FREQUENCY OFFSET ESTIMATION METHOD FOR OFDM-IM SYSTEM

Abstract

The invention provides a frequency offset estimation method for an OFDM-IM system. The method includes: S1. performing preliminary frequency offset compensation on a received signal subjected to non-uniform frequency offset by using a two-step method of: (1) resampling and down conversion; and (2) unified compensation for residual frequency offset ε, wherein in the step (2), a sum of energy of null sub-carriers is used as a cost function, an initial estimation value of ε is obtained by one-dimensional search, and the preliminary compensation is performed; S2. estimating positions of non-activated sub-carriers in the OFDM-IM system by using the signal subjected to the preliminary compensation; and S3. assigning certain weights to the estimated sub-carriers, adding energy of the estimated sub-carriers into the cost function according to different weights, obtaining a final estimation value of ε by the one-dimensional search performed on ε, and performing secondary compensation.

Description

BACKGROUND

Technical Field

The present invention relates to the fields of OFDM, orthogonal frequency division multiplexing with index modulation, digital communication and parameter estimation, and more particularly, to a frequency offset estimation method for an OFDM-IM system.

Description of Related Art

Orthogonal Frequency Division Multiplexing (OFDM) technology is a multi-carrier modulation technology employed in many existing wireless communication systems, and is most typically represented by Wi-Fi which is a wireless local area network technology under Standard IEEE 802.11, and 4G LTE which is a fourth generation mobile communication system that has been put into commercial use. OFDM has higher spectrum efficiency and data transmission rate, and is able to effectively confront selective fading of frequency domain of a channel and a multipath effect of the channel respectively by dividing broad-band signals into a series of narrow-band signals with nearly flat fading, and adding a cyclic prefix.

The OFDM technology has the disadvantages as follows: the OFDM is sensitive to frequency offset and phase noise; a peak-to-average power ratio (PAPR) is large; a complexity of a system may be increased by a load algorithm and an adaptive modulation technology, and complexities of a transmitter and a receiver may be increased by using the load algorithm and the adaptive modulation technology.

An Orthogonal Frequency Division Multiplexing with Index Modulation (OFDM-IM) technology is a novel OFDM technology proposed in the past two years. OFDM-IM refers to an idea of a spatial modulation technology and carries a part of transmission information by using a degree of freedom of positions of activated sub-carriers, and a receiving terminal obtains this part of information by detecting the positions of the activated sub-carriers. An OFDM-IM system is better in performance than a traditional OFDM system, and a spectrum efficiency of the system can be flexibly configured by selecting different numbers of activated sub-carriers.

The sub-carriers of the OFDM-IM employed by the present invention are partially activated, which means that in addition to initially set null sub-carriers, energy of non-activated sub-carriers may also be used. Therefore, the present invention optimizes estimation of Doppler frequency shift by using the non-activated sub-carriers.

SUMMARY

An objective of the present invention is to provide a frequency offset estimation method for an OFDM-IM system. A main idea of the present invention is to improve a performance of a frequency offset estimation by using null carriers of OFDM-IM. In the present invention, a number of sub-carriers of the OFDM-IM system is N, a MPSK modulation mode is used in the system, in addition to a pilot frequency, a guard interval and N_null preset null sub-carriers, the system contains a plurality of OFDM sub-blocks, and each block has N_total sub-carriers, wherein N_active sub-carriers are activated sub-carriers, a position and a symbol of the pilot frequency are known, and positions of the preset null sub-carriers are known.

The frequency offset estimation method for the OFDM-IM system provided by the present invention includes the following steps.

S1. Performing preliminary frequency offset compensation on a received signal subjected to non-uniform frequency offset by using a two-step method of: 1. resampling and down conversion to obtain a discrete baseband received signal z=Γ(∈(ε)y; and 2. unified compensation for residual frequency offset, a sum of energy of null sub-carriers being used as a cost function in the step 2:

$J_1\left(ϵ\right)=\sum _{m\in S_N}{f_m^H{\Gamma }^H\left(ϵ\right)z}^2$

wherein y is the discrete baseband received signal when the frequency offset is 0, ε is a residual normalized frequency offset, S_N is a null sub-carrier set, f_m is an m^th column of an inverse Fourier transformation matrix, Γ(ε) is a frequency offset matrix diag(1,e^j2πTcε, . . . , e^{j2πTc(k−1)ε}), j=√{square root over (−1)}, (.)^H is a conjugating operation, and T_c is a sampling interval; and

• • an initial estimation value {circumflex over (ε)}₁=arg min_εJ₁(ε) of ε is obtained by one-dimensional search performed on ε.

S2. Using a signal z₁ subjected to the preliminarily estimated compensation of ε, with specific calculation formula:

z ₁=Γ^H({circumflex over (ε)}₁)z;

and estimating positions of non-activated sub-carriers in the OFDM-IM system through the signal subjected to the frequency offset compensation, wherein sub-carriers meeting the following criteria are determined to be the non-activated sub-carriers:

• • {circle around (1)} a power P_null of the present sub-carriers is lower;
  • {circle around (2)} one of two adjacent sub-carriers with a higher power P_side is larger;
  • {circle around (3)} a specific value P_side/P_null is large enough; and
  • at most N_max sub-carriers meeting the above criteria are used in the following steps for the system.

S3. Sorting the estimated sub-carriers in a descending order of a size of P_side, assigning a larger weight to energy of the sub-carrier with a larger P_side, and forming the weights of each sub-carrier into a diagonal matrix W;

• • adding energy of the estimated sub-carriers into the cost function according to different weights, the cost function being modified as:

$J_2\left(ϵ\right)=\sum _{m\in S_N}{f_m^H{\Gamma }^H\left(ϵ\right)z}^2+\sum _{m\in \widehat{S_N}}{{\mathrm{Wf}}_n^H{\Gamma }^H\left(ϵ\right)z}^2$

• • wherein is an estimated non-activated sub-carrier set; and
  • obtaining a final estimation value {circumflex over (ε)}₂=arg min_εJ₂(ε) of ε by the one-dimensional search performed on ε, a signal subjected to final compensation being

$z_2={\Gamma }^H\left({\widehat{ϵ}}_2\right)z.$

Further, in the step S2, the positions of the non-activated sub-carriers are estimated through power detection by using the signal subjected to the preliminary compensation.

Further, in the step S3, the energy of the sub-carriers estimated to be non-activated sub-carriers is added into a final cost function according to different weights.

Compared with a traditional OFDM technology, the sub-carriers of the OFDM-IM employed by the present invention are partially activated, which means that in addition to initially set null sub-carriers, the energy of the non-activated sub-carriers may also be used. Therefore, the present invention optimizes estimation of Doppler frequency shift by using the non-activated sub-carriers.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a flow chart of a method according to the present invention.

FIG. 2 and FIG. 3 are respective trend charts of mean square errors of different {circumflex over (ε)}_norm changing with a signal-to-noise ratio.

DESCRIPTION OF THE EMBODIMENTS

The present invention is further described in detail hereinafter with reference to the accompanying drawings and the specific embodiments, the implementation and protection of the present invention are not limited to the accompanying drawings and the specific embodiments. It should be noted that all the processes which are not specifically described hereinafter can be implemented or understood by those skilled in the art with reference to the prior art.

The present invention provides an OFDM-IM method based on multiple modes, wherein 128 sub-carriers are provided in an OFDM-IM system, and a QPSK modulation mode is used in the system. In addition to a pilot frequency, a guard interval and 4 preset null sub-carriers, the system contains 28 OFDM sub-blocks, and each block has 4 sub-carriers, wherein 3 of the 4 sub-carriers are activated sub-carriers. A position and a symbol of the pilot frequency are known, and positions of the preset null sub-carriers are known. The specific implementation includes the following steps.

In step S1, preliminary frequency offset compensation is performed on a received signal subjected to non-uniform frequency offset by using a two-step method of: 1. resampling and down conversion to obtain a discrete baseband received signal z=Γ(ε)y; and 2. unified compensation for residual frequency offset. A sum of energy of null sub-carriers is used as a cost function in the step 2:

$J_1\left(ϵ\right)=\sum _{m\in S_N}{f_m^H{\Gamma }^H\left(ϵ\right)z}^2$

wherein y is the discrete baseband received signal when the frequency offset is 0, ε is a residual normalized frequency offset, S_N is a null sub-carrier set, f_m is an m^th column of an inverse Fourier transformation matrix, Γ(ε) is a frequency offset matrix diag(1,e^j2πTcε, . . . , e^{j2πTc(k−1)ε}), j=√{square root over (−1)}, (.)^H is a conjugating operation, and T_c is a sampling interval.

An initial estimation value {circumflex over (ε)}₁=arg min_εJ₁(ε) of ε is obtained by one-dimensional search performed on ε.

In step S2, positions of non-activated sub-carriers in the OFDM-IM system are estimated by using a signal z₁=Γ^H({circumflex over (ε)}₁)z subjected to the preliminarily estimated compensation, with criteria that:

• • {circle around (1)} a power P_null of the present sub-carriers is lower that an average value of a power of the non-activated sub-carriers;
  • {circle around (2)} one of two adjacent sub-carriers with a higher power P_side is larger than an average value of the power of the activated sub-carriers;
  • {circle around (3)} a specific value is P_side/P_null>16; and
  • at most 8 sub-carriers meeting the above criteria are used in the following steps for the system.

In step S3, the estimated sub-carriers are sorted in a descending order of a size of P_side, a larger weight is assigned to energy of the sub-carrier with a larger P_side, a sub-carrier with a serial number k is W_k=√{square root over ((9−k)/4)}, and the weights of each sub-carrier are formed into a diagonal matrix W.

Energy of the estimated sub-carriers is added into the cost function according to different weights, wherein the cost function is modified as:

$J_2\left(ϵ\right)=\sum _{m\in S_N}{f_m^H{\Gamma }^H\left(ϵ\right)z}^2+\sum _{m\in \widehat{S_N}}{{\mathrm{Wf}}_n^H{\Gamma }^H\left(ϵ\right)z}^2$

wherein is an estimated non-activated sub-carrier set.

A final estimation value {circumflex over (ε)}₂=arg min_εJ₂(ε) of ε is obtained by the one-dimensional search performed on ε. A signal subjected to final compensation is z₂=Γ^H({circumflex over (ε)}₂)z.

In step S6, Monte Carlo Simulation is performed on the above method. The following 3 situations are stimulated, as shown in Table 1.

TABLE 1
		Number of null carriers
Method	Modulation mode	used for estimation
{circle around (1)} Method of Li	OFDM QPSK	4
{circle around (2)} Method of Li	OFDM-IM QPSK	4
{circle around (3)} Method of the	OFDM-IM QPSK	Initial: 4
present invention		Supplemental: not more than 8

Performances of the above methods are measured in the stimulation through comparing mean square errors of an estimation value {circumflex over (ε)}_norm={circumflex over (ε)}/Δƒ. As shown in FIG. 2 and FIG. 3, graphic symbols from top to bottom in a stimulation result diagram are trend charts of mean square errors of {circumflex over (ε)}_norm in methods {circle around (1)}, {circle around (2)}, and {circle around (3)} changing with a signal-to-noise ratio.

Stimulation results show that {circle around (2)} has a slightly better performance than {circle around (1)}, and {circle around (3)} has a better performance than {circle around (1)} and {circle around (2)}. Under a high signal-to-noise ratio, {circle around (3)} has a more obvious performance advantage. The stimulation results verify the effectiveness of the method according to the present invention.

The above specific embodiment is merely one implementation of the present invention, and is not used for limiting the scope of patent of the present invention. Any equivalent structure or equivalent process transformation made by using the spirit and principle of the present invention and the contents of the accompanying drawings should fall within the protection scope of patent of the present invention.

Claims

1. A frequency offset estimation method for an OFDM-IM system, the method comprising following steps: S1. performing preliminary frequency offset compensation on a received signal subjected to non-uniform frequency offset by using a two-step method of: (1) resampling and down conversion to obtain a discrete baseband received signal z=Γ(∈(ε)y; and (2) unified compensation for residual frequency offset, a sum of energy of null sub-carriers being used as a cost function in the step (2):

2. The frequency offset estimation method for the OFDM-IM system according to claim 1, wherein in the step S2, the positions of the non-activated sub-carriers are estimated through power detection by using the signal subjected to the preliminary compensation.

3. The frequency offset estimation method for the OFDM-IM system according to claim 1, wherein in the step S3, the energy of the sub-carriers estimated to be non-activated is added into a final cost function according to different weights.