KR101800959B1 - Non-orthogonal Multiple Access with Prediction-based MMSE Interference Cancellation Schemes - Google Patents

Abstract

A non - orthogonal multiple access method using prediction - based MMSE interference cancellation is proposed. In the non-orthogonal multiple access method proposed by the present invention, interference cancellation of a minimum mean square error is applied to a first user to remove interference from a second user, Obtaining an IPMMSE weight factor in consideration of the interference, estimating a channel of a minimum mean square error by the IPMMSE weight factor, re-modulating the demodulated signal in the first user, Obtaining a received signal after interference cancellation of a least mean square error with respect to a transmission signal for a user, obtaining power of residual interference with respect to a transmission signal for the first user, calculating a residual interference power by using a RIPMMSE weight factor .

Description

[0002] Non-orthogonal multiple access with Prediction-based MMSE Interference Cancellation Schemes [

The present invention relates to a non-orthogonal multiple access method applying a predictive minimum mean square error (MMSE) interference cancellation technique.

The concept of Non-Orthogonal Multiple Access (NOMA) has been proposed as a candidate radio access technology for a fifth generation mobile communication system. The NOMA technique utilizes the power domain for user multiplexing on the transmitter side, and a continuous interference cancellation (SIC) receiver is applied as a basic receiver technique considering the expected future mobile device evolution. However, recent research has focused on analyzing NOMA performance on the assumption that perfect continuous interference cancellation is possible at the receiver.

Orthogonal frequency division multiple access (OFDMA) or single carrier frequency division multiple access (SC-FDMA) based orthogonal access schemes have been employed in fourth generation mobile communication systems such as LTE, WiMAX and LTE-Advanced. The orthogonal access scheme is a reasonable choice to keep the complexity of the receiver low and to obtain an improved system throughput. However, as the demand for high-capacity services such as image transmission, video streaming and cloud-related services is rapidly increasing, a next generation (5G) mobile communication system capable of improving performance such as system throughput is required. In order to realize this requirement, a non-orthogonal multiple access (NOMA) receiver using continuous interference cancellation (SIC) in the downlink is presented as one of several promising radio access technologies. For downlink NOMA, non-orthogonality is obtained by introducing a power domain into one of the time / frequency / code regions for user multiplexing. In this case, the user demultiplexing is obtained by making a significant difference in the user-to-user power allocation at the transmitter side and applying the SIC to the receiving end. By doing this, all users can use the entire transmission bandwidth to achieve higher spectral efficiency, and by allocating more power to users with poor channel conditions compared to the conventional orthogonal multiple access (OMA) scheme, better user equality Can be achieved. Also, because it can support more simultaneous user connections, large-scale access is also possible by NOMA.

NOMA is a candidate technology that can improve the performance of LTE and LTE-Advanced systems. This enhancement is possible by applying perfect SIC to the receiver. Therefore, we analyze the system performance of NOMA applying realistic continuous interference cancellation techniques (ie zero-forcing or minimum mean square error-based continuous interference cancellation) and propose a method to improve NOMA performance in this realistic case.

SUMMARY OF THE INVENTION An object of the present invention is to analyze the effect of interference caused by interference cancellation by considering a realistic continuous interference cancellation technique for NOMA. Based on this analysis, a new interference prediction minimum mean square error (IPMMSE) interference cancellation technique for NOMA downlink based on the Minimum Mean Square Error (MMSE) . In addition, based on IPMMSE removal and interference cancellation error analysis, we propose residual interference predicted MMSE (RIPMMSE) interference cancellation for residual interference cancellation, which can further improve system performance.

In one aspect, a non-orthogonal multiple access method employing a prediction-based MMSE interference cancellation scheme proposed in the present invention includes applying a minimum mean square error interference elimination to a first user to remove interference from a second user, Estimating a channel of the minimum mean square error by the IPMMSE weight factor, calculating a weighted mean square error factor for the first user, Modulating a demodulated signal in the received signal for the first user and obtaining a received signal after interference cancellation of a least mean square error with respect to the transmitted signal for the second user, And obtaining a RIPMMSE weight factor to remove the residual interference.

Wherein the removing of the minimum mean square interference from the first user is applied to remove interference from the second user comprises demodulating the transmission signal for the first user after first demodulating the transmission signal for the second user, A weight factor of a minimum mean square error with respect to a channel coefficient of a transmission signal for the second user is obtained.

The step of obtaining the interference with the transmission signal plus noise for the second user assumes that the background noise for demodulating the transmission signal for the second user is the interference to the transmission signal plus noise for the second user.

The step of obtaining the IPMMSE weight factor in consideration of the interference is obtained by using the variance of the interference with the transmission signal plus noise for the second user.

Modulating the demodulated signal by the first user and obtaining a received signal after removing the interference of the minimum mean square error with respect to the transmission signal for the second user, And using the channel coefficients for the second user to remove the demodulated signal from the modulated first user.

Wherein the step of obtaining the power of residual interference with respect to the transmission signal for the first user comprises the step of calculating a residual interference power of the transmission signal for the first user using the SNR of the transmission signal for the second user after the interference cancellation of the minimum mean square error, Obtains the power of the interference, and then demodulates the transmission signal for the first user.

The step of obtaining the RIPMMSE weight factor to remove the residual interference updates the weight factor of the minimum mean square error by using the variance of residual interference plus noise and by taking the power of the residual interference by differentiating the RIPMMSE weight factor.

According to embodiments of the present invention, the effect of errors caused by interference cancellation can be analyzed by considering a realistic continuous interference cancellation technique for NOMA. Based on this analysis, a new interference prediction minimum mean square error (IPMMSE) interference cancellation technique for NOMA downlink based on the Minimum Mean Square Error (MMSE) And proposed a residual interference predicted MMSE (RIPMMSE) interference cancellation for residual interference cancellation, which can further improve system performance, based on IPMMSE interference cancellation and interference cancellation error analysis.

FIG. 1 is a diagram for explaining a basic NOMA technique using an SIC in a downlink according to an embodiment of the present invention. Referring to FIG.
2 is a diagram illustrating a transmission technique scenario for a NOMA-MIMO scheme in a downlink with two users according to an embodiment of the present invention.
3 is a block diagram of a transmitter and a receiver for a downlink NOMA-MIMO scheme according to an embodiment of the present invention.
FIG. 4 is a graph comparing PER performance when NOMA is applied and when it is not applied under an AWGN channel according to an embodiment of the present invention.
FIG. 5 is a graph comparing the throughputs when NOMA is applied and when the NOMA is applied under an AWGN channel according to an embodiment of the present invention.
6 is a graph illustrating the simulation results of a first user UE1 under NOMA with different modulation techniques in accordance with an embodiment of the present invention.
7 is a graph illustrating a data rate threshold in UE ₁ NOMA using interference cancellation under an AWGN channel in accordance with an embodiment of the present invention.
8 is a block diagram of zero forcing interference cancellation under a Rayleigh channel in accordance with an embodiment of the present invention.
9 is a diagram illustrating a data rate threshold and simulation results in UE ₁ NOMA using zero forcing interference cancellation under a single path Rayleigh channel according to an embodiment of the present invention.
10 is a flowchart illustrating a non-orthogonal multiple access method using a prediction-based MMSE interference cancellation technique according to an embodiment of the present invention.
11 is a block diagram of an MMSE IC under a Rayleigh channel according to an embodiment of the present invention.
12 is a graph showing a data rate limit value and simulation result of UE ₁ NOMA using realistic zero forcing and minimum mean square error interference cancellation under a single path Rayleigh channel according to an embodiment of the present invention.
13 is a performance comparison BER graph of IPMMSE versus IPMMSE for a first user UE1 using NOMA using other interference cancellation techniques under the conventional OMA and single path Rayleigh fading channels according to an embodiment of the present invention.
FIG. 14 is a performance comparison BER graph of RIPMMSE versus RIPMMSE for a first user UE1 when using NOMA of the conventional OMA and other interference cancellation techniques under a single path Rayleigh fading channel according to an embodiment of the present invention.

The non-orthogonal multiple access method using the proposed prediction-based MMSE interference cancellation technique is used to clarify the difference between the perfect SIC application and the realistic NOMA technique SIC and to test the possibility of applying the NOMA using the SIC to the future cellular system We analyze the performance of NOMA with multiple input multiple output (MIMO) by considering zero-forcing (ZF) and minimum mean square error (MMSE) successive interference cancellation (SIC) The proposed method provides an analysis of the error effect according to the realistic SIC method for NOMA. Based on this analysis, we propose a new IPMMSE interference cancellation technique that modifies the weight of MMSE using interference signal information. In addition, we propose residual interference predicted MMSE (RIPMMSE) interference cancellation technique to remove residual interference, which can further improve system performance, based on IPMMSE interference cancellation and interference cancellation error analysis. The simulation results show that by considering the actual interference cancellation technique, the BER performance is lower than that of the conventional orthogonal multiple access (OMA) due to the interference cancellation error. However, compared with the conventional NOMA using zero forcing and MMSE interference cancellation, the proposed residual interference cancellation interference cancellation technique provides more improved performance. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a diagram for explaining a basic NOMA technique using an SIC in a downlink according to an embodiment of the present invention. Referring to FIG.

A basic NOMA scheme using successive interference cancellation for the case of the second user terminal UE ₂ in the cellular downlink is schematized in FIG. In case of UE _i (i = 1,2), if the transmission information at base station (BS) is s _i and the transmission power is p _i , the transmission signal of UE _i is as follows.

(1)

(One)

The sum of transmit power is limited to p. Therefore, the transmission signals are added as follows.

(2)

The received signal at UE _i is as follows.

(3)

(3)

Where h _i is the complex channel coefficient between UE _i and the base station. And n _i denotes the Gaussian noise of the receiver including the inter-base-station interference. The output density of n _i is N _{o, i} . In the NOMA downlink, the decoding order is a value obtained by normalizing the channel gain by noise and inter-cell interference power | h _i | ² / N _{o, i} in ascending order. In the case of UE ₂ in Fig. 1, | h ₁ | ² / N _{o, 1} > | h ₂ | ² / N _{o, 2} , so that the first user UE ₁ first decodes x ₂ and then removes it from the received signal y ₁ . And a second user (UE ₂₎ decodes the _second signal x directly without removing the interference is much x1 signal path loss less transmission power. UE _i of the throughput R _i is:

(4)

(4)

2 is a diagram illustrating a transmission technique scenario for a NOMA-MIMO scheme of a UE ₂ downlink according to an embodiment of the present invention.

Assume orthogonal frequency division multiplexing (OFDM) signaling despite considering non-orthogonal user multiplexing. 2 shows a transmission technique scenario for downlink NOMA-MIMO in the case of two UEs where the first user UE ₁ is a cell center user (i.e., a near user) , And the second user UE2 is a cell edge user (i.e., far user).

3 is a block diagram of a transmitter and a receiver for a downlink NOMA-MIMO scheme according to an embodiment of the present invention.

Assuming that there are two transmit antennas in the base station, each antenna transmits one UE signal. The transmission signal x _i for UE _i is as follows.

(5)

(5)

Where p _i is the allocated power and S _i is the data transmitted to UE _i . After being transmitted over the 2x2 channel H, the system with two UEs shown in FIG. 2 is expressed as follows.

(6)

(6)

Where n _i (i = 1, 2) denotes the Gaussian noise of the receiver in UE _i . Y _i of the received signal UE _i is expressed as follows.

(7)

(7)

를 얻는다. 2명의 사용자가 존재하는 경우, 수학식(6)에서의 제1 사용자(UE₁)와 제2 사용자(UE₂)의 수신 신호는 각각 다음과 같다.Here, H _ji denotes a channel between the j-th antenna and the i-th receiver of the base station, x _{j, which} is a transmission signal of UE _j , acts as an interference signal of UE _i , and n _i denotes Gaussian noise. After receiving, the signals are ranked in power descending order. Channel estimation (CE) is the power p _j> p _i is applied to the interference signal y _ji (desired power of the signal), is a continuous interference cancellation (SIC) until all of the interference signal is removed using, received estimate of the UE _i Signaling

. When there are two users, the reception signals of the first user (UE ₁ ) and the second user (UE ₂ ) in Equation (6) are as follows.

(8)

(8)

Since the NOMA allows users (UEs) to share the same resources and differentiates the UEs by power, less power UEs perform interference cancellation to eliminate inter-user interference. Since the second user UE2 is a remote user with higher power, the H ₂₁ x ₂ element is removed from y ₁ by interference cancellation. While y ₂ is directly demodulated without interference rejection. The proposed method considers realistic zero-forcing and continuous interference cancellation based on MMSE criterion. As assumed in the system model, by assuming power p ₂ > P ₁ , the second user UE 2 detects the signal directly without eliminating interference. The received signal y ₁ of the first user UE1 is as follows.

(9)

(9)

s₂이다. 채널H₂₁의 가중치 팩터(weight factor)인

와

는 제로포싱 또는 MMSE 기준에 의해 각기 다음과 같이 구해지며, 이를 위하여 추정채널치인

을 구하여 사용한다.Where the interference is H ₂₁

s ₂ . The weight factor of channel H ₂₁

Wow

Is obtained by zero-forcing or MMSE criterion as follows. For this purpose,

.

(10)

(10)

In this equation, the superscript H signal is the Hermitian transpose, and σ ² _n1 is the variance of the noise n ₁ . Then, the estimated interference signal can be obtained as follows.

(11)

(11)

The received signal is obtained by subtracting the estimated interference signal and updating it as follows.

(12)

(12)

After eliminating the high power interference, UE ₁ can detect the desired signal s ₁ from the updated received signal.

FIG. 4 is a graph comparing PER performance when NOMA is applied and when it is not applied under an AWGN channel according to an embodiment of the present invention.

LLS compares packet throughput performance (PER) with conventional orthogonal multiple access (OMA) and NOMA throughput performance. We use the system model for the two-user (UE) case mentioned above. Conventional OMA without NOMA allocates 0.8 to power p ₁ of first user UE ₁ (proximity user) and second user UE 2 (far user) according to a water filling (WF) in the p ₂ allocates 0.2. And NOMA assigns p ₁ = 0.2 and p ₂ = 0.8 to power control (PC). Link-level simulation (LLS) results were used to simulate system-level simulation (SLS), where power allocation was used instead of path loss for power allocation with and without NOMA. Table 1 summarizes the simulation parameters according to LTE specification.

<Table 1>

FIG. 4 shows the PER performance of NOMA (case 1) and NOMA (case 2). In case 2, realistic zero-forcing interference cancellation is applied to the first user UE1.

FIG. 5 is a graph comparing the throughput of NOMA with and without NOMA under an AWGN channel according to an embodiment of the present invention. Referring to FIG.

Depending on the PER, the throughput C can be calculated as follows.

(13)

(13)

Thus, the throughputs of cases 1 and 2 are shown in FIG. If at the link level simulation does not account for path loss compared to the _{case1 (p 1 = 0.2, p} 2 = 0.8) and _{case2 (p 1 = 0.8, p} 2 = 0.2) under a scenario there is no If NOMA, a first user in case1 throughput (UE ₁₎ it will be equal to the second user (UE2) at the case2. And case1 second user (UE2) the throughput is the same as that of the first user (UE ₁₎ of case2. Therefore, it is reasonable to compare case1 with NOMA and case2 without NOMA.

When measuring the performance NOMA is from a simulation shown in Figure 4, the second user (UE2) may sense information with greater power than the interference from the first user (UE _1). On the other hand, does not apply when the NOMA case2, the second user (UE2) will not be able to detect the information, because of the lower power than the interfering signal from the first user (UE _1). On the other hand, the first user (UE ₁ ) using the NOMA can benefit from applying interference cancellation even if the power of the interference signal is much higher than the desired signal. On the other hand, in the case 2 in which the NOMA is not applied, the first user UE ₁ has similar performance to the second user UE 2 using the NOMA because of the existing interference caused by the low power user. As shown in FIG. 5, the NOMA can increase the total throughput as compared with the case where the NOMA is not applied, and can improve the equality between the near and far users. In order to measure the performance of the NOMA in the simulation, the LLS results of the first user (UE ₁ ) are compared under different modulation schemes using the limit values of the first user (UE ₁ ) obtained in the above-mentioned equation (4) .

Figure 6 is a graph illustrating the frequency efficiency of a first user (UE ₁ ) under NOMA using other modulation techniques in accordance with an embodiment of the present invention.

The convolutional code for the coding scheme, the single path Rayleigh fading for the channel environment, and other simulation parameters are the same as those in Table 1. From these comparisons, all simulation curves show lower threshold values and higher frequency efficiencies when using higher level modulation techniques. This confirms the validity of the LLS results and analysis for future research.

7 is a graph illustrating a data rate threshold in UE ₁ NOMA using interference cancellation under an AWGN channel in accordance with an embodiment of the present invention.

For convenience, the NOMA-MIMO transmission scenario as described above is used. At the first user (UE ₁ ) (neighboring user), interference cancellation is applied to eliminate interference from the second user (UE2). In the AWGN channel, the signal received at the first user (UE ₁ ) is as follows.

(14)

(14)

Since x ₁ higher than the power of the x _2, x ₂ is demodulated first. When x ₂ is demodulated in the first user (UE ₁ ), x ₁ becomes an interference element.

(15)

(15)

라고 가정한다. 이 경우 잡음 증폭 비 (Noise Enhancement Ratio)는 다음과 같다.Where i is the sum of x _{2 1_AWGN} interference and noise, the first user (UE ₁₎ to x ₂ in the demodulation signal,

. In this case, the noise enhancement ratio is as follows.

(16)

(16)

는 y₁으로부터 제거되기 위해서 재변조되고, 간섭제거 이후의 x₂ 수신 신호는 다음과 같다.after

Is remodulated to be removed from y ₁ , and the x ₂ received signal after interference cancellation is:

(17)

(17)

는 재변조된 신호이고,

성분은 간섭제거 이후 x₁의 잔여 간섭이다. x₁의 잔여간섭 전력은 다음과 같다.

Lt; / RTI &gt; is a remodulated signal,

The component is the residual interference of x ₁ after interference cancellation. The residual interference power of x ₁ is as follows.

(18)

(18)

는 신호 대 간섭과 노이즈의 합의 비(SINR)이고,

는 완벽한 간섭제거를 이용한 경우 x₂의 신호 대 잡음비(SNR)를 의미한다.here

(SINR) of the sum of signal-to-interference and noise,

Means the signal-to-noise ratio (SNR) of x ₂ when perfect interference cancellation is used.

Therefore, when the practical interference cancellation technique is applied to x ₂ , the limit value of the first user (UE ₁ ) is as follows.

(19)

(19)

Figure 7 shows loss due to interference cancellation error in UE ₁ NOMA under realistic interference cancellation compared to complete interference cancellation.

8 is a block diagram of a system applying zero forcing interference cancellation under a Rayleigh channel in accordance with an embodiment of the present invention.

In contrast to the AWGN channel, various interference cancellation schemes can be used in the Rayleigh channel due to the nature of channel estimation (CE). First, we analyze the performance of zero-forcing interference cancellation in a single-path Rayleigh channel. Similar to the AWGN case, zero-forcing interference cancellation is performed by the second user UE2, depicted in FIG. 8, To remove interference from the signal. The signal received for the first user UE1 is as follows.

(20)

(20)

로 가정한다. 제1 사용자(UE1)에서 복조할 때 잡음 증폭 비 (Noise Enhancement Ratio)는 수학식 (22)와 같다.Where H ₁₁ and H ₂₁ are the channel coefficients for x ₁ and x ₂ , respectively. And because of the relatively large power x _2, x ₂ is demodulated first. When x ₂ is demodulated in the first user UE1, _H11x1 becomes interference. The signal after (perfect) zero forcing interference cancellation for x ₂ is as shown in equation (21). Where i is the sum of x _{2 1_ZF} interference and noise, the demodulated signal x ₂ in a first user (UE1)

. The noise enhancement ratio when demodulating in the first user UE1 is expressed by Equation (22).

(21)

(21)

(22)

(22)

가 재변조 되고, y₁으로부터 이 신호를 제거하기 위해 채널 계수 H₂₁을 이용해 곱하면, x₂를 위한 제로포싱 간섭제거 이후 수신된 신호는 다음과 같다.to the next

Is modulated and multiplied by the channel coefficient H ₂₁ to remove this signal from y ₁ , the received signal after zero forcing interference cancellation for x ₂ is:

(23)

(23)

는 재변조된 신호,

성분은 제로포싱 간섭제거 이후 x₁의 잔여 간섭이다. x₁에서의 잔여 간섭의 전력은 다음과 같다.here

Modulated signal,

The component is the residual interference of x ₁ after zero forcing interference cancellation. The power of the residual interference at x ₁ is

(24)

(24)

은 SINR, 그리고

은 완벽한 간섭제거 시의 x₂의 SNR이다. 그러므로 x₂의 제로포싱 간섭제거 이후 제1 사용자(UE1)에서의 한계값은 다음과 같다.here

Is the SINR, and

Is the SNR of x ₂ at perfect interference cancellation. Therefore, the limit value of the interference cancellation after the forcing of x ₂ zero first user (UE1) is as follows.

(25)

(25)

9 is a diagram illustrating a data rate threshold and simulation results in UE ₁ NOMA using zero forcing interference cancellation under a single path Rayleigh channel according to an embodiment of the present invention.

Referring to FIG. 9, the difference between the data rate threshold and the simulation curve is due to the loss caused by applying realistic interference cancellation as compared to the perfect interference cancellation case.

FIG. 10 is a flowchart illustrating a non-orthogonal multiple access method applying an interference prediction based MMSE (IPMMSE and RIPMMSE) interference cancellation technique according to an embodiment of the present invention.

In the non-orthogonal multiple access method using the proposed interference prediction based MMSE interference cancellation scheme, interference cancellation of a minimum mean square error is applied (1010) in a first user to remove interference from a second user, A step 1030 of obtaining an IPMMSE weight factor in consideration of the interference, a step 1030 of estimating a channel of a minimum mean square error by the IPMMSE weight factor, Modulating the demodulated signal by the first user and obtaining a received signal after removing the interference of the minimum mean square error with respect to the transmission signal for the second user, Obtaining a power of residual interference for the signal (1060), and obtaining (1070) a RIPMMSE weighting factor to remove the residual interference.

At step 1010, the minimum mean square interference cancellation at the first user is applied to cancel the interference from the second user. At this time, the transmission signal for the second user is first demodulated and then the transmission signal for the first user is demodulated. Then, the weight factor of the minimum mean square error with respect to the channel coefficient of the transmission signal for the second user is obtained.

11 is a block diagram of a system applying MMSE interference cancellation under a Rayleigh channel according to an embodiment of the present invention.

Similar to zero-forcing interference cancellation analysis, we analyze the performance of least mean square error interference cancellation under a single-path Rayleigh channel. In the first user UE1, the minimum mean square interference cancellation is applied to eliminate interference from the second user UE2 as described in FIG. The received signal of the first user UE1 is as follows.

(26)

(26)

Where H ₁₁ and H ₂₁ are the channel coefficients for x ₁ and x ₂ , respectively. And because of the relatively large power x _2, x ₂ is demodulated first. When x ₂ is demodulated in the first user UE1, _H11x1 becomes interference. Here, assuming perfect interference estimation, the minimum mean square error weight factor for channel H ₂₁ is:

(27)

(27)

The (perfect) minimum mean square error channel estimate for x ₂ is

(28)

(28)

라 가정하면, 제1 사용자(UE1)에서 x₂를 복조할 때 잡음 증폭 비 (Noise Enhancement Ratio)는 다음과 같다.Where i is the sum of x _{2 1_MMSE} interference and noise, of x ₂ in a first user (UE1) Demodulated signal

La Assuming, first noise amplification ratio (Enhancement Noise Ratio) is as follows: When demodulating the x ₂ from the user (UE1).

(29)

(29)

가 재변조 되고 y₁로부터의 이 신호를 제거하기 위해 채널 계수 H₂₁을 이용해 곱하면, x₂를 위한 최소평균자승에러 간섭제거 이후 수신된 신호는 다음과 같다.to the next,

The re-modulated and multiplied with the channel coefficients H ₂₁ in order to remove this signal from the y _1, the received signal interference, the minimum mean square error after removal for x ₂ is as follows.

(30)

(30)

는 재변조된 신호,

성분은 최소평균자승에러 간섭제거 이후 x₁의 잔여 간섭이다. x₁에서의 잔여 간섭의 전력은 다음과 같다.here

Modulated signal,

The component is the residual interference of x ₁ after the least mean square error interference cancellation. The power of the residual interference at x ₁ is

(31)

(31)

은 SINR, 그리고

은 완벽한 간섭제거 시의 x₂의 SNR이다. 그러므로 x₂의 최소평균자승에러 간섭제거 이후 제1 사용자(UE1)에서의 한계값은 다음과 같다.here

Is the SINR, and

Is the SNR of x ₂ at perfect interference cancellation. Therefore, a minimum mean square error after removing the interference threshold value of x ₂ in a first user (UE1) is as follows.

(32)

(32)

12 is a graph showing a data rate limit value and simulation result of UE ₁ NOMA using realistic zero forcing and minimum mean square error interference cancellation under a single path Rayleigh channel according to an embodiment of the present invention.

Referring to FIG. 12, both the data rate limits and the simulation curves show the benefit of the least mean square error interference cancellation compared to the case of zero forcing interference cancellation.

By using the information of the interference signal to improve the link level performance of the NOMA based on the sum of interference and noise for the second user and based on the minimum mean square error interference cancellation analysis in step 1020, We propose an average squared error weight factor. From equation (28), after the minimum mean square error channel estimate for x ₂ , the sum of x ₂ interference and noise can be obtained as:

(33)

(33)

In step 1030, it is possible to assume that the background noise for demodulating a i x _{2 1_MMSE,} therefore IPMMSE weight factor is:

(34)

(34)

Where σ ² _{i1_MMSE} can be obtained from the variance of i _{1_MMSE} .

(35)

(35)

Here, the symbol E indicates an expected value.

Then, in step 1040, the channel of the minimum mean square error by the IPMMSE weight factor is estimated.

Then, in step 1050, the demodulated signal is re-modulated by the first user and a received signal is obtained after interference cancellation of the minimum mean square error is performed on the transmission signal for the second user.

At this time, the demodulated signal is multiplied by the channel coefficient for the second user to remove the demodulated signal from the first user remodulated from the received signal for the first user.

After differentiating equation (35), the minimum mean square error weight factor can be modified by considering interference. This makes these algorithms practical. The equation (28) can be replaced as follows.

(36)

(36)

FIG. 13 is a graph of performance comparison BER versus IPMMSE for a first user UE1 using NOMA using conventional interference cancellation schemes under OMA and other interference cancellation schemes according to an embodiment of the present invention.

Referring to FIG. 13, the modulation and coding scheme used in the simulation is QPSK using a 1/3 convolutional code. From these simulation results, we found that OMA has the best BER performance because there is no interference for OMA signaling. In the case of NOMA, zero-forcing interference cancellation exhibits the worst BER even with the lowest complexity. And, consistent with previous analysis, minimum mean square error interference cancellation showed better BER performance than zero forcing interference cancellation.

Finally, the proposed IPMMSE interference cancellation provides the best BER performance among the NOMA interference cancellation schemes because it takes into account the effect of interference signals. When the target BER is 10 ^-3 , the IPMMSE interference cancellation technique is about 1dB compared to the zero-forcing interference cancellation technique and about 0.2dB compared with the MMSE interference cancellation technique.

In step 1060, the power of the residual interference is determined for the transmission signal for the first user. The residual interference power for the transmission signal for the first user is obtained using the SNR of the transmission signal for the second user after the interference cancellation of the minimum mean square error and then the transmission signal for the first user is demodulated .

Based on the analysis of the interference cancellation effect, after the IPMMSE channel estimation, the noise enhancement ratio when demodulating x ₂ in the first user UE1 is as follows.

(37)

(37)

Then the power of the residual interference for x ₁ is

(38)

(38)

은 SINR,

은 완벽한 간섭제거를 한 경우 x₂의 SNR이다. here

Lt; / RTI &gt;

Is the SNR of x ₂ for perfect interference cancellation.

In step 1070, a RIPMMSE weight factor is calculated to remove interference. At this time, the weight factor of the minimum mean square error is updated by taking the residual interference and the variance of the noise sum, and by taking the residual interference power into account by differentiating the RIPMMSE weight factor.

In other words, the next step is to demodulate x ₁ and suggest MMSE (RIPMMSE) interference cancellation where residual interference is predicted to eliminate residual interference. The RIPMMSE weight factor for channel H ₁₁ is:

(39)

(39)

는 잔여간섭과 잡음 합의 분산으로부터 획득된다.here

Is obtained from the variance of residual interference and noise sum.

(40)

(40)

After differentiating (40), the weight factor of the MMSE can be updated by considering the power of the residual interference. This makes the algorithm feasible.

FIG. 14 is a graph illustrating a comparison of BER performance for RIMMs other than RIPMMSE for a first user UE1 using a conventional OMA and other interference cancellation techniques under a single path Rayleigh fading channel according to an embodiment of the present invention.

Referring to FIG. 14, the modulation and coding scheme used in the simulation is QPSK using a 1/3 convolutional code. Simulation results show that RIPMMSE interference cancellation can improve BER performance more than IPMMSE interference cancellation for NOMA.

When the target BER is 10 ^-3 , the RIPMMSE interference cancellation technique provides a gain of about 1.5 dB compared to the zero-forcing interference cancellation technique and about 0.5 dB compared to the MMSE interference cancellation technique. And based on this theory, for a NOMA system with three users, the minimum mean square error weight factor can be updated three times to further improve BER performance.

In the case of the proposed perfect or realistic continuous interference cancellation technique, the performance of the non - orthogonal multiplexing access (NOMA) - multiple - input multiple - output (MIMO) system is maximized and the need for research on the interference cancellation technique for the NOMA is clarified. Since many prior studies have focused on NOMA, studies such as some research themes, the use of realistic continuous interference cancellation, and error effects due to interference cancellation are still at an early stage or not fully progressed. In the proposed method, we perform error analysis related to zero forcing and minimum mean square successive interference cancellation in NOMA, and propose a new IPMMSE interference cancellation technique by predicting the MMSE weight factor using the interference signal information based on this analysis . In addition, based on analysis of IPMMSE interference cancellation technique and interference cancellation error effect, RIPMMSE interference cancellation is proposed to improve system performance. For a single path Rayleigh channel 2 under the users provide a link-level performance evaluation results for NOMA-MIMO system, the simulation results show that the IC technique RIPMMSE each target BER of 10 ^-3 compared technique zero forcing and MMSE interference removal when 1.5 dB and 0.5dB, respectively.

The apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the apparatus and components described in the embodiments may be implemented within a computer system, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA) A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device , Or may be permanently or temporarily embodied in a transmitted signal wave. The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI &gt; or equivalents, even if it is replaced or replaced.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (8)

In a non-orthogonal multiple access method,
Applying interference cancellation of a minimum mean square error at a first user to remove interference from a second user;
Obtaining interference for a transmission signal and a noise sum for the second user;
Obtaining an IPMMSE weight factor in consideration of the interference;
Estimating a channel of a minimum mean square error by the IPMMSE weight factor;
Modulating a signal demodulated by the first user and obtaining a received signal after eliminating interference of a minimum mean square error with respect to a transmission signal for the second user;
Obtaining a power of residual interference for a transmission signal for the first user; And
Obtaining a RIPMMSE weight factor to remove the residual interference
Lt; / RTI &gt;
The step of obtaining a RIPMMSE weight factor to remove the residual interference comprises:
The RIPMMSE weight factor is calculated by using the residual interference and the noise sum dispersion, the weight factor of the minimum mean square error is updated by considering the residual interference power by differentiating the RIPMMSE weight factor, and the RIPMMSE weight factor for the channel H ₁₁ is calculated using the following equation ,

Where H _ji is the channel between the jth antenna and the i th receiver of the base station,

Is obtained from the variance of residual interference and noise sum
Non - orthogonal multiple access method. The method according to claim 1,
Wherein the minimum mean square interference cancellation at the first user is applied to remove interference from the second user,
Demodulates the transmission signal for the first user after first demodulating the transmission signal for the second user and obtains a weight factor of the minimum mean square error with respect to the channel coefficient of the transmission signal for the second user
Non - orthogonal multiple access method. The method according to claim 1,
The method of claim 1, wherein the obtaining of the interference of the transmission signal and the noise sum for the second user comprises:
The background noise for demodulating the transmission signal for the second user is assumed to be the interference for the transmission signal for the second user and the noise sum
Non - orthogonal multiple access method. The method according to claim 1,
The step of obtaining an IPMMSE weight factor in consideration of the interference includes:
Using the variance of the interference of the transmission signal and the noise sum for the second user
Non - orthogonal multiple access method. The method according to claim 1,
Modulating the demodulated signal by the first user and obtaining a received signal after removing the interference of the minimum mean square error with respect to the transmission signal for the second user,
Using a channel coefficient for the second user to remove a demodulated signal from the remodulated first user in a received signal for the first user
Non - orthogonal multiple access method. The method according to claim 1,
Wherein the step of determining the residual interference power for the transmission signal for the first user comprises:
The residual interference power for the transmission signal for the first user is obtained using the SNR of the transmission signal for the second user after the interference cancellation of the minimum mean square error and then the transmission signal for the first user is demodulated
Non - orthogonal multiple access method. delete delete

Images