KR100833703B1 - Scheduling Method and Device Using Selection Diversity in Communication System through Nonnormal Orthogonal Matrix and Vector Modulation - Google Patents

Abstract

The present invention provides at least one channel of K transmission channels k = 1, .., K over each N _{t , k} transmission interfaces and each N _{r , k} reception interfaces for transmission of matrix or vector modulated data symbols. A method for scheduling, comprising: calculating a respective channel quality indicator (CQI) q _k for at least one of the K transmission channels, and the K for transmission of the matrix or vector modulated data symbols Scheduling (scheduling) at least one of the transport channels, wherein the scheduling is based at least in part on the calculated CQIs q _k . The invention further relates to apparatus, transmitting stations, wireless communication systems, computer programs and computer program products.

Description

Exploiting selection diversity in communications systems with non-orthonormal matrix and vector modulation

The present invention provides a method for scheduling at least one of K transmission channels k having N _{t , k} transmission interfaces and _{Nr , k} receiving interfaces _, respectively _, for transmission of data symbols modulated according to a nonnormal orthogonal matrix or vector modulation. It is about.

For example, in the context of a wireless communication system, such as the Universal Mobile Telecommunications System (UMTS), scheduling is a group of substantially orthogonal resources (eg, frequency, subcarriers, spreading codes, Time slots, spatial / polar eigenmodes) to be expressed as an equation converted to a preference function f _k calculated for at least one of the K transport channels. Can be. Transport channels or channels having a maximum value of f _k , or a value of f _k above some predefined limit, are used to transmit data using scheduled resources. In this regard, the transmission channel k may for example represent a physical propagation channel between a single or multi-antenna transmitter and a single or multi-antenna receiver. There may be only one or several transmitters, and there may be one or several receivers correspondingly. The transmitters and receivers may represent base stations, mobile stations, or relay stations in a wireless communication system. The transport channel k may be defined for both uplink and downlink. The transmit channel k may further be associated with a channel between a subgroup of transmit antenna elements of a multi-antenna transmitter and a single or multi-antenna receiver, or a channel between subgroups of single or multi-transmitter and receiver antenna elements.

The preferred function f _k for transport channel k can be:

f _k = f (u, d _k , z _k , h _k , c _k , CQIm _k , CQIe _k , CQIs _k , ...)

Here, u is a parameter representing the general state of the transmission queue, and the remaining parameters are unique for each transmission channel k = 1, ..., K. Thus, d _k is the delay experienced by transport channel k, i.e., the time that the next packet from / from transport channel k is spent in the queue, and z _k is scheduled for the user (or scheduled by the user for transmission during uplink operation). Is the size of the next packet, h _k represents the history taking into account the amount of data transmitted over the transport channel k immediately before, and c _k represents the subscription type, terminal, and / or data type It is a potential priority class. Link-specific Channel Quality Indicators (CQI) that scheduling can be based on are as follows.

CQIm _k , which determines the transmission mode of the transport channel k. The mode selection may be based on side information based on measurements at the transmitter or feedback from the receiver. In particular, CQIm _k is the rate of transmission channel k, and possibly feedback based beamforming, selected adaptive space-time modulation (matrix / vector modulation, space-time coding), concatenated channel code. Rate, and / or modulation alphabets.

Indicates the expected error rate of the transmission mode indicated by CQIe _k , CQIm _k .

CQIs _k , the speed of the transport channel k, ie the channel coherence time.

Based on these parameters (or a subset thereof), the scheduler can determine which transport channel is suitable for data transmission in consideration of the overall throughput, fairness, delay, etc. in the application. A portion of f _k that is not dependent on the radio link is the domain of higher layer protocols. However, CQIs are closely related to the physical layer and must be designed according to physical layer algorithms.

Given the physical layer, the preference functions f _k can be used to minimize transmit power, maximize rate (rate), or optimize performance. Often, the physical layer and its higher layer algorithms include an automatic retransmission request (ARQ) protocol. In such cases, maximizing system capability typically requires that the frame or block error rate be close to a given optimal value. Thus, the preference functions f _k support the highest rate with the maximum allowable transmit power and can be used to select a transport channel that reaches the desired error performance.

When scheduling is based on CQIe _k , ie, CQIs that correlate with expected error performance, the most obvious shortcut to measuring scheduling performance is to measure select diversity performance. This means that K transport channels with some type of channel statistics are considered. Due to the statistical nature of these K transport channels, it is advisable to presume a particular transport channel among the K transport channels as long as it has substantially the best transport channel characteristics, such as the lowest fading or path loss in data transmission. Excellent performance can be achieved compared to transmitting on the selected transport channel. Thus, multi-user diversity is possible in which a single transceiver of a communication system transmits and receives signals, or a signal transmitted by a transmitter using only a subset of multiple receiver elements and a signal using only a subset of multiple transmitter elements. It is possible to use antenna diversity in a receiver with the proper placement of multiple antenna elements to transmit, or frequency diversity when signals are transmitted in different parts of the available channel bandwidth, so that various transceivers or elements thereof can be used. Selective divergence can be used in the scheduler when given a set of freedoms to choose from among the various transport channels.

For optimal performance of select diversity, a transport channel with the lowest predicted bit error rate (BER) can be scheduled for transmission. This results in a reduction in the average error rate. However, for multiple input multiple output (MIMO) channels, the scheduling based on the direct evaluation of BER requires excessive computation.

A.G., a prior art document published by IEEEcomm in June 2001. Kogiantis, N.Joshi, and O. Sunay, "Transmit Diversity and Scheduling over Wireless Packet Data," Vol. 8, pages 2433-2437, discloses scheduling on multiple input single output (MISO) channels. First of all, the orthogonal diversity transmission scheme (space-time transmission diversity, STTD) is compared with the single antenna transmission scheme without diversity. In both schemes, a transmitter (either with two transmit antennas and STTD, or with a single antenna) schedules one of the K transmit channels, with each individual transmit channel being two or more of the transmitters for data transmission. Defined by the physical propagation channel between one transmit antenna and one antenna of each receiver, the scheduling is based on the maximum carrier-to-interference power ratio (C / I) for each transmission channel.

STTD schemes, such as those used in Kogiantis citations, can be considered matrix modulation schemes, which are, for example, modulation symbols, such as time slots or symbol periods, frequency carriers, orthogonal codes, or a combination thereof. It can be defined as a mapping onto non-orthogonal and spatial resources and orthonormal resources.

In matrix modulation schemes, diversity can be applied. For example, in a space-time matrix modulation scheme, at least one of the modulation symbols may be mapped to a first antenna element in a first symbol period and to a second antenna element in a second symbol period. Similarly, in the space-frequency matrix modulation scheme, at least one of the modulation symbols may be mapped to a first antenna element and transmitted over a first carrier frequency, and mapped to a second antenna element and transmitted over a second carrier frequency. have.

If modulation symbols are mapped to only non-normal orthogonal spatial resources, so-called vector modulation is performed and only spatial diversity is available.

As a follow up to this concept introduction, space-time matrix modulation methods can be proposed as examples of normal orthonormal matrix modulation. Thus, orthogonal resources are represented in T symbol periods (or time slots) to which a block of data symbols is mapped. However, the proposed matrix modulation methods can be easily applied to matrix modulation using frequency domain, code domain, eigenmode domain or polarization domain as orthogonal resources instead of time domain.

A space-time matrix modulator using N _t transmit antennas and T symbol periods is defined as TXN _t modulation matrix X. The modulation matrix X is the Q complex-valued modulation symbols x _n to be transmitted by the N _t antenna transmitters during the T symbol periods. It is a linear function of (n-1, .., Q). The modulation symbols are, for example, according to Binary Phase Shift Keying (BPSK), Quaternary Phase Shift Keying (QPSK) or Quadrature Amplitude Modulation (QAM) symbol alphabet. Thus, modulation matrix X is basically the modulation symbols x _n where n = 1, .., Q And / or -x _n , x _n * Or-x _n The same functions of the modulation symbols are defined when transmitted from transmit antennas n _t = 1, .. N _t at time instance t1, .. T. In this regard, the superscript "*" denotes one complex conjugate. Now matrix modulation, the modulation symbols and the function of, the N T symbol periods _t transmitted by the respective transmitting antenna elements of the N _t transmit antennas in It can be understood as mapping to individual data streams. In the STTD scheme as applied in the Kogiantis reference, the modulation matrix X is defined according to the so-called Alamouti space-time code, such as the following equation of T = 2 and N _t = 2.

At the N _{k , r} receive antennas at the receive side of transmit channel k, the signal model is read as follows:

Y _k is a TXN _{k , r} matrix of signals y _{k , ij} received at receive antenna elements j = 1, .. N _{k , r} during symbol period i = 1, .. T, i = 1, .. The elements h _{k , ij} of N _t XN _{k , r} channel matrix H _{k with} , N _t and j = 1, .., N _{k , r} are flat fading between transmit antenna element i and receive antenna element j Creating a propagation channel, TXN _{k , r} matrix N _{k , r} represents the noise received within symbol period t = 1, .., T at receive antenna elements n _k = 1, .., N _{k , r} .

The matrix modulation effects defined by matrix X and the propagation effects defined by channel matrix H _k are put together to yield a complex value TN _{k, r} XQ equivalent channel matrix G _k having the following equivalent signal model:

The vector y _k in the TN _{k , r} dimension is the signals received during the T symbol periods, or functions thereof, in the N _{k , r} receiving antenna elements at the receiving side of the transmission channel k, i.e. the conjugate complex received signal, negative A received signal, or a negative conjugate complex received signal, or the like; Q-dimensional vector x includes Q complex-valued modulation symbols x _n (n = 1, .., Q) modulated by matrix modulation on the N _t transmit antenna elements within T symbol periods; TN _{k , r} dimensional vector n _k includes noise received at receive antenna elements during T symbol periods.

In the STTD scheme, the equivalent channel matrix G _k has the following form:

And the equivalent system model is read as:

From Equation 5, both effects of matrix modulation and channel propagation are now included in the equivalent channel matrix G _k , thus the sickle of the original modulation symbols x _n and this modulation symbol x _n in the equivalent signal model of Equation 5 It is easy to see that the (not) functions are multiplied by the equivalent channel matrix G _k . Through the equivalent channel matrix G _k , the equivalent channel correlation matrix (ECCM) R _k can be defined as follows.

The superscript operator "H" represents the Hermitian conjugate of the matrix.

In the STTD scheme, it can be easily seen that ECCM is a diagonal matrix as follows.

The ECCM can be interpreted as applying a matched filter to the equivalent channel matrix G _k , so that the matched filter estimate of the vector x containing the Q modulated symbols x _n with n = 1, .. Q is Obtained together.

는 기본적으로

항에 의해 주어진다, 즉, 정합 필터 추정치들

은, 단순히, i=1,..,N_t 송신 안테나들과 i=1,..,N_{k ,r} 수신 안테나들 사이의 채널들 h_{k , ij}의 세기들의 합으로서 정해지는 실수 팩터

에 의해 스케일링된 변조 심볼들 x_n이다. Matched filter estimate when the magnitude of the noise contribution n _k is negligible compared to the magnitude of the received signal y _k

Is basically

Given by the term, i.e. matched filter estimates

Is simply a real factor determined as the sum of the intensities of the channels h _{k , ij} between i = 1, .., N _t transmit antennas and i = 1, .., N _{k , r} receive antennas

Is the modulation symbols x _n scaled by.

내 변조 심볼들 x_n 사이의 자기-간섭(self-interference)을 초래하지 않는다. 게다가, 대각선 상의 요소들은 모두 동일하므로, ECCM은 단위 매트릭스에 비례한다. 단위 매트릭스에 비례하는 ECCM을 통한 매트릭스 변조 방식들이 결국 정규직교(orthonormal) 매트릭스 변조 방식들로서 표시되는데 반해, 단위 매트릭스에 비례하지 않는 ECCM들을 통한 매트릭스 변조 방식들은 비정규직교(non-orthonormal) 매트릭스 변조 방식들로 표시된다.ECCM according to the STTD matrix modulation scheme includes only nonzero entries on the main diagonal, so matched filter estimates

Does not result in self-interference between my modulation symbols x _n . In addition, since the elements on the diagonal are all the same, the ECCM is proportional to the unit matrix. Matrix modulation schemes with ECCM proportional to the unitary matrix are eventually represented as orthonormal matrix modulation schemes, whereas matrix modulation schemes with ECCMs not proportional to the unitary matrix are non-orthonormal matrix modulation schemes. Is displayed.

1 and 2 show practical features related to scheduling in setting STTD adoption selection diversity as disclosed in Kogiantis reference. In particular, the scheduling performance was investigated with N _{k , r} = 1 and matrix modulation (N _t = 1) for several transport channels K and normal orthogonal matrix modulation (STTD, N _t = 2), Here, the transmission channel selection for data transmission is based on the maximum C / I at the receiving side of the transmission channel.

Figure 1 shows the scheduling performance based on the ratio E _b / N _o log ₁₀ (BER) as a function of dB unit of energy E _b and noise power density N _o per bit (additive white Gaussian noise (AWGN) on the receiving side transmission channels Process). The results of QPSK modulation and independent identically distributed flat Rayleigh fading channels, respectively, do not employ STTD (dotted line) or do not (solid line) K = 1, 2, 4, And as separate figures for 8 transport channels.

In the presence of transport channels up to K = 4, we observe that STTD adoption plans are superior to scheduling without transmit diversity, which is not (coded) by the STTD adoption BER (uncoded) over the entire E _b / N _o regime. Because it is much smaller than STTD non-adopted BER. However, K = 8 in the transport channel, the scheduling is not employed for the STTD STTD based scheduling BER 10 ^- shows a better performance ^through. In the high E _b / N _o regime, the transmit diversity of STTD still leads to better performance.

인 i.i.d. 레이리 페이딩 채널들에 대해 그려지고 있다. 줄어든 외연과 PDF들의 현저한 최대치들로부터 어떻게 STTD가 채널 전력의 변동(fluctuation)을 줄이는지를 명확히 알 수 있다. 또, STTD는 채널의 평균 SNR을 바꿀 수 없고, 다만 양호한 채널들을 악화시키는 것을 감수하면서 불량 채널들을 개선시킨다는 것이 분명하다. 따라서, 한 송신 안테나로부터 강한 채널을 가지고, 다른 것으로부터 약한 채널을 가지는 전송 채널들의 주요 부분은 MRC 합성 후에 더 악화된 수신 SNR을 가지게 된다 (도 2의 위 좌측 참조). 증가하는 개수 K의 전송 채널들을 통해, 비 STTD 전송은 STTD가 적용되었다면 더 악화된 채널을 가졌을 전송 채널로 항상 스케줄링될 수 있다. 따라서, 다중 전송 채널들 K에 있어, STTD를 채택하지 않는 스케줄링이 더 양호하게 작동한다.The reason for this can be seen in FIG. 2, where the probability distribution functions (PDFs) of the signal-to-noise ratio (SNR) at the scheduled receiver (adopting the maximum ratio synthesis method (MRC) for the STTD), again. For STTD once adopted (dotted line) and not adopted (solid line) and for a different number of transport channels K = 1, 2, 4, and 8 where scheduling is performed

Iid is drawn for the Rayleigh fading channels. From the reduced margins and the significant maximums of the PDFs it is clear how STTD reduces the fluctuations in channel power. Moreover, it is clear that STTD cannot change the average SNR of a channel, but improves bad channels while taking the deterioration of good channels. Thus, the main portion of the transmission channels with a strong channel from one transmit antenna and a weak channel from the other will have a worse receive SNR after MRC synthesis (see upper left in FIG. 2). With an increasing number of K channels, non-STTD transmissions can always be scheduled to transport channels that would have had worse channels if STTD was applied. Thus, for multiple transport channels K, scheduling that does not employ STTD works better.

The situation related to the synthesis of Orthonormal Matrix Modulation (STTD) and scheduling as disclosed in Kogiantis Reference can be summarized as follows:

For more than K = 5 transmission channels, orthogonal matrix modulation (STTD) yields worse performance than single antenna transmission, a component of the select diversity scheme with maximum C / I based scheduling at BER 10 ^-3 . see.

If more than K = 3 transport channels are scheduled at the same time, STTD and other orthonormal matrix modulations are virtually useless.

In view of the above-mentioned problems, an object of the present invention is to improve a communication system for performing scheduling by combining matrix and vector modulation in selecting diversity.

는 단위 매트릭스에 비례하지 않을 때, 상기 방법은, 상기 K 전송 채널들 중 적어도 하나에 대해 각자의 채널 품질 표시자 (CQI) q_k를 계산하는 단계, 및 상기 매트릭스 변조된 데이터 심볼들의 전송을 위해 상기 K 전송 채널들 중 적어도 하나를 예정하는 단계를 포함하고, 상기 예정은 적어도 부분적으로 상기 계산된 CQI들인 q_k에 기반한다. _{K over} each N _{t , k} transmission interfaces and each N _{r , k} receive interfaces for transmission of data symbols modulated according to a nonnormal orthogonal matrix modulation that modulates data symbols in both the non-orthogonal spatial domain and at least one orthogonal domain. A method for scheduling at least one channel in k = 1, ..., K is proposed, wherein the data symbols are transmitted through N _{t , k} transmission interfaces after matrix modulation and thus the transport channels k = At least one equivalent channel matrix (ECM) G _k may be defined, which translates into data symbols received at the N _{r , k} receive interfaces of one channel of 1, .. K, wherein the at least one ECM G equivalent channel correlation matrix of _k (ECCM)

Is not proportional to a unitary matrix, the method further comprises: calculating a respective channel quality indicator (CQI) q _k for at least one of the K transmission channels, and for transmitting the matrix modulated data symbols. Scheduling at least one of the K transport channels, the scheduling being based at least in part on the calculated CQIs q _k .

The transmission channel k represents, for example, a physical propagation channel between a single or multiple antenna transmitter and a single or multiple antenna receiver. There may be only one or several transmitters and correspondingly there will be one or several receivers. The transmitters and receivers may refer to base stations, mobile stations, or relay stations in a wireless communication system. The transport channel k may be defined for both uplink and downlink. The transmission channel k may also be associated with a channel between subgroups of only the transmit antenna elements of a multi-antenna transmitter or between subgroups of antenna elements of a single or multiple antenna transmitter and receiver.

Data symbols, which may represent, for example, BPSK, QPSK, or QAM modulation symbols, are matrix modulated and transmitted over respective transmission interfaces of the transmission channel k. The matrix modulation is the mapping of data symbols to (real) orthogonal and non-orthogonal spatial resources such as time slots, frequency carriers, orthogonal codes, spatial or polarity eigenmodes, or a combination thereof. Indicates. The matrix modulation will be defined by a modulation matrix that is a linear function of the data symbols to be transmitted over the N _{t , k} transmission interfaces of the transmission channel k. In space-time matrix modulation, the modulation matrix may specify when data symbols and / or functions of the data symbols are transmitted from any transmission interface during any symbol period. Space-time modulation matrix is now, the data symbols and their function, may be understood as the mapping to the N _t of each data stream to be transmitted by N _t of each transport interface of the T symbol periods. Similarly, in space-frequency matrix modulation, the modulation matrix defines which symbols are sent from which transmission interface on which frequency. Instead of time and frequency, additional orthogonal or substantially orthogonal domains may be used, such as code domains, polar domains, eigenmode domains, and so forth.

The transmission over the transmission channels k = 1,..., K of the matrix modulated data symbols is indicative of the attenuation and phase shift experienced when data symbols are transmitted on individual channels between a transmission interface and a reception interface. It can be defined by a transport channel matrix containing individual entries. The transmission channel may be a wired or wireless transmission channel.

By combining the effects of the modulation matrix and the transmission channel matrix or within the equivalent channel matrix G _k , the equivalent channel matrix now uses the matrix modulation of data symbols and the transmission channel k between transmission interfaces and receiving interfaces. Represents transmission and reception of modulated data symbols. The transmission interfaces of the transmission channel may be connected with one or multiple transmitters, and correspondingly, the reception interfaces of the transmission channel may be connected with one or more receivers.

Since the ECCM R _k of the equivalent channel is not proportional to the unitary matrix, the matrix modulation scheme is classified as a non-normal orthogonal matrix modulation scheme. Thus, when a matched filter is used at the receiving side of transmission channel k, its matched filter estimate of data symbols undergoes self-interference between the data symbols. Non-normal orthogonal matrix modulation schemes can be made through linear combinations of orthonormal matrix modulation techniques, such as space-time codes, space-frequency codes, and the like.

The scheduling may include scheduling scheduling at the transmitting side or the receiving side of the transmission channel to determine which of the K transmission channels is most suitable for use in transmitting data symbols that are matrix modulated and subsequently transmitted to the receiving side of the transmission channel. instance).

The scheduling is based, at least in part, on the CQIs q _k that have been calculated for at least one of the K individual transport channels. CQIs q _k are preferably calculated for at least two of the K individual transport channels. The scheduling can then be determined according to the comparison of the CQIs which transport channel is more suitable for the transmission of data symbols. Scheduling may also be done in such a way that if a CQI for a transport channel exceeds a defined threshold, that transport channel is used—thus in some cases only one CQI may be sufficient.

The CQIs do not need to be calculated for every K transport channels for which selection is made: prior knowledge of the CQIs of the transport channels or any other means of avoiding repeated calculations may also be applied.

The K transmission channels to be selected may not all use the same modulation scheme, for example, some of them may use orthonormal matrix modulation, others may be non-orthogonal matrix modulation, vector modulation, single antenna transmission or diver. City transport can be used. Correspondingly, the CQIs for the individual transport channels can be calculated in their own way according to the modulation technique applied to the data symbols to be transmitted on the respective transport channel.

At least one ECM G _k may be defined to represent non-normal orthogonal matrix modulation of the data symbols and their transmission over a particular transmission channel, which may all be defined by the channel matrix H _k . Due to the fact that matrix modulation is nonnormal orthogonal, the ECCM R _k of ECM G _k is now not proportional to the unit matrix. The ECCM can be a non-diagonal matrix or a diagonal matrix containing entries that are not equal on the diagonal. Non-diagonality reflects the fact that interference occurs between matrix modulated data symbols that can be simply eliminated through matched filtering at the receiving side of a particular transmission channel. If matrix modulation is performed on two or more of the K transport channels, two or more ECM G _k may be defined, each with ECCM R _k . However, ECCMs R _k of transmission channels on which non-normal orthogonal matrix modulation is performed Only will not be proportional to the unit matrix, and ECCMs of the transmission channels using orthonormal matrix modulation will be proportional to the unit matrix.

Depending on the definition of the transmission channels, scheduling may be applied to select between different users in the uplink or downlink of a wireless communication system or to select between different groups of Tx antennas, where matrix modulation is a subgroup of selected antennas. It is used in the phase. In addition, if any receiver has fewer RF-circuits than the antenna elements, this receiver will select the best antenna / antenna to receive the transmission. When the transmitter uses matrix modulation, each possible choice for receive antennas at the receiver is interpreted as a transmission channel according to the present invention. The receiver may then select (a group of) receive antenna (s) based on the CQIs disclosed in this invention.

Scheduling may be performed at the time, frequency, or code level, or matrix modulation may be performed at frequency, for example, but scheduling may be performed through a combination of these, such as at both time and frequency. For example, based on the calculated at least one CQI, the scheduler can determine which transport channel is scheduled for the next time slot or which transport channel is scheduled for subcarrier, polarity, spatial eigenmode, and so on.

Scheduling may also be applied in handover situations where a selection is made between several base stations in uplink or downlink transmissions. In the downlink, this will require cooperation between several base stations.

One or more transmission channels may be scheduled for transmission of matrix modulated data symbols simultaneously. At the receiving side of the transmission channel, the receiver may have one or more receiving interfaces and apply receive diversity synthesis techniques.

Thus, in contrast to prior art schemes for performing scheduling in combination with orthogonal matrix modulation in setting selectivity, which has proven inferior to scheduling without transmit diversity for an increasing number of transport channels, The predestination is performed in combination with non-normal orthogonal matrix modulation in setting select diversity. The non-normal orthogonal matrix modulation scheme can be thought of as introducing additional diversity into equivalent channels, thereby proving better than both the normal orthogonal matrix modulation scheme and the no transmit diversity scheme.

According to the method of the present invention, it is preferred that the respective CQI q _k for at least one of the K transport channels is derived from the ECCM R _k .

At least one ECCM R _k corresponding to at least one ECM G _k includes both effects of matrix modulation and transport channel and is therefore particularly suitable as a basis for scheduling. It is preferred that equivalent channel matrices G _k are known by the scheduler when scheduling is performed.

According to the method of the present invention, it is preferred that the respective CQI q _k for at least one of the K transport channels is calculated as a function that is a determinant of the linear function of the ECCM R _k .

을 고려하고, 사용된 전송 및 검출 방식에 따라 상수

를 조정함으로써 조금 더 개선될 수 있다. 상수

는 0과 같은 것으로도 선택될 수 있다. 상기 행렬식의 함수는 이를테면 상기 행렬식의 거듭제곱이나 근, 또는 그 배수, 혹은 수학 함수들일 수 있다.The determinant of ECCM R _{k has} a strong correlation with the BER obtained when data symbols are matrix modulated and transmitted over transmission channel k, as defined by equivalent channel matrix G _k . The strong correlation to BER makes the scheduling based on the expected error rate of the transmission, allowing the scheduling to reduce the average error rate of the wireless communication system. Correlation is a determinant of a constant matrix multiplied by ECCM R _k , that is,

Taking into account the constants

It can be improved further by adjusting. a constant

May also be selected to be equal to zero. The determinant function may be, for example, the power or root of the determinant, or a multiple thereof, or mathematical functions.

According to the method of the invention, it is also preferred that for each of the K transport channels the respective CQI q _k is calculated as the ECCM R _k trace function.

The trace of the ECCM also shows a strong correlation with the BER obtained when the data symbols are matrix modulated and transmitted over transmission channel k, as defined by equivalent channel matrix G _k . In addition, unstructured ECCM R _k The computational cost of producing a trace is relatively low.

According to the method of the present invention, for at least one of the K transport channels, it may be desirable for each CQI q _k to be calculated as a trace function of the inverse ECCM R _k .

The inverse ECCM trace also exhibits a strong correlation with the BER obtained when the data symbols are matrix modulated and transmitted over transmission channel k, as defined by equivalent channel matrix G _k .

의 행렬식이 이러한 방식으로 해석될 수 있는데, 여기서 I는 단위 매트릭스이고

는 상수이다. 이것은 캐리어 대 간섭 및 잡음 전력 비율

로서 확장될 수 있고, 이러한 확장 계수들이 R_k의 트레이스, R_k의 행렬식, 및 기타 R_k의 다른 대칭적 다항식들이다. 상수

는 사용된 전송 및 수신 방식에 따라 조정될 수 있다. Alternatively, any combination of the above-described calculation schemes for CQIs, such as a determinant function, ECCM trace or inverse ECCM trace, or any other ECCM function can be taken as CQI. For example, the matrix

The determinant of can be interpreted in this way, where I is the unit matrix

Is a constant. This is the carrier-to-interference and noise power ratio

It can be extended as and, these expansion coefficients R _k are the traces, the matrix R _k, and other symmetric polynomial of other R _k. a constant

Can be adjusted according to the transmission and reception scheme used.

According to the method of the present invention, it may be desirable for each of the K transport channels given at least one CQI q _k to be calculated as a function of the elements of the channel matrix H _k defining the at least one transport channel and , in which the function is derived from the ECCM R _k under utilization of the structural property of the ECCM R _k.

The channel matrix H _k defines a physical propagation channel between the respective transmission interfaces and the respective reception interfaces of the at least one transport channel. The channel matrix H _k preferably defines a transport channel on which the at least one ECM G _k can be defined. The equivalent channel matrix G _k then represents a combination of the modulation matrix and the channel matrix H _k . Due to the structure of the modulation matrix inherited by the equivalent channel matrix G _k , ECCM R _k has a structure that can be utilized to reduce the amount of computation when calculating the determinant or trace or other functions of ECCM R _k . In particular, a closed form representation of CQI q _k can be derived which can only depend on single elements of the channel matrix H _k . ECCM R _k when calculating CQI q _k for transport channel k during scheduling It calculates its own and must be, instead, is only closed form representation of the elements of an effective, the channel matrix H _k into the calculated phase calculation to calculate the CQI from the ECCM q _k R _k. In preferred embodiments, the simple closed form representations greatly reduce the complexity in computing the CQI q _k for matrix modulators to below 2Q square root, where Q is the number of data symbols in the matrix modulator.

The structure of ECCM R _k can also be equally well utilized at the receiving side of the transmission channel when estimating received matrix modulated data symbols using equalization or multi-user detection. In particular, the calculation of the determinant and reciprocal of ECCM R _k is greatly simplified by utilizing the structure.

If all K transmission channels k = 1, ..., K are suitable for transmission of non-normally orthogonal matrix modulated data symbols, then each of the transmission channels k can be characterized by its respective channel matrix H _k , and then Each ECM G _k , each ECCM R _k and each CQI q _k will be calculated for each of these K transport channels from each R _k or H _k . The schedule may then be based on comparing all of their respective CQIs q _k .

According to a first embodiment of the present invention, said non-normal orthogonal matrix modulation maps a block of four data symbols onto N _{t , k} = 4 transmission interfaces and spaces two spaces within four units of said at least one orthogonal domain. It is a so-called "ABBA" nonnormal orthogonal matrix modulation based on nonnormal orthogonal synthesis of time transfer diversity (STTD) codes. Since the units of the at least one orthogonal domain may be time slots or frequency subcarriers, the non-normal orthogonal matrix modulation is either space-time or space-frequency matrix modulation, respectively. An example of a normal orthogonal STTD code is given as a transpose matrix of the Alamouti code of equation (1).

According to the first preferred embodiment of the present invention, the ECCM R _k has the following form.

p _k and n _k are real functions of the elements of the channel matrix H _k .

This structure of ECCM R _k is encountered in so-called “ABBA” non-normal orthogonal matrix modulation, for example. When applied as a space-time matrix modulation, the modulation matrix includes two STTD blocks X _A and X _B (cf. transpose of Equation 1) that modulate 4 data symbols onto 4 transmission interfaces during a 4 symbol period. By doing so, one symbol is actually transmitted per symbol period (so-called matrix modulation of symbol rate 1).

의 함수로서 산출됨이 바람직하고, 여기서

는 상수이다. 따라서, 상기 각자의 CQI q_k가, 상기 채널 매트릭스 H_k의 요소들에만 의존하는 실수 함수들인 p_k 및 n_k로부터 효과적으로 계산된다. 상수

는 사용된 전송 및 수신 방식에 따라 조정될 수 있다.According to the first preferred embodiment of the present invention, the respective CQI q _k is equal to at least one of the K transport channels.

Preferably calculated as a function of

Is a constant. Thus, the respective CQI q _k is effectively calculated from p _k and n _k , which are real functions that depend only on the elements of the channel matrix H _k . a constant

Can be adjusted according to the transmission and reception scheme used.

According to a second preferred embodiment of the present invention, said non-normal orthogonal matrix modulation maps a block of eight data symbols onto N _{t, k} = 4 transmission interfaces and four STTDs within four units of the at least one orthogonal domain. It is a so-called DABBA nonnormal orthogonal matrix modulation based on nonnormal orthogonal synthesis of codes. Since the units of the at least one orthogonal domain may be time slots or frequency subcarriers, the non-normal orthogonal matrix modulation is either space-time or space-frequency matrix modulation, respectively. An example of a normal orthogonal STTD code is given as a transpose matrix of the Alamouti code of equation (1).

According to a second preferred embodiment of the present invention, the ECCM R _k has the following form.

p _{k, 1} , p _{k, 2} , n _{k , 1} and n _{k , 2} are real functions of the elements of the channel matrix H _k and s _k is a complex function of the elements of the channel matrix H _k . This structure of ECCM R _k meets in a so-called “DABBA” non-normal orthogonal matrix modulation scheme, for example. When applied as a space-time matrix modulation scheme, the modulation matrix comprises a linear combination of 4 STTD blocks X _A , X _B , X _C and X _D which modulate 8 data symbols onto 4 transmission interfaces during a 4 symbol period. By doing this, two symbols are actually transmitted per symbol period (the so-called matrix modulation of symbol rate 2).

의 함수로서 산출됨이 바람직하고, 여기서

는 상수이다. 따라서, 상기 각자의 CQI q_k가, 상기 채널 매트릭스 H_k의 요소들에만 의존하는 실수 함수들인 p_k,1, p_k,2, n_{k ,1}, 및 n_{k ,2} 그리고 복소수 함수 s_k로부터 효과적으로 계산된다. 상수

는 사용된 전송 및 수신 방식에 따라 조정될 수 있다.According to the second preferred embodiment of the present invention, the respective CQI q _k is equal to at least one of the K transport channels.

Preferably calculated as a function of

Is a constant. Accordingly, from the CQI q _k of the respective, p _{k, 1,} which are real number function that depends only on the elements of the channel matrix H _k p _{k, 2,} n _{k, 1,} and n _{k, 2,} and the complex function s _k Is calculated effectively. a constant

Can be adjusted according to the transmission and reception scheme used.

According to a third preferred embodiment of the present invention, the non-normal orthogonal matrix modulation maps a block of four data symbols onto N _{t , k} = 2 transmission interfaces and two STTDs in two units of the at least one orthogonal domain. TSTTD nonnormal orthogonal matrix modulation based on nonnormal orthogonal synthesis of codes.

Since the units of the at least one orthogonal domain may be time slots or frequency subcarriers, the non-normal orthogonal matrix modulation is either space-time or space-frequency matrix modulation, respectively. An example of a normal orthogonal STTD code is given as a transpose matrix of the Alamouti code of equation (1).

According to a third preferred embodiment of the present invention, the ECCM R _k has the following form.

p _{k, 1} and p _{k, 2} are real functions of the elements of the channel matrix H _k and s _k is a complex function of the elements of the channel matrix H _k . This structure of ECCM R _k is experienced, for example, in a so-called "TSTTD (Twisted or Twice STTD)" nonnormal orthogonal matrix modulation scheme. When applied as a space-time matrix modulation scheme, the modulation matrix is represented as the sum of STTD block X _A and STTD block X _B that modulates 4 data symbols on 2 transmission interfaces during a 2 symbol period (cf. Transpose matrix), actually 2 symbols per symbol period are to be transmitted (the so-called symbol rate 2 matrix modulation).

의 함수로서 산출됨이 바람직하고, 여기서

는 상수이다. 따라서, 상기 각자의 CQI q_k는 ECCM R_k의 실제 계산을 요구하지 않고 채널 매트릭스 H_k로부터 직접 결정될 수 있다. 상수

는 사용된 전송 및 수신 방식에 따라 조정될 수 있다.According to a third preferred embodiment of the present invention, the respective CQI q _k is equal to at least one of the K transport channels.

Preferably calculated as a function of

Is a constant. Thus, the respective CQI q _k can be determined directly from the channel matrix H _k without requiring the actual calculation of ECCM R _k . a constant

Can be adjusted according to the transmission and reception scheme used.

According to the method of the present invention, said non-normal orthogonal matrix modulation maps a block of four data symbols onto N _{t , k} = 4 transmission interfaces in two units of said at least one orthogonal domain and denormal orthogonality of two STTD codes. It is desirable to be a DSTTD nonnormal orthogonal matrix modulation based on the synthesis.

Since the units of the at least one orthogonal domain may be time slots or frequency subcarriers, the non-normal orthogonal matrix modulation is either space-time or space-frequency matrix modulation, respectively. An example of a normal orthogonal STTD code is given as a transpose matrix of the Alamouti code of equation (1).

According to the method of the invention, said non-normal orthogonal matrix modulation comprises a combination of at least two orthonormal matrix modulations. For example, the modulation matrix defining the matrix modulation can be the sum of two orthonormal STTD blocks, resulting in a rate 2 non-orthonormal matrix modulation instead of rate 1 orthonormal matrix modulation represented by one STTD block.

의 선형 함수의 행렬식 함수로서 산출되는 단계, 및 상기 데이터 심볼들의 전송을 위해 상기 K 개의 전송 채널들 중 적어도 하나를 스케줄링하고, 이때 상기 예정은 상기 산출된 CQI들 q_k에 부분적으로 기반하는 단계를 포함한다.Further proposed is a method for scheduling at least one from K transport channels k = 1, .. K via respective N _{t , k} transmission interfaces and respective N _{r , k} reception interfaces for transmission of data symbols and At least two of the data symbols are transmitted side by side from at least one of the K transmission channels N _{t , k} transmission interfaces, and when this is defined by a channel matrix H _k , the method transmits the K transmission channels. Calculate a respective channel quality indicator (CQI) q _k for at least one of the above, wherein the respective CQIs q _k At least one of the channel correlation matrix of the channel matrix _Hk

Calculating as a determinant function of the linear function of and scheduling at least one of the K transmission channels for transmission of the data symbols, wherein the scheduling is based in part on the calculated CQIs q _k . Include.

At least two of the data symbols are transmitted side by side from the respective N _{t , k} transmission interfaces, eg, one data symbol may be transmitted from the first transmission interface and the second data symbol is transmitted from the second transmission interface. Can be. If the transmission interfaces are antenna elements of the transmitter, the two data symbols interfere with each other when received at the reception interfaces of transmission channel k. Transmitting the at least two data symbols side by side can be considered a vector modulation scheme, where the data symbols are mapped to non-orthogonal spatial resources. A representative of such a modulation scheme is BLAST (Bell Laboratories Layered Space Time Architecture). In contrast to matrix modulation schemes, diversity is used only within the spatial domain without using orthogonal domains such as the time, frequency or polar domains.

의 행렬식, 또는 그것의 다른 함수들은, 데이터 심볼들이 채널 매트릭스 H_k에 의해 규정된 대로 전송 채널 k를 이용해 송신기로부터 전송될 때 얻어지는 BER과의 강한 상관관계를 보인다.The matrix

The determinant of, or its other functions, shows a strong correlation with the BER obtained when data symbols are transmitted from the transmitter using transmission channel k as defined by channel matrix H _k .

Based on the interpretation of the transport channels and scheduling scenarios, the above description of transport channels and scheduling scenario types is equally true for vector modulation.

According to the method of the present invention, at least one receiver is transmitted over the predetermined transmission channel and received by the receiver over the receiving interfaces of the predetermined transmission channel using a Maximum Likelihood algorithm or a linear estimator. It is desirable to estimate the data symbols.

Due to the use of nonnormal orthogonal matrices or vector modulation schemes, estimation of data symbols based on matched filtering of matrix symbols transmitted and received with matrix / vector modulation is self-sufficient to make correct estimation of data symbols impossible. May include interference. Thus, it is desirable to perform linear equalization techniques of maximum possible or reduced complexity, such as a zero-forcing algorithm, a minimum mean square error algorithm, or the like. In both schemes, it would be desirable to be based on a sufficient set of statistics obtained through matched filtering on data symbols received over the receive interfaces of the transmission channel.

According to the method of the present invention, it may be preferred that a transmission channel k = 1, .. K having a maximum CQI q _k is intended for the transmission of the data symbols.

In particular, when all K transport channels use the same matrix modulation, it is preferable that transport channels k = 1, .., K having a maximum CQI q _k are expected. BER is substantially close to the strict monotone decreasing function of the trace of the matrix equation R _k significance of ECCM linear function and ECCM R _k. Thus, when the target BER is determined, the lowest BER or lowest power consumption is obtained when transmitting data over the transmission channel with the maximum CQI q _k . Alternatively, one may want to minimize the transmit power used for the expected error potential. This is done in a similar manner using the same CQI q _k .

When different transport channels use different matrix modulation schemes, perhaps at different rates, the calculated CQIe _k characterizing the error probability of the transport channels is used with modulation types and rates characterized by CQIm _k . In order to reach the expected error probability as the transmission channel providing the highest rate with the given power and the expected error probability, or the transmission channel providing the lowest error probability with the expected rate and the given power, or Select the transmission channel that uses the lowest power. To do this, the error probabilities for each transmission mode can be approximated on a log-log scale as linear functions of CQIe _k with appropriate parameters that will provide the best suitability. Approximation functions are very accurate and can be used to find solutions to minimization problems.

According to the method of the present invention, the transmission channels are transmission channels of a wireless communication system, and the transmission and reception interfaces of the transmission channels are transmission and reception antenna elements of one or several transmitters and one or several receivers, respectively. This is preferable.

The wireless communication system may be, for example, a system according to Universal Mobile Telecommunications System (UMTS), Global System for Mobile Communications (GSM), or Interim Standard 95 (IS-95). Alternatively, the system may be a wireless local area network (W-LAN) system, such as a HIPERLAN / 2 or 802.11 system.

Further proposed is a computer program comprising instructions operative to cause a processor to perform the steps of the method described above.

Further proposed is a computer program product having a computer program with instructions operable to cause a processor to perform the steps of the method described above.

는 단위 매트릭스에 비례하지 않을 때, 상기 장치는, 상기 K 개의 전송 채널들 중 적어도 하나에 대해 각자의 채널 품질 표시자 (CQI) q_k를 산출하는 수단, 및 상기 매트릭스 변조된 데이터 심볼들의 전송을 위해 상기 K 개의 전송 채널들 중 적어도 하나를 예정하는 수단을 포함하고, 상기 예정은 적어도 부분적으로 상기 계산된 CQI들인 q_k에 기반한다. _{K over} each N _{t , k} transmission interfaces and each N _{r , k} receive interfaces for transmission of data symbols modulated according to a nonnormal orthogonal matrix modulation that modulates data symbols in both the non-orthogonal spatial domain and at least one orthogonal domain. Further proposed is a device for scheduling at least one channel in k = 1, ..., K, wherein the data symbols are matrix modulated and then transmitted over N _{t , k} transmission interfaces so that the transmission channels k At least one equivalent channel matrix (ECM) G _k may be defined, which converts into data symbols that were received at the N _{r , k} receive interfaces of one channel of = 1, .. K, wherein the at least one ECM Equivalent Channel Correlation Matrix (ECCM) of G _k

Is not proportional to a unitary matrix, the apparatus further comprises means for calculating a respective channel quality indicator (CQI) q _k for at least one of the K transmission channels, and transmitting the matrix modulated data symbols. Means for scheduling at least one of the K transport channels, wherein the scheduling is based at least in part on the calculated CQIs, q _k .

는 단위 매트릭스에 비례하지 않을 때, 상기 송신국은, 상기 K 개의 전송 채널들 중 적어도 하나에 대해 각자의 채널 품질 표시자 (CQI) q_k를 산출하는 수단, 및 상기 매트릭스 변조된 데이터 심볼들의 전송을 위해 상기 K 개의 전송 채널들 중 적어도 하나를 예정하는 수단을 포함하고, 상기 예정은 적어도 부분적으로 상기 계산된 CQI들인 q_k에 기반한다. _{K over} each N _{t , k} transmission interfaces and each N _{r , k} receive interfaces for transmission of data symbols modulated according to a nonnormal orthogonal matrix modulation that modulates data symbols in both the non-orthogonal spatial domain and at least one orthogonal domain. Further proposed is a transmitting station in a wireless communication system scheduling at least one channel at k = 1, ..., K, wherein the data symbols are matrix modulated and then transmitted over N _{t , k} transmission interfaces to At least one equivalent channel matrix (ECM) G _k may be defined that translates to data symbols that were received at the N _{r , k} receive interfaces of one of the transmission channels k = 1, .. K, An equivalent channel correlation matrix (ECCM) of the at least one ECM G _k

Is not proportional to a unitary matrix, the transmitting station further comprises means for calculating a respective channel quality indicator (CQI) q _k for at least one of the K transmission channels, and transmitting the matrix modulated data symbols. Means for scheduling at least one of the K transport channels for the scheduling being based at least in part on the calculated CQIs, q _k .

는 단위 매트릭스에 비례하지 않을 때, 상기 적어도 한 송신국은 상기 K 개의 전송 채널들 중 적어도 하나에 대해 각자의 채널 품질 표시자 (CQI) q_k를 산출하고, 상기 매트릭스 변조된 데이터 심볼들의 전송을 위해 상기 K 개의 전송 채널들 중 적어도 하나를 예정하며, 상기 예정은 적어도 부분적으로 상기 계산된 CQI들인 q_k에 기반한다.Further proposed is a wireless communication system comprising at least one transmitting station and at least one receiving station, wherein non-normal orthogonal matrix modulation modulates data symbols in both the non-orthogonal spatial domain and the at least one orthogonal domain, the matrix modulated data symbols being K transmission channels on each N _{t , k} transmission interfaces and each N _{r , k} reception interfaces are transmitted on at least one channel at k = 1, .. K, and the data symbols are matrix modulated and then N at least one equivalent channel matrix (ECM) that is transmitted over _{t , k} transmission interfaces and converts into data symbols received at the N _{r , k} reception interfaces of one of the transmission channels k = 1, .. K Matrix) G _k may be defined and an equivalent channel correlation matrix (ECCM) of the at least one ECM G _k .

Is not proportional to a unitary matrix, the at least one transmitting station calculates a respective channel quality indicator (CQI) q _k for at least one of the K transmission channels, and transmits the matrix modulated data symbols. Schedule at least one of the K transport channels, the schedule being based at least in part on the calculated CQIs q _k .

의 선형 함수 행렬식의 함수로서 산출하는 수단, 및 상기 데이터 심볼들의 전송을 위해 상기 K 개의 전송 채널들 중 적어도 하나를 예정하는 수단을 포함하고, 상기 예정은 적어도 부분적으로 상기 계산된 CQI들인 q_k에 기반한다.Further proposed by the apparatus for scheduling at least one channel in K transmission channels k = 1, .., K over each N _{t , k} transmission interfaces and each N _{r , k} receiving interfaces for transmission of data symbols And at least two of the data symbols are transmitted side by side from N _{t , k} transmission interfaces of at least one of the K transmission channels defined by channel matrix H _k . Calculate a respective channel quality indicator (CQI) q _k for at least one of, wherein at least one of the respective CQIs q _k is a channel correlation matrix of the channel matrix H _k

Means for calculating as a function of a linear function determinant of and means for scheduling at least one of the K transmission channels for transmission of the data symbols, wherein the scheduling is at least partially in q _k , the computed CQIs. Based.

의 선형 함수 행렬식의 함수로서 산출하는 수단, 및 상기 데이터 심볼들의 전송을 위해 상기 K 개의 전송 채널들 중 적어도 하나를 예정하는 수단을 포함하고, 상기 예정은 적어도 부분적으로 상기 계산된 CQI들인 q_k에 기반한다.Further proposed by the transmitting station scheduling at least one channel in K transmission channels k = 1, .., K over each N _{t , k} transmission interfaces and each N _{r , k} receiving interfaces for transmission of data symbols and, the data symbol at least two of the time the side transmitted from the K transmission channels, at least one of the N _{t, k} transport interface of which is defined by the channel matrix H _k, the transmitting station has the K transmitted channels Calculate a respective channel quality indicator (CQI) q _k for at least one of the above, wherein at least one of the respective CQIs q _k is a channel correlation matrix of the channel matrix H _k

Means for calculating as a function of a linear function determinant of and means for scheduling at least one of the K transmission channels for transmission of the data symbols, wherein the scheduling is at least partially in q _k , the computed CQIs. Based.

의 선형 함수 행렬식의 함수로서 산출하고, 상기 적어도 한 송신국은 상기 데이터 심볼들의 전송을 위해 상기 K 개의 전송 채널들 중 적어도 하나를 예정하며, 상기 예정은 적어도 부분적으로 상기 계산된 CQI들인 q_k에 기반한다.Further proposed is a wireless communication system comprising at least one transmitting station and at least one receiving station, wherein at least two data symbols comprise at least K transmission channels over each N _{t , k} transmission interfaces and each N _{r , k} reception interfaces. When the at least one transport channel is defined by a channel matrix H _k , the transmitting station is configured to transmit at least one of its respective CQIs q _k when the at least one transport channel is defined side by side from the N _{t , k} transport interfaces of one channel. channel correlation matrix of _k

Calculating at least one of the K transmission channels for transmission of the data symbols, the schedule being at least partially at q _k , the computed CQIs. Based.

These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments to be described hereinafter.

 1: According to the prior art, different numbers of (solid lines) cooperate with normal orthogonal matrix modulations with two transmit antenna elements (space time transmit diversity, STTD, dotted lines) and with single antenna transmission without STTD (solid lines) Bit error rate (BER) performance obtained when scheduling transport channels K;

Figure 2: Different numbers of transmissions (solid lines) in cooperation with regular orthogonal matrix modulations with two transmit antenna elements (space time transmission diversity, STTD, dotted lines) and in conjunction with a single antenna transmission without STTD according to the prior art. Probability density function (PDF) of the signal-to-noise ratio (SNR) of the predetermined transmission channel when scheduling of channels K is performed;

3 is a table showing the reduction in the complexity of the calculations required to calculate R _k and det R _k for different non-normal orthogonal matrix modulation schemes in accordance with the present invention;

4: Correlation between BER and CQI q _k = det R _k for different orthonormal and nonnormal orthogonal matrix modulation schemes in accordance with the present invention;

5A: E _b / N for different normal orthogonal (STTD) and non-normal orthogonal (TSTTD) matrix modulation schemes and different CQI selections when prescribing one of K = 2 transport channels, in accordance with the present invention. BER as a _zero function;

5B: BER, which is an E _b / N ₀ function for different non-normal orthogonal (ABBA, DSTTD) matrix modulation schemes and different CQI selections when prescribing one of K = 2 transport channels, in accordance with the present invention. ;

FIG. 5C: BER, which is an E _b / N ₀ function for non-normal orthogonal DABBA matrix modulation scheme and different CQI selections when scheduling one of K = 2 transport channels, in accordance with the present invention; FIG.

6: BER, which is an E _b / N ₀ function in scheduling in coordination with non-orthogonal BLAST and TSTTD matrix modulations for different numbers of transport channels K;

7: a flowchart of the method according to the invention; And

8: an apparatus according to the invention.

The present invention proposes combining non-normal orthogonal matrix modulation with scheduling in selecting diversity. This approach is based on the definition of a channel quality indicator (CQI) which serves as the basis for the determination of which of the K transmission channels are matrix modulated and scheduled for use in transmitting data symbols transmitted by the transmitter. It may vary.

The need for CQI definition usually occurs in situations where transmission schemes need to be adjusted according to channel conditions, and require different CQIs for different types of adjustment methods. Two different types of adjustment can be identified:

1. Scheduling methods in TDMA systems that allocate multiple time slots to a transport channel having a "best" channel, for example, in systems comprising multiple transport channels as the basis of resource allocation. In this case, CQIe _k should be correlated with the performance that the channel can support.

2. Depending on the channel condition, choose the correct coding and modulation mode to maximize the rate or throughput of a single transport channel. The decision may be based on some CQIm _k threshold at which the coding / modulation is changed accordingly. Since this means comparing the performance of two different codes / modulations for one fixed channel, CQIm _k should be sensitive to the number of data streams the channel can support.

Preferably, the CQI for select diversity is preferably given by the bit error rate after decoding of the channel codes, or by the corresponding frame error rate (FER). Often, after demodulation, a CQI based on BER is used. The ideal CQI has a one-to-one correspondence with respect to bit error rates (BER) or frame or block error rates. In single stream (and orthogonal multi-stream) channels, the CQI with the best correlation for BER / FER is the received signal-to-noise ratio (SNR), which is generalized to the maximum ratio combining (MRC) channel power for orthogonal transmissions. . This is the overall maximum C / I scheduler discussed in Kogiantis Demonstration.

의 트레이스에 비례한다. 같은

을 갖는 MIMO 채널들은, 전송 채널의 사양과 선택된 전송 방식에 따라, 여전히 매우 상이한 비트/프레임 에러율들에 이르게 할 것이다.The situation is no longer simple for non-normal orthogonal transmissions, in particular what is considered by the present invention, for MIMO transmission of multiple data. Total channel power to generalize overall SNR or C / I is now channel correlation matrix

Is proportional to the trace of. same

MIMO channels with will still lead to very different bit / frame error rates, depending on the specification of the transport channel and the selected transmission scheme.

Finding a good CQI is even more complicated when matrix modulation is used to encode the signals. Matrix modulators are designed to provide more diversity for transmission and to maximize mutual information between transmit and receive signals. As a result, they have inherent symmetry that limits their structure. Thus, the channel quality indicator does not depend solely on the transmission channel, but also on how the symbols are matrix modulated and detected. This means that in principle different matrix modulation and / or detection methods may require different channel quality indicators.

Accordingly, according to the present invention, CQIs suitable for use in multiple transport channel resource allocation (type 1 in the above division) in non-normal orthogonal matrix modulated systems are proposed. In particular, relatively simple CQIs are proposed, which in one mode of transmission—to the simulation accuracy—have a one-to-one correspondence with BER / FER in the reduced search maximum likelihood detection performed at a receiver receiving data transmissions over a predetermined transmission channel. There is a correspondence.

Before discussing the results of Figs. 3 to 7, non-normal orthogonal matrix modulation schemes in accordance with preferred embodiments of the present invention, orthonormal Alamouti space-time code of Equation 1 or its transpose matrix (the latter is now space-time transmission) Diversity (STTD) modulation matrix), the signal model of Equation 2, the equivalent signal model of Equation 3 and the equivalent channel correlation matrix (ECCM) R _k of Equation 6 will be defined briefly.

The following embodiments are focused on space-time matrix modulation techniques, but it should be noted that it is also directly applicable to other non-normal orthogonal matrix modulation techniques such as space-frequency or space-polar matrix modulation techniques. Therefore, the scope of the present invention should not be construed as limited to the following embodiments.

Irregular orthogonality ABBA  Matrix modulation scheme

The modulation matrix of this scheme is identified as follows.

Where X _A is the STTD modulation matrix, the transpose matrix of Equation 1 with two data symbols x ₁ and x ₂ , and X _B is the transpose matrix of Equation 1 with two additional data symbols x ₃ and x ₄ STTD modulation matrix. For optimal performance, these symbols can be selected as a linear combination and / or rotation of the entire modulation symbols. This applies to all matrix modulations discussed in this invention. The modulation matrix X _ABBA thus represents a rate 1 matrix modulator that maps 4 data symbols onto 4 transmit antenna elements for 4 data periods. The ECCM of the ABBA modulation scheme can be identified as follows.

p _k is the total power of all channels, and n _k represents the strength of the interference between two different STTD blocks, X _A and X _B.

For example for N _t = 4 transmit antennas and N _{k , r} = 2 receive antennas, the real functions p _k and n _k are defined as follows:

It can be seen that R _k has only two different eigenvalues p _k ± n _k . It should be noted that the number of different eigenvalues does not depend on the size of the original channel matrix, only the explicit forms of elements p _k , n _k . Since the determinant is given as a product of eigenvalues, for ABBA we get:

의 계산을 128 개의 복소수 곱셈식과 112 개의 복소수 덧셈식 (128 Cx, 112 C+)에서 12 개의 복소수 곱셈식과 10 개의 실수 덧셈식 (12 Cx, 10 R+)으로 우선 단순화하였고, 그런 다음 det R_k 산출의 복잡도를 (40 Cx, 30 C+)로부터 (3 Rx, 1 R+)로 더 줄였음을 알아야 한다. 이것은 실제로 커다란 복잡도 감소에 해당한다. 추가적 복잡도 감소는 R_k의 역수가 스케줄링 또는 검출에 필요로 될 때 볼 수 있다; 그 역수는 단순히 다음과 같다.The above formula is | diagonal part of R _k ] ^2- | part other than diagonal part | Proportional to ² Using the structure of the matrix modulator, for N _{k , r} = 2 receive antennas

We first simplified the calculation of from 128 complex multiplication and 112 complex addition (128 Cx, 112 C +) to 12 complex multiplication and 10 real addition (12 Cx, 10 R +), and then det R _k Note that the complexity was further reduced from (40 Cx, 30 C +) to (3 Rx, 1 R +). This is actually a big reduction in complexity. Further complexity reduction can be seen when the inverse of R _k is needed for scheduling or detection; The inverse is simply

In particular, it can be seen that the diagonal components of the reciprocal are proportional to the trace and inversely proportional to the determinant.

으로 단순화되고, 이때

는 성능에 대한 상관성을 최대화하도록 수치적으로 결정될 수 있는 상수이다. 예를 들어, QPSK 변조 및 감축된 검색 최대 가능성 검출 (reduced search maximum likelihood deection)에 있어서, 두 수신 안테나들을 이용할 때 최상의 적합성은

인 경우에 보여졌다.Instead of the determinant of ECCM, a more generalized CQI for the determinant of linear functions of ECCM can also be used. For ABBA, this is

Is simplified to

Is a constant that can be determined numerically to maximize the correlation for performance. For example, in QPSK modulation and reduced search maximum likelihood deection, the best suitability when using two receive antennas is

Was shown.

A variation of the ABBA matrix modulation is the so-called diagonal ABBA. The modulation matrix is written as follows.

ECCM is now a diagonal matrix, but not a normal orthogonal matrix. The determinant of ECCM is the same as for conventional ABBA.

Irregular orthogonality DABBA  Matrix modulation scheme

The modulation matrix of this scheme is identified as follows.

In this case, X _A , X _B , X _C , and X _D are STTD modulation matrices as the transpose matrix of Equation 1, each having two data symbols. The modulation matrix X _DABBA thus represents a rate 2 matrix modulator that maps eight data symbols onto four transmit antenna elements in four symbol periods. The ECCM of the DABBA modulation scheme can be identified as follows.

p _{k, 1,} p _{k, 2} are each antenna pair, which are (1,2) and (3,4) all received deulyigo total channel power to the antennas, n _{k, 1,} n _k, from ₂ of H _k Are real combinations of elements, and s _k is a complex term. Again, p _{k, 1} + p _{k, 2} on the diagonal of R _k is the total channel power, and other terms other than the diagonal portion represent interference between blocks. Again, R _k has only two different eigenvalues:

The determinant is thus understood as follows.

The inverse of R _k can be formulated in a similar manner as in the ABBA mode. The complexity reduction in this case is indeed huge and can be seen in the table of FIG. Thus, when using the disclosed methods, high rate, high diversity modulators of the DABBA type become practical in practice. A similar simple representation of det R _k can also be obtained for other matrix modulators, due to the broad internal symmetries of the modulators.

에 대해,

About,

는 여전히 수치적으로 결정될 수 있으며, 두 수신 안테나들을 이용한 QPSK 변조 및 감축된 검색 최대 가능성 검출의 최선의 값은 0.04임이 확인되었다.a constant

Can still be determined numerically, and it was found that the best value of QPSK modulation and reduced search maximum likelihood detection using two receive antennas is 0.04.

Irregular orthogonality DSTTD  Matrix modulation scheme

The modulation matrix of this scheme is identified as follows.

X _A and X _B are two modulation symbols, each of which is a modulation matrix given by the transposed Alamouti code of Equation (1). The modulation matrix X _DSTTD thus represents a rate 2 matrix modulator that maps four data symbols _onto four transmit antenna elements in a two symbol period.

Irregular orthogonality TSTTD  Matrix Modulation Method

The modulation matrix of this scheme is identified as follows.

X _A is the STTD modulation matrix, which is the transpose of Equation 1 with two data symbols x ₁ and x ₂ , and X _B is the STTD modulation, which is the transpose of Equation 1 with two additional data symbols x ₃ and x ₄ Matrix. The modulation matrix X _TSTTD thus represents a rate 2 _mateX modulator that maps 4 data symbols onto two transmit antenna elements within a two symbol price. The ECCM of the TSTTD modulation scheme can be identified as follows.

p _{k, 1} and p _{k, 2} are the total channel powers from transmit antennas 1 and 2 to all receive antennas, respectively, and s _k is given as follows when H _k is 2 × 2.

ECCM R _k of TSTTD has the following characteristics.

This indicates that in this case, equivalent channel based scheduling and matrix channel based scheduling are equivalent to the preferred CQI, resulting in further simplification in calculating equivalent channel based CQIs.

FIG. 3 depicts the computational complexity experienced in calculating ECCM R _k and its determinant det R _k in ABBA, DABBA and TSTTD schemes, either directly or through simple closed form representations proposed by this command. The computational complexity is given as real (Rx) and complex (Cx) multiplications and real (R +) and complex (C +) additions.

In the TSTTD scheme, det R _k can be calculated directly from det H _k . The complexity of the proposed algorithms for calculating the determinant is the 2Q product root of the complexity of the direct method, where Q is the number of complex symbols in modulation, i.e. 4, 8, and 4 for ABBA, DABBA, and TSTTD, respectively. to be.

Figure 4 shows the correlation between BER and preferred CQI qk = det R _k for different normal orthogonal (STTD) and irregular orthogonal (ABBA, DSTTD, TSTTD and DABBA) matrix modulation schemes according to the present invention. The simulations assumed an additional white Gaussian noise (AWGN) at the receiving side of the transmission channel of E _b / N _o fixed at 6 dB, QPSK modulated symbols were used as data symbols, and entries in the transmission channel matrix were iid Rayleigh. It is assumed to be distributed.

Clearly, the proposed CQI q _k = det R _k shows a strong correlation to BER in all simulated orthonormal and nonnormal orthogonal matrix modulation schemes. Because the BER is a monotonic function of E _b / N _o, the quantitative results obtained in the operating point should not be changed according to E _b / N _o. These results are repeated for higher order modulation schemes, since the pairwise error probability can be expressed in terms of the pairwise error probabilities of the BPSK modulated system. For example, other linear detection algorithms, such as a zero forced block linear detector or a minimum mean square error block linear detector, should also be considered.

5A, 5B, and 5C show different normal orthogonal (STTD) and nonnormal orthogonal (TSTTD, ABBA, DSTTD, and DABBA) matrix modulation schemes when scheduling one of the K = 2 transport channels in accordance with the present invention; Simulated BER is shown as a function of E _b / N _o in different CQI selections. Convenience scheduling of a transport channel with better CQI was performed irrespective of the state of potential temporal scheduling constraints such as Rayleigh fading transport channels, AWGN at the receive interfaces of the intended transport channel, and average packet delay. . For STTD and TSTTD, 16-QAM modulated data symbols were sent, whereas for ABBA, DSTTD and DABBA, QPSK modulated symbols were sent. Except for STTD, which has a bit rate of 2 bps, all non-orthogonal schemes have a bit rate of 4 bps.

이고, 스케줄링 없는 (K=2 전송 채널들 중 하나가 자신의 CQI와 무관하게 임의로 선택됨)는 BER 성능이 도시된다. STTD, DABBA 및 TSTTD 방식들에 있어서, 부가적으로 CQI q_k=tr R_k에 대한 BER 성능이 도시된다.For each matrix modulation scheme, CQI q _k = det R _k And CQI

BER performance is shown without scheduling (one of the K = 2 transport channels is randomly selected irrespective of its CQI). For the STTD, DABBA and TSTTD schemes, additionally the BER performance for CQI q _k = tr R _k is shown.

의 행렬식에 기반하는 CQI들에 대해 스케줄링 없는 시스템의 BER 성능에 대한 E_b/N_o의 상당한 이익들 또한 얻어진다.As can be readily seen from FIGS. 5A, 5B and 5C, scheduling based on CQI q _k = det R _k works best for all studied matrix modulation schemes. However, the trace of R _k and

Significant benefits of E _b / N _o for BER performance of a system without scheduling for CQIs based on a determinant of are also obtained.

Possible choices of further CQI include the trace of the inverse of R _k , the _number of states of H _k , and the ratio or difference between the diagonal and non-diagonal parts of R _k .

For example, the performance of vector modulation, such as the Bell Laboratories Layered Space Time Architecture (BLAST) scheme combined with scheduling in the selection diversity set on K = 2 and K = 8 transport channels, is shown in FIG. Are compared.

In Figure 6, both schemes use N _t = 2 transmit and N _{k , r} = 2 receive antenna elements and QPSK modulated data symbols, noise is AWGN, and the transmit channels are iid flat lay fading channels. In each scheme, four data symbols are transmitted during two time instances, where the modulation matrix in the BLAST scheme is identified as follows.

That is, two data symbols are mapped to non-orthogonal spatial resources during each symbol period.

를 갖는 전송 채널이 BLAST 경우에 대해 예정되었다.6 shows both the scheduling results of K = 2 and K = 8 transport channels and the results when no scheduling is performed. In scheduling results, a transport channel with a maximum CQI q _k = det R _k has been reserved for the TSTTD scheme, and the maximum CQI

A transport channel with is scheduled for the BLAST case.

이고, 상수

가 CQI 및 BER 사이의 최선의 상관성을 보장하기 위해 각 방식마다 조정될 때, 이익들은 더 증가될 것이라는 것을 알 수 있다. 이 경우, TSTTD의 최적 상수는 0.04였고, BLAST에 대해서는 0.0이었다.From Fig. 6, it can be clearly seen that when the number of transport channels K increases, scheduling increases the performance of both TSTTD (dashed curves) and BLAST (solid curves). Since the gains from the scheduling are larger in the BLAST scheme, the performance difference will decrease as more transport channels are added. However, even in transmission channels with K = 8, TSTTD still has better performance at BER = 10 ⁻³ , and the difference of E _b / N ₀ is about 1.5 dB. Again, this simulation is for convenience scheduling, and the results show only an upper bound on the scheduling gains. In practice, the gains can be smaller, which means that TSTTD can function better for much more transport channels than expected from the convenient scheduling results. In addition, the CQI becomes proportional to the determinant of the linear function, that is,

, Constant

It can be seen that the benefits will be further increased when is adjusted for each scheme to ensure the best correlation between CQI and BER. In this case, the optimum constant of TSTTD was 0.04 and 0.0 for BLAST.

7 shows a flowchart according to the first preferred embodiment of the present invention (ABBA non-normal orthogonal matrix modulation scheme).

In step 10, the transport channel index is initialized to 1 and the variable q _max is set to 0. The channel matrix H _k is then called from storage or estimated from signals that have been previously received on the transmission channel. The functions p _k and n _k are calculated from the channel matrix H _k (step 12). From the functions p _k and n _k , CQI q _k = det R _k is calculated (step 13). This CQI q _k is compared with q _max (step 14), and if q _k is q _max If greater than, q _max is set equal to q _k and the variable k _max is set equal to k (step 15). The transport channel index k is increased by 1 (step 16), and it is checked whether the transport channel index k is equal to the number of transport channels K where scheduling is being performed (step 17). If not equal, steps 11 to 16 are repeated until the transport channel index k is equal to K. Matrix modulation according to the ABBA scheme is performed on data symbols to be transmitted on the transmission channel with index k _max (step 18). The matrix modulated data symbols are then transmitted (step 19), and the flowchart begins again from step 10.

8 shows an apparatus according to the invention. The device comprises respective data packet buffers 20-1..20-3 containing data packets to reach the device and to be transmitted on each of K transport channels, in the example of FIG. 3 Each packet buffer 20-1..20-3 is controlled by a buffer controller 21-1..21-3, respectively. The buffer controller 21-1..21-3 signals the status of the buffers 20-1..20-3 to the scheduler 22. Also, when triggered by the scheduler 22, the buffer The controllers 21-1..21-3 may move the data packets to the matrix modulator 26, where the data symbols contained in the data packets are matrix modulated and then the transmission modules 27-1..27 -4) and via transmit antenna elements 28-1..28-4, the scheduler 22 from the CQI calculation instance 23 which fetches the transmit channel data from the channel storage 24; Is further received, and the transmission channel data of the channel storage 24 is now updated by the channel estimation instance 25. The CQI calculation instance 23 has its own transmission channel taken from the channel storage 24. Determine the CQIs for the K transport channels based on the data, eg ABBA matrix modulation scheme applied Time, determining the CQI calculation instance 23 of the channel matrix H _k entries h _k, to calculate the from _ij function, p _k and n _k, and then the function, p _k and n CQI q _k from _k The calculated CQIs are then sent to the scheduler 22. Based on the input from the CQI calculation instance 23 and the input from the buffer controllers 21-1..21-3, the scheduler sends a K. Determine which of the channels is desired to be scheduled for transmission of matrix modulated data symbols, so that the BER of the transmission is as low as possible This scheduling is based, at least in part, on the CQIs produced by the instance 23 However, although the CQI ₁ of the transmission channel k = ₁ is larger than the CQI _{2 of the} transmission channel k = 2, the data packet in the data packet buffer 20-2 associated with the transmission channel k = 2 is matrix-modulated and transmitted. I can imagine being May it reasons geuphada more than the data packet in the data packet buffer 20-1 associated with the data packet transfer channel k = 1. The result of the scheduling is signaled to the corresponding buffer controller 21-1..21-3, which forwards the matrix modulator 26 to matrix modulate and transmit the data packet.

The invention has been described above using the preferred embodiments. It should be understood that there are other selectable methods and variations apparent to those skilled in the art and capable of implementing without departing from the scope and concept of the appended claims. In particular, the proposed CQIs can be used in part within any scheduling algorithm operating in a MIMO / MISO system, where link specific information is taken into account in scheduling. The proposed CQI is valid for any non-normal orthogonal MIMO or MISO transmission scheme. The proposed algorithms are any concatenated via any duplexing (FDD / TDD / mixed), multiplexed access (CDMA, TDMA, SDMA, OFDMA) single carrier / multicarrier (MC-CDMA / OFDMA) systems. It is also valid for feedback / beamforming. Any channel code (trellis, convolution, turbo, block, LDPC, TCM), any modulation (PAM, PSK, QAM, high-dimensional spherical / lattice) and joint / separate channel coding / space-time coding / modulation, Joint / separation detection / decoding (linear, decorrelation, LMMSE, maximum likelihood, reduced search maximum likelihood, sphere) can be used. At some stage the transmission is actually non-orthogonal spatial resources (antenna / Beam / polarity) and a matrix (of non-orthogonal space resources) and a matrix (of non-orthogonal space and substantially orthogonal resources) transmitted from substantially orthogonal resources (time, frequency, subcarrier, code, spatial / polar eigenmodes). Can only be represented by the only requirement.

Claims (30)

_{K over} each N _{t , k} transmission interfaces and each N _{r , k} receive interfaces for transmission of data symbols modulated according to a nonnormal orthogonal matrix modulation that modulates data symbols in both the non-orthogonal spatial domain and at least one orthogonal domain. A method for scheduling at least one channel of k transport channels k = 1,... The data symbols are matrix-modulated _and transmitted over N _{t , k} transmission interfaces to convert them into data symbols that were received at the N _{r , k} reception interfaces of one of the transmission channels k = 1, .. K. At least one equivalent channel matrix (ECM) G _k may be defined and an equivalent channel correlation matrix (ECCM) of the at least one ECM G _k .

When is not proportional to the unit matrix, the method Calculating a respective channel quality indicator (CQI) q _k for at least one of the K transport channels, and Scheduling (scheduling) at least one of the K transmission channels for transmission of the matrix modulated data symbols, The scheduling is based at least in part on the calculated CQIs, q _k . The method of claim 1, wherein the respective CQI q _k is derived for at least one of the K transport channels from the ECCM R _k . 3. The method of claim 2, wherein each CQI q _k is calculated as a function of the determinant of the ECCM R _k function for at least one of the K transport channels. The method of claim 2, wherein the respective CQI q _k is calculated as a trace function of the ECCM R _k for at least one of the K transport channels. 3. The method of claim 2, wherein the respective CQI q _k is calculated as a trace function of the inverse of the ECCM R _k for at least one of the K transport channels. The method of claim 1, wherein the respective CQI q _k is calculated as a function of channel matrix H _k elements defining at least one transport channel, for at least one of the K transport channels, the function being the ECCM R _k. Derived from the ECCM R _k under the utilization of the structural features of. The method of claim 1, wherein the non-normal orthogonal matrix modulation maps a block of 4 data symbols on N _{t , k} = 4 transmission interfaces and at least two space-time transmit diversity (STTD) within 4 units of the at least orthogonal domain. ) ABBA nonnormal orthogonal matrix modulation based on a nonnormal orthogonal combination of codes. The method of claim 6, wherein the ECCM R _k is

ego, Wherein p _k and n _k are real functions of the channel matrix H _k elements. 9. The method of claim 8, wherein each CQI q _k for at least one of the K transport channels is

Calculated as a function of

Is a constant. The method of claim 1, wherein the non-normal orthogonal matrix modulation maps a block of eight data symbols onto N _{t , k} = 4 transmission interfaces within four units of the at least one orthogonal domain and provides four space-time transmission diversity. DABBA nonnormal orthogonal matrix modulation based on a nonnormal orthogonal combination of (STTD) codes. The method according to claim 6, wherein the ECCM R _k is of the form

P _{k, 1} , p _{k, 2} , n _{k , 1} and n _{k , 2} are real functions of the elements of the channel matrix H _k , and s _k is a complex function of the elements of the channel matrix H _k . How to. 12. The method of claim 11, wherein each CQI q _k for at least one of the K transport channels is

Calculated as a function of

Is a constant. The non-normal orthogonal matrix modulation according to claim 1, wherein the non-normal orthogonal matrix modulation maps a block of four data symbols onto N _{t , k} = 2 transmission interfaces in two units of the at least one orthogonal domain and a non-normal orthogonal combination of two STTD codes. TSTTD non-normal orthogonal matrix modulation based on the method. The method of claim 6, wherein the ECCM R _k is of the form

Wherein p _{k, 1} and p _{k, 2} are real functions of the elements of the channel matrix H _k , and s _k is a complex function of the elements of the channel matrix H _k . 15. The method of claim 14, wherein each CQI q _k for at least one of the K transport channels is

Calculated as a function of

Is a constant. The non-normal orthogonal matrix modulation of claim 1, wherein the non-normal orthogonal matrix modulation maps a block of four data symbols onto N _{t , k} = 4 transmission interfaces within two units of the at least one orthogonal domain and a non-normal orthogonal combination of two STTD codes. And based on DSTTD non-normal orthogonal matrix modulation. 2. The method of claim 1, wherein the non-normal orthogonal matrix modulation comprises space-time or space-frequency coding. 2. The method of claim 1, wherein the non-normal orthogonal matrix modulation comprises a combination of at least two orthonormal matrix modulation. A method for scheduling at least one of K transport channels k = 1, .. K over each N _{t , k} transmission interfaces and each N _{r , k} receiving interfaces for transmission of data symbols _, the method comprising: When at least two of the data symbols are transmitted side by side from N _{t , k} transmission interfaces of at least one of the K transmission channels defined by channel matrix H _k , Calculate a respective channel quality indicator (CQI) q _k for at least one of the K transport channels, wherein the respective CQIs q _k At least one of the channel correlation matrix of the channel matrix _Hk

Calculated as a determinant function for the linear function of, and Scheduling (scheduling) at least one of the K transmission channels for transmission of the data symbols, The scheduling is based in part on the calculated CQIs q _k . 20. The receiver according to any one of the preceding claims, wherein at least one receiver is transmitted over the predetermined transport channel and uses the maximum likelihood algorithm or a linear estimator. Estimating the data symbols received by the receiver via the receiver. 20. A method according to any one of the preceding claims, characterized in that a transmission channel k = 1, .. K having a maximum CQI q _k is intended for the transmission of the data symbols. The transmission channel of claim 1, wherein the transmission channels are transmission channels of a wireless communication system, and the transmission and reception interfaces of the transmission channels are transmission and reception antenna elements of one or several transmitters and one or several receivers, respectively. How to feature. 20. A storage medium comprising computer program codes comprising instructions operative to cause a processor to perform the steps of the method according to any one of claims 1 and 19. delete _{K over} each N _{t , k} transmission interfaces and each N _{r , k} receive interfaces for transmission of data symbols modulated according to a nonnormal orthogonal matrix modulation that modulates data symbols in both the non-orthogonal spatial domain and at least one orthogonal domain. An apparatus for scheduling at least one channel of k transport channels k = 1,. The data symbols are matrix-modulated _and transmitted over N _{t , k} transmission interfaces to convert them into data symbols that were received at the N _{r , k} reception interfaces of one of the transmission channels k = 1, .. K. At least one equivalent channel matrix (ECM) G _k may be defined, An equivalent channel correlation matrix (ECCM) of the at least one ECM G _k

When is not proportional to the unit matrix, the apparatus, Means for calculating a respective channel quality indicator (CQI) q _k for at least one of the K transport channels, and Means for scheduling (scheduling) at least one of the K transmission channels for transmission of the matrix modulated data symbols, And the schedule is based at least in part on the calculated CQIs, q _k . For transmission of data symbols modulated according to nonnormal orthogonal matrix modulation, which modulates data symbols in both the non-orthogonal spatial domain and at least one orthogonal domain, through each N _{t , k} transmission interfaces and each N _{r , k} reception interfaces. A transmitting station in a wireless communication system that schedules at least one of K transport channels k = 1, ..., K, The data symbols are matrix-modulated and then transmitted over N _{t , k} transmission interfaces to convert them into data symbols received at N _{r , k} reception interfaces of one of the transmission channels k = 1, .. K. At least one equivalent channel matrix (ECM) G _k may be defined, An equivalent channel correlation matrix (ECCM) of the at least one ECM G _k

When is not proportional to the unit matrix, the transmitting station, Means for calculating a respective channel quality indicator (CQI) q _k for at least one of the K transport channels, and Means for scheduling at least one of the K transmission channels for transmission of the matrix modulated data symbols, And the schedule is based at least in part on the calculated CQIs, q _k . In a wireless communication system, At least one transmitting station, and At least one receiving station, Non-normal orthogonal matrix modulation modulates the data symbols in both the non-orthogonal spatial domain and at least one orthogonal domain, wherein the matrix modulated data symbols are K numbered over each N _{t , k} transmission interfaces and each N _{r , k} receive interfaces. Transmitted on at least one of transmission channels k = 1,..., K, and the data symbols are transmitted through N _{t , k} transmission interfaces after matrix modulation. At least one equivalent channel matrix (ECM) G _k may be defined, which converts into data symbols that have been received at N _{r , k} receive interfaces of one channel of K, and the equivalent channel of the at least one ECM G _k . Correlation Matrix (ECCM)

Is not proportional to the unit matrix, The at least one transmitting station calculates a respective channel quality indicator (CQI) q _k for at least one of the K transmission channels and at least one of the K transmission channels for transmission of the matrix modulated data symbols. Schedule one, wherein the schedule is based at least in part on the calculated CQIs, q _k . An apparatus for scheduling at least one channel in K transmission channels k = 1, .., K over each N _{t , k} transmission interfaces and each N _{r , k} receiving interfaces for transmission of data symbols, At least two of the data symbols are transmitted side by side from N _{t , k} transmission interfaces of at least one of the K transmission channels defined by channel matrix H _k , and the apparatus further comprises: Calculate a respective channel quality indicator (CQI) q _k for at least one of the K transport channels, wherein at least one of the respective CQIs q _k is a channel correlation matrix of the channel matrix H _k

Means for calculating as a function of the linear function determinant of, and Means for scheduling at least one of the K transmission channels for transmission of the data symbols, And the schedule is based at least in part on the calculated CQIs, q _k . A transmitting station for scheduling at least one channel in K transmission channels k = 1, .., K over each N _{t , k} transmission interfaces and each N _{r , k} receiving interfaces for transmission of data symbols, At least two of the data symbols are transmitted side by side from N _{t , k} transmission interfaces of at least one of the K transmission channels defined by channel matrix H _k , and Calculate a respective channel quality indicator (CQI) q _k for at least one of the K transport channels, wherein at least one of the respective CQIs q _k is a channel correlation matrix of the channel matrix H _k

Means for calculating as a function of the linear function determinant of, and Means for scheduling at least one of the K transmission channels for transmission of the data symbols, And the schedule is based at least in part on the calculated CQIs, q _k . In a wireless communication system, At least one transmitting station, and At least one receiving station, At least two data symbols are transmitted side by side from N _{t, k} transmission interfaces of at least one of the K transmission channels over each N _{t, k} transmission interfaces and each N _{r, k} reception interfaces, the at least one transmission channel Is defined by the channel matrix H _k , The transmitting station calculates a respective channel quality indicator (CQI) q _k for at least one of the K transport channels, The transmitting station assigns at least one of the respective CQIs q _k to the channel correlation matrix of the channel matrix H _k

As a function of the linear function determinant of The at least one transmitting station schedules at least one of the K transmission channels for transmission of the data symbols, the scheduling being based at least in part on the calculated CQIs q _k .

Images