JP2570158B2 - Fading simulator - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a versatile fading simulator capable of simulating a communication path in consideration of a shadowing correlation between an interference wave of a desired wave and an N wave.

[0002]

2. Description of the Related Art The following three phenomena can be cited as factors that characterize a radio wave propagation environment in land mobile communication.

[0003] 1. 1. A signal in which a wave having a difference in propagation distance on the order of the carrier wavelength is received, and the reception level greatly fluctuates when the transmitting station or the receiving station moves on the order of the carrier wavelength (Rayleigh fading). 2. When a wave whose propagation distance difference is greater than or equal to the reciprocal order of the signal bandwidth is synthesized, a phenomenon occurs in which ghost or intersymbol interference causes distortion (frequency selective fading). Fluctuations in the received signal level (shadowing, short-term median fluctuations) due to the signal being blocked by buildings and trees around the transmitting or receiving station In land mobile communications, these phenomena are extremely complex, Complex propagation models need to be considered. In order to simulate such a propagation environment, one that considers frequency selective fading and Rayleigh fading (for example, Ichiro Tsujimoto, “Fading Simulator”, Japanese Patent Application No. 2-86352) is known. on the other hand,
As shown in Fig. 2, a simulator that takes into account the effect of shadowing uses data on fluctuations in the received signal level due to shadowing obtained by conducting experiments to actually emit radio waves, and uses this value to estimate Rayleigh fading. Known simulating configurations are known (Toshihito Kanai, Masanori Takeji, Seiji Kondo, Nozuna Furuya, "High-speed Handoff Method Experimental System for Digital Mobile Communications," IEICE, Technical Committee on Wireless Communication Systems Report, RCS89
-37, 1989).

[0004]

As described above, only a fading simulator using data based on experiments is known as a fading simulator in consideration of shadowing, and its versatility is extremely low. Further, in a cellular system, a mobile station receives radio waves from a plurality of base stations existing in various directions due to radio wave overreach or the like. Because shadowing is caused by buildings and trees around the transmitting and receiving points, radio waves from base stations in the same direction as the receiving point have highly correlated shadowing, and have different directions. It is thought that low-correlation shadowing has occurred in the radio wave from the base station in the above, and in order to obtain a highly versatile simulator, it is necessary to consider the shadowing correlation.

[0005]

SUMMARY OF THE INVENTION In a fading simulator according to the present invention which handles N + 1 waves composed of a desired wave and N interference waves, a) N + 1 pieces of independent white normal random number processes are generated. N is a natural number) and b) outputs from the (N + 1) normal random number generation circuits are input, and the respective input signals are weighted and synthesized to generate N + 1 correlated random number sequences. C) N + 1 low-pass filter groups each having N + 1 correlated random number sequence groups obtained from the weighting synthesis circuit, and d) the desired wave and N
N + that stores the variance of the short-term median variation for a wave
One first storage circuit group; e) N + 1 multiplier groups for multiplying a value stored in the first storage circuit group by an output of the N + 1 low-pass filter groups; A variable conversion circuit for performing variable conversion on the outputs of the N + 1 multiplier groups; g) N + 1 second storage circuit groups for storing long-term median values for the desired wave and the N-wave; ) N +
N + 1 additions for adding the long-range median values of the desired wave and the N-wave stored in the N + 1 second storage circuit groups to the respective outputs of one variable conversion circuit group Instrument group.

In the normal random number generating circuit group of the fading simulator according to the present invention, a) the independent white normal random number processes generated from the (N + 1) normal random number generating circuit groups have the same variance. I have.

In the weighting synthesis circuit of the fading simulator of the present invention, a) an input signal R _i for N + 1 (N is a natural number) input signals, R _i , i = 1, 2 _,. , (I = 1, 2,..., N +
1) a storage circuit group 1 for storing a first weighting coefficient group a _{i, i} , and b) the first weighting coefficient group a _{i, i} and the input signal R _i , (i = 1,2) ,..., N + 1) as inputs.
ti _{i, i} and c) the ith (i = 1, 2,..., N +
N + 1−i second weighting coefficient groups a _{i, j} , (j = i + 1, i + 2,..., N) for the input signal R _i of 1)
+1), d) the i-th (i = 1, 2,..., N + 1) input signal R _i and the second weighting coefficient group a _{i, j} , (J = i + 1, i + 2,
.., N + 1) as inputs.
ti _{i, j} and e) k of the (N + 1) first multiplier circuits.
(K = 2, 3,..., N + 1)
lti _{k, k} ) and Mu of the second multiplication circuit group
til _{n, k} , (n = 1, 2,..., k−1).

In the weighting coefficient group of the weighting synthesis circuit of the fading simulator of the present application, the normalized correlation coefficient between the i-th output and the j-th output among the weighting synthesis circuit outputs is ρ _{i, j} (= ρ _{j, i} ), when each root mean square is 1.0,

[0009]

(Equation 2)

B) j is fixed in the order of j = 3, 4,..., N + 1,

[0011]

(Equation 3)

In accordance with the [0012], i = 2,3, ···, recurrently determined for the j-1, _further, a _j, the _j

[0013]

(Equation 4)

A _{i, j} , j =
1, 2,..., N + 1 i = 1, 2,.

[0015]

It is known that if the median of the short interval of the reception level is R, R fluctuates in accordance with a lognormal distribution due to shadowing (for example, supervised by Moriya Kuwahara, car phone, published by the Institute of Electronics, Information and Communication Engineers). , Corona, 1985).
here,

[0016]

(Equation 5)

X is a random variable that follows a normal distribution with a mean of zero. However,

[0018]

[Outside 1]

Is the median of R.

In the present invention, a variable representing shadowing is converted into a variable as shown in Equation 1 and considered as a normally distributed variable.
In consideration of the case where there are N + 1 waves of the interference wave of the desired wave and the N wave, and in order to simulate shadowing for each wave, first, the N + 1 independent averages are zero and the variance is one. White normal random numbers, X ₁ , X ₂ , ...,
Generate X _{N + 1} . Further, in order to consider the correlation of each shadowing, X ₁ , X ₂ ,..., X _{N + 1}
The linear combination, mean zero, shadowing Y ₁ which are variables respectively converted a correlation variance 1, Y _2, ···, Y
Consider finding _{N + 1} . Here, assuming that the correlation coefficient between the i-th wave and the j-th wave of the N + 1 wave is ρ _{i, j} , ρ _{i, j} = ρ _{j, i} (2) holds.

[0021]

(Equation 6)

It is sufficient to consider the equation The problem is that in this equation, a _{i, j} is determined so as to satisfy the condition of <Y _i · Y _j > = ρ _{i, j} (4) <Y _i · Y _i > = 1.0 (5) It comes down to what you can do. Here, <Y _i · Y _j > indicates the average of Y _i and Y _j . If a _{i, j} can be determined, variable transformed correlated shadowing can be obtained. Equations (3) and (5), X ₁ + X ₂ ,.
・, X _{N + 1}

[0023]

(Equation 7)

Is obtained. Equations (3) and (4)
And X ₁ , X ₂ ,..., X _{N + 1} ,

[0025]

(Equation 8)

Is obtained. Here, from equation (6), a _1,1 = 1.0 for i = 1 (8)

[0027]

(Equation 9)

Is obtained. From equation (9),

[0029]

(Equation 10)

Is required. In Expression (10), the sign of ± can take different values, but a positive value is used here for simplicity. On the other hand, from equation (7),

[0031]

[Equation 11]

Is obtained. Further, from equation (12),

[0033]

(Equation 12)

Is obtained. From equations (6) to (13),
a _{i, j} can be obtained recursively according to the flow shown in FIG. X ₁ , using the coefficients given by a _{i, j}
By weighting and combining X ₂ ,..., X _{N + 1} ,
Given correlation coefficient ρ _{i, j} , i, j = 1,2, ...,
Has a correlation defined by N + 1, obtained mean zero, the dispersion 1 normal distribution variable Y _1, Y _2, · · ·, a Y _{N + 1.} Y ₁ , Y ₂ , and Y _{N + 1} obtained as described above are converted into variables that simulate actual shadowing. That is, the variance σ _i and the median of the desired wave to be simulated and the (N + 1) -wave interference wave are set to meet the simulation conditions

[0035]

[Outside 2]

Are multiplied and added, respectively, and the inverse transformation of Equation 1 is performed. This operation

[0037]

(Equation 13)

Can be realized. As described above, the shadowing R _i , i having the cross-correlation with respect to any combination of the desired desired wave and the N + 1 interference wave
= 1, 2,..., N + 1.

[0039]

FIG. 1 shows an embodiment in which the present invention is applied in consideration of a desired wave and two interference waves. In the figure, 100
Is a random number generation circuit group which is a basis of shadowing for a desired wave and two interference waves, 101 is a weighting synthesis circuit, 10
2 is a filter group, 103 is a memory group for storing a long-range median value of shadowing for a desired wave and two interference waves,
104 is a variable conversion circuit group, 105 is an adder group, 106 is a multiplier group, 107 is a memory group for storing the variance of shadowing with respect to a desired wave and two interference waves, 108 is an output terminal group, and 109 is an instantaneous variation This is a fading simulator group. In the random number generation circuit group 100, the mean zero and the variance 1 as the basis of shadowing for the interference wave of the desired wave and the two waves are set.
Thus, three normal random numbers, each of which is independent, are generated. These random number groups are input to the weighting synthesis circuit 101, and given a predetermined correlation.

When considering a desired wave and two interference waves, the weighting synthesis circuit 101 can be configured as shown in FIG. 4, for example. 4, reference numerals 400 to 402 denote input terminals, 403 to 407 denote multipliers, 408 and 409 denote adders, 410 to 414 denote memories, and 415 to 417 denote output terminals. The random numbers output from the random number generation circuit group 100 and serving as the basis for shadowing the desired wave, the interference wave 1 and the interference wave 2 are input from the input terminals 400 to 402, respectively.
Weighted synthesis is performed as shown in FIG. Where ρ
_D-I1 , ρD _-I2 , and _ρI1-I2 are predetermined correlation coefficients generated in the desired wave-interference wave 1, the desired wave-interference wave 2, and the interference wave 1-interference wave 2, and are stored in the memory. In 410 to 413, ρ
_D-I1 , ρD _-I2 , _ρI1-I2 as variables

[0041]

[Equation 14]

The value given by is stored. The random number group generated from the random number generation circuit group 100
414, adders 408 and 409, and multipliers 403 to 40
6 and a desired correlation is added, and output from output terminals 415 to 417 as random numbers indicating correlated shadowing for the desired wave, the interference wave 1 and the interference wave 2, respectively.

In general, when N + 1 waves of an interference wave between a desired wave and N waves are considered, j, (j = 1, 2,..., N + 1)
The part of the weighting synthesis circuit for the input random number can be configured as shown in FIG. In FIG. 5, 50
0-1 to 500-j are input terminal groups, 501-1 to 501
-(J-1) and 502 are multipliers, 503 is an adder, 5
04-1 to 504-j are memories, and 505 is an output terminal. When N + 1 waves are considered, random numbers are generated from N + 1 random number generation circuits.
Signals from the 2,..., J-th random number generation circuit are input from input terminals 500-1 to 500-j, respectively. The input j random numbers are weighted _{using j} weight coefficients a _{1, j} , a _{2, j} ,..., A _{j, j} as shown in FIG. Here, a _{1, j} , a _{2, j} , ...
_{, A j} , j are given by the cross-correlation coefficient ρ _{i, j} , (i = 1, 2,..., J
If (j = 1, 2,..., N + 1) is given, it can be obtained recursively according to the flow shown in FIG.

The output from the weighting synthesis circuit 101 obtained as described above is a normal random number having a correlation of 0 mean and 1 variance. Here, in order to give an appropriate time variation to the shadowing of each wave, the weighting synthesis circuit 10
1 are smoothed by the filter group 102. Here, each transfer function H _j (ω) of the filter group 102 satisfies the condition of | H _j (ω) | ² = 1.0 (20), and by using a filter that does not cause a DC offset. , The variance and average of each random number are preserved. According to the output of the filter group 102, the memory group 1
07 is multiplied by the variance of shadowing for each wave, and the median of the long sections of each wave stored in the memory group 103 is added. Further, in the variable conversion circuit group 104, the variance and the median are The multiplied and added normal random numbers are converted into log-normally distributed random numbers based on equation (14) to obtain shadowing in which a desired cross-correlation occurs. By inputting the shadowing of each wave obtained as described above to, for example, the existing instantaneous fluctuation fading simulator group 109 (Rayleigh fading simulator) and giving instantaneous fluctuation based on the given shadowing, A precise propagation environment can be simulated.

The present invention can be implemented using software. Further, in the present embodiment, the weighting synthesis circuit 101, the filter group 102, the median value adder (the memory group 103 for storing the long-range median value of the shadowing with respect to the desired wave and the two interference waves, and the adder group 10)
5), the dispersion multiplying units (the memory group 107 and the multiplier group 106 for storing the dispersion of shadowing for the interference wave of the desired wave and the two interference waves) are all linear operations, so that they are equivalent even if they are switched in order. The effect of can be obtained.

[0046]

According to the present invention, it is possible to easily perform a simulation of shadowing occurring in land mobile communication in consideration of a cross-correlation for an arbitrary combination of a desired wave and an interference wave.

[Brief description of the drawings]

FIG. 1 is a system diagram showing an embodiment in which the first invention of the present application is applied when a desired wave and two interference waves are considered.

FIG. 2 is a system diagram showing a conventional technique.

FIG. 3 is a flowchart showing a flow for obtaining a weight coefficient of a weighting synthesis circuit 101;

FIG. 4 is a system diagram showing an embodiment of a weighting synthesis circuit when a desired wave and two interference waves are considered.

FIG. 5 is a system diagram showing a configuration of a weighting circuit for an output of a j-th random number generation circuit when N + 1 waves are generally considered.

[Explanation of symbols]

REFERENCE SIGNS LIST 100 random number generation circuit group serving as a basis for shadowing for desired wave and two interference waves 101 weighting synthesis circuit 102 filter group 103 Memory group for storing a long interval median value of shadowing for desired wave and two interference waves Reference Signs List 104 Variable conversion circuit group 105 Adder group 106 Multiplier group 107 Memory group for storing dispersion of shadowing with respect to desired wave and two interference waves 108 Output terminal group 109 Instantaneous fluctuation fading simulator group 200 Shadow obtained in measurement experiment Memory circuit group for storing the inching data 201 Rayleigh fading simulator group for simulating instantaneous fluctuation fading 202 output terminal group 400-402 input terminal 403-407 multiplier 408,409 adder 410-414 memory 415-417 output terminal 500- 1 to 00-j input terminal group 501-1~501- (j-1), 502 multiplier 503 adder 504-1~504-j memory 505 output terminal

Claims (4)

(57) [Claims] 1. A fading simulator that handles N + 1 waves consisting of a desired wave and N interference waves, wherein: a) N + generating independent white normal random number processes;
One (N is a natural number) random number generation circuit group; and b) the outputs from the N + 1 normal random number generation circuit groups are input, and respective input signals are weighted and synthesized to obtain N + 1
A weighted synthesis circuit that outputs a group of correlated random number sequences; c) N + 1 low-pass filter groups each receiving as input the N + 1 correlated random number sequence groups obtained from the weighted synthesis circuit; D) N + 1 first storage circuits that store the variance of the short-term median variation for the desired and N waves; e) the values stored in the first storage circuits and the N + 1
N + 1 multiplier groups for multiplying the outputs of the low-pass filter groups; f) N + 1 second storage circuit groups for storing long-range median values for the desired wave and the N-wave; g) N + 1 additions for adding the long-range median values of the desired wave and the N-wave stored in the N + 1 second storage circuit group to the respective outputs of the N + 1 multiplier groups And h) a variable conversion circuit for performing variable conversion on the outputs of the N + 1 adder groups. 2. The N + 1 normal random number generation circuit group according to claim 1, wherein: a) independent white normal random number processes generated from the N + 1 normal random number generation circuit groups have equal variances, respectively. The fading simulator according to claim 1, wherein: 3. The weighted synthesis circuit according to claim 1, wherein N + 1 (N is a natural number) input signals, R _i , i =
For 1, 2,..., N + 1: a) A storage circuit group for storing the first weighting coefficient group a _{i, i} for the input signal R _i (i = 1, 2,..., N + 1) B) the first weighting coefficient group a _{i, i} and the input signal R
_i (i = 1, 2,..., N + 1) as inputs, and N + 1 first multiplication circuit groups Multi _{i, i} ; c) the ith (i = 1, 2,..., N + 1) )), N + 1−i second weighting coefficient groups a for the input signal R _i
a second storage circuit group for storing _{i, j} , (j = i + 1, i + 2,..., N + 1); and d) the ith (i = 1, 2,..., N + 1) input signal R _i and the second weighting coefficient group a _{i, j} , (j = i
+ 1, i + 2, ··· , the second multiplication circuit group Multi _i to N + 1) inputs _{a, j} and, e) k -th (N + 1) of the first multiplier circuits (k =
2, 3,..., N + 1) multiplication circuit (Multi _{k, k} )
And Multi _{n, k} , among the second multiplication circuit group
(N = 1, 2,..., K−1)
3. The fading simulator according to claim 1, further comprising a group of adders. 4. The first and second weighting coefficient groups according to claim 3, wherein a _{i, j} , (i = 1, 2,..., N + 1 j =
i, i + 1,..., N + 1), the normalized correlation coefficient between the i-th output and the j-th output of the weighted synthesis circuit output is ρ _{i, j} (= ρ _{j, i} ), When the root mean square is 1.0, The fading simulator according to claim 1, 2 or 3, wherein: