JPH0795253A - Differential offset frequency detection method - Google Patents

Abstract

(57) [Summary] [Purpose] Fading phase characteristics do not follow a uniform distribution,
Even if the transmission signal is exposed to the phase change due to irregular disturbance,
Detect accurate offset frequency. A receiver that receives a transmission signal in which a carrier wave of a predetermined frequency is modulated by a digital signal and demodulates based on a reference carrier wave signal detects a difference between the predetermined frequency and the frequency of the reference carrier wave signal as an offset frequency. The phase ψ (t) of the demodulated digital signal is detected, and d ² ψ (t) / d ² t
When the value of is approximately zero, the offset frequency is determined by the value of (1 / 2π) · dψ (t) / dt. Systems 71 and 72 for detecting the phase of a predetermined symbol in the time slot of the demodulated digital signal for each same time slot, and a system 7 for calculating the inter-symbol phase change amount of the phase for each same time slot.
3 and a system 74 that calculates the inter-symbol phase acceleration for each same time slot based on the phase change amount, and the phase change amount when the phase acceleration is the minimum is the offset frequency.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an offset frequency detection system used in general digital mobile communication systems and for detecting a difference in carrier frequency between transceivers.

[0002]

2. Description of the Related Art In a mobile communication system, the carrier frequency is not the same between the transmitter and the receiver because of differences in usage conditions such as temperature. Therefore, in the mobile communication system, the difference in carrier frequency between the transceiver and
That is, it is general to detect the offset frequency by some method and perform the correction. It should be noted that this correction is called AFC (Automatic Frequency Control).

FIG. 1 shows a simple conceptual diagram of such a principle. A transmitter Tx modulates a carrier wave having a carrier frequency fc with an information signal s (t) and outputs it.
This transmission output passes through multiple propagation paths, and is therefore subjected to Rayleigh fading including disturbance noise (thermal noise, etc.) and a carrier error (offset frequency f0) with respect to the carrier frequency fc of the transmitter is generated. Reach Rx. The receiver Rx receives the transmission output and performs demodulation with the reference carrier signal having a frequency obtained by adding the carrier frequency fc of the transmitter to the antiphase component -f0 of the offset frequency f0 of the carrier frequency fc of the transmitter. To compensate for the difference.

In digital mobile communication, this offset frequency is detected based on received digital data in the base band region. Generally, the offset frequency is detected using known data between the transmitter and the receiver (for example, a sync symbol in a block called a preamble added to the head of each slot as shown in the figure).

FIG. 2 is a block diagram of a general receiver that performs such offset frequency detection. In the figure,
An RF (high frequency) signal received from an antenna is subjected to signal processing such as frequency conversion, sampling, and quantization in the down-conversion / sampling / quantization unit 1 to drop it to the baseband. The signal-processed baseband signal is subjected to preprocessing for timing extraction in the symbol rate conversion unit 2 and supplied to the mixer 3. The output signal of the mixer 3 is supplied to a symbol sampler 5 for sampling each symbol in the slot after passing through a pulse shaping filter 4 which is a bandpass filter for removing noise. The output signal of the symbol sampler 5 is supplied to a decoding system (not shown) as a demodulation signal.

The preprocessed output of the symbol rate conversion section 2 is also supplied to the timing extraction section 6. The timing extraction unit 6 estimates frame synchronization based on the pre-processed output, establishes the synchronization, generates the sample timing in the symbol sampler 5, and supplies this to the symbol sampler 5. The symbol sampler 5 performs sampling processing based on the given sample timing.

The output signal of the symbol sampler 5 is also supplied to the AFC system 7 for detecting the offset frequency. In the AFC system 7, first, the synchronization symbol sampler 7a extracts the preamble from the supply signal, the angle deviation estimator 7b detects the phase shift of the carrier (phase shift between transmission and reception), and the frequency deviation estimator The detected phase deviation is converted into a frequency deviation by 7c, and the converted output is supplied to the mixer 3 through the loop filter.

The mixer 3 multiplies the output baseband signal of the symbol rate converter 2 with the supplied signal as an AFC signal, that is, a signal corresponding to the offset frequency. Thereby, the carrier frequency deviation between the transmitter and the receiver is corrected. It should be noted that Δω corresponds to the offset frequency. By the way, since the received digital data (baseband signal) is affected by the above fading, it is necessary to devise a method for detecting the offset frequency.

[0009] For example, International Publication No. WO91 / 2014
In the method described in the specification of the patent application relating to No. 0, the offset frequency f _offset is obtained using the phase angle obtained by adding the differences between the adjacent data of the received complex sync signals. If you write this with a mathematical formula,

[0010]

[Equation 1]

[0011]

[Equation 2]

[0012] According to this, the value of the offset frequency is detected by using the raw data that has been affected by fading. However, because of the statistical nature of the fading phase (where [-π, π] is uniformly distributed), the sum (y) of the differences in the above cancels the phase component deteriorated by fading and only the offset frequency component is cancelled. Is being extracted. If the fading phase distribution is well and uniformly distributed within a limited number of samples, this method is acceptable, but this is not always the case. In this case, an error occurs in the estimation of the offset frequency, and the BER
(Bit Error Ratio) deteriorates, which is not desirable.

[0013]

SUMMARY OF THE INVENTION The present invention has been made in view of the above-mentioned points, and an object of the present invention is whether or not the fading phase characteristics follow a uniform distribution, and An object of the present invention is to provide an offset frequency detecting method capable of detecting an accurate offset frequency even when a transmission signal is exposed to a phase change due to the irregular disturbance of.

[0014]

An offset frequency detecting method according to the present invention is a receiver for receiving a transmission signal in which a carrier having a predetermined frequency is modulated by a digital signal and demodulating the digital signal based on a reference carrier signal. An offset frequency detecting method for detecting a difference between the predetermined frequency and the frequency of the reference carrier signal as an offset frequency, wherein a phase ψ (t) of a demodulated digital signal is detected, and d ² ψ (t) / At the time when the value of d ² t becomes substantially zero, (1 / 2π) · dψ (t) / d
The offset frequency is determined by the value of t.

The offset frequency detection circuit according to the present invention is a receiver for receiving a transmission signal in which a carrier having a predetermined frequency is modulated by a digital signal and demodulating the digital signal based on a reference carrier signal.
An offset frequency detection circuit for detecting a difference between the predetermined frequency and the frequency of the reference carrier signal as an offset frequency, the symbol phase detection detecting a phase of a predetermined symbol in a time slot of a demodulated digital signal for each time slot. Based on the phase change amount calculated by the phase change amount calculation unit, and a phase change amount calculation unit that calculates the phase change amount between the symbols of the phase detected by the symbol phase detection unit for each time slot. And a phase acceleration calculating means for calculating the phase acceleration between symbols for each time slot, and the offset frequency is calculated by the phase change amount corresponding to the minimum phase acceleration calculated by the phase acceleration calculating means. It is characterized by determining.

[0016]

According to the offset frequency detecting method of the present invention,
The phase ψ (t) of the demodulated digital signal is detected, and d
^{At the} time when the value of ² ψ (t) / d ² t becomes substantially zero,
The offset frequency is determined by the value of (1 / 2π) · dψ (t) / dt. According to the offset frequency detection circuit of the present invention, the phase of the predetermined symbol in the time slot of the demodulated digital signal is detected for each time slot, and the phase change amount between the symbols of the detected phase is calculated for each time slot. The phase acceleration between symbols is calculated for each time slot based on the calculated phase change amount, and the offset frequency is set by the corresponding phase change amount when the calculated phase acceleration is minimum.

[0017]

The present invention will be described in detail below with reference to the drawings. First, the present invention is to improve the AFC block 7 shown in FIG. Before explaining the circuit configuration of this AFC system, the principle of the offset frequency detection method, which is the core of the present invention, will be described using mathematical expressions.

Now, assuming that the offset frequency is f0 [Hz], the signal r (t) received by the receiver includes fading effects.

[0019]

[Equation 3]

Can be expressed as Where s (t) is the sync signal, ρ (t) is the Rayleigh distribution function, and φ (t) is the uniform distribution function.

[0021]

[Equation 4]

Is the complex gain due to Rayleigh fading. This allows

[0023]

[Equation 5]

It is However, although s (t) is assumed to be a real signal, generality is not lost. Here, if the phase signal of the reception sink is ψ (t),

[0025]

[Equation 6]

The linear phase (2πf0t) corresponding to the offset frequency and the random phase φ due to fading
(T) (However, the distribution follows a uniform distribution [−π,
π]). The point of this system is to estimate the offset frequency f0 by paying attention to the equation (5).

From equation (5),

[0028]

[Equation 7]

And becomes 2πf0 from the offset frequency
Note that it is the sum of (= const.) And the time change rate of the random phase. if,

[0030]

[Equation 8]

If

[0032]

[Equation 9]

Therefore, the offset frequency is

[0034]

[Equation 10]

Is calculated as

[0036]

[Equation 11]

The above condition is satisfied by observing within a certain time because φ (t) is random. To find that time, for example,

[0038]

[Equation 12]

It suffices to adopt the time at which And

[0040]

[Equation 13]

It suffices to detect as the offset frequency. As described above, the phase of the received signal is affected by fading, and therefore, the phase according to the random uniform distribution is mixed in addition to the offset frequency. Therefore, this method is to detect the offset frequency by finding a point where the random phase change due to fading becomes zero in an instantaneous value.

FIG. 3 is a block diagram showing an offset frequency estimating circuit to which the offset frequency detecting method is applied, and the same parts as those in FIG. 2 are designated by the same reference numerals. In FIG. 3, the sampling output from the symbol sampler 5 is given the phase normalization coefficient {ak: k = 0,1,2, ..., N} corresponding to each symbol in the preamble in the preamble phase normalization circuit 71. And is supplied to the delay circuit 72 and the phase change amount calculation circuit 73.
The delay circuit 72 delays the input symbol by one symbol and sends it to the phase change amount calculation circuit 73. The phase change amount calculation circuit 73 calculates the phase change amount between each symbol based on the output of the preamble phase normalization circuit 71 and the output of the delay circuit 72 {b (k): k = 0,1,2, ... , N-
1} [rad / sec]. Each of the obtained phase change amounts is sent to the phase acceleration calculation circuit 74 and the offset frequency determination circuit 75. Phase acceleration calculation circuit 74
On the basis of the transmitted phase change amount, the degree of change between these change amounts, that is, the phase acceleration is {c
(K): k = 1, 2, 3, ..., N-1} [rad / sec ² ]
Ask as. The offset frequency determination circuit 75 selects, as the offset frequency f0, the phase change amount b (k) corresponding to the smallest value of the obtained phase acceleration. The selected offset frequency f0 is applied to the AFC signal e ^-j2 π in the next stage offset correction circuit 76.
^It is adopted for ^{(f0 / fsym) t} , and this AFC signal is supplied to the mixer 3. The mixer 3 multiplies the received data on which the offset frequency from the symbol rate conversion unit 2 is added by the AFC signal. As a result, the data from which the variation due to the offset frequency has been removed can be sent to the pulse shaping filter 4.

FIG. 4 shows a more specific configuration example of the offset frequency estimating circuit block, and FIG.
The same reference numerals are given to the portions equivalent to, and the number of N is 8. In FIG. 4, the preamble phase normalization circuit 71 and the delay circuit 72 are composed of coefficient giving sections 710 to 718 and 1-symbol delay sections 721 to 728, and preamble symbol data is input to the coefficient giving section at every 1-symbol delay. The configuration is as follows.

The output of each coefficient assigning unit outputs the complex conjugate computing units 7300 to 730 in the phase change amount calculating circuit 73.
It becomes 7 inputs. The output of each complex conjugate calculation unit is the multiplication unit 7
It becomes one of the multiplication inputs of 310 to 7317. To the other multiplication input of each multiplication unit, the output of the coefficient addition unit based on the symbol data one symbol before that of the one multiplication unit is directly introduced. These multiplying units multiply one and the other multiplying input and send it to the tan ⁻¹ computing units 7320 to 7327. The tan ⁻¹ operation unit outputs its operation output b (k) {k =
Each of 0, 1, 2, ..., 7} is sent to the phase acceleration calculation circuit 74 and the offset frequency determination circuit 75 as a phase change amount signal between symbols.

In each phase velocity calculation circuit 74, subtraction units 7401 to 7407 first obtain the arithmetic output from the tan ⁻¹ arithmetic unit, that is, individual differences in the inter-symbol phase change amount, and further the absolute values of these differences. Is calculated by the absolute value calculation units 7411 to 7417, and each of the calculated values c (k) {k = 1, 2, 3, ..., 7} is sent to the offset frequency determination circuit 75 as a phase acceleration.

The offset frequency judging circuit 75 is a minimum value judging unit 7 for judging the smallest of the input signal values.
51, the minimum value indicating signal sw indicating the smallest phase acceleration c (k) is obtained. The determination circuit 75 also causes the selection unit 752 to selectively output one of the phase change amount signals b (k) from the phase change amount calculation circuit 73 to the offset correction circuit 76 according to the instruction signal sw. For example, if the instruction signal sw indicates k = 5, the phase change amount signal b (5) of the phase change amount calculation circuit 73 is selectively output.

In this way, the offset frequency is estimated from the amount of phase change when the instantaneous value of the phase acceleration (angular acceleration) is closest to zero. On the other hand, when the principle of the present system as described above is depicted by a diagram, it becomes as shown in FIG. 5, for example. In FIG. 5, the vertical axis F represents the frequency [Hz] and the horizontal axis t.
× 1/3000 indicates [sec], and the characteristic curve in the graph is r (n) = {f _sym / (2π)} · ar
g [y (n) / y (n-1)] is shown. The symbol frequency f _sym is 3000 Hz, and the moving body velocity is 70
Km / h. The carrier frequency is 150
It is calculated as kHz.

In this characteristic curve, the present system adopts the value near the most dense data, that is, the value indicated by the dotted line in the figure, as the offset frequency. Therefore, in this case, it is possible to accurately obtain the offset frequency of 350 Hz. For reference, FIG. 6 shows a simulation result of offset frequency estimation under fading when the offset frequency is 350 Hz and the carrier frequency fc = 150 MHz according to the above-described embodiment.

In FIG. 6, the vertical axis F indicates the frequency [Hz].
The horizontal axis v represents the speed [Km / h] of the receiver which is a moving body, and the dependence on the speed is also shown. From this, it can be seen that the same or higher effect can be expected as compared with the conventional method. Although the AFC signal is applied in the next stage mixer 3 of the symbol rate converter 2 in the above embodiment, it may be applied in the down conversion block 1 to adjust the center frequency. It suffices if the automatic frequency control is performed by the AFC signal thus obtained. Further, in the above embodiment, the symbol number N for phase detection is set to 8. However, it is needless to say that the number is not limited to this number and can be changed appropriately.

[0050]

As described above in detail, according to the offset frequency detecting method of the present invention, the phase ψ (t) of the demodulated digital signal is detected and the value of d ² ψ (t) / d ² t is detected. At the time when is almost zero, (1 / 2π) · dψ (t) /
The offset frequency is determined by the value of dt. Further, according to the offset frequency detection circuit of the present invention, the phase of a predetermined symbol in the time slot of the demodulated digital signal is detected for each time slot, and the phase change amount between the symbols of the detected phase is calculated for each time slot. The phase acceleration between symbols is calculated for each time slot based on the calculated phase change amount, and the offset frequency is set by the corresponding phase change amount when the calculated phase acceleration is the minimum. .

As described above, in the differential offset frequency detection method of the present invention, the offset frequency is estimated from the phase change amount when the instantaneous value of the phase acceleration (angular acceleration) is closest to zero. Even if the phase characteristic of fading does not follow a uniform distribution, and also when the transmission signal is exposed to a phase change due to other irregular disturbance, the accurate offset frequency can be detected.

[Brief description of drawings]

FIG. 1 is a diagram for explaining a general principle of detecting an offset frequency and correcting the offset frequency.

FIG. 2 is a block diagram of a general receiver that detects an offset frequency according to the principle of FIG.

FIG. 3 is a block diagram showing an offset frequency estimation circuit of a receiver to which the offset frequency detection method of the present invention is applied.

FIG. 4 is a circuit diagram showing a more specific configuration example of an offset frequency estimation circuit block in FIG.

FIG. 5 is a diagram for describing and explaining the principle of offset frequency detection of the present invention from the other side.

FIG. 6 is a reference characteristic diagram for confirming the operation effect of the offset frequency detection of the present invention.

[Explanation of symbols for main parts]

1 Down-conversion / sampling / quantization unit 2 Symbol rate conversion unit 3 Mixer 4 Pulse shaping filter 5 Symbol sampler 6 Timing extraction unit 7 AFC system 7a Synchronous symbol sampler 7b Angle deviation estimation unit 7c Frequency deviation estimation unit 7d Loop filter 7A AFC system 71 Preamble Phase Normalization Circuit 72 Delay Circuit 73 Phase Change Amount Calculation Circuit 74 Phase Acceleration Calculation Circuit 75 Offset Frequency Judgment Circuit 76 Offset Correction Circuit 710-718 Coefficient Assigning Section 721-728 1 Symbol Delay Section 7300-7307 Complex Conjugate Computation Section 7310 -7317 Multiplier 7320-7327 tan- ¹ Calculator 7401-7407 Subtractor 7411-7417 Absolute value calculator 751 Minimum value determiner 752 Selector

Claims (4)

[Claims] 1. A receiver for receiving a transmission signal in which a carrier wave having a predetermined frequency is modulated by a digital signal and demodulating the digital signal based on a reference carrier wave signal, wherein the predetermined frequency and the frequency of the reference carrier wave signal An offset frequency detecting method for detecting a difference as an offset frequency, wherein a phase ψ (t) of a demodulated digital signal is detected, and d
^{At the} time when the value of ² ψ (t) / d ² t becomes substantially zero,
An offset frequency detecting method, wherein the offset frequency is determined by a value of (1 / 2π) · dψ (t) / dt. 2. A receiver for receiving a transmission signal in which a carrier wave having a predetermined frequency is modulated by a digital signal and demodulating the digital signal based on a reference carrier wave signal, wherein the predetermined frequency and the frequency of the reference carrier wave signal An offset frequency detection circuit for detecting a difference as an offset frequency, which detects a phase of a predetermined symbol in a time slot of a demodulated digital signal for each time slot, and a symbol phase detection means for detecting the phase. Phase change amount calculation means for calculating the phase change amount between the symbols of the phase for each time slot, and phase acceleration between symbols for each time slot based on the phase change amount calculated by the phase change amount calculation means. A phase acceleration calculating means for calculating Offset frequency detecting circuit, characterized in that determining the offset frequency by the phase change amount corresponding to the time phase acceleration calculated by the acceleration calculation means is minimum. 3. The offset frequency detection circuit according to claim 2, wherein the predetermined symbol is a synchronization symbol existing in a preamble in the time slot. 4. A receiver characterized in that automatic frequency control is performed according to an offset frequency detected by the method according to claim 1.