This application claims the benefit of Taiwan application Serial No. 107106723, filed on Mar. 1, 2018, the subject matter of which is incorporated herein by reference.
BACKGROUND OF THE INVENTION
Field of the Invention
The invention relates to carrier frequency offset (CFO) estimation, and more particularly to CFO estimation that cancels echo in advance.
Description of the Related Art
Each of a transmitting end and a receiving end in a communication system is provided with at least one oscillation signal source (e.g., a quartz oscillator) for providing clock signals as reference for circuit operations thereof. A difference between the frequency of clock signals at a receiving end and that at a transmitting end is referred to as a carrier frequency offset (CFO), which may cause the receiving end to be unable to correctly read and parse signals from the transmitting end. To avoid the above issue, a receiving end needs to estimate the CFO so as to perform compensation accordingly.
FIG. 1 shows a block diagram of a conventional CFO estimation device. The CFO estimation device 100 includes a fourth-power circuit 110, a fast Fourier transform (FFT) circuit 120, a maximum peak frequency determining circuit 130 and a division circuit 140. The fourth-power circuit 110 performs a fourth-power calculation on a received signal R to generate a fourth-power calculation result R4. The FFT circuit 120 performs FFT on the fourth-power calculation result R4 to generate a spectrum Z of the fourth-power calculation result R4, as shown in FIG. 2. The maximum peak frequency determining circuit 130 then identifies a maximum peak PMAX from the spectrum Z, and outputs a peak frequency Ω corresponding to the maximum peak PMAX. The division circuit 140 divides the peak frequency Ω by four to obtain a calculation result, and uses the calculation result as an estimated CFO E. However, when the received signal R is transmitted to a receiving end through a multipath, the received signal R is frequently mixed with echo, such that the peak frequency Ω corresponding to the maximum peak PMAX of the spectrum Z significantly deviates from a value four times the actual CFO, causing inaccurate CFO estimation of the CFO estimation device 100.
SUMMARY OF THE INVENTION
Therefore, it is an object of the present invention to provide a carrier frequency offset (CFO) estimation device and a CFO estimation method for preventing inaccurate CFO estimation.
A CFO estimation device disclosed by the present invention includes: an echo cancelling unit, performing echo cancellation to generate an echo-cancelled signal; and a CFO estimating unit, performing CFO estimation according to the echo-cancelled signal to generate an estimated CFO.
A CFO estimation method further disclosed by the present invention includes: performing echo cancellation on a received signal to generate an echo-cancelled signal; and performing CFO estimation according to the echo-cancelled to generate an estimated CFO.
The above and other aspects of the invention will become better understood with regard to the following detailed description of the non-limiting embodiments. The following description is made with reference to the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of a conventional carrier frequency offset (CFO) estimation device;
FIG. 2 is a schematic diagram of an exemplary spectrum of a fourth-power calculation result;
FIG. 3 is a block diagram of a CFO estimation device according to an embodiment of the present invention;
FIG. 4 is a flowchart of a CFO estimation method according to an embodiment of the present invention;
FIG. 5 is a block diagram of an echo cancelling unit according to an embodiment of the present invention;
FIG. 6 is a flowchart of an echo cancelling method according to an embodiment of the present invention;
FIG. 7 is a block diagram of a CFO estimating unit according to an embodiment of the present invention;
FIG. 8 is a block diagram of a coefficient generating unit according to an embodiment of the present invention; and
FIG. 9 is a flowchart of a coefficient generating method according to an embodiment of the present invention.
DETAILED DESCRIPTION OF THE INVENTION
FIG. 3 shows a block diagram of a carrier frequency offset (CFO) estimation device 30 according to an embodiment of the present invention. The CFO estimation device 30 includes an echo cancelling unit 32 and a CFO estimating unit 34. FIG. 4 shows a flowchart of a CFO estimation method 40 according to an embodiment of the present invention. The echo cancelling unit 32 first performs echo cancelling on a received signal R to generate an echo-cancelled signal Rec (step S410). The CFO estimating unit 34 performs CFO estimation according to the echo-cancelled signal Rec to generate an estimated CFO E (step S420).
FIG. 5 shows a block diagram of an echo cancelling unit 32 according to an embodiment of the present invention. In this embodiment, the echo cancelling unit 32 is a whitening filter; however, the present invention is not limited thereto. The echo cancelling unit 32 includes a shift register 322, a multiplication unit 324, a coefficient generating unit 325, a summing unit 326 and a subtraction unit 328. FIG. 6 shows a flowchart of an echo cancelling method 60 according an embodiment of the present invention. The shift register 322 stores a plurality of symbols of the received signal R (step S610). For example, the shift register 322 includes registers 322-1 to 322-N, which receive symbols ri−1, ri−2, . . . and ri−N of the received signal R, respectively. The multiplication unit 324 includes multipliers 324-1 to 324-N, which multiply the symbols ri−1, ri−2, . . . and ri−N stored in the registers 322-1 to 322-N by coefficients w1 to wN, respectively, to obtain products (ri−1*w1), (ri−2*w2), . . . and (ri−N*wN) (step S620), wherein the coefficients w1 to wN are provided by the coefficient generating unit 325. In one embodiment, the coefficients w1 to wN may be a default value, e.g., 0, or be generated by the coefficient generating unit 325 according to the echo-cancelled signal Rec and the symbols ri−1, ri−2, . . . and ri−N stored in the shift register 322. The summing unit 326 then sums up the plurality of products above to obtain a product sum S (step S630). The subtractor 328 subtracts the product sum S from a selected symbol rsel of the received signal R to generate a symbol in the echo-cancelled signal Rec (step S640). The subtractor 328 may be implemented by an adder.
In one embodiment, the subtractor 328 is coupled to the register 322-N, and the selected symbol rsel is the symbol ri−N stored in the register 322-N, wherein the symbol ri−N is the symbol received earliest among the symbols ri−1 to ri−N, so as to cancel post-echo in the received signal. In another embodiment, the subtractor 328 is coupled to an input end of the shifter register 322 and the selected symbol is the next symbol rN, so as to cancel pre-echo in the received signal.
FIG. 7 shows a block diagram of a CFO estimating unit 34 according to an embodiment of the present invention. In this embodiment, the CFO estimating unit 34 is a fourth-power CFO estimator; however, the present invention is not limited thereto. The CFO estimating unit 34 includes a fourth-power circuit 342, a spectrum generating circuit 344, a maximum peak frequency determining circuit 346, and a division circuit 348. The fourth-power circuit 342 receives the echo-cancelled signal Rec from the echo cancelling unit 32, and performs a fourth-power calculation on the echo-cancelled signal Rec to generate a fourth-power calculation result Rec4. The spectrum generating circuit 344 performs FFT on the fourth-power calculation result Rec4 to generate a spectrum Zec of the fourth-power calculation result Rec4. The maximum peak frequency determining circuit 346 identifies a maximum peak PMAX from the spectrum Zec, and outputs a peak frequency Ω corresponding to the maximum peak PMAX. The division circuit 348 eventually divides the peak frequency Ω by four to obtain a calculation result, and uses the calculation result as an estimated CFO E.
The coefficient generating unit 325 updates the coefficients w1 to wN according to the echo-cancelled signal Rec and the symbols ri−1, ri−2, . . . and ri−N stored in the shift register 322. FIG. 8 shows a block diagram of a coefficient generating unit 325 according to an embodiment of the present invention. The coefficient generating unit 325 includes a conjugate unit 3252, a multiplication unit 3254 and an adding unit 3256. FIG. 9 shows a flowchart of a coefficient generating method according to an embodiment of the present invention. The conjugate unit 3252 includes conjugators 3252-1 to 3252-N, which respectively perform a conjugate operation on the symbols ri−1, ri−2, . . . and ri−N to generate conjugate symbols r*i−1, r*i−2, . . . and r*i−N (step S910). The multiplication unit 3254 includes multipliers 3254-1 to 3254-N, which respectively multiply the conjugate symbols r*i−1, r*i−2, . . . and r*i−N by the echo-cancelled signal Rec and a step size (not shown) to generate coefficient adjustment values a1 to aN (step S920). The step size is used for adjusting magnitudes of the adjustment values a1 to aN, e.g., 0.1, and may be designed according to actual requirements in practice. More specifically, the adjustment values a1 to aN get larger as the step size increases, and the adjustment amount of the coefficients w1 to wN also increases. The adding unit 3256 includes adders 3256-1 to 3256-N, which respectively add the adjustment values a1 to aN to the coefficients w1 to wN to generate updated coefficients w1′ to wN′ (step S930). The updated coefficients w1′ to wN′ are respectively multiplied with updated r*i, r*i−1, . . . and r*i−(N−1) in the shift register 322 to generate a next symbol in the echo-cancelled signal Rec.
In conclusion, the received signal R first passes through the echo cancelling unit 32 to undergo echo cancellation and becomes the echo-cancelled signal Rec, and then enters the CFO estimating unit 32 to undergo CFO estimation, that is, the input signal of the CFO estimating unit is the echo-cancelled signal Rec. Thus, the peak frequency Ω corresponding to the maximum peak PMAX of the spectrum Zec can be prevented from significantly deviating from four times the echo and thus from causing inaccurate CFO estimation of the CFO estimating unit 34. In other words, first performing echo cancellation on the received signal R by the echo cancelling unit 32 effectively enhances the accuracy of the CFO estimation device 30.
While the invention has been described by way of example and in terms of the embodiments, it is to be understood that the invention is not limited thereto. On the contrary, it is intended to cover various modifications and similar arrangements and procedures, and the scope of the appended claims therefore should be accorded the broadest interpretation so as to encompass all such modifications and similar arrangements and procedures.
Patent Application
20190273532
