KR20160028334A - Blind Detection Method for detecting Presence of Interference Data Transmission and estimating Interference Traffic-to-Pilot Power Ratio - Google Patents

Abstract

The present invention provides interference canceling techniques based on blindly detected NAICS interference parameters. The purpose of the present invention is to check out a blind detection method for estimating a TPR to be used by a base station (BS) which provides interference. Especially, estimating the TPR is suggested by joint-classifying interference transmission modes into a set of Grassmannian vectors wherein the interference transmission modes can be obtained through projections of a signal, in which information is received. In addition, the suggested TPR detection method can be operated as a blind estimator for detecting whether PDSCHs providing interference exist in a given RB pair. Link-level simulation results are provided to verify effects of the suggested interference detection algorithms and interference canceling techniques.

Description

[0001] Blind Detection Method for Detecting the Presence of Interference Data Transmission and Estimating Interference Traffic-to-Pilot Power Ratio [0002] Blind Detection Method for detecting Interference Data Transmission and Estimating Interference Traffic-to-Pilot Power Ratio [

The present invention relates to a blind detection method for detecting the presence of an interference data transmission and estimating an interference traffic-to-pilot power ratio.

This network-assisted interference cancellation and suppression (NAICS) feature has been studied by the 3rd Generation Partnership Project (3GPP) [1]. A user equipment (UE) may be configured to use a traffic-to-pilot power ratio (TPR), a rank indicator (RI), a precoding matrix indicator (PMI) modulation level (MOD) is known to the UE through the support of network signaling. However, this assumption is based on the assumption that the interference parameters are dynamically changed from one resource block (RB) to another RB in the frequency domain every transmission time interval (TTI) according to the channel conditions , Which results in an excessive increase in the network-signaling load as well as limiting the scheduling performance.

A method for eliminating interference of a terminal in a cellular communication system is provided.

There is provided a method of canceling interference of a terminal in a cellular communication system, comprising: receiving a signal including an interference signal and a desired signal from at least one base station; detecting, based on at least one of a transmission parameter of the desired signal and a transmission parameter of the interference signal, Determining a stellar diagram, blind detection of an additional transmission parameter of the interference signal using the determined constellation diagram, and removing the interference signal from the received signal using the detected additional transmission parameter Lt; / RTI >

를 추정하는 다른 검출 방법들의 성능.Fig.

The performance of the other detection methods to estimate.

In order to meet the stringent requirements of the International Telecommunication Union Radio Communication Sector (ITU-R) requirements, the Long Term Evolution-Advanced (LTE-A) Order spatial multiplexing of up to eight layers in wider bandwidths with up to 100 MHz carrier aggregation (CA) and up to 100 MHz carrier aggregation (CA). More importantly, by adding more cells it has been shown that spatial frequency reuse provides greater capacity gain compared to the increased spatial order or the increased spatial bandwidth. In LTE-A, heterogeneous networks use low-power small cells that increase cell density in high-power macrocell environments. When the cells are arranged very densely in such heterogeneous networks, the intercell interference, which can cause enormous problems for cellular networks, becomes even more severe. Enhanced inter-cell interference coordination (eICIC) has been proposed to mitigate macro interference to the user equipment (UE) closer to the small cells. Meanwhile, when the macro is being transmitted through the DL data channel, i.e., a physical downlink shared channel (PDSCH), excluding pilot signals, i.e., a cell-specific reference signal (CRS) ABS (almost blank subframes) concepts have been introduced. In addition to the base station (BS) -based approach, a UE-based interference mitigation scheme has also been considered by using the acknowledgment of the CRS sequence on the UE side. Particularly, a further enhanced inter-cell interference coordination (FeICIC) granting CRS interference cancellation (CRS-IC) by the UE is disclosed in the 3rd Generation Partnership Project 3GPP) in LTE Release 11 (LTE Release 11).

Now, the point of natural inquiry for normal subframes including the pilot signals as well as traffic signals; We will extend the UE-based approach to handling interference issues for such a typical LTE downlink. In response to this inquiry, research items called network-assisted interference cancellation and suppression (NAICS) have been initiated in the LTE Release 12 period and are currently being considered for inclusion in the LTE Release 12 Standard Are considered [1]. In 3GPP, the above study is based on the assumption that the interference parameters associated with unwanted interfering transmissions are known to the UE by broadcast signaling, e.g. radio-resource control (RRC) signaling, Lt; RTI ID = 0.0 > NAICS < / RTI > feature can achieve significant performance gains. However, this assumption is not always maintained in real systems because the backhaul capacity between BSs and the control channel capacity from the BS to the UE are generally limited. The CRS interference cancellation (IC) includes fixed interference parameters, such as physical cell identity (CID), CRS antenna port (AP), multimedia broadcast multi It is worth noting that the signaling overhead for enabling the FeICIC is manageable because it requires only the MBSFN subframe configuration. In recent years, a similar approach has been discussed for NAICS to support RRC signaling for fixed interference parameters, as summarized in Table I. However, unlike FeICIC, NAICS also provides a UE-specific traffic-to-pilot power ratio (TPR), a rank indicator (RI), a precoding matrix indicator (PMI) And addresses the interference by the data signals in the PDSCH requiring recognition of dynamic interference parameters including a modulation level (MOD). Further, the success of NAICS based on the signaling of the dynamic parameters indicates that the interfering BSs can use the signaled values of TPR, RI, PMI, and MOD to potentially limit the scheduling flexibility in neighboring cells It is based on commitment.

In order to overcome the drawbacks such as network signaling overhead and scheduling constraints, the UE may estimate the dynamic interference parameters from the received signals by blind [1]. In this disclosure, we will look at blind detection (BD) methods for estimating a UE-specific TPR to be used by interfering base stations. As shown in Table I, a subset of three candidate values for the UE-specific TPR is signaled to reduce detection complexity and improve detection performance. The conventional method removes the effect of other interfering transmission modes on the received signal power by inverting the effect of the channel matrix and performs an estimation of the TPR based on the received signal observations. Unlike the conventional method, we propose to jointly classify the interference transmission modes based on the difference in the received signal power to estimate the TPR. In particular, we project the received signal vectors into Grassmannian vectors and estimate the interference TPR by exploiting the difference in the projection output between different interfering transmission modes. On the other hand, we will show that the proposed TPR detection method can also operate as a blind estimator to detect whether or not the PDSCHs that are interfering in a given RB pair are present.

The remainder of this disclosure is organized as follows: First, Section II presents a system model and describes the baseline LTE and advanced NAICS LTE-A receivers. Section III describes a new blind detection method for estimating the interference TPR. In section IV, we will provide simulation results comparing the Advanced NAICS receivers based on blindly detected interference parameters with the baseline LTE receiver. Finally, a conclusion is reached in Section V.

≪ II. System Models and Advanced LTE Receivers>

This section presents the system model and describes the Advanced NAICS demodulators as defined in [3]. For performance comparison with the existing baseline LTE receiver in Section IV, we will briefly review the legacy linear MIMO demodulator as specified in [4]. We N _t transmission antennas two BS to each of the unique message to be transmitted to the desired UE having the N _r receive antennas downlink MIMO Orthogonal Frequency Division Multiplexing comprising the (MIMO orthogonal frequency division multiplexing: MIMO-OFDM ) Systems.

In the NAICS study, the single set of RI, PMI and MOD consists of 12 consecutive subcarriers during a period of one TTI corresponding to 14 OFDM symbol intervals in the case of a cyclic prefix (CP) Lt; RTI ID = 0.0 > RB < / RTI > pair. Therefore, blind detection and data detection will be performed in a unit of RB pair consisting of 168 resource elements (REs). In this section, we assume that the UE is aware of the NAICS interference parameters, i.e. TPR, RI, PMI and MOD.

라고 나타내기로 하며, 여기서

는 l번째 공간 계층을 나타내며, l_i는 송신 계층들의 개수, 즉 RI를 나타내며,

는 벡터의 트랜스포즈(transpose)를 나타낸다. 심볼

는 그 기수(cardinality)가

와 같이 나타내지는 성상도 집합(constellation set)

로부터 선택된다. 상기

의 평균 송신 전력은

로 주어지고, 여기서

는 상기 기대 연산자(expectation operator)를 나타내고,

는 복소수의 절대값을 나타낸다. 일반성의 손실없이, 우리는 BS i = S는 상기 서빙 BS이고, BS i = I는 상기 간섭을 주는 BS라고 가정하기로 한다. We use the l _i -dimensional complex signal vector transmitted from BS i in the k th RE

, Where < RTI ID = 0.0 >

Denotes the lth spatial layer, l _i denotes the number of transmission layers, i.e. RI,

Represents the transpose of the vector. symbol

Has its cardinality

(Constellation set)

. remind

The average transmit power of

Lt; / RTI >

Represents the expectation operator,

Represents the absolute value of the complex number. Without loss of generality, we assume that BS i = S is the serving BS and BS i = I is the BS giving the interference.

We define r _k as the signal vector received at the desired UE at RE k. Then, r _k can be written as shown in Equation 1 below.

는 상기 실제 채널 행렬과 상기 프리코딩 행렬을 포함하는 유효 채널 행렬(effective channel matrix)을 나타내고, n_k 는 분산

를 가지는, 그 엘리먼트들이 독립적이고 균일하게 분포된(identically-distributed: i.i.d.) 복소 가우시안(Gaussian)이고, K는 각 RB 페어에서 사용되는 코딩된 자원 엘리먼트(resource element: RE)들의 개수를 나타낸다.In the above equation (1)

Denotes an effective channel matrix including the real channel matrix and the precoding matrix, and n _k denotes an effective channel matrix

(K) is the number of coded resource elements (REs) used in each RB pair, the elements being independent and uniformly distributed (iid) complex Gaussian.

<A. Baseline IRC Receiver>

In LTE Release 11, the linear baseline LTE receiver, referred to herein as the term interference rejection combiner (IRC), has been introduced in [4]. The IRC can suppress the inter-stream interference as well as the inter-stream interference. The weight matrix for the IRC can be expressed by the following equation (2).

In Equation (2), the covariance matrix R _k including the desired signal and the undesired signal and the noise vectors may be expressed by Equation (3).

Since the CRS sequence of the interfering cell is known to the UE [4] [5], the interference noise covariance matrix R ^{I + N} may include elements of the interference and noise vectors in the serving CRS- It is estimated by averaging. Then, R ^{I + N} is given by the following equation (4).

는 상기 서빙 셀의 CRS 시퀀스에 상응하는 송신 벡터이다.In Equation (4), K _crs is the number of serving CRS-REs per RB pair,

Is a transmission vector corresponding to the CRS sequence of the serving cell.

It should be noted that the baseline IRC receiver does not require prior knowledge of the interfering transmission mode since the interference covariance matrix can be directly evaluated in the serving CRS REs.

<B. Advanced NAICS receivers: Enhanced IRC and ML>

Unlike the baseline LTE receiver, the Advanced NAICS receivers will use the information on the interference, i.e. interference channel matrix and interference parameters, which improves MIMO performance. To this end, in this subsection, we will generalize the model of Equation (1) above in terms of the known interference information and describe two advanced NAICS receivers based on enhanced IRC and ML demodulations.

이라고 정의하고, 상기 N_r X N_t 채널 행렬

의 (m, n) 엔트리(entry)는 BS i의 안테나 n으로부터 상기 UE에서의 안테나 m으로의 경로 이득을 나타낸다. 상기 엔트리들은 제로 평균(zero mean)과 유닛 분산(unit variance)을 가지는 독립적인 복소 가우시안(Gaussian) 랜덤(random) 변수들, 즉 레일레이 페이딩(Rayleigh fading)으로 모델링된다.

를 RE k에서 BS i에 의해 사용되는 N_t X l_i 프리코딩 행렬(혹은 벡터)이라고 하기로 한다. 표 II는 2개의 송신 안테나들을 가지는 MIMO LTE 시스템들을 가정할 경우의, 각 송신 모드에 대해 [6]에서 명시되어 있는 상기 프리코딩 행렬들(혹은 벡터들)을 나타내고 있다. We denote the channel model from BS i to the desired UE in RE k as N _r XN _t channel matrix

, And the N _r X N _t channel matrix

(M, n) entry of the BS indicates the path gain from antenna n of BS i to antenna m at the UE. The entries are modeled as independent complex Gaussian random variables with zero mean and unit variance, i.e., Rayleigh fading.

Is the N _t X l _i precoding matrix (or vector) used by BS i at RE k. Table II shows the precoding matrices (or vectors) specified in [6] for each transmission mode, assuming MIMO LTE systems with two transmit antennas.

를 채널 행렬

과 프리코딩 행렬

의 곱으로서 연산할 수 있다. 그리고 나서,

와 같이 나타냄으로써, 우리는 수학식 1을 하기 수학식 5와 같이 다시 쓸 수 있다. The UE determines the effective channel

Channel matrix

And a precoding matrix

As shown in FIG. Then the,

, We can rewrite Equation 1 as Equation 5 below.

를 획득하고, 상기 상응하는 웨이트 행렬(weight matrix)은 하기 수학식 6과 같이 계산된다.The enhanced IRC defined in the NAICS study uses the recognition of interfering RI and PMI

And the corresponding weight matrix is calculated as: &lt; EMI ID = 6.0 &gt;

에 대해서 소프트 비트 정보(soft bit information)만을 필요로 한다. 우리는 상기 성상도 심볼

의 m번째 비트

를

라고 나타내기로 하며, 또한 우리는

를 하기 수학식 7과 같이 정의되는

에 대한 상기 로그-우도 비(log-likelihood ratio: LLR) 값으로 나타내기로 한다.In comparison to the NAICS IRC, the NAICS ML can realize the full benefit of the NAICS feature by considering non-linear interference cancellation, which requires additional recognition of the interference MOD as well as RI and PMI. From the definition of the NAICS,

Only soft bit information is required. We use the constellation symbol

The mth bit of

To

And we also

Is defined as &lt; EMI ID = 7.0 &gt;

(Log-likelihood ratio (LLR) value).

는 상기 랜덤 변수

가 상기 값 b를 가질 확률을 나타낸다(b = 0 혹은 1). In Equation (7)

Lt; RTI ID = 0.0 &gt;

Represents the probability of having the value b (b = 0 or 1).

가

와

에서 조건화되는 r_k의 조건 확률 밀도(conditional probability density: pdf) 함수를 나타낼 경우, 상기

는 하기 수학식 8과 같이 나타내진다.

end

Wow

The conditional probability density (pdf) function of r _k conditioned in

Is expressed by the following equation (8).

의 LLR 값은 하기 수학식 9와 같이 획득된다. Then,

LLR &lt; / RTI &gt;

는 l_i-폴드(l_i-fold) 데카르트 곱(Cartesian product)

로서 획득되는 모든 가능한 심볼 벡터들

의 집합을 나타내고,

는

(b = 0 혹은 1) 인

의 서브 집합을 나타내고,

는 유클리디안 놈(Euclidean norm)을 나타낸다. 상기 수학식 8에서

가 상수이기 때문에, 우리는 다음 섹션들에서

를 무시할 것이다.In Equation (9)

Are l _i - fold (l _i -fold) a Cartesian product (Cartesian product)

Lt; RTI ID = 0.0 &gt;

, &Lt; / RTI &gt;

The

(b = 0 or 1)

, &Lt; / RTI &gt;

Represents the Euclidean norm. In Equation (8)

Because we are constant, we have

.

Of the LTE receivers as described above, it is clear that the NAICS ML receiver has the highest complexity. The trade-off between performance and complexity can be achieved by applying the max-log approximation to Equation (9). This approximation allows the LLR computation based on the minimum euclidean distance according to Equation (10).

&Lt; III. Detection of traffic-to-pilot power ratio>

In this section, we will look at blind detection methods that estimate the interference TPR based on the received signal observations.

와

라고 나타냄으로써, 우리는 k번째 데이터 RE에서 상기 수신된 신호 벡터를 하기 수학식 11과 같이 explicitly 표현할 수 있다.In LTE systems, the CRS is used by the UE to estimate the channel from the BS. The estimated channel matrixes for the serving BS and the interfering BS are expressed as

Wow

, We can explicitly express the received signal vector in the k-th data RE as: &lt; EMI ID = 11.0 &gt;

는 상기 데이터 RE 송신 전력 대 상기 CRS RE 송신 전력의 비를 나타낸다.In Equation (11), for i = S or I,

Represents the ratio of the data RE transmission power to the CRS RE transmission power.

는 각 TTI 내에서 OFDM 인덱스(index)를 참조하여 데이터 RE 위치들을 기반으로 [7]에 명시된, P_a 와 P_b로 나타내지는 상기 2개의 TPRs들의 함수로서 주어진다. 표 I에 나타낸 바와 같이, 고정적 셀-특정 파라미터 P_b 의 정확한 값이 RRC 시그널링에 의해 상기 UE에서 유용한데 반해, 상기 다이나믹 셀-특정 파라미터 P_a 는 상기 시그널된 3개의 후보 값들로부터 검출될 필요가 있다. The transmission power ratio

Is given as a function of the two TPRs indicated by P _a and P _b specified in [7] based on the data RE positions with reference to the OFDM index in each TTI. As shown in Table I, while the exact value of the fixed cell-specific parameter P _b is useful in the UE by RRC signaling, the dynamic cell-specific parameter P _a needs to be detected from the signaled three candidate values have.

가 행렬(혹은 벡터)의 프로베니우스 놈(Frobenius norm)이라고 나타내기로 하고, 하기 수학식 12와 같은 수신 신호 전력의 기대값을 고려하기로 한다. When you start,

(Frobenius norm) of the matrix (or vector), and the expectation value of the received signal power as shown in Equation (12) below is considered.

의 함수임을 나타낸다.(12) &lt; / RTI &gt; where the expected value is the interference precoding matrix (or vector)

.

와

가 아닌, 상기

및

에서의 정보가 네트워크 시그널링에 의해 상기 UE에게 제공되기 때문에, 상기 수신 신호 전력을 기반으로 하는

의 블라인드 검출은

의 인지를 필요로 한다. As can be seen from Equation (12) above,

Wow

,

And

Is provided to the UE by the network signaling,

Blind detection of

And the like.

<A. Conventional method based on channel inversion >

의 인지에 관련되는

의 블라인드 검출의 어려움을 극복하기 위해서, 상기 종래의 방법은 상기 수신 신호 벡터에 Z로 표현되는, 간섭 채널 행렬의 의사-역(pseudo-inverse), 즉

으로 획득되는 행렬을 적용하고,

를 만족시키는 프리코딩 행렬

의 유닛-놈 특성(unit-norm property)을 사용하여

를 추정한다.

Related to the recognition of

In order to overcome the difficulties of blind detection of the interfering channel matrix, the prior art method is based on a pseudo-inverse of the interference channel matrix,

&Lt; / RTI &gt; is applied,

A precoding matrix &lt; RTI ID = 0.0 &gt;

Using the unit-norm property of &lt; RTI ID = 0.0 &gt;

.

Applying the channel inverse matrix to the received signal vector results in Equation (13).

We then obtain Equation 14 below.

이라는 사실을 사용했다.In the above equation (14), we

.

의 추정이 하기 수학식 15와 같이 상기 수신 신호 벡터들로부터의 관측들을 평균내는 것을 기반으로 수행될 수 있다는 것을 나타낸다. Equation (14)

May be performed based on averaging observations from the received signal vectors as shown in Equation (15): &lt; EMI ID = 15.0 &gt;

<B. Proposed Method>

를 추정하는 것을 제안하기로 한다. 이전 서브 섹션에서 나타낸 바와 같이, 상기 종래의 방법은 상기 채널 행렬의 효과를 반전시키고, 상기 LTE 프리코딩 행렬들의 유닛-놈 특성을 사용하여 상기 평균 수신 신호 전력에서 다른 송신 모드들의 효과를 제거하는 반면에, 우리는 상기 결과 복합 채널 행렬(composite channel matrix)의 놈(norm)에 따라 상기 송신 모드들을 분류하는 것을 통해 상기 송신 모드들간의 차를 활용할 것이다. In this sub-section, we classify the interference transmission modes

Is estimated. As shown in the previous subsection, the conventional method reverses the effect of the channel matrix and removes the effect of other transmission modes on the average received signal power using the unit-norm property of the LTE precoding matrices , We will use the difference between the transmission modes by classifying the transmission modes according to the norm of the resulting composite channel matrix.

에 따라 다음과 같은 5개의 TM 그룹들: g₁) TM2, TM3, TM4, g₂) PMI=0을 가지는 TM6, g₃) PMI=1을 가지는 TM6, g₄) PMI=2를 가지는 TM6, g₅) PMI=3을 가지는 TM6로 분할될 수 있다. 그리고 나서, 우리는

의 추정시, 다른 그룹들은 다른 놈 값들을 가진다는 사실을 사용할 수 있다. 상기 제1 그룹 g₁에 속하는 상기 송신 모드는 상기 프리코딩 행렬

가 상기 전력 제한에 부합하기 위해서

에 의해 스케일된(scaled) 유니터리 행렬(unitary matrix)이기 때문에,

에 상관없이

와 같은 상기 동일한 제곱 놈을 가진다는 것에 유의하여야 할 것이다. 이와 비교하여, 나머지 그룹들

에 대해서는, 상기

의 제곱 놈이

의 함수로 남아 있고, 여기서

는 또한 N_t-차원 복소 벡터 공간

에서

와

의 행 벡터들에 의해 스팬(span)되는 2개의 서브 공간들 각각 간의 각도를 실제로 기반으로 하는 상기 빔포밍(beamforming) 이득으로 알려져 있다. As can be seen from Table II, all of the CRS-based transmission modes specified for LTE systems correspond to the squared norm values of the corresponding composite channel matrix, i. E.

The following five TM group as _{per: g 1) TM2, TM3,} TM4, g 2) PMI = with _{0 TM6, g 3) TM6,} g having a PMI = 1 ₄₎ TM6 with the PMI = 2, g ₅ ) can be divided into TM6 having PMI = 3. And then,

, Other groups can use the fact that they have different norm values. Wherein the transmission mode belonging to the first group g ₁ is a pre-

In order to meet the power limit

Is a unitary matrix scaled by &lt; RTI ID = 0.0 &gt;

Regardless

Quot; has the same &lt; RTI ID = 0.0 &gt; squared &lt; / RTI &gt; In comparison, the remaining groups

, The above-

The square of

, Where &lt; RTI ID = 0.0 &gt;

_Nt -dimensional complex vector space _&lt; RTI ID = 0.0 _&gt;

in

Wow

Is known as the beamforming gain that is based essentially on the angle between each of the two subspaces spanned by the row vectors of &lt; RTI ID = 0.0 &gt;

에 의해 결정되는 상기 TM 그룹이 상기 UE에서 알려질 경우, 이

검출 문제는 매우 간단하고, 일 예로, 상기 원하는 신호 전력 대 상기 나머지 신호 잡음 전력의 비를 최대화시키는 측면에서 최적인 상기 결과 채널 행렬

에 관해 최대 비 결합(maximum ratio combing: MRC) 벡터를 사용함으로써 해결될 수 있다.

If the TM group determined by the UE is known in the UE,

The detection problem is very simple, and in one example, the result channel matrix, which is optimal in terms of maximizing the ratio of the desired signal power to the residual signal noise power,

Can be solved by using a maximum ratio combing (MRC) vector.

를 추정하는 것을 제안한다. 나머지 문제는 어떻게 간섭 TM이 어떤 그룹에 속하는지를 식별하는 지에 있다. 이 연구에서, 우리는 상기 수신 신호 벡터를 m = 1, … , M 에 대해서 p_m 으로 나타내지는 M개의 프로젝션 벡터들의 집합으로 프로젝팅함으로써 상기 TM 그룹 분류에 대해서 살펴보기로 한다. 상기 복합 채널 행렬

에 대한 정보는 상기 M개의 벡터들의 집합에 대한 상기 복합 채널 행렬

의 프로젝션들로 획득될 수 있다. 우리는 랜덤 변수에 의한 p_m 에 대한 상기 수신 신호 벡터의 프로젝션을 나타내기로 하며, 상기 랜덤 변수는 하기 수학식 16과 같이 나타낼 수 있다. Unfortunately, this is not the case, so we have TM group g _i , i = 1, ... Joint with five.

. The remaining problem is how to identify which group the interference TM belongs to. In this study, we consider the received signal vector as m = 1, ... And projecting the M projection vectors, denoted as p _m for M, into the TM group classification. The complex channel matrix

Information on the set of M vectors is the complex channel matrix &lt; RTI ID = 0.0 &gt;

Lt; / RTI &gt; We denote the projection of the received signal vector with respect to p _m by random variables, and the random variable can be expressed as: &lt; EMI ID = 16.0 &gt;

의 열 벡터(column vector)들은

에서 고유하게 분포된다고 가정될 수 있다. 다른 관측들에 의해 동기가 부여되어, 우리는 [8]에서 제시된 바와 같은 그라스마니안 벡터들의 집합을 사용하는 것을 제안하기로 한다. 게다가,

와

의 엘리먼트들이 독립적이고, 제로 평균 및 각각 분산

와

으로 가우시안 분포되고, 주어진 채널 실현들 및 프리코딩 행렬들이라는 가정 하에서, 상기 랜덤 변수

는 제로 평균, 즉

와 하기 수학식 17과 같은 분산을 가지는 가우시안 분포를 가진다고 나타낼 수 있다.Under the iid MIMO Rayleigh Fading Scenarios, the resulting composite channel

The column vectors &lt; RTI ID = 0.0 &gt;

As shown in Fig. We are motivated by other observations, and we will propose to use a set of Grassmannian vectors as presented in [8]. Besides,

Wow

Elements are independent, zero mean and variance

Wow

Gaussian distributions are given, and under the assumption that given channel realizations and precoding matrices, the random variable &lt; RTI ID = 0.0 &gt;

Is zero mean, that is,

And a Gaussian distribution having a variance as shown in Equation (17).

를 하기 수학식 18과 같이 정의하기로 할 경우,

은 K개의 자유도들을 가지는 카이 제곱 분포를 따른다. It is well known that the sum of squares of K independent Gaussian variables with zero means and unit variances is a chi-square random variable with K degrees of freedom. That is,

Is defined as the following equation (18)

Follows a chi-square distribution with K degrees of freedom.

One of the fundamental properties of the chi-square distribution is that the mean and variance of the chi-square distribution are each the same as doubling the degrees of freedom and degrees of freedom, respectively, and the mean and variance of the chi-square distribution are: And Equation (20).

은 상기 샘플(sample) 번호 K가 상기 중심 극한 정리(central limit theorem)에 의해 증가되기 때문에 가우시안 분포를 따르려는 경향이 있다. 이에 따라, 상기

의 분포는 하기 수학식 21과 같이 나타내질 수 있다.At the same time,

Tends to follow the Gaussian distribution because the sample number K is increased by the central limit theorem. Accordingly,

Can be expressed by the following equation (21).

를 explicitly 고려하고,

의 확률은 하기 수학식 22와 같이 전체 M개의 프로젝션들을 고려함으로써 획득될 수 있다.The mathematics and conventional methods such as given in equation 15 is again a look at other Equation 16 and Equation 17, we said precoded sequence to determine the TM group g _i in a matrix

Consider explicitly,

Can be obtained by considering all M projections as shown in the following Equation (22).

To obtain a low complexity metric, we can process the algorithm described in Equation 22 above and remove the constant terms and then define the metric as: &lt; EMI ID = 23.0 &gt;

를 검출할 수 있다.Finally, we obtain the estimate &lt; RTI ID = 0.0 &gt;

Can be detected.

를 추정하는 다른 검출 방법들의 성능을 나타낸다.Fig.

Lt; RTI ID = 0.0 &gt; a &lt; / RTI &gt;

는 가능한 TPR들의 집합을 지시한다.In Equation 24,

Indicates a set of possible TPRs.

에 의해 나타내지는 바와 같은 상기 후보 TPR들의 상응하는 집합은 상기 시그널링된 P_a 및 P_b 값들의 함수로서 획득될 수 있고,

, 즉

를 위해 사용될 수 있다.As described above, in 3GPP, the candidate P _a and P _b values to be used in adjacent cells can be signaled to the UE through RRC signaling not only to improve the performance of the TPR detection but also to reduce the complexity of the TPR detection There have been discussions. For example,

The corresponding set of candidate TPRs as represented by the signalized P _a and P _b values may be obtained as a function of the signaled P _a and P _b values,

, In other words

. &Lt; / RTI &gt;

와 같이 나타내질 수 있다. 요약하면, 우리는

의 후보 송신 전력 비들의 집합을 사용함으로써 간섭을 주는 PDSCH들의 검출과 함께 조인트하게 TRP의 블라인드 검출을 수행할 수 있다.Before proceeding to the next section we will look at the above RI, PMI and MOD detection problems, the proposed TPR detection method also operates as a blind estimator to detect if it exists or does not exist as interfering PDSCHs in a given RB pair It is worth noting that you can. In particular, the absence of a PDSCH

As shown in FIG. In summary, we

Lt; RTI ID = 0.0 &gt; TRP &lt; / RTI &gt; with the detection of interfering PDSCHs by using a set of candidate transmit power ratios.

IV. Numerical Results>

In this section, we will provide link level simulation results to demonstrate the effectiveness of the proposed detection algorithms to estimate the interference TPR. Unless otherwise noted, we assume LTE 3GPP specifications as a baseline for simulations. We use the above multipath channel model based on an extended vehicular A (EVA) power delay profile (PDP) for vehicle users [9], a maximum Doppler frequency of 70 Hz frequency). As previously described, interference parameter detection is performed for each RB pair of every TTI.

<A. Detection of traffic-to-pilot power ratio>

Figure 1 presents Monte Carlo simulations illustrating the detection success probability of the proposed method as a function of the size of the applied grusmanian set. For comparison, we also provide the results for the conventional method based on the channel inversion. We can observe that the proposed method achieves a substantial gain through the conventional method. It can also be seen that most of the gain is obtained by applying a set of Grmaniac vectors of size 2 ² = 4.

Our simulation shows that the interference transmission mode should be considered in the interference TPR detection based on the received signal power observations. Due to the space limitations, we present simulation results only in terms of detection success probabilities, but we believe that the proposed TPR detection is negligible degradation in the BLER performance as compared to when the actual TPR is known at the UE . In one example, the proposed TPR detection maintains a performance loss of no greater than 0.2 dB when assuming a target BLER of 10% in all scenarios being considered in the following figures, Channel inversion-based detection will experience a performance loss of up to 1.0 dB.

From the above simulation results as presented in this section, we conclude that the Advanced NAICS receivers, based on blind detected interference parameters, can significantly improve the performance of LTE-A UEs in interference limited cellular environments I will.

<V. Conclusion>

In the present disclosure, we have developed interference cancellation techniques based on blind detected interference parameters for advanced NAICS receivers. We have proposed a new method of detecting traffic-to-pilot power ratio by applying granny projection vectors to the received signal and utilizing the difference in the projections between the different interfering transmission modes. In conclusion, we have shown that advanced NAICS receivers based on blind detection can be a promising candidate for additional high performance and low complexity UE devices.

<References>

[1] 3rd Generation Partnership Project (3GPP), "Network Auxiliary Interference Cancellation and Suppression for LTE," TSG RAN Meeting 63, RP-140519, March 2014 (3rd Generation Partnership Project (3GPP) &Quot; Network Assistance Interference Cancellation and Suppression for LTE, &quot; TSG RAN Meeting 63, RP-140519, March 2014).

[2] Third Generation Partnership Project (3GPP), "Evolved Universal Terrestrial Radio Access (E-UTRA); Radio Resource Control (RRC) Protocol Specification, "TS 36.331, January 2014 (3GPP)," Evolved Universal Terrestrial Radio Access (E-UTRA) , &Quot; TS 36.331, January 2014).

[3] Third Generation Partnership Project (3GPP), "Network-assisted interference cancellation and suppression for LTE," TR 36.866 v12.0.1, March 2014 (3GPP, Assisted Interference Cancellation and Suppression for LTE, &quot; TR 36.866 v12.0.1, March 2014).

[4] 3rd Generation Partnership Project (3GPP), Technical Description Group Access Network; Improved performance requirements for LTE User Equipment (UE), "TR 36.829 v11.1.0, December 2012 (3rd Generation Partnership Project (3GPP)," Technical Specification Group Radio Access Network Equipment (UE), &quot; TR 36.829 v11.1.0, December 2012).

[5] Y. Ohwatari, N. Miki, T. Asai, T. Abe, and H. Taoka at the Vehicular Technology Conference (VTC) "Suppress intercell interference in the LTE- &Quot; Performance of Advanced Receivers Using Interference Rejection Combinations to &quot; Advanced Downlink Combining to Suppress Inter-Cell Interference in LTE-Advanced Downlink, "Performance of Advanced Receiver Employing Interference Rejection Combining to Suppress Inter-Cell Interference in LTE-Advanced Downlink, Y. Ohwatari, N. Miki, T. Asai, T. Abe, and H. Taoka, &Quot; in Vehicular Technology Conference (VTC), 2011 IEEE, pp. 1-7, September 2011).

[6] Third Generation Partnership Project (3GPP), "Evolved Universal Terrestrial Radio Access (E-UTRA); Physical Channels and Modulation (Release 10), &quot; TS 36.211, March 2012 (3rd Generation Partnership Project (3GPP), &quot; Evolved Universal Terrestrial Radio Access (E-UTRA) 36.211, March 2012).

[7] Third Generation Partnership Project (3GPP), "Evolved Universal Terrestrial Radio Access (E-UTRA); Physical layer procedures (Release 10), &quot; TS 36.213, March 2012 (3GPP), &quot; Evolved Universal Terrestrial Radio Access (E-UTRA) March 2012).

[8] D. J. Love, R. W. Heath and T. Strohmer, "Grassmannian beamforming for multiple input multiple output wireless systems," IEEE Transactions in Information Theory, pp. 2735-2747, October 2003 (D. J. Love, R. W. Heath, and T. Strohmer, "Grassmannian Beamforming for Multiple-Input Multiple-Output Wireless Systems," IEEE Transactions on Information Theory, pp. 2735-2747, October 2003).

[9] Third Generation Partnership Project (3GPP), "Evolved Universal Terrestrial Radio Access (E-UTRA); User Equipment (UE) Radio Transmission and Reception (Release 11), "TS 36.101, October 2013 (3GPP)," Evolved Universal Terrestrial Radio Access (E-UTRA) ) radio transmission and reception (Release 11), &quot; TS 36.101, October 2013).

Claims (1)

A method for interference cancellation of a terminal in a cellular communication system,
Receiving a signal including at least one desired signal and an interference signal from at least one base station;
Determining a constellation diagram based on at least one of a transmission parameter of the desired signal and a transmission parameter of the interfering signal;
Blind detection of additional transmission parameters of the interfering signal using the determined constellation diagram; And
And removing the interfering signal from the received signal using the detected additional transmission parameter.

Images