JPS59194542A - Method for detecting interference quantity of same frequency - Google Patents

Abstract

PURPOSE:To detect quantitatively an interference quantity during communication by sampling an envelope square detecting output of a received wave and finding out low frequency and high frequency components to process these components. CONSTITUTION:A receiver 10 receives fading FD produced by a required wave D wave and an interference wave U wave, the intermediate frequency signal IF is amplified by an amplifier 11 and the envelope square of the received wave is detected by a detector 12. The output of the detector 12 is sampled and the average value is found out to derive a low frequency component having a frequency component almost equal to the FD frequency. The envelope square detected output is sampled by A/D converters 13, 14 respectively at time (t) and time t+DELTAt, a high frequency component to be varied at the frequency higher than the FD frequency is sampled at delay time DELTAt which makes the product of two sampled values zero and the square average of the difference of two sample values at the time t and t+DELTAt is found out to derive the high frequency component. After finding out the average of the low frequency components and the high frequency component, a computer 15 calculates an interference D/U ratio from these values.

Description

DETAILED DESCRIPTION OF THE INVENTION (Technical Field) The present invention relates to a method for detecting the amount of co-frequency interference in a wireless communication system.

(Background Art) In a conventional interference detection method, the transmitting side sends out radio waves by putting signals of different frequencies outside the communication band every round in order to detect interference. On the receiving side, a receiver capable of receiving multiple interference detection frequencies outside the communication band is prepared, and when an interference detection frequency other than that of the communication partner is received, it is determined that there is interference.

This system has the disadvantage that, in addition to the communication signal, an interference detection signal must be transmitted and received outside the communication band, so that several 710 devices must be attached to the transmitter.

In addition, conventional interference detection could only determine the presence or absence of interference, so if the amount of interference that falls below a certain communication quality occurs, in order to provide high-quality communication, it is necessary to quantitatively detect the amount of interference. Advanced control such as switching channels was not possible.

(Problems to be solved by the invention) In order to eliminate these drawbacks, the present invention is capable of quantitatively detecting interference during communication. In a wireless communication system whose envelope fluctuates at a frequency higher than the fading frequency due to modulation or carrier frequency offset, it is possible to sample the envelope output and calculate its average. By calculating the low JM component having a frequency component almost equal to the fading frequency, the values obtained by sampling the envelope square law detection output at times t and t + Δt can be regarded as the same value for fading, and the above fading frequency For high-frequency components that fluctuate at higher frequencies, the two sampling points in one rotation are assumed to be zero, l+!: Sampling is performed with a different delay time Δt, and the square mean of the difference between the sampling values of t(!: t+Δt This is a co-frequency interference detection method in which a high frequency component is obtained by taking the following, and the amount of interference is detected from these low frequency components and high frequency components.

(Structure and operation of the invention) The principle and examples of interference amount detection are shown below.

The re-modulated desired wave (D wave) and interference wave (U) are expressed by the formula (
1) and (2).

7'2 Here, E;+ (t) and E2(t) are the amplitudes of the D wave and the U wave, and the amplitudes have a Rayleigh distribution. ω1. ω2 and Δω1. Δω2 is the angular velocity of D wave and U wave, frequency deviation,
Pl and P2 are the frequencies of the D and U wave modulation signals. φ
is the phase difference between the U-wave modulated wave and the D-wave modulated wave, and θ is the phase difference between the U-wave modulated wave and the D-wave modulated wave.

When D and U are input to the receiver, the combined wave e becomes Equation (3).

e=e(10e2-CE,"Ct)+E:(t)+2'EICt)E2(
t)rmψ<t)〕”World life ψ(t)−(ω2−ωl)
t+φA =E1/E2
(5) IF ratio output 2 of the receiver that received the composite wave
When multiplicative detection is performed, the envelope source output becomes equation (6).

R<t>-E"<t>+g (t)→-2E+ (t)
E2(t)co! +ψ(t) (6) Receiver IF
Figure 1 shows the envelope detection output. Curve 1 is the detection output, curve 2 is E (t) + E"(t> is a low frequency component that changes on the order of fading determined by vehicle speed, radio wave wavelength, etc.). In the case of a 900 MHz band car phone, the vehicle speed is 4. OK.
When running at +n/h, the frequency is about 30Hz. 2 EI
(t)L(t)cosψ(t)uAs seen from equation (4), depending on the modulation degree, beat frequency (ω2-ω1)/2π, etc.! Generally, it has a high frequency component that is sufficiently higher than Rayleigh fading.')'I/(Angle modulation on the transmitting side'
! It can be constructed by offsetting the carrier by l-.

The present invention improves this frequency characteristic by using two AlD
High-speed A/I) sampling by F converter (1-J) with one high-speed converter, average of low frequency <E;
:(t) 10E:(t) >, average of high frequency components <EI
Shun, who found (t)g2(t)>, the amount of interference D to the operator
This is a device that adjusts the lU ratio.

The sampling method is as shown in Fig. 1, after sampling at time tKA/D converter I, the A/D is delayed by ΔtF￥f.
Sample with KmaII. Let each value be R(t
), R(t+Δt).

If the carrier wave is angularly modulated, or if the sampling interval of A/D converter I is set randomly, <part ψ> = OK'lZ
Taking this into consideration, sampling R(t) N times and taking the average value yields X.

N <R(t)>knee ΣR・(t) N, -summer <E>10<E>=X (8) Ku>
Hail the average.

Next, find the high frequency component.

If we take the root mean square of the difference between R(t) and R(with Δt), we get -<
((E,z −E”,,l + F2−E2A)+
2 (Esb2%ψ−El,,E2□fishψΔt)]2
(9) Here, the sampling delay time ΔL is
El+' that makes a huge difference! 'I for 2
#EIJ ”2 #E2.+ (10) For a residence ψ with a fast fluctuation period (cosψ・ωSψΔ〉fly o (1
If we choose Δt rather than 1), equation (9) becomes K such that 0<[7i! (t)-yx(t+Δt)]2>-4<
E”, ><E:>=y (12) The square ratio of the amplitudes of D and U (DlU ratio) is Fr=<p:
”,>i<E:n (13), we obtain equation (III) for F from equations (8), (+2), and (+3).

7”-2Kr+ 1 = O(1'4) Therefore r-, - to 1\, : to 2-1
(15) However, since =2X2/Y-1 is ≧1, equation (15) is not a real number, and the interference fluorescence p can be obtained.

Let us consider the delay time Δt.

Equation (14) for F was derived under the condition that Equations (10) and (11) hold when sampling is performed with a sampling delay time Δt. Therefore, formula (10)
, (11) does not hold, a PK error occurs.

Therefore, it is necessary to consider the influence of Δt on F, and to set the range of Δt to be .

Here, the following conditions are used. These conditions can be easily met in mobile communications.

As can be seen from equation (8), X = <R(t) is not affected by Δt.

Y is affected by Δt, and let Y at this time be Y'.
z2>+4<F::><E:n (1-2<Residence ψ・(欝φ
Δ〉)...(18) However, Z=E", -E:, +E
: -E:□Y formula (12) when there is no influence of ΔL and Δ
Influence of t: If the ratio of Y' equation (18) at a certain time is 1v,
Y'hak Z2n.

−8<g:><'g:><expletive ψ・possible ψΔ>term〃; Effect when 4t is large 7
5 can be done, so y=<z”>-t4kuE;n (747z, n
(19). Therefore, let's consider <Z2>
<z"> is 0 <J"n-8b; [F2(1-ρ:)-1-(1-ρ:
)] (20) However, b,”==<E:n/8.
ρ is the autocorrelation cousin of El (t) and F2(t), and ρ-・shi, (2π, f and Δ, (21) f
Is the fading frequency f77L−υ (speed)/λ (wavelength)? ? ? From equation (21), Δt affects Y in combination with f and In. <Z”> normalized by Y when there is no influence of Δt
/Y is shown in FIG. The influence of f-.Δt differs depending on F.The shadow # is large in 11 where pF is large.

Further, the influence of Δt on F is shown by a solid line in FIG.

The horizontal axis is true D/U=rtX, and the vertical axis is D/U2FM when there is an influence of Δt.

FM appears smaller as f7n·Δt increases, and its influence becomes particularly strong where F is large. r
In order to suppress the detection error to within 1 dB when t=15 dB or less, fm·Δt<0.03 must be satisfied. In addition, FIG. 3 was calculated as ρ1−ρ2.

When Δt is small, y = 4 <g: ><g:><1-2<□□□
ψ・part ψΔ>) (22). Therefore 〈fusa ψ・
- Consider φΔ〉.

<Fish ψ・〉ψΔ> is given by equation (23).

Kusaki ψ, animal ψ, l>--Jo (Z+), Jo (Z2)
+Σ[cos(2nθ2−2mθ+)J2n(Z2)J
2-(Z+)]-Σ[cos((2n+1)θ2-(
2m+1)θ1)J2n+1(Z2)12m+1(Zl
)]...(23) However, the second term P2-rn, P1, and the third term are (2n+
1) It is effective when P2=(2m+1) Pl, a constant of n, m (m=n)O) holds true.

However, θ+ "jan-' (5inP1Δt / (cns
P1Δt-1)) θ2-2-'((cOsθ-cos
(P2Δ to 10θ))/(s=θ−81n(Pzt10θ)
乃Z + "J-2'('-1-coSP+Δ't
,)-Δωl/PIZz 2J2 (] cmP2t
) Δω2/P21 ne 14 is 2〈focus ψ・fusa ψΔ〉
An example showing the relationship between and ΔL, Δω1” 1000 Hz
, Δω, , -1010H2, fg wave is a sine drum signal and P
1=200 fiz, P2=] 95~205
fiz lid, this is the case when it is morphed. 2〈House ψ・Fish ψ
Δ> decreases with Δt, but is not affected by changes in the frequency of P2, becomes O near Δt: 0.4 ms 11, and thereafter decreases with increasing amplitude. Generally, a modulated wave does not have a single frequency, but when the low frequency region of the frequency spectrum of the f harmonic is small relative to the degree of modulation, it exhibits the characteristics shown in the figure. In other words, if Δt is small, the fish ψ cos ψ
Δ〉 is close to [, but as Δt increases, it becomes smaller.The situation depends on the spectrum of the modulated wave and the degree of modulation;
l,! ,ru. .

As can be seen from equation (18), 2ku fish ψ・house ψΔ> is Δ
It represents the decrease in Y when t is small.

FIG. 5 shows the influence of Δt on F under the conditions shown in FIG. When Δt becomes smaller, F appears to be larger, and this effect is particularly large when F is small. F
The reason why Y appears to be large is because Y becomes small when Δt is small.

FIG. 6 shows an indoor experiment system for confirming the present invention, and the broken line portion is the interference amount detection unit of the present invention. In the figure, 3.6 is a standard signal generator for generating D waves and U waves, and 4 and 7 are fading simulators for creating pseudo propagation paths. 5 and 8 are attenuators, 9 is a combiner, 10 is a receiver, 11 is a filter, 12 is a detector, 13 and 14 are analog-to-digital converters (A7v converters), and 15 is an arithmetic unit.

FIG. 7 shows a flowchart of the arithmetic unit 15. The number of samples N1 and the delay time Δt are set, and i for counting sampling is set to i=1. The A/D converters 13 and 14 sample the sum of squares Σ(R
Take (t)-R(t+Δt))2. If i=N, further sampling is performed using an A/D converter.

When i -, N Vc, calculate X and Y.

If h2-1>o, the amount of interference r2 is calculated. Moshi 2-1 (
If O, F is not calculated, the value is returned to 1-=l, and the calculation of bvfr is started.

O and UW are received after being synthesized in a high-fidelity manner through all fading simulators and attenuators.

The receiver IF output is square-law detected, sampled by an A/D converter, and input to an arithmetic unit. The modulation signal (PN signal) has a maximum frequency deviation Δω1−Δω2−1, KIJz, D
, the experiment was conducted with the U wave carrier difference fb=O(H2).

Regarding the influence when the size difference Δt is large, Δt=0.4w,
Experiments were conducted with S constant and the fading frequency M2 varied. Figure 3 shows the measured values. Each plot point has sampling number N=1000, sampling frequency 18=
The average value of approximately 10 pieces of β data measured at 80 Hz is shown.

The amount of interference can be detected, and the theoretical value and the actual measurement 11σ are in good agreement. However, the actual measured value of Ft, which is larger, is around f, which is p larger than the theoretical value. This is due to the S/H problem of the measurement system, and the beat due to the geothermal noise of the interference beat is added to Y.
It is considered that the value of <Z">/Y shown by becomes smaller.

The S/N of the measurement system used is approximately 45 dB.

From the above, when fm・Δt increases, the actual measurement of D/U +w
rlMh,X(D D/U1M I't, 1: f:
＞small＜ Qb, and for F required from the system design,
Δt can be set.

Next, an experiment was conducted to examine the effect when Δt is small.

In this case, the fading frequency fm was kept constant at 20 Hz, a value was chosen where the influence of <Z2> due to m·Δt could be almost ignored, and experiments were conducted by varying Δt.

The results are shown in Figure 5. The measured values are in good agreement with the theoretical values9.
, Δt is small, the actual measured value rM appears larger than the true F4, but as Δt becomes larger, the error between the actual measured value and the established value becomes smaller and the influence of ΔL disappears.

, In order to keep the measurement error within 1 dB, it is necessary to set Δt > 0.3 mS (Effect of the invention) As explained above, the interference layer detection method V according to the present invention: can be detected constant, so the interference f
If the i Karuru setting value or more is reached, the i1k channel V (
-) On the other hand, there is an advantage in that high quality communication can be continued without any deterioration in call quality.

In a system where the wireless zone is set to a 3 to 5 small zone system to improve the usage rate of the frequency band, Mal ()! i
1. Because the number repetition distance can be reduced by 1'ill. :!
The usage rate of flv R4 has improved, and it is possible to be disappointed in J stomach people of 7/1.1 people valley 11j.

4.1 Figure 1 River's 1h゛] 4i's most self-explanatory 1 So II is the cri of the f anti-vessel output, and the evil 2 is Z'/Y and f
・A diagram showing the relationship of Δt. Figure 3 is a diagram showing the effect of the delay time Δt given by the interference effect VC when Δt is small.
1ba2〈4ψ・HouseφΔ〉and Δt, 13! Figure 1 shows the =j coefficient, and Figure 5 shows the J4 effect on interference detection when Δt is small.
7z4 is a sign of the shadow of time, 1 and 6 is the present invention.
Block diagram of 4M recognition/C presentation actual # device, Figure 7
7 is a diagram showing an operation flow of the arithmetic unit 15 in FIG. 6. FIG.

Patent applicant Nippon Telegraph and Telephone Public Corporation Patent agent Megumi Yamamoto - Curtain 1 Figure ground 2 concave fm・ΔL Tail A - Figure 0.0 (II!2 θ4 θ
6 0. f3/:θ delay MPJ A
L Crn Sec] 1f-5 diagram

Claims (1)

[Claims] When the desired wave and interference wave are received simultaneously, angle modulation or )E
Due to the offset of the transmission frequency, the envelope is the fading frequency! In a wireless communication system in which the frequency varies with the number of waves, the envelope square detection output of the receiving board is sampled and the average (1i:j) is taken, so that it has a frequency component approximately equal to the fading frequency. Find the low frequency component and output the squared envelope detection at time t.
The values sampled at and t + Δt can be considered to be the same value for fading, and for high frequency components that fluctuate at a frequency lower than the above fading frequency, the value of the two sampling gains ρ can be considered to be zero. By sampling at ΔL and taking the average of the difference between the sampled values of t and t + Δt, we obtain the Ichichu component.
Low frequency! Interference-1if by processing the spectral and high-frequency components
, ' A method for detecting the amount of same-frequency interference, which is characterized by detecting f.