conventional wireless device, the received signal includes a noise component and thus becomes larger than the actual signal power. In particular, signals with a low transmission rate, such as RMC 12.2K (12.2 kbps, 3GPP TS25.141) and SDCCH 13.6K (13.6 kbps, 3GPP TS25.141), have a large influence on noise power, and noise power is reduced. It cannot be ignored. The offset value (offset [dB]) shown in the following equation is generated in the SIR when Equation 2 is adopted as the signal power as in the conventional case and when Equation 6 is adopted as the signal power as in the present embodiment. .

That is, when Equation 2 is used for calculating the signal power, the SIR is increased by 1 + ERRPOW _slot / (SIGPOW _slot × N_pilot) times as compared with the case where Equation 6 is adopted.

FIG. 5 is a graph showing the relationship between SIR and offset value. The horizontal axis of FIG. 5 shows the SIR when the received power is the signal power, that is, the SIR calculated based on the prior art. As shown in FIG. 5, when the SIR is less than 0 [dB], the offset value is remarkably large, and it can be seen that particularly in the region where the SIR is small, the influence of the noise component on the SIR calculation is large. Further, the offset value increases as the number of pilot symbols N_pilot decreases.

In the SIR calculation unit 8 described above, the SIR is calculated in slot units. However, for example, the SIR may be calculated in frame units. An example in which there are 15 slots per frame as shown in FIG. 2 will be described. When the SIR calculation unit 8 calculates the SIR in units of frames, the signal power / interference power calculation unit 7 calculates the signal power (SIGPOW _flame ) for each frame and the interference power (ERRPOW _flame ) for each frame. Alternatively, the signal power / interference power calculation unit 7 calculates signal power (SIGPOW _slot ) and interference power (ERRPOW _slot ) for each _slot , and the SIR calculation unit 8 calculates signal power (SIGPOW _flame ) and interference power (ERRPOW) for each frame. After calculating _flame ), SIR (SIR _flame ) for each frame may be calculated. The interference power for each frame (ERRPOW _flame ) is expressed by the following equation.

The signal power (SIGPOW _flame ) for each frame is expressed by the following equation.

Accordingly, the SIR (SIR _flame ) for each frame is expressed by the following equation.

The central limit theorem described above defines that the sample mean Xm approaches a normal distribution when the size n of the random sample becomes infinitely large. In practice, the size of the random sample n is about 30. But it approaches a normal distribution. Even if the number N_pilot of pilot symbols per slot is 3 to 8, it is 45 to 120 in one frame. Therefore, the value σ ² / (N_pilot × 15) obtained by dividing the variance σ ² for one frame by the total number of pilot symbols for one frame is close to a normal distribution, and is close to Gaussian noise corresponding to thermal noise. . In calculating the signal power for each frame, the noise power with high accuracy is removed from the received power, so that the calculation accuracy for the signal power for each frame is high. Accordingly, the calculation accuracy of the SIR for each frame is also increased.

The SIR (SIR _slot ) for each slot is calculated using the signal power (SIGPOW _slot ) and the interference power (ERRPOW _slot ) calculated for each _slot , and the signal power is not calculated instead of calculating the sum total in one frame of the SIR _slot. After calculating the sum of one frame (SIGPOW _flame , ERRPOW _flame ) for each interference power, the SIR is calculated last. By calculating the SIR (SIR _flame ) for each frame in this way, variations in thermal noise for each slot can be eliminated, and the accuracy of SIR calculation is improved. By improving the accuracy of the SIR calculation, the scheduler unit 10 can accurately grasp the quality of the wireless transmission path, set an appropriate transmission rate in the transmitter 9, and notify the communication partner station of the appropriate transmission rate. You can also

In the above description, the method for calculating the signal power and the interference power for each slot or for each frame has been described. However, it may be calculated for each predetermined number of slots. If the number of slots used to calculate signal power and interference power is N_slot, the signal power and interference power for each predetermined number of slots are expressed by the following equations.

However, RXPOW _{N_slot} , AVE_I ₂ and AVE_Q ₂ are as follows.

In the W-CDMA system, since one frame has 15 slots, if N_slot = 15 is substituted in Equations 12 to 16, the average signal power and interference power of one frame can be calculated. Since sym_I [isym] and sym_Q [isym] are averaged, the calculation accuracy of signal power and interference power is improved.

When calculating the average signal power and interference power of one

frame using Equations

12 and 13, the dispersion of pilot signals for one frame is obtained as interference power, but the dispersion of pilot signals in as many slots as possible is obtained. Interference power may be used. For example, instead of one frame, seeking the variance of the pilot signals for four frames (60 slots), and the population variance sigma _S60 ^2. And specimens withdrawn population variance sigma _S60 ² from one frame of the number of pilot symbols (N_pilot × 15), the average in a sigma _S60 ² ÷ of the variance of the received signal at that sample a (N_pilot × 15) as the noise power Good. Since the calculated noise power is closer to the normal distribution when the pilot signal variance for 4 frames is used as the variance of the pilot signal for 1 frame, the noise power is calculated. Increases accuracy. If the signal power is calculated using highly accurate noise power, the accuracy of the signal power is also increased.

In addition, in the above description, as the averaging of sym_I [isym], sym_Q [isym] and sym_I [isym] ² + sym_Q [isym] ² , the method of adding the number of samples and then dividing and averaging by the number of samples Although used, a method of multiplying time constants and sequentially adding them may be used like a loop filter. When the division process is performed, the circuit scale becomes large. However, by using the sequential addition method, the division process becomes unnecessary, so that the circuit scale can be reduced.

The calculation of the loop filter is performed as follows, for example. α and β are time constants (such as 1/256). accum_sym_I [N-1] ² is an average obtained by sequentially adding N sym_I [isym] ² from the symbol number isym from 0 to (N-1) using a time constant, and accum_sym_Q [ N-1] ² is obtained by averaging N sym_Q [isym] ² by sequentially adding them using a time constant. Similarly, accum_sym_I [N-1] is obtained by averaging N sym_I [isym] by sequential addition using a time constant, and accum_sym_Q [N-1] is N sym_Q [ isym] is averaged by sequential addition using a time constant.

The signal power / interference power calculation unit 7 uses Equation 17 instead of the first term on the right side of Equation 5, and uses Equation 18 and Equation 19 instead of AVE_I and AVE_Q in Equation 5 and Equation 6, to calculate signal power and interference power. May be calculated. Further, the signal power and the interference power may be calculated using Equation 17 instead of the first term on the right side of Equation 12, and using Equation 18 and Equation 19 instead of AVE_I ₂ and AVE_Q ₂ of Equation 12 and Equation 13. .

According to the first embodiment of the present invention, signal power is obtained by subtracting noise power from the power of a received signal including a noise component, so that signal power calculation accuracy is improved and SIR measurement accuracy is also improved. . Further, in calculating the signal power and the interference power, the calculation accuracy of the signal power and the interference power can be further improved and the measurement accuracy of the SIR can be improved by increasing the population used for the calculation and increasing the number of samples. By improving the measurement accuracy of the SIR, the scheduler unit 10 can accurately grasp the quality of the radio transmission path, so that the accuracy of the scheduling process according to the quality of the radio transmission path is increased and radio communication can be performed at an appropriate transmission rate. For this reason, waste of power consumption due to wireless communication at an excessive transmission rate can be suppressed.

Embodiment 2. FIG.
In Embodiment 1, the interference power is calculated using the sample variance as the variance, but the interference power may be calculated using unbiased variance. The sample variance has a smaller expected value than the population variance. Unbiased variance is obtained by correcting the expected value to be equal to the variance of the population. If the number of samples is sufficiently large, sample variance = unbiased variance. However, when the number of samples (= the number of pilot symbols per slot N_pilot) is small as in the case where a pilot signal for one slot is used as a sample, accuracy is higher when unbiased dispersion is used for interference power. The configuration of the radio apparatus according to the second embodiment is the same as that of the first embodiment, and a description thereof will be omitted.

In Embodiment 2, in calculating the interference power, instead of using the variance averaged by the number of samples, that is, the number of pilot symbols N_pilot, the variance averaged by (number of pilot symbols N_pilot-1) is used. This is because, when the number of samples is small, the unbiased variance corrected so that the expected value is equal to the population variance can express the degree of data dispersion more correctly than the sample variance. That is, when the number of pilot symbols N_pilot is small, interference power calculation accuracy is improved by using unbiased variance rather than sample variance of the received signal.

The signal power and interference power calculated by the signal power / interference power calculation unit 7 in the second embodiment will be described. The interference power (ERRPOW2 _slot ) for each slot calculated by the signal power / interference power calculation unit 7 is expressed by the following equation.

FIG. 6 is a graph showing the difference between the interference power (ERRPOW2 _slot ) when using unbiased variance and the interference power (ERRPOW _slot ) when using sample variance. As shown in the following equation, when displayed in decibels, the value obtained by subtracting the interference power (ERRPOW _slot ) using sample variance from the interference power (ERRPOW2 _slot ) using unbiased variance is a signal similar to the calculation of SIR. If power is used, it matches the value obtained by subtracting the SIR calculated using unbiased variance from the SIR calculated using sample variance.

As shown in FIG. 6, the interference power is higher by about 0.5 to 1.0 [dB] when using unbiased dispersion than when using sample dispersion. That is, the SIR is about 0.5 to 1.0 [dB] smaller when the unbiased variance is used than when the sample variance is used.

When the interference power (ERRPOW2 _slot ) is calculated by unbiased dispersion, the signal power can also be expressed as follows using the unbiased dispersion.

FIG. 7 is a graph showing a difference ΔSIR [dB] between the SIR calculated based on the prior art and the SIR calculated based on the second embodiment. 8 to 11 are tables showing ΔSIR [dB] for each number of pilot symbols N_pilot. A difference ΔSIR [dB] between the SIR calculated based on the prior art and the SIR calculated based on the second embodiment is expressed by the following equation.

If the same signal power is used for the calculation of SIR, the first term in the last row of Equation 23 is equal to the value obtained by subtracting the SIR calculated using unbiased variance from the SIR calculated using sample variance. Also, the second term in the last row of Equation 23 is the SIR offset value when using unbiased dispersion.

In the second embodiment, the signal power / interference power calculation unit 7 calculates the signal power (SIGPOW2 _slot ) and the interference power (ERRPOW2 _slot ) for each slot using Expression 22 and Expression 20. By removing the noise component from the received power, the calculated signal power becomes smaller than before. In the calculation of interference power, when the number of samples (that is, the number of pilot symbols per slot N_pilot) is small, the calculated interference power becomes larger than before by using unbiased variance without using sample variance. . Since SIR is signal power ÷ interference power, the SIR decreases as the signal power decreases, and the SIR decreases as the interference power increases. Therefore, the SIR calculated based on the second embodiment is smaller than the SIR calculated based on the conventional technique.

As shown in FIG. 7, the difference ΔSIR is about 5 to 7 [dB] in the region where the SIR is small, and the difference ΔSIR is about 0.5 to 1 [dB] in the region where the SIR is large. In a region where the SIR is small, the SIR is significantly smaller when the second embodiment is used than when the conventional technique is used, and the tendency is more remarkable as the number of pilot symbols N_pilot, that is, the number of samples is smaller. become. Even in a region where the SIR is large, there is an error of about 0.5 to 1 [dB], so that the SIR is calculated more accurately when the second embodiment is used than when the conventional technique is used. be able to.

In the second embodiment, the method of calculating the signal power (SIGPOW2 _slot ) and the interference power (ERRPOW2 _slot ) for each _slot has been described. However, using the SIGPOW2 _slot and the ERRPOW2 _slot , as in the first embodiment, each frame. The signal power (SIGPOW2 _flame ) and the interference power (ERRPOW2 _flame ) may be calculated. Since the variation of the thermal noise for each slot can be removed, the SIR calculation accuracy is improved.

When performing wireless communication at a high transmission rate such as HSUPA (High Speed Uplink Packet Access), a high-quality wireless transmission path, that is, a high SIR is required. However, there is a problem that SIR does not increase while maintaining linearity due to quantization error, signal power, and noise power calculation errors, and the accuracy of SIR calculation deteriorates in a region where SIR is high. Further, high-transmission-rate wireless communication such as HSUPA has the following characteristics (1) and (2).
(1) It is difficult to obtain a diffusion gain because of a large amount of data and a low spreading factor.
(2) Since the amount of data for encoding cannot be increased at the time of encoding, the encoding gain cannot be obtained.
For this reason, in wireless communication at a high transmission rate, reception performance and scheduling performance are more dependent on demodulation performance such as SIR estimation than at low transmission rates. In particular, the linearity of the SIR cannot be maintained in a region where the SIR is large, and the SIR reaches the middle. However, the scheduler unit 10 cannot determine whether to request the highest transmission rate from the communication partner station, and the highest transmission rate. There was a problem that it was impossible to maintain the reception. However, according to the second embodiment, since the SIR can be calculated with high accuracy even in a region where the SIR is large, the scheduler unit 10 can accurately grasp the state of the wireless transmission path and perform wireless communication at a high transmission rate. Even in such a case, it can be controlled to transmit and receive at an appropriate transmission rate. In addition, waste of power consumption due to wireless communication at an excessive transmission rate can be suppressed.

Embodiment 3 FIG.
In order to satisfy reception performance at a high transmission rate such as HSUPA, it is necessary to take a large bit range in A / D conversion (conversion from analog to digital). For example, when the A / D converter 3 having a resolution of 16 bits is used, the demodulator 4 performs the multiplication after performing the averaging process in the multiplication process used for calculating the signal power and the like. A bit × 32 bit multiplier may be required, which increases the circuit scale. In order to reduce the circuit scale while preventing the reception performance from being deteriorated, it is effective to extract only necessary portions from the large bit range every time the arithmetic processing is performed. In the third embodiment, a method of extracting 8 bits from a received signal that has been A / D converted at 16 bits before calculating signal power will be described. In the third embodiment, a method of extracting 8 bits from 16 bits will be described, but the number of bits is not limited to these.

FIG. 12 is a block diagram showing a configuration of a radio apparatus according to Embodiment 3 of the present invention. A description of the same configuration as that of the first embodiment will be omitted, and differences from the first embodiment will be described. An input bit width adjustment unit 20 is added to the demodulation unit 4. A 16-bit complex signal (I component and Q component signals) is sent from the detection unit 6 to the input bit width adjustment unit 20 in a state in which the timing of the received signal that has been A / D converted into a digital signal is synchronized. Entered. The input bit width adjustment unit 20 extracts 8 bits from the input 16-bit complex signal and outputs the extracted signal to the signal power / interference power calculation unit 7.

FIG. 13 is a block diagram showing the configuration of the input bit width adjustment unit 20 according to Embodiment 3 of the present invention. FIG. 14 is an explanatory diagram showing the operation of the input bit width adjustment unit 20 according to Embodiment 3 of the present invention. With reference to FIGS. 13 and 14, the operation of the input bit width adjustment unit 20 in the third embodiment will be described. In the DPCCH shown in FIG. 2, the operation of the input bit width adjustment unit 20 will be described by taking the case of the slot format # 1 shown in FIG. 3 as an example.

The leading symbol averaging unit 21 adds the absolute values of the leading eight symbols (corresponding to the pilot signal) of the slot for the input 16-bit signal and then averages them. Since the averaging is division by 8 which is the number of pilot symbols, the circuit scale can be reduced by using a 3-bit shift without using a divider.

Based on the result averaged by the head symbol averaging unit 21 (hereinafter referred to as “RX _AVE” ), the bit position determination unit 22 selects the 8 bits at any position out of the bit width of the 16-bit input signal. Determine whether to extract. RX _AVE is an average value of the amplitude of the received signal, and the bit position determination unit 22 determines a bit position to be extracted based on the amplitude of the received signal. The bit position determination unit 22 holds eight parameters S1 to S8 in advance, compares RX _AVE with the parameters S1 to S8, and determines an 8-bit extraction position. The parameters S1 to S8 are set to increase gradually, for example, S1 = 16, S2 = 32, S3 = 64, S4 = 128, S5 = 256, S6 = 512, S7 = 1024, S8 = 2048. The Further, the parameters S1 to S8 are set as variables, and can be changed by setting from software.

As shown in FIG. 14, the bit position determination unit 22 determines that the bit extraction position is 8 bits from the least significant bit when RX _AVE <S1, and sets the bit extraction position to 1 when S1 ≦ RX _AVE <S2. Bit shift is performed and it is determined that the lower 2 bits are 8 bits. As RX _AVE increases, the bit extraction position is shifted to the upper side. When RX _AVE ≧ S8, the bit extraction position is determined to be 8 bits from the most significant bit.

The extraction unit 23 extracts 8 bits from the 16-bit input signal based on the bit extraction position determined by the bit position determination unit 22, and replaces the extracted 8 most significant bits with a sign bit. The bit position determination unit 22 determines a bit extraction position for each slot, and the extraction unit 23 does not change the bit extraction position in the same slot section. The signal power / interference power calculation unit 7 calculates signal power and interference power based on the 8-bit signal (I component amplitude and Q component amplitude for each symbol) extracted by the input bit width adjustment unit 20. Since the operations of the signal power / interference power calculation unit 7 and the SIR calculation unit 8 are the same as those in the first embodiment or the second embodiment, description thereof will be omitted. 12 shows the configuration in which the input bit width adjustment unit 20 is added to the wireless device shown in FIG. 1, the input bit width adjustment unit 20 may be added similarly to the wireless device shown in FIG. . In the demodulator 4, the input bit width adjuster 20 may be disposed at least before the signal power / interference power calculator 7.

In Embodiment 3, by changing the bit extraction position based on the amplitude of the pilot signal for each slot, bit overflow can be suppressed and appropriate bit extraction can be performed. It is possible to prevent the reception performance from being deteriorated due to saturation of the received signal and missing bits, to reduce the circuit scale, and to reduce the power consumption. In addition, since the optimum bit extraction position is determined based on the amplitude of the pilot signal, the calculation accuracy of the signal power and the interference power is improved in spite of using a reception signal with a small bit width, and the embodiment The effect equivalent to 1 or Embodiment 2 can be obtained.

Embodiment 4 FIG.
In actual wireless communication, a received signal has a frequency offset that oscillates with a certain period. Also, in the case of bit extraction described in the third embodiment, even if an overflow is unavoidable, the received signal input to the signal power / interference power calculation unit 7 similarly has a frequency offset. Arise. In the fourth embodiment, the calculation accuracy of the signal power and the interference power is improved by calculating the signal power and the interference power using the received signal whose frequency offset is corrected.

FIG. 15 is a graph showing signal power, interference power, and SIR errors caused by frequency offset. When a frequency offset occurs, an error occurs in the SIR as shown in FIG. 15, but after correcting the frequency offset, the calculation accuracy of both powers and the calculation accuracy of SIR can be improved by calculating the signal power and the interference power. Can be improved.

FIG. 16 is a block diagram showing a configuration of a demodulation unit in the fourth embodiment of the present invention. The frequency offset correction unit 30 estimates the frequency offset from the input I component and Q component signals and corrects the frequency offset of the received signal. The corrected signal is output to the signal power / interference power calculation unit 7. The signal power / interference power calculation unit 7 calculates the signal power and the interference power by one of the methods described in the first embodiment or the second embodiment. Also, the received power represented by Equation 2 may be used as the signal power. By calculating the signal power and the interference power using the received signal whose frequency offset is corrected, the calculation accuracy of the signal power and the interference power is improved. Even when the input bit width adjustment unit 20 has an insufficient bit width and an overflow occurs, it is possible to prevent deterioration in calculation accuracy of signal power and interference power by correcting the frequency offset.

The frequency offset correction unit 30 estimates the frequency offset by the following method. Hereinafter, the values expressed by Equation 3 and Equation 4 are referred to as estimated transmission path characteristics. Estimated transmission line characteristics including the I component and the Q component are expressed by the following equations.

islot indicates the slot number. AVE_I [islot] and AVE_Q [islot] are the average of the I component and the average of the Q component in the slot number islot, respectively, and are expressed using

Equations

3 and 4.

The frequency offset correction unit 30 first obtains the estimated transmission path specific change amount AVE_dif [islot] in the slot number islot.

Here, AVE [islot-1] ^* is a complex conjugate of AVE [islot-1]. The change amounts AVE_I_dif [islot] and AVE_Q_dif [islot] for each of the I component and the Q component are expressed by the following equations.

Next, the frequency offset correction unit 30 averages the amount of change in the estimated transmission path characteristics expressed by Expressions 25 to 27. The averaging process is performed as follows using the forgetting factor γ.

γ can also be said to be a time constant. The averaging process for each of the I component and the Q component is expressed by the following equations.

Using the average of the estimated amount of change in transmission path characteristics, the frequency offset is expressed by the following equation.

The frequency offset correction unit 30 corrects the frequency offset of the received signal using the frequency offset estimated by Equation 31. Since esfo is a phase rotation amount in one slot, when one slot is composed of 10 symbols as in the DPCCH shown in FIG. 2, the phase rotation amount of the frequency offset per symbol is esfo / 10. . AVE_I [islot] and AVE_Q [islot] are corrected based on the rotation amount at the center position of the pilot signal. Since the number of pilot signal symbols per slot is N_pilot, correction is performed by rotating by 0- (N_pilot / 2) symbols.

The frequency offset correction unit 30 corrects the received signal using the estimated frequency offset. A signal AVEcor [islot] obtained by correcting AVE [islot] is expressed by the following equation.

Further, signals AVE_Icor [islot] and AVE_Qcor [islot] obtained by correcting each of the I component and the Q component are calculated as follows.

In the fourth embodiment, since the frequency offset correction unit 30 corrects the frequency offset of the received signal and then calculates the signal power and the interference power, the calculation accuracy of the signal power and the interference power is improved. The fourth embodiment can be used even when overflow does not occur. In this case, since the frequency offset that is not the cause of overflow is corrected, it is effective in improving the calculation accuracy of the signal power and the interference power.

Embodiment 5 FIG.
In mobile communication, the amplitude of a received signal is not constant but varies. When the amplitude of the received signal is small, the signal power and the interference power cannot be calculated correctly, and the SIR calculation accuracy deteriorates. In addition, when the amplitude of the received signal is large, the calculation accuracy of the signal power, interference power, and SIR deteriorates due to the occurrence of overflow. Therefore, in the fifth embodiment, the calculation accuracy of the signal power, the interference power, and the SIR is improved by calculating the signal power and the interference power using the reception signal subjected to the automatic gain control. In the fifth embodiment, a method for obtaining a desired wave level and interference wave level with high accuracy based on the signal power calculated by the signal power / interference power calculation unit 7, interference power, and control values used for automatic gain control. explain. The desired wave level is the power of the signal component at a predetermined measurement location, and the interference wave level is the power of the interference component at the predetermined measurement location. The predetermined measurement location may be, for example, an antenna end. On the other hand, the signal power and the interference power described above are the signal component power and the interference component power in the signal power / interference power calculation unit 7, respectively. Thus, the specified points differ between the desired wave level and the signal power, and between the interference level and the interference power.

FIG. 17 is a block diagram showing a configuration of a receiver in the fifth embodiment of the present invention. In FIG. 17, the AGC control unit 40 controls the gain of the variable amplifier 60 based on a gain control amount (AGC_GAIN) described later, and outputs the gain control amount (AGC_GAIN) to the desired wave / interference wave level calculation unit 50. . The desired wave / interference wave level calculation unit 50 calculates the desired wave level and the interference wave level.

FIG. 18 is a block diagram showing a configuration of a radio apparatus according to Embodiment 5 of the present invention. In the receiver 2 shown in FIG. 18, the I component and the Q component signal are separated from the received signal by the mixer 15, and the I component and the Q component signal are converted into digital signals by the A / D converter 3, respectively. Input to part 4. The AGC control unit 40 and the desired wave / interference wave level calculation unit 50 are the same as those of the receiver 2 shown in FIG. Note that the input bit width adjustment unit 20 and the frequency offset correction unit 30 shown in FIGS. 17 and 18 may be provided as necessary, and are not essential components.

Next, the operation will be described with reference to FIG. A signal received by the antenna 1 is amplified by an amplifier 14 and separated into an I component and a Q component by a mixer 15. The I component and Q component signals are each amplified or attenuated by the variable amplifier 60 based on the gain control amount (AGC_GAIN) obtained by the AGC control unit 40. The amplified or attenuated I component and Q component signals are converted into digital signals by the A / D converter 3 and input to the demodulator 4.

The I component and Q component signals converted into digital signals by the A / D converter 3 are input to the demodulator 4 and despread in the case of, for example, the W-CDMA system. The quadrant conversion unit 13 multiplies the despread signal by the inverse characteristic of the pilot signal for each symbol. Thereby, a pilot signal having positive and negative amplitudes can be shifted to the first quadrant of the IQ plane. The signal power / interference power calculation unit 7 calculates the signal power (SIGPOW) and the interference power (ERRPOW) using any of the methods described in the first to fourth embodiments. Also, the received power represented by Equation 2 may be used as the signal power. By calculating the signal power and the interference power using the reception signal subjected to the automatic gain control, the calculation accuracy of the signal power and the interference power is improved. The calculated signal power (SIGPOW) and interference power (ERRPOW) are output to the SIR calculation unit 8 and the desired wave / interference wave level calculation unit 50.

The AGC control unit 40 receives I component and Q component signals. In FIG. 18, the signal immediately after being converted into a digital signal by the A / D converter 3 is input to the AGC control unit 40, but the input signal to the AGC control unit 40 is a signal indicating an I component and a Q component. It may be sufficient, and may be a signal after frequency offset correction as shown in FIG. Hereinafter, the I component signal used in the AGC control unit 40 is referred to as I, and the Q component signal is referred to as Q.

The AGC control unit 40 first calculates Σ (I ² + Q ² ) as the RMS measurement value. When Σ (I ² + Q ² ) exceeds a preset upper limit value (AGC_UP), a gain control amount (AGC_GAIN) that attenuates the I component signal and the Q component signal is obtained, and Σ When (I ² + Q ² ) falls below a preset lower limit (AGC_DN), a gain control amount (AGC_GAIN) that amplifies the I component signal and the Q component signal, respectively, is obtained.

The gain control amount (AGC_GAIN), the upper limit value (AGC_UP), and the lower limit value (AGC_DN) are input from the AGC control unit 40 to the desired wave level / interference wave level calculation unit 50. Instead of the upper limit value and the lower limit value, an RMS convergence value (RMSconv) that is an intermediate value between the upper limit value and the lower limit value may be input to the desired wave level / interference wave level calculation unit 50. The RMS convergence value (RMSconv) is expressed by the following equation.

The AGC control unit 40 controls the gain of the variable amplifier 60 so that the RMS measurement value converges to the RMS convergence value (RMSconv). The RMS convergence value (RMSconv) is a convergence value for automatic gain control. The gain control amount (AGC_GAIN) is set to a value that decreases the gain of the variable amplifier 60 when the RMS measurement value exceeds the upper limit value (AGC_UP), and when the RMS measurement value falls below the lower limit value (AGC_DN). The gain of the variable amplifier 60 is set to a value that increases.

The desired wave / interference wave level calculation unit 50 includes the signal power (SIGPOW) and interference power (ERRPOW) calculated by the signal power / interference power calculation unit 7, the gain control amount (AGC_GAIN) calculated by the AGC control unit 40, Based on a preset RMS convergence value (RMSconv) and a gain (GAIN_sp) from a predetermined measurement location to the signal power / interference power calculation unit 7, the desired wave level and the interference wave level are calculated. The desired wave level and the interference wave level are expressed by the following equations.

Strictly speaking, GAIN_sp indicates the gain of the portion excluding the variable amplifier 60 between the predetermined measurement location and the signal power / interference power calculation unit 7.

If the SIR calculated by the SIR calculating unit 8 is configured to be input to the desired wave / interference wave level calculating unit 50, the interference wave level may be calculated as follows.

The calculations of Expressions 39 to 41 are based on the assumption that all the various parameters used in the calculation are true values. However, if the various parameters are values obtained by dB conversion, addition (+) instead of multiplication (×) is performed. ) Is calculated using subtraction (−) instead of division (÷).

By using the gain control amount (AGC_GAIN) calculated by the AGC control unit 40 and the signal power (SIGPOW) calculated by the signal power / interference power calculation unit 7 by the above calculation method, a desired wave level with high accuracy is obtained. Can be sought. In calculating the desired wave level, only the signal power SIGPOW, the gain control amount (AGC_GAIN), and the fixed values (RMSconv and GAIN_sp) are used, and the interference power (ERRPOW) is not used. Since the interference power (ERRPOW) is obtained using dispersion, the accuracy deteriorates unless there is a sufficient number of samples. For this reason, the calculation accuracy of the desired wave level is higher when the interference power (ERRPOW) is not used than when the desired wave level is calculated using it.

When the interference power (ERRPOW) is used to calculate the desired wave level, the gain control amount (AGC_GAIN), the RMS convergence value (RMSconv), and the RMS measurement value (Σ (I ² + Q ² )) are used to calculate The desired wave level is calculated as follows.

Here, ε is the amplitude ratio of the measurement target channel. More specifically, ε is the amplitude ratio of the channel used for signal power calculation with respect to the channel multiplexed with the channel used for signal power calculation. For example, in the W-CDMA system, if a channel used for signal power calculation is DPCCH and a channel multiplexed on DPCCH is DPDCH (Dedicated Physical Data Channel), ε = DPDCH amplitude / DPCCH amplitude. When the channel used for automatic gain control is different from the channel used for calculation of signal power and interference power, the calculation accuracy of the desired wave level can be improved by taking into account the power ratio of both channels.

In the fifth embodiment, a desired wave / interference wave level is calculated using values (AGC_GAIN, RMSconv, etc.) used for automatic gain control and the signal power (SIGPOW) calculated by the signal power / interference power calculator 7. The unit 50 obtains a desired wave level. Since the interference power (ERRPOW) is not used for calculating the desired wave level, the calculation accuracy of the desired wave level is improved. Even when the interference power (ERRPOW) is used for calculation of the desired wave level, the calculation accuracy of the desired wave level is improved by considering the amplitude ratio of the measurement target channel.

Since the radio apparatus according to the present invention is configured to improve the accuracy of SIR calculation, it is suitable for use in a radio apparatus in the mobile communication field.

Claims

In a wireless device that measures SIR using a known predetermined received signal,
A variable amplifier that amplifies or attenuates the input signal;
An AGC control unit for controlling a gain of the variable amplifier to automatically control a received signal;
A signal power / interference power calculation unit that calculates signal power and interference power using a predetermined reception signal that is automatically gain controlled; and
An SIR calculation unit for calculating the SIR based on the signal power and the interference power calculated by the signal power / interference power calculation unit;
A wireless device comprising:
The radio apparatus according to claim 1, wherein the signal power / interference power calculation unit calculates a value obtained by subtracting noise power from reception power in the predetermined reception signal as the signal power.
A frequency offset correction unit for correcting the frequency offset of the received signal is provided.
The signal power / interference power calculation unit is a predetermined reception signal subjected to automatic gain control, and the signal power and the interference using the predetermined reception signal whose frequency offset is corrected by the frequency offset correction unit. The radio apparatus according to claim 1, wherein power is calculated.
The radio apparatus according to claim 1, wherein the signal power / interference power calculation unit calculates a sample variance of the predetermined received signal as the interference power.
The radio apparatus according to claim 1, wherein the signal power / interference power calculation unit calculates an unbiased variance of the predetermined received signal as the interference power.
3. The radio apparatus according to claim 2, wherein the signal power / interference power calculation unit sets a value obtained by dividing the interference power by the number of symbols of the predetermined reception signal as the noise power.
An A / D converter for converting a received signal from an analog signal to a digital signal;
An input bit width adjustment unit that extracts a bit width smaller than the bit width of the digital signal from the input signal;
With
The signal power / interference power calculation unit calculates the signal power and the interference power using a digital signal extracted by the input bit width adjustment unit, which is a predetermined reception signal subjected to automatic gain control. The wireless device according to claim 1, wherein:
The radio apparatus according to claim 7, wherein the input bit width adjustment unit determines a bit position to be extracted based on an amplitude of the predetermined reception signal.
A desired wave / interference wave level calculation unit for calculating a desired wave level and an interference wave level at a predetermined measurement location,
The AGC control unit controls the gain of the variable amplifier based on the gain control value so that the power of the received signal converges to the convergence value of the automatic gain control;
The desired wave / interference wave level calculation unit includes the gain control value, the convergence value of the automatic gain control, the signal power, and the gain from the predetermined measurement location to the signal power / interference power calculation unit. The radio apparatus according to claim 1, wherein the desired wave level is calculated.
In a SIR measurement method for measuring SIR using a known predetermined received signal,
An AGC control step for automatically gain-controlling the received signal;
A signal power / interference power calculation step of calculating signal power and interference power using a predetermined received signal that has been automatically gain-controlled in the AGC control step;
Measuring the SIR based on the signal power and the interference power;
SIR measurement method comprising:
The signal power / interference power calculation step calculates a variance of the predetermined reception signal subjected to the automatic gain control as interference power, and subtracts noise power from the reception power in the predetermined reception signal subjected to the automatic gain control. The SIR measurement method according to claim 10, wherein the SIR measurement method is calculated as the signal power.