RU2278471C2 - Method for directional transmission with check connection - Google Patents

Abstract

FIELD: radio engineering, in particular, technologies for receiving and transmitting radio signal with use of adaptive antenna array, possible use in cell phone communication systems.

SUBSTANCE: in accordance to invention, in first variant of method base station transmits signal to client station, switching direction of signal transmission between several dominant directions, selecting direction with currently maximal transmission coefficient. In accordance to second variant, base station transmits signal to client station simultaneously in all dominant directions, setting a phase difference between them, providing for coherent addition of appropriate signal components in point of receipt.

EFFECT: increased capacity of system due to improvement of main parameters of directional transmission with check connection: speed of tracking changes of spatial communication channel and trustworthiness of controlling information transmissions.

2 cl, 17 dwg, 1 tbl

Description

The invention relates to the field of radio engineering, in particular to methods for transmitting and receiving a radio signal using an adaptive antenna array, and can be used in cellular communication systems.

The need to increase capacity is one of the most pressing problems of modern cellular communication systems. To a large extent, this relates to the capacity of the direct channel — from the base station to the user — through which large amounts of data are transmitted at high speed. One of the most effective ways to increase the capacity of the forward channel is to use a multi-element transmit antenna or antenna array (AR).

Two main methods of signal transmission using an antenna array are known: a directional transmission method and a diversity transmission method. The directional transmission method is based on the use of an antenna array configuration with a distance between antenna elements approximately equal to half the wavelength (d≈λ / 2). This method is described in Downlink Beamforming Concepts in UTRA FDD, Mirko Schacht, Armin Dekorsy, Peter Jung, Lucent Technologies, 5th November 2002 [1]; "Transformation Based Downlink Beamforming Techniques For Frequency Division Duplex Systems", Klaus Hugl, Juha Laurila, Ernst Bonek, Institut für Nachrichtentechnik und Hochfrequenztechnik, Technische Universität, Wien, Austria [2]; "Downlink Beamforming Avoiding DOA Estimation For Cellular Mobile Communications", T.Aste, P.Forster, L. Fety, S. Mayrargue, Laboratoire Electronique & Communication, CNAM, France [3]; "Downlink Beamforming for DS-CDMA Mobile Radio with Multimedia Services Ying-Chang Liang", Senior Member, IEEE, Francois PSChin, and KJRay Liu, IEEE Transactions on Communications, Vol.49, No.7, July 2001 [4] .

The diversity transmission method uses antenna configurations with far-spaced elements (d is much greater than λ). The potentially most effective diversity transmission method is the so-called coherent diversity transmission, in which the amplitudes and phases of the signal in different antennas are set so as to provide the maximum signal-to-noise ratio at the receiving point for a given distribution channel implementation. However, the practical implementation of coherent diversity transmission involves knowing the parameters of the propagation channel on the transmitting side. In communication systems with frequency duplex, this implies the use of feedback from the subscriber station. The principles of diversity transmission are described in "Transmission diversity with limited feedback information in fading channels". Chau, Y.A .; Shi-Hong Yu. Vehicular Technology Conference Proceedings. 2000, VTC 2000-Spring Tokyo. 2000 IEEE 51st, Volume: 2, 2000 [5]; "Physical layer procedures (FDD). 3GPP TS 25.214" [6].

With diversity transmission, a signal is formed at the receiving end, which is the sum of the signals transmitted through the independent propagation channels. Due to this, the signal fading is reduced, the noise immunity of the reception and, accordingly, the capacity of the communication channel are increased.

One of the effective methods of using two far-spaced transmit antennas is presented in the 3GPP standard as closed-loop diversity transmission [6]. In this case, the mobile (subscriber) station receives the pilot signals transmitted from various antennas, estimates the phase difference between them, quantizes this difference by two values and transmits it via the feedback channel using a one-bit command. The base station receives this command and transmits the signal intended for the given subscriber station through both antennas, setting the phase difference in accordance with the received command (the phase difference of the signals in the transmitting antennas should approximately compensate for the phase difference of the signals arising in the propagation channel).

This method is one of the most effective ways to use two transmitting antennas in terms of implementation complexity and noise immunity. However, the growing need to increase direct channel capacitance dictates the need to increase the number of antennas. In the case of diversity transmission, this leads, on the one hand, to an increase in the amount of control data transmitted via the feedback channel and the need to modify the existing standard, and, on the other hand, to a significant increase in the dimensions of the antenna system, which is recognized by system designers as highly undesirable.

Therefore, the attention of researchers is attracted by directional transmission, which uses an antenna array with closely spaced elements and increases the capacity of the direct channel by limiting the spatial area of the signal radiation. The advantage of directional transmission is a significant reduction in the level of intra-system interference, high noise immunity in the region of low signal-to-noise ratios, and also the small dimensions of the antenna system, see, for example, [1] and "Advanced closed loop Tx diversity concept (eigenbeamformer)". Siemens TSGR1 # 14 (00) 0853; "Description of the eigenbeamformer concept (update) and performance evaluation". Siemens TSGR1 # 19 R1-01-0203 [7].

For directional transmission, the transmitter must have information about the communication channel. In communication systems with frequency duplex, the instantaneous characteristics of the forward and reverse communication channels are different, and therefore the transmitter does not have such information. One of the relatively simple methods of directional transmission is based on the fact that some averaged characteristics of the forward and reverse channels are the same, for example, the spatial direction of signal propagation. With this approach, a BS receives a user signal, estimates the angle of its arrival, and when transmitting a direct channel signal for a given user, generates a maximum radiation pattern in this direction (see "Smart antennas for Wireless Communications: IS-95 and Third Generation CDMA Applications", JCLiberti , JR.TSRappaport./ Prentice Hall PTR. 1999 [8]). The disadvantage of this method is that its effectiveness is significantly reduced in a multipath situation. Due to the multipath, as a rule, there are several spatial directions along which the signal transmitted from the base station reaches the subscriber station. In this case, signals propagating in these directions are subject to independent fading. In a communication system with a frequency duplex, the directions of the best propagation of the signals of the forward and reverse channels at the current time interval may not coincide. Accordingly, the signal may be transmitted in a non-optimal direction and, as a result, the signal-to-noise ratio at the receiving point will be low. In this regard, the efficiency of directional transmission based on an estimate of the angle of arrival of the signal in the return channel is significantly lower than the potentially achievable level.

The solution to this problem is possible by evaluating the necessary parameters of the communication channel at the subscriber station and transmitting the indicated estimate to the base station using the feedback channel. Moreover, under the conditions of a rapidly changing distribution channel, it is required to transmit significant amounts of control data at a high speed via the feedback channel. However, the feedback channel capacity is typically limited. For example, the feedback channel provided by 3GPP has a throughput of only 1,500 bits per second. This circumstance is a serious obstacle to the implementation of the potentially high efficiency of directional transmission in the forward channel.

One of the ways to solve this problem was proposed by Siemens [7] as a concept of the formation of its own rays. At the same time, the parameters of the communication channel are divided into slow (long-term) and rapidly changing (short-term). This allows you to use a large time period for the transfer of the most voluminous information - slowly changing parameters. Tracking rapidly changing channel parameters is performed using less voluminous information that is transmitted over a shorter time interval. This method is the closest in technical essence to the proposed and adopted as a prototype.

The prototype method is that

- from the base station through the transmitter transmit the information signal of the subscriber station, forming a radiation pattern based on the control information received from the subscriber station via the feedback channel,

- while at the base and subscriber stations carry out slow and fast signal processing, and the slow signal processing is performed periodically with a period of T ₁ and fast signal processing is performed periodically with a period of T ₂ , while T _{2 is} less than T ₁ .

At the subscriber station periodically with a period T _{1 of} slow signal processing

- evaluate the spatial communication channel between the base and subscriber stations, and this assessment is carried out using pilot signals that are generated by the transmitter of the base station and transmit in addition to the information signal,

- evaluate the covariance matrix R of the spatial communication channel,

- determine the set of eigenvectors and eigenvalues of the covariance matrix R of the spatial channel,

- determine the set of signal eigenvectors of the spatial channel as M eigenvectors corresponding to M the largest eigenvalues of the covariance matrix R,

- a plurality of signal eigenvectors is ordered, quantized, and transmitted to the base station via a feedback channel during the period of slow processing T ₁ as long-term control information,

- at the base station, on the return channel from the subscriber station, the ordered quantized values of the signal eigenvectors are received and stored in the order in which they were received,

- at each fast processing period T ₂ at the subscriber station, the signal energy corresponding to each of the signal eigenvectors that are already transmitted to the base station is estimated,

- during the fast processing period T ₂ transmit to the base station the serial number of the dominant eigenvector corresponding to the maximum value of the signal energy, as short-term information,

- at the base station, the serial number of the dominant eigenvector is received, and when transmitting the information signal of the subscriber station, a radiation pattern is generated in accordance with the dominant eigenvector.

The disadvantage of this solution is the very large amount of transmitted long-term control information. In the case of a subscriber moving at high speed, updating long-term information about the parameters of the communication channel may not keep up with the actual change in this information. As a result, directional transmission efficiency will noticeably decrease. In addition, the feedback channel is not ideal, that is, it has a certain probability of error when transmitting information through it. For example, in the literature it is customary to consider the probability of an error in the feedback channel equal to 4% (see "Recommended simulation parameters for Tx diversity simulations". Nokia. TSGR1 # 14 (00) 0867 [9]). With a large amount of transmitted long-term control information, it is likely to be damaged. Own vectors transmitted by feedback in the prototype are especially sensitive to errors, and even an error in one bit can completely distort the transmitted information. This factor leads to a significant decrease in the efficiency of directional transmission performed by the prototype method in practice.

Another disadvantage is that the use of eigenvectors in the multipath channel as weight coefficients often leads to the formation of sufficiently wide radiation patterns, which increases the scattering region and, accordingly, increases the level of multipath interference of the forward channel.

The problem that the proposed method solves is to increase the throughput of the communication system by improving the basic parameters of directional transmission with feedback: the speed of tracking changes in the spatial communication channel and the reliability of the transmission of control information.

The technical result is achieved by two variants of the directional transmission method with feedback.

In both versions, the subscriber station estimates the dominant (prevailing main) spatial directions of signal propagation in the direct communication channel (from the base station to the subscriber station), corresponding to the highest values of the signal power transmission coefficient. The subscriber station transmits information about the dominant spatial directions of signal propagation to the base station via the reverse channel.

In the first version of the feedback directional transmission method, the base station transmits a signal to the subscriber station, forming as narrow a radiation pattern as possible with a maximum in the direction characterized by the currently maximum transmission coefficient.

In the directional feedback feedback method according to the first embodiment, the information signal of the subscriber station is transmitted from the base station via the transmitter, forming a radiation pattern based on the control information received from the subscriber station via the feedback channel, while at the base and subscriber stations performed slow and fast signal processing, and a slow processing is performed periodically with a period T _1, the signal processing is performed fast and periodic signal period T _2, wherein T ₂ less than T _1, to the subscriber station periodically with a period of slow processing of the signal T ₁ is evaluated spatial channel connection between the base and subscriber stations, wherein this assessment is performed using pilot signals which are formed by the base station transmitter and transmit in addition to the information signal, according to the invention, the following operations are introduced:

on each period T ₁ slow signal processing at the subscriber station

- evaluate the spatial directions of signal propagation from the base station to the subscriber station,

- choose M dominant signal propagation directions from the indicated estimated spatial signal propagation directions, which are characterized by the highest received signal energy,

- the obtained M estimates of the dominant directions of signal propagation are ordered, quantized, subjected to error-correcting coding, thus forming a sequence of ordered estimates of M dominant directions of signal propagation,

- a sequence of ordered estimates of M dominant signal propagation directions is transmitted via a reverse channel from a subscriber to a base station as long-term control information,

at the base station, a sequence of ordered estimates of M dominant signal propagation directions is obtained from the subscriber station on the return channel, decode them, and remember in the order in which they were received,

on each period of fast processing T ₂ at the subscriber station

- evaluate the signal energy corresponding to each of the M dominant signal propagation directions, estimates of which have already been transmitted to the base station,

- evaluate the energy value of the received signal in the next period of fast processing for each of the M dominant directions of signal propagation, using for this assessment of the signal energy generated on several previous periods of fast processing,

- determine the best direction for signal transmission as the dominant signal propagation direction, which corresponds to the maximum estimated value of the signal energy, and the number of this direction in an ordered sequence of estimates M of the dominant signal propagation directions is transmitted via the return channel to the base station as short-term control information,

- at the base station, take the serial number of the best direction for signal transmission, and transmit the information signal to the subscriber station, forming a radiation pattern with a maximum in this direction.

The base station can receive the serial number of the best direction for signal transmission and transmit the information signal to the subscriber station, forming a radiation pattern with a maximum in this direction.

The base station can take the serial number of the best direction for transmitting the signal and transmit the bulk of the power of the information signal of the subscriber station, forming a radiation pattern with a maximum in this direction, and evenly distribute the rest of the power for transmission in the remaining M minus 1 dominant directions.

Evaluation of M dominant directions of signal propagation from the base station to the subscriber station relative to the antenna array of the base station is performed at the subscriber station, for example, based on the reception of pilot signals transmitted from various antenna elements of the antenna array of the base station and the known configuration of the antenna array, or based on reception K pilot signals transmitted from a base station with different radiation patterns.

When M is 1 at the mobile station, one obtained estimate of the dominant direction of signal propagation is quantized and transmitted over the return channel from the subscriber to the base station as control information, at the base station, an estimate of the dominant direction of signal propagation is received on the return channel from the subscriber station, and the subscriber information signal is transmitted stations, forming a radiation pattern with a maximum in a given direction.

In the second variant of the directional transmission method, the base station transmits the signal to the subscriber station simultaneously in all dominant directions, establishing amplitude-phase relations between them, which ensure maximum signal power at the receiving point.

In the directional feedback feedback method according to the second embodiment, the information signal of the subscriber station is transmitted from the base station via the transmitter, forming a radiation pattern based on the control information received from the subscriber station via the feedback channel, while on the base and subscriber stations carry out slow and fast signal processing, and slow signal processing is performed periodically with a period of T ₁ , and fast signal processing is performed periodically with period T ₂ , with T ₂ less than T ₁ , at the subscriber station periodically with a period of slow signal processing T ₁ evaluate the spatial communication channel between the base and subscriber stations, and this assessment is carried out using pilot signals that are generated by the transmitter of the base station and transmit in addition to the information signal, according to the invention, the following operations are introduced:

on each period T ₁ slow signal processing at the subscriber station

- evaluate the spatial directions of signal propagation from the base station to the subscriber station,

- choose M dominant signal propagation directions from the indicated estimated spatial directions, which are characterized by the highest received signal energy,

- the obtained M estimates of the dominant directions of signal propagation are ordered, quantized, subjected to error-correcting coding, thus forming a sequence of ordered estimates of M dominant directions of signal propagation,

- a sequence of ordered estimates of M dominant signal propagation directions is transmitted along the reverse channel from the subscriber station to the base station as long-term monitoring information - at the base station, a sequence of ordered estimates of M dominant signal propagation directions is transmitted via the return channel from the subscriber station, decode them and store them in that order in which they were adopted,

on each period T ₂ fast signal processing at the subscriber station

- evaluate the values of the relations of amplitudes and phase differences between the signal components, adopted from various dominant directions of signal propagation,

- the resulting estimates are quantized and transmitted over the reverse channel to the base station as short-term control information,

at the base station

- get estimates of the values of the relations of amplitudes and phase differences and transmit the information signal of the subscriber station in the M dominant directions of signal propagation, setting the values of the relations of amplitudes and phase differences between the signal components transmitted in different directions, in accordance with the obtained estimates in such a way as to maximize the signal power at the point of reception.

Evaluation of the M dominant directions of signal propagation from the base station to the subscriber station is performed based on the reception of pilot signals transmitted from various antenna elements of the antenna array of the base station and the known configuration of the antenna array, or based on the reception of K pilot signals transmitted from the base station with various radiation patterns.

When M is 1, at the mobile station, one obtained estimate of the dominant direction of signal propagation is quantized and transmitted over the return channel from the subscriber to the base station as control information, at the base station, an estimate of the dominant direction of signal propagation is received on the return channel from the subscriber station, and information signal of the subscriber station, forming a radiation pattern with a maximum in this direction.

At each period of fast processing of the T ₂ signal at the subscriber station, only the phase difference between the signal components received from different dominant signal propagation directions can be estimated, the obtained estimates are quantized and transmitted along the reverse channel to the base station as short-term control information, while on the base stations receive phase difference estimates and transmit the information signal of the subscriber station in M dominant signal propagation directions, setting the separation values and phase between the signal components transmitted in different directions in accordance with the estimates obtained, so as to maximize the signal power at the receiving point.

A comparative analysis of the first and second variants of the directional transmission method with feedback from the prototype shows that the proposed solutions differ significantly from the prototype, as it allows to increase the throughput of the communication system by improving the main parameters of the directional transmission: the tracking speed of changes in the spatial communication channel and the reliability of transmission control information.

A comparative analysis of the claimed first and second variants of the method with other technical solutions in the art did not reveal the features claimed in the characterizing part of the claims. Therefore, the claimed method of directional transmission with feedback meets the criteria of "novelty", "technical solution of the problem", "significant differences" and has a non-obvious solution.

To clarify the essence of the proposed solution, the following graphic materials are given:

Figure 1, a is an illustration of the proposed method of directional transmission.

Figure 1, b is an illustration of the relationship between long and short time intervals.

Figure 2 - structure of the transmitter of the base station in the first embodiment of the assessment of dominant directions.

Figure 3 - structure of the receiver of the subscriber station.

4 is a functional diagram of a block of slow signal processing in the first method for assessing dominant directions.

Figure 5 is a functional diagram of a communication channel estimation node.

6 is an illustration of the estimation (prediction) process of the first embodiment of fast signal processing.

7 is an example of the radiation patterns of the pilot signals.

Fig - the structure of the transmitter of the base station in the second version of the assessment of dominant directions.

Fig.9 is a functional block diagram of the slow signal processing in the second method for assessing dominant directions.

Figure 10 - noise immunity characteristics.

11 - simulation results in a non-ideal feedback channel.

Fig - simulation results in conditions as close as possible to the real conditions of the cellular communication system 3GPP.

Fig is a variant of the timing diagram of the proposed method.

Fig - conveyor organization of the algorithm for slow processing.

Fig - a method of reducing the required number of omnidirectional pilots.

The proposed method consists in the fact that the subscriber station estimates the dominant (prevailing) spatial directions of the direct communication channel (from the base station to the subscriber), corresponding to the highest values of the signal power transmission coefficient. The subscriber station transmits the received information about the dominant spatial directions to the base station.

In the first version of the feedback directional transmission method, the base station transmits a signal to the subscriber station, switching the direction of signal transmission between the dominant directions, choosing the direction with the maximum transmission coefficient at the moment.

The directional feedback feedback method of the first embodiment is that

- from the base station through the transmitter transmit the information signal of the subscriber station, forming a radiation pattern based on the control information received from the subscriber station via the feedback channel,

- while at the base and subscriber stations carry out slow and fast signal processing, and the slow signal processing is performed periodically with a period of T ₁ and fast signal processing is performed periodically with a period of T ₂ , while T _{2 is} less than T ₁ ,

- periodically with a period of slow signal processing T ₁ at the subscriber station, the spatial communication channel between the base and subscriber stations is evaluated, and this assessment is carried out using pilot signals that are generated by the base station transmitter and transmitted in addition to the information signal,

- evaluate the spatial directions of signal propagation from the base station to the subscriber station,

- choose M dominant signal propagation directions from the indicated estimated spatial directions, which are characterized by the highest received signal energy,

- the obtained M estimates of the dominant directions of signal propagation are ordered, quantized, subjected to error-correcting coding, thus forming a sequence of ordered estimates of M dominant directions of signal propagation,

- a sequence of ordered estimates of M dominant signal propagation directions is transmitted via a reverse channel from a subscriber to a base station as long-term control information,

- at the base station, a sequence of ordered estimates of M dominant signal propagation directions is received on the return channel from the subscriber station, they are decoded and stored in the order in which they were received,

at each period of fast signal processing T ₂ at the subscriber station

- evaluate the signal energy corresponding to each of the M dominant signal propagation directions, estimates of which have already been transmitted to the base station,

- evaluate the energy value of the received signal in the next period of fast processing for each of the M dominant directions of signal propagation, using for this assessment of the signal energy generated on several previous periods of fast processing,

- determine the best direction for signal transmission as the dominant signal propagation direction, which corresponds to the maximum estimated signal energy value, and the number of this direction in an ordered sequence of estimates of M dominant signal propagation directions is transmitted via the reverse channel to the base station as a short-term control information - at the base station receive the serial number of the best direction for transmitting the signal and transmit the information signal to the subscribers second station to form a radiation pattern with a maximum in this direction.

In this case, the base station can either receive the serial number of the best direction for transmitting the signal and transmit the information signal to the subscriber station, forming a radiation pattern with a maximum in this direction, or the base station can receive the serial number of the best direction for transmitting the signal and transmit the bulk of the power of the subscriber information signal stations, forming a radiation pattern with a maximum in this direction, and evenly distribute the rest of the power for transfers in the remaining M minus 1 dominant directions.

Evaluation of M dominant directions of signal propagation from the base station to the subscriber station relative to the antenna array of the base station is performed at the subscriber station, for example, based on the reception of pilot signals transmitted from various antenna elements of the antenna array of the base station and the known configuration of the antenna array, or based on reception K pilot signals transmitted from a base station with different radiation patterns.

When M is 1, at the mobile station, one obtained estimate of the dominant direction of signal propagation is quantized and transmitted over the return channel from the subscriber to the base station as control information, at the base station, an estimate of the dominant direction of signal propagation is received on the return channel from the subscriber station, and information signal of the subscriber station, forming a radiation pattern with a maximum in this direction.

In the second variant of the directional transmission method, the base station transmits a signal to the subscriber station simultaneously in all dominant directions, setting the phase difference between them, ensuring coherent addition of the corresponding signal components at the receiving point.

The feedback directional transmission method of the second embodiment is that

- from the base station through the transmitter transmit the information signal of the subscriber station, forming a radiation pattern based on the control information received from the subscriber station via the feedback channel,

- while at the base and subscriber stations carry out slow and fast signal processing, and the slow signal processing is performed periodically with a period of T ₁ and fast signal processing is performed periodically with a period of T ₂ , while T _{2 is} less than T ₁ ,

- periodically with a period of slow processing of the signal T ₁ at the subscriber station,

- evaluate the spatial communication channel between the base and subscriber stations, and this assessment is carried out using pilot signals that are generated by the transmitter of the base station and transmit in addition to the information signal,

- evaluate the spatial directions of signal propagation from the base station to the subscriber station,

- choose M dominant signal propagation directions from the indicated estimated spatial directions, which are characterized by the highest received signal energy,

- received M estimates of the dominant directions of signal propagation are ordered, quantized, subjected to noise-resistant coding, thereby forming a sequence of ordered estimates of M dominant directions of signal propagation,

- a sequence of ordered estimates of M dominant signal propagation directions is transmitted via a reverse channel from a subscriber to a base station as long-term control information,

- at the base station, a sequence of ordered estimates of M dominant signal propagation directions is received on the return channel from the subscriber station, they are decoded and stored in the order in which they were received,

on each period T ₂ fast signal processing at the subscriber station

- evaluate the values of the relations of amplitudes and phase differences between the signal components, adopted from various dominant directions of signal propagation,

- the resulting estimates are quantized and transmitted over the reverse channel to the base station as short-term control information,

at the base station, estimates of the values of the ratios of the amplitudes and phase differences are obtained and the information signal of the subscriber station is transmitted in the M dominant directions of signal propagation, setting the values of the ratios of the amplitudes and phase differences between the signal components transmitted in different directions in accordance with the obtained estimates, so that Maximize signal strength at the receiving point.

Evaluation of M dominant directions of signal propagation from the base station to the subscriber station relative to the antenna array of the base station is performed at the subscriber station, for example, based on the reception of pilot signals transmitted from various antenna elements of the antenna array of the base station and the known antenna array configuration, or based on reception K pilot signals transmitted from a base station with different radiation patterns.

When M is 1, at the mobile station, one obtained estimate of the dominant direction of signal propagation is quantized and transmitted over the return channel from the subscriber to the base station as control information, at the base station, an estimate of the dominant direction of signal propagation is received on the return channel from the subscriber station, and information signal of the subscriber station, forming a radiation pattern with a maximum in this direction.

At each period of fast processing of the T ₂ signal at the subscriber station, only the phase difference between the signal components received from different dominant signal propagation directions can be estimated, the obtained estimates are quantized and transmitted along the reverse channel to the base station as short-term control information, while on the base stations receive phase difference estimates and transmit the information signal of the subscriber station in M dominant signal propagation directions, setting the separation values and phase between the signal components transmitted in different directions in accordance with the estimates obtained, so as to maximize the signal power at the receiving point.

The inventive method uses the estimation of the dominant (prevailing) spatial directions of signal propagation from the base station to the subscriber station, corresponding to the highest values of the signal power transmission coefficient.

Figure 1, a illustrates the situation in which the signal propagates from the base station to the subscriber station in two beams, due to the presence of two reflective objects. Two dominant angular directions θ ₁ and θ _{2 of the} signal propagation relative to the antenna array of the base station correspond to these rays. Due to the movement of the subscriber station, the set of these directions changes during the communication session. However, due to fading, the transmission coefficients of the communication signal in the directions θ ₁ and θ ₂ change. Therefore, with directional transmission, it is necessary to monitor both the set of dominant directions and the currently best transmission direction from the set of dominant directions.

The set of directions θ ₁ and θ ₂ changes relatively slowly and can serve as long-term control information about the communication channel. Therefore, the evaluation and transmission of this relatively stable information to the base station is performed on a relatively long time interval T ₁ , which is a period of slow signal processing. The volume of this information is small and, significantly, does not depend on the number of antenna elements. For example, when representing the angle θ with 5 bits and using two dominant directions, the amount of long-term data is only 10 bits.

The present invention can be used in various radio communication systems, but for the convenience of description, we will further consider its use in a communication system according to the promising 3GPP standard.

In a 3GPP communication system, data is transmitted using a frame format: 1 frame is 10 ms or 15 slots of 0.6667 ms each. For the transmission of feedback data in the 3GPP system, 1 bit is allocated in each slot. Under these conditions, when transmitting long-term data at a speed of 1 bit / frame, the length of the long time interval will be T ₁ = 10 frames.

After a set of dominant directions has been transmitted to the base station, it is enough to tell her only the number of the best transmission direction at the current time, which serves as a short-term control information about the communication channel. Therefore, the selection of the best direction and transmission of the corresponding number to the base station is performed on a shorter time interval T ₂ , which represents a period of fast signal processing. With two dominant directions, 1 bit is required to transmit the best direction number. In the 3GPP system, the short interval duration will be T ₂ ≈1 slot.

1b illustrates the relationship between periods of slow and fast signal processing. Moreover, T ₁ / T ₂ = 150.

To increase the reliability of the transmission of long-term data on the reverse control channel, they can be additionally encoded using an error-correcting code. For example, to encode long-term information, you can use the well-known block-frequency BCH code (see, for example, Prokis J. “Digital Communications”, M., Radio and Communications, 2000, pp. 372-375 [10]). In order to correct, for example, up to three errors in the received data using the BCH code, it is necessary to add 15 check bits to 10 bits of long-term data. That is, the total amount of long-term information, taking into account its encoding, will be 25 bits. In this case, the long time interval will be T ₁ = 25 frames.

The timing diagram for the simplest sequential execution of the functions of the proposed algorithm is presented in Fig. 13, where the notation is accepted: CI, CI - long-term and short-term control information, respectively.

The time spent on the formation of long-term information at a subscriber station (MS) and its transmission to the BS leads to a delay between the estimation of dominant directions and their use for directional transmission. In particular, this leads to an initial time delay of approximately equal to T _beg = T ₁ + T ₁ . Here, the first term T ₁ is the duration of the formation of CI (period of slow signal processing). The second term T ₁ corresponds to the transmission time of the CI. This sum does not include the propagation time of the signal τ, which is small compared with the other terms.

It is significant that the initial delay T _beg takes place only at the beginning of the algorithm (see Fig.13). During this period, the information signal for a given MS can, for example, be transmitted uniformly, that is, without using directional transmission.

In one of the options for implementing the prototype, the delay value of long-term information T _beg ≈2 · 640 = 1280 frames. In the aforementioned implementation option of the proposed solution, T _beg ≈2 · 10 = 20 frames (or 200 ms) if the coding of long-term control information is not applied, and T _beg ≈2 · 25 = 50 frames (or 500 ms) if the coding of long-term control information.

Since the spatial directions of the propagation of the signal change relatively slowly (the lifetime of the spatial directions of the propagation of the radio signal is usually at least 1 second), the delay of long-term information in the steady state does not lead to a deterioration in the characteristics of the proposed method.

However, the implementation of the slow processing algorithm described above is not the only possible one. Its other options are possible in the framework of the invention. For example, the delay of long-term information in the proposed method can be significantly reduced due to the application of the pipeline organization of the slow processing algorithm. The implementation of this principle on the example of two (M equals 2) dominant directions is presented in Fig. 14. The main idea here is that the updating of long-term control information occurs gradually, i.e. instead of updating simultaneously all M dominant directions using one large block of MDIs (as in the previous case), M dominant directions are updated in turn. First, the 1st dominant direction is updated, then the 2nd, ..., Mth direction, then again the 1st direction, etc. To do this, the assessment of M dominant directions is carried out at overlapping intervals. In this case, the overlap value corresponds to the duration of transmission on the return channel of one dominant direction. The newly obtained estimates of M dominant directions are correlated with M dominant directions that are already transmitted and stored on the BS. As a result, one dominant direction is selected, which replaces the dominant direction stored on the BS to be updated.

Let, for example, the transmission interval of each dominant direction be 9 frames (5 bits of useful information and 4 redundant bits when using encoding), i.e. 90 ms Accordingly, the estimation time for a set of dominant directions is 2 · 90 = 180 ms and, accordingly, the delay time of long-term information is 2 · 90 + 90 = 270 ms. Thus, the delay in long-term information was almost halved compared to the first method for implementing the slow processing algorithm.

The most sensitive to delay is the short-term processing algorithm. Due to the fast fading of the signal, the transmission coefficient of the signal in each of the dominant directions (which themselves change very slowly) can change rapidly. Therefore, to obtain the greatest gain from the application of the proposed method of directional transmission, it is necessary to minimize the delay in monitoring the best direction of transmission (the time interval between the moments of the beginning of the estimation of the best direction of transmission and the beginning of directional signal transmission in this direction). In the general case, the delay in the short-term processing algorithm is the sum of the duration of the short time interval T ₂ and the delay in the propagation of the signal from the BS to the MS. Due to the relative smallness of the latter, one can usually neglect (Fig. 13). During a short time interval, the current best transmission direction is evaluated, and the corresponding short-term control information is generated and transmitted to the BS. When M is 2, in the case of the 3GPP communication system, the interval of short-term information is equal to the duration of one slot (0.667 ms) and is quite small. The efficiency of tracking the best spatial direction in the present invention is further enhanced by predicting the best transmission direction based on its previous estimates, which in most cases can almost completely eliminate the delay in the short-term processing algorithm.

The base station receives and stores a set of dominant directions θ ₁ and θ ₂ . In the first version of the fast processing algorithm for each of the short time intervals T _{2, the} base station receives the number of the best transmission direction and generates an information signal for this user in this direction. In fact, the base station transmits a signal to the subscriber station, periodically switching the direction of signal transmission between the dominant directions, thereby monitoring the preferred direction of signal transmission.

However, another way is possible to transmit the signal to the base station at each short time interval. According to this method, at each of the short time intervals T _{2, the} base station receives the number of the best transmission direction and transmits the main part of the power of the information signal of the subscriber station, forming a radiation pattern with a maximum in this direction, and transfers the rest of the power, distributing the remaining M minus 1 dominant directions. This method may have an advantage due to its greater resistance to errors in determining dominant directions at a mobile station and to errors in a feedback channel.

The second variant of the fast signal processing algorithm is that BS transmits the signal simultaneously in all dominant directions, setting the ratios between the amplitudes and phases of the signal components transmitted in different directions, ensuring the maximum signal power at the receiving point

The second option provides high efficiency directional transmission in the case when the multipath components are insoluble in time, that is, they have close values of the time delays. In this case, fading in the communication channel is predominantly flat.

When implementing the proposed method, the second fast processing option is used along with the first option to increase the efficiency of directional transmission. This is carried out, for example, as follows. In the process of evaluating the communication channel at the subscriber station, the power values of the multipath components are compared. If the relative fraction of the maximum beam power does not exceed a certain threshold (for example, 90%), then the first version of the fast processing algorithm is performed. Otherwise, perform the second option. At the same time, the base station is informed about the selected mode, adding one bit to each packet with long-term information, the value of which corresponds to one of two transmission modes.

The inventive method provides two options for assessing the dominant directions of signal propagation: 1) based on the reception of non-directional (transmitted throughout the working angular sector) pilot signals transmitted from various antenna elements of the antenna array of the base station and the known configuration of the antenna array, and 2) based on the reception K pilot signals transmitted from a base station with different radiation patterns.

A possible way to implement the alleged invention is presented in relation to the 3GPP standard system when using a four-element antenna array at the base station.

Consider the first option for assessing dominant directions (non-directional pilot signals).

The generalized transmitter structure of the base station in the first version of the assessment of dominant directions is presented in Figure 2.

The base station transmitter comprises a physical data channel generation unit 1, multipliers 2-5, adders 8-11, directional transfer weight generator 6, pilot signal generator 12, a reverse channel reception unit and feedback information extraction unit 7. An output of the physical data channel generation unit 1 is connected to the first input of each of the multipliers 2-5, the second input of each of the multipliers 2-5 is connected to the corresponding output of the load shaper of the directional transmission 6. The output of each of the multipliers 2-5 is connected to the first input Ohm the corresponding adder 8-11. The second input of each of the adders 8-11 is connected to the corresponding output of the pilot signal generator 12. The output of each adder is connected to its corresponding antenna. The input of the shaper of the weight of the directional transmission 6 is connected to the output of the receiving unit of the return channel and the allocation of feedback information 7.

2, for simplicity, a signal conditioning channel of only one user is shown. For other users, identical channels are used, each of which consists of a block for the formation of a physical data channel 1, four multipliers 2-5 and a weight shaper for directional transmission 6.

The signal of the subscriber station transmitted in the physical data channel allocated for it is generated in the corresponding block for the formation of the physical data channel 1 and is supplied simultaneously to the first input of each of the four multipliers 2, 3, 4, 5. The second input of each of the multipliers 2-5 receives the corresponding weight of the directional transmission, previously formed in the shaper of the weight of the directional transmission 6. Information about the dominant directions is received at the input of the shaper of the weight of the directional transmission 6 and the best transmission direction, formed in the block for receiving the return channel and extracting feedback information 7. In the shaper of the scales of directional transmission 6, a slow and fast signal processing is performed.

The signal from the output of each of the multipliers 2-5 is fed to the first input of the corresponding adder 8-11. On the second input of each of the adders 8-11 receives the pilot signal of the corresponding transmitting antenna from the pilot signal generator 12. In the 3GPP standard, this is a signal from one of the common pilot channels. The output signal of each of the adders 8-11 enters the channel of the corresponding transmitting antenna A1, ..., A4. Figure 2 does not show the signal processing blocks at the radio frequency.

Figure 3 presents a variant of the receiver structure of the subscriber station. The receiver of the subscriber station consists of an information signal receiver 13, a slow signal processing unit 14, a fast signal processing unit 15, and a feedback message shaper 16. An input signal x is input simultaneously to an information signal receiver 13, a slow signal processing unit 14, and a fast signal processing unit 15. The output of the slow signal processing unit is connected to the second input of the fast signal processing unit 15 and the first input of the feedback shaper 16. The output of the fast processing unit with I drove 15 is connected to a second input of the feedback messages 16, from which output signal is supplied to the transmitter of the subscriber station.

The receiver of the information signal 13 performs the usual operations necessary to isolate the information received at the subscriber station during the communication session. In the block of slow signal processing 14, estimates of dominant directions are formed. These estimates are sent to the second input of the signal processing unit 14, as well as to the input of the feedback shaper 16. In the block of fast signal processing 15, an estimate of the best direction of directional transmission is generated, which is fed to the second input of the feedback shaper 16. Feedback shaper 16 generates information symbols for the feedback message, which are transmitted to the transmitter of the subscriber station for transmission to the base station via the reverse channel.

A variant of the functional diagram of the slow signal processing unit 14 in the first method for evaluating dominant directions is presented in FIG. 4. The slow signal processing unit 14 comprises a communication channel estimation node 17, a dominant signal propagation direction estimation node 18, and a memory node 19.

The slow signal processing unit 14 performs communication channel estimation using the communication channel estimation node 17, an embodiment of which is shown in FIG. The communication channel estimation node 17 consists of a set of quadrature correlators 20-1-20-R, an element for generating vector estimates of channel 21 and a copy generator of the pilot signals 22. In this case, the signals from the outputs of the correlators 20-1-20-R are fed to the inputs of the forming element channel 21 estimation vectors, and the control signal from the output of channel 21 channel estimation vector generation element is fed to the input of the pilot signal generator 22, the signals from the output of the pilot signal generator 22 are fed to the reference inputs of the quadrature correlators 20-1-20-R. The functional diagram of the quadrature correlator is described, for example, in Andrew J. Viterbi "CDMA Principles of Spread Spectrum Communication". Addison-Wesley Wireless Communications Series, p.p. 41-42 [11].

In the element of forming the vector of estimates of channel 21, the L most powerful multipath components (or rays) of the input signal are detected and their time delays are estimated. Methods of signal detection and time synchronization are described, for example, in [11] on pages 39-48. The maximum number of beams detected is often determined by the complexity limits of the receiver. For example, it can be 4, that is, L is less than or equal to 4. Then, using the quadrature correlators 20-1-20-R and the copy generator of the pilot signals 22, the complex envelope h _{l, n of} each of the 4 pilot signals is estimated ( n = 1, ..., 4) for each temporal ray (l = 1, ..., L):

и

- значения синфазной и квадратурной составляющих корреляции между входным сигналом и n-м опорным пилот-сигналом с l-м значением временной задержки,

. Набор значений h_l,n образует вектор оценок пространственного канала для l-го временного лучаWhere

and

- the values of the in-phase and quadrature components of the correlation between the input signal and the nth reference pilot signal with the l-th value of the time delay,

. The set of values h _{l, n} forms a vector of spatial channel estimates for the l-th time ray

The formation of the vector of spatial channel estimates is repeated on each of the successive sub-intervals of a long time interval. Such a sub-interval may be, for example, a short time interval T ₂ . The generated estimation channel vectors of the spatial channel h ₁ , ..., h _L go to the input of the functional unit for estimating dominant directions 18.

The node for evaluating the dominant directions 18 can operate according to various algorithms for estimating the angular directions of signal propagation relative to the antenna array of the base station (see, for example, [8]).

Consider the method described in [8], which consists in performing the following sequence of operations.

The covariance matrix of the communication channel is formed for each of the L rays (l = 1, ..., L):

where h _l (k) is the spatial channel estimation vector generated at the kth subinterval of a long time interval;

(.) ^H is the sign of complex conjugation and transposition.

The covariance matrix of the communication channel is formed by summing the matrices R _l :

и соответствующих собственных значений

ковариационной матрицы R.Define N eigenvectors

and corresponding eigenvalues

covariance matrix R.

Select D signal eigenvectors. There are two possible ways of choosing:

1) D is determined in advance, D is selected with the largest eigenvalues, and the corresponding eigenvectors are considered as signal ones.

, где α - некоторое число.2) All eigenvalues are compared with the Thresh threshold. The maximum eigenvalue and values exceeding the threshold, and the corresponding eigenvectors are considered as signal ones. The threshold is formed, for example, by the minimum eigenvalue: Thresh = αmin (λ _i ),

, where α is a certain number.

The remaining N-D smallest eigenvalues are considered noise. A matrix is formed from the corresponding noise eigenvectors

The crucial function is formed in accordance with the expression

where w (θ) = [w ₁ (θ), w ₂ (θ), ..., w ₄ (θ)] ^T is the vector of weight coefficients of the antenna array of the base station, corresponding to the angular direction 9 of the signal propagation.

.D local maxima of the decision function P (θ) are selected and the corresponding angles are determined

.

оценивают мощность сигнала, принятого с этого направления по формулеChoose M dominant signal propagation directions. For this, for each of the corners

estimate the power of the signal received from this direction according to the formula

where h _l (k) is the vector of spatial channel estimates for the l-th time ray at the k-th subinterval of the long time interval T ₁ ;

K is the total number of sub-intervals in a long time interval.

Estimates of M dominant directions are formed as angles θ _i , to which corresponds M the largest estimates of power p _j .

If M is greater than D, then the angle with the highest power rating is reused as the missing estimate of the dominant direction. The generated estimates of the dominant directions θ ₁ , ..., θ _M are quantized.

In the case of the implementation of the first variant of fast signal processing in the block of fast signal processing 15 at each k-th sub-interval, the total power of all L multipath signal components received from each of the M dominant directions, information about which has already been transmitted to the base station, is estimated:

.Using estimates of the power of the mth dominant direction on the S previous subintervals p _m (k-S + 1), ..., p _m (k), the value for the next k + 1st subinterval is estimated (predicted) by extrapolation (see 6). This process is repeated for all angles θ _m ,

.

The angle corresponding to the largest predicted value of the received power is taken as an estimate of the best direction for signal transmission.

To increase the efficiency of directional transmission, along with the first one, the second fast processing option can be used.

This is carried out, for example, as follows. In the process of evaluating the communication channel at the subscriber station, the values of the power L of the time-resolved multipath components are evaluated:

The relative part of the maximum beam power is estimated

The threshold is determined for the relative part of the maximum beam power, for example, δ ₀ is 0.9, and the value δ is compared with it.

If δ is less than δ ₀ , then the first version of the fast processing algorithm is executed, if δ is greater than or equal to δ ₀ , then the second version of the fast processing algorithm is executed.

In this case, one bit of feedback information is used to inform the base station of the selected option. This bit is transmitted, for example, at the beginning of each packet with long-term information.

When the second variant of the fast processing algorithm is executed, the following operations are performed in the fast processing unit 15.

A vector of spatial channel estimates h is formed for the strongest temporal ray.

, соответствующие М доминантным направлениям распространения сигнала с помощью, например, следующих выражений:Calculate the complex amplitudes of the signals at _m ,

corresponding to M dominant signal propagation directions using, for example, the following expressions:

Estimate the ratio of signal amplitudes to the largest signal amplitude

The value of the phase difference between each signal and the signal corresponding to the first dominant direction is estimated

where arg (x) is the argument of a complex number x;

(x) * is the complex conjugate of x.

поступают на выход блока быстрой обработки сигнала 15. С выхода блока быстрой обработки сигнала 15 полученные значения отношений амплитуд и разностей фаз поступают на вход формирователя сообщения обратной связи 16, где их квантуют и передают на базовую станцию.Δ _m and / or Δφ _m

arrive at the output of the fast signal processing unit 15. From the output of the fast signal processing unit 15, the obtained values of the ratios of the amplitudes and phase differences are fed to the input of the feedback message generator 16, where they are quantized and transmitted to the base station.

At the base station, a signal is transmitted simultaneously in M dominant directions, and the amplitude of the transmitted signal in each of the M directions is set proportionally to the amplitude ratios obtained on the return channel. In this case, the largest signal amplitude is transmitted in the dominant direction with the highest transmission coefficient, and the smallest - in the direction with the lowest transmission coefficient.

, то есть таким образом, чтобы передаваемые сигналы в точке приема сложились по возможности синфазно.The phases of the signals transmitted in M dominant directions except the first, are rotated by -Δφ _m , respectively

, that is, in such a way that the transmitted signals at the receiving point are as common as possible.

A preferred embodiment of the present invention corresponds to the case M is 2. When M is 2, the fast processing algorithm of the second embodiment is preferably performed in accordance with closed loop diversity transmission algorithms from the 3GPP standard mode 1 and mode 2 (see [6]). Moreover, the signals at ₁ and ₂ , formed using expression (11), should be used as input signals for these algorithms at the mobile station. Those. these diversity transmission algorithms are used to maximize the sum of the signals arriving at the receiving point in various dominant propagation directions, instead of the signals transmitted by the diversity antennas, as in the case of the 3GPP standard.

In the mode 1 diversity transmission algorithm, only the phase difference between the two input signals is determined at the mobile station. This phase difference is quantized into two possible values. Therefore, to transmit short-term information to the base station requires only one bit. Accordingly, the short-term interval is only 1 slot direct channel 3GPP.

In the mode 2 diversity transmission algorithm at the mobile station, the phase difference of the two input signals is quantized to a larger number of values, and in addition, the amplitude ratio between the two signals is calculated and quantized. Therefore, 4 bits are used to transmit short-term information to the base station. Accordingly, the short-term interval is 4 3GPP forward channel slots.

When implementing the second variant of the estimation of dominant directions, the base station transmits K pilot signals, forming for each of them a radiation pattern with a maximum in the corresponding direction and the corresponding angular width, so as to "cover" the entire area of the angular working sector of the base station (see Fig. .7).

The structure of the transmitter of the base station in the second version of the assessment of the dominant directions of signal propagation is presented in Fig. 8.

The base station transmitter comprises a physical data channel generation unit 1, multipliers 2-5, adders 8-11, a directional transfer weight generator 6, a pilot signal generator 12, a reverse channel reception and feedback information extraction unit 7, a beam former 23. An output the unit for forming the physical data channel 1 is connected to the first input of each of the multipliers 2-5, the second input of each of the multipliers 2-5 is connected to the corresponding output of the shaper of the weight of the directional transmission 6. The output of each and of multipliers 2-5 connected to the first input of the corresponding adder 8-11. The second input of each of the adders 8-11 is connected to the corresponding output of the beam former 23, the inputs of which are connected to the corresponding outputs of the pilot signal generator 12. The output of each of the adders 8-11 is connected to its corresponding antenna A1-A4. The fifth output of the shaper of the directional transmission weights 6 is connected to the input of the reverse channel receiving unit and information extraction 7.

8, for simplicity, a signal conditioning channel of only one user is shown. For other users, identical channels are used, each of which is represented by a physical data channel generation unit 1, four multipliers 2-5 and a directional transfer weight shaper 6. The signal processing units at the radio frequency are also not shown in Fig. 8 for reasons of simplicity.

The pilot signals generated in the pilot signal generator 12 are fed to the input of the beam former 23, where each of them in the channel of each antenna is multiplied with the corresponding weight coefficient, after which the results of multiplication corresponding to the same antenna are summed and fed to the output of the beam former 23. The output signals of the beam former 23 are fed to the inputs of the adders of the respective antennas.

This option for assessing the dominant directions of signal propagation can be implemented based on the same receiver structure of the subscriber station (Figure 3). The functional block circuit of the slow signal processing 14 in this case can be performed as shown in Fig.9.

The slow processing unit consists of a set of R quadrature correlators, an element for estimating the power of pilot signals 24, a generator of copies of the pilot signals 22, and an element for generating estimates of dominant directions 25. The signals from the outputs of the correlators 20-1-20-R are fed to the inputs of the evaluation element power of the pilot signals 24, and from the first output of this element for evaluating the power of pilot signals 24, the control signal is input to the generator of copies of the pilot signals 22, from the output of which the generated copies of the pilot signals are supplied to the reference inputs of the quad Rural correlators 20-1-20-R. The signals from the outputs of the power estimation P (θ ₁ ), ..., P (θ _K ) of the pilot power estimation element 24 are fed to the input of the dominant direction estimation generating element 25. For a functional diagram of the quadrature correlator, see, for example, Andrew J. Viterbi "CDMA Principles of Spread Spectrum Communication". Addison-Wesley Wireless Communications Series, pp41-42.

In the power estimation element of the pilot signals 24, delays of each of the L most powerful multipath components (or time beams) of the signal are detected and evaluated. Evaluation methods are presented, for example, in [11] p. 39-48. The maximum number of rays detected is determined by the restrictions on the complexity of the receiver. For example, it can be 4, that is, L is less than or equal to 4. Then, the power p _{l, k of} each of the K pilot signals (k = 1, ..., K) is estimated for each time beam (l = 1 ,. .., L):

и

- синфазная и квадратурная составляющие корреляции между входным сигналом и k-м пилот-сигналом с l-м значением временной задержки, формируемые в квадратурном корреляторе 20. Мощность пилот-сигнала, принятого с направления θ_k, оценивается, какWhere

and

- in-phase and quadrature components of the correlation between the input signal and the k-th pilot signal with the l-th time delay value generated in the quadrature correlator 20. The power of the pilot signal received from the θ _k direction is estimated as

In the element of forming estimates of dominant directions, the set of values of P (θ _k ) is considered as a function of the angle θ _k .

M dominant directions are determined, for example, as θ _k values corresponding to M local function maxima. Other methods for determining M dominant directions are also possible. For example, first determine the dominant direction, which corresponds to the maximum of the function P (θ _k ). Then modify the function P (θ _k ), excluding from it the found maximum and values in some neighborhood of the found maximum. The width of the excluded region is determined by the angular resolution of the antenna array. Then, the maximum of the modified function is determined and the corresponding propagation direction is taken as the second dominant direction. The described procedure is repeated until all M dominant directions are determined.

The output of the slow signal processing unit 14 receives the numbers n ₁ , ..., n _{M of the} corresponding pilot signals.

When implementing the first fast processing option, the fast processing unit 15 determines the best direction for signal transmission in each short time interval. To do this, evaluate the power values of the pilot signals corresponding to the dominant directions, information about which has already been transmitted to the BS:

Select a pilot signal with maximum power. The number of this pilot in the sequence of M pilot signals already transmitted at the base station is transmitted to the base station for a short time interval T ₂ .

To increase the efficiency of directional transmission, along with the first one, the second fast processing option can be used.

This is carried out, for example, as follows. In the process of evaluating the communication channel at the subscriber station, the total power of the pilot signals corresponding to each of the L time beams is determined.

. Оценивают относительную часть мощности наиболее сильного лучаCompare power

. The relative part of the power of the strongest beam is estimated

A threshold is determined for the relative part of the power of the strongest beam, for example, δ ₀ is 0.9, and the quantity δ is compared with it.

If δ is greater than or equal to δ ₀ , then the second version of the fast processing algorithm is executed, if δ is less than δ ₀ , then the first version of the fast processing algorithm is executed.

In this case, one bit of feedback information is used to inform the base station of the selected option. For example, one bit representing the option number is added at the beginning of each packet of long-term information. In the case of the implementation of the second variant of fast processing in the quick processing unit 15 perform the following operations.

, соответствующие М доминантным направлениям распространения сигнала, например, следующим образом:Calculate the complex amplitudes of the signals at _m ,

corresponding to M dominant signal propagation directions, for example, as follows:

и

- синфазное и квадратурное значения корреляции между входным сигналом и копией n-го пилот-сигнала со значением временной задержки наиболее сильного луча.Where

and

- in-phase and quadrature correlation values between the input signal and the copy of the nth pilot signal with the time delay value of the strongest beam.

Estimate the ratio of signal amplitudes to the largest signal amplitude

The value of the phase difference between each signal and the signal corresponding to the first dominant direction is estimated

where arg (x) is the argument of a complex number x;

(x) * is the complex conjugate of x.

поступают на выход блока быстрой обработки сигнала 15. С выхода блока быстрой обработки сигнала 15 полученные значения отношений амплитуд и разностей фаз поступают на вход формирователя сообщения обратной связи 16, где их квантуют и передают на базовую станцию.Δ _m and / or Δφ _m

arrive at the output of the fast signal processing unit 15. From the output of the fast signal processing unit 15, the obtained values of the ratios of the amplitudes and phase differences are fed to the input of the feedback message generator 16, where they are quantized and transmitted to the base station.

At the base station, a signal is transmitted simultaneously in M dominant directions, while the amplitude of the transmitted signal in each of the M directions is set proportionally to the amplitude ratios obtained on the return channel. In this case, the largest signal amplitude is transmitted in the dominant direction with the highest transmission coefficient, and the smallest - in the direction with the lowest transmission coefficient.

, то есть таким образом, чтобы передаваемые сигналы в точке приема сложились по возможности синфазно.The phases of the signals transmitted in M dominant directions except the first, are rotated by -Δφ _m , respectively

, that is, in such a way that the transmitted signals at the receiving point are as common as possible.

A preferred embodiment of the present invention corresponds to the case M is 2. When M is 2, the fast processing algorithm of the second embodiment is preferably performed in accordance with closed loop diversity transmission algorithms from the 3GPP standard mode 1 and mode 2 (see [6]). Moreover, the signals at ₁ and ₂ , formed using expression (11), should be used as input signals for these algorithms at the mobile station. Those. these diversity transmission algorithms are used to maximize the sum of the signals arriving at the receiving point in various dominant propagation directions, instead of the signals transmitted by the diversity antennas, as in the case of the 3GPP standard.

In the mode 1 diversity transmission algorithm, only the phase difference between the two input signals is determined at the mobile station. This phase difference is quantized into two possible values. Therefore, to transmit short-term information to the base station requires only one bit. Accordingly, the short-term interval is only 1 slot direct channel 3GPP.

In the mode 2 diversity transmission algorithm at the mobile station, the phase difference of the two input signals is quantized to a larger number of values, and in addition, the amplitude ratio between the two signals is calculated and quantized. Therefore, 4 bits are used to transmit short-term information to the base station. Accordingly, the short-term interval is 4 3GPP forward channel slots.

To increase the accuracy of estimation, the number of pilot signals must be increased. However, the resources of the pilot channels in any system, including 3GPP, are limited.

When using directional pilots to solve this problem, temporary multiplexing of the direction in which each pilot signal is transmitted can be used. For example, it is possible to transmit K pilot signals in 2 * K directions by changing the transmission direction of each pilot signal in two consecutive slots. In this case, the method of multiplexing should be known at the subscriber station.

In the case of using non-directional pilot signals, the problem of their limited number can also be solved, for example, as follows.

In the general case, to evaluate the communication channel, an appropriate pilot signal should be transmitted from each antenna. However, the present invention is characterized in that it uses only the most important information about the distribution channel — the dominant signal propagation directions.

To assess the dominant directions of signal propagation, you can use an antenna array with fewer antenna elements. This means that the BS can transmit pilot signals not from each antenna element, but, for example, only from a few central antenna elements. Due to the use on the MS of an algorithm for estimating the angle of arrival with high resolution (for example, MUSIC), the accuracy of estimating the dominant directions of signal propagation even when using an antenna array with a smaller number of antenna elements will be quite high.

Consider the principle described above with reference to an eight-element antenna array. At the same time, four central elements of the array can be used to evaluate dominant directions, that is, pilot signals can be transmitted from four central antenna elements (see Fig. 15). The directional signal transmission to the BS is carried out using all eight elements in order to form the narrowest beam of the radiation pattern.

The proposed feedback directional transmission method is based on direct estimation of the spatial signal propagation channel and therefore is superior in efficiency to similar methods based on estimating the angle of arrival of the return channel signal (see, US patent # 6347234. "Practical space-time radio method for CDMA communication capacity enhancement ". Adaptive Telecom, Inc. (Campbell, CA), Metawave Communications Corporation (Redmond, WA). February 12, 2002 [12], as well as [8]), as it is guaranteed to choose the optimal direction of signal propagation.

The proposed method effectively reduces fading of the signal, as it uses various propagation directions for transmission, which, as a rule, correspond to independent signal propagation paths.

Compared with the prototype, the level of intra-system interference is minimized, since the transfer of useful power is performed in a narrow spatial direction.

The proposed method can be implemented in the conditions of a feedback channel with a low transmission rate, since compact and efficient information is transmitted through it. For example, the method is quite effective if 2 dominant directions are used, and the angle θ is represented by five bits. The volume of slowly changing data is 10 bits, while the prototype requires 54 bits for transmitting two eigenvectors with four transmitting antennas and representing a complex number of eight bits.

With an increase in the number of transmitting antennas, the volume of long-term monitoring information in the prototype increases linearly, and for the proposed method, this volume is saved or increases slightly. This property makes it possible to increase the number of transmitting antennas without a noticeable increase in the amount of data transmitted over the feedback channel.

A small amount of long-term control information allows you to transmit it with less delay, as a result of which the proposed method provides quick tracking of the parameters of the distribution channel.

The minimum amount of transmitted data gives an advantage in conditions of an imperfect feedback channel, which allows a certain probability of error. Firstly, with a smaller volume, the block of transmitted data is less likely to be distorted. Secondly, a small amount allows you to apply the coding of slowly changing data and thereby almost completely eliminate errors when transmitting data via the feedback channel.

The proposed algorithm with the frequency of a short time interval performs switching between M dominant directions. To eliminate the effect of feedback delay, prediction of the best transmission direction for the next short time interval can be used.

In order to evaluate the characteristics of the proposed algorithm, computer simulation was performed. Figure 10 presents the dependencies of the bit probability of the BER error to the decoder for the following algorithms: for the prototype, the angle of return of the return channel and the proposed solution. Modeling conditions: the number of transmitting antennas N is 4, the Doppler frequency shift is 50 Hz; propagation channel with two time-insoluble rays, angular directions of signal propagation -20 and +20 degrees and angular expansion of 3 degrees.

Analysis of the presented dependence indicates the high efficiency of the proposed method in a multipath channel. For example, at the BER level of 10%, the gain of the proposed method relative to the prototype is approximately 2 dB.

The simulation results in a non-ideal feedback channel are presented in Fig.11. The estimated parameter here is the transmit power of the base station relative to the level of thermal noise. The curves show that when the probability of a feedback channel error of more than 4%, the gain of the proposed algorithm relative to the prototype exceeds 1.5 dB.

A comparison of the effectiveness of the proposed method and the prototype was also performed under conditions as close as possible to the real conditions of the direct channel of the 3GPP communication system. One connection was simulated (one sector of the base station and one mobile station). Gaussian noise was modeled as interference. The remaining simulation parameters are given in the table.

Table
Simulation parameters Information transfer rate 10 kbps Broadband Signal Width 3.84 MHz Coding rate 1/3 Carrier frequency 2 GHz Quick Power Adjustment Speed 1500 Hz Transmit Power Adjustment Step 1 dB Target Packet Probability one% errors in the power adjustment algorithm Reverse control channel bit rate 1500 bps Delay back channel information 1 slot Bit error rate in the reverse control channel 5% Signal propagation channel model Spatial Link Model Developed by 3GPP / 3GPP2 Standardization Groups

The parameters of the proposed method during modeling were as follows: M is 2 (two dominant directions), the amount of useful long-term data is 10 bits (5 bits per transmission direction), 15 redundant bits (allows you to fix up to three errors), the total amount of long-term data (useful data plus redundant) is 25 bits, the long-term processing interval is 375 slots, the short-term processing interval is 1 slot (one-bit command).

Modeling of the proposed method and prototype was performed for a four-element antenna array with a distance between the elements at half the wavelength.

The simulation of the diversity transmission algorithm with feedback (see [6]) was performed for two antenna elements spaced at ten wavelengths.

During the simulation, the ratio of the total (over all antennas) transmitted signal power to the thermal noise dispersion in the communication channel was estimated. This ratio is designated as TX ROT. The TX ROT parameter is one of the most important characteristics of transmission methods, since through it the throughput of the communication system is directly estimated. The lower the TX ROT value, the higher the throughput of the communication system.

The Тх ROT parameter (dB) was obtained for 50 random implementations of the distribution channel for each of the three possible types of distribution media (microcell, urban macrocell, suburban macrocell) according to the spatial distribution channel model developed by the 3GPP / 3GPP2 standardization group.

12 is a diagram showing the average values of the TX ROT parameter of the transmission methods under consideration obtained for the three types of distribution medium indicated above.

The obtained simulation results show that the proposed method has a significant gain in the parameter Tx ROT among the compared methods. Therefore, due to the practical implementation of the proposed method, the throughput of the communication system will be significantly increased.

The obtained high efficiency of the invention in comparison with the prototype is explained by a more complete use of the capabilities of a lattice with closely spaced elements, directional transmission using such lattices, the most informative parameter is the direction of transmission. Using this parameter can significantly reduce the amount of control data transmitted on the return channel. In this case, the antenna array implements a fairly efficient diversity in the multipath channel by adapting the transmission direction in accordance with fading in the communication channel.

Claims (10)

1. The method of directional transmission with feedback, which consists in the fact that the base station transmits the information signal of the subscriber station from the base station, forming a radiation pattern based on the control information received from the subscriber station via the feedback channel, while at the base and subscriber stations slow and fast signal processing, wherein the slow signal processing is performed periodically with a period T _1, a fast signal processing is performed periodically with a period T _2, et m T ₂ less than T _1, to the subscriber station periodically T ₁ signal with a period of slow processing evaluate spatial channel communication parameters between the base and subscriber stations, wherein this assessment is performed using pilot signals which are formed by a base station transmitter and transmitting the addition information signal characterized in that in each period T ₁ slow signal processing on the subscriber station evaluates spatial propagation direction of the signal from the base station to the CALLER At the station, choose M dominant signal propagation directions from the indicated estimated spatial directions, which are characterized by the highest received signal energy, obtained M estimates of dominant signal propagation directions are ordered, quantized, and subjected to error-correcting coding, thereby forming a sequence of ordered estimates of M dominant signal propagation directions, a sequence ordered estimates of M dominant signal propagation directions The channels are transmitted over the return channel from the subscriber station to the base station as long-term monitoring information; at the base station, a sequence of ordered estimates of M dominant signal propagation directions are received from the subscriber station via the return channel, decoded and stored in the order in which they were received, in each period T ₂ fast processing signal to the subscriber station evaluates the energy of the signal corresponding to each of the M dominant directions of propagation, which estimate already transmitted to the base station, estimate the energy value of the received signal in the next period of fast signal processing for each of the M dominant signal propagation directions, using for this purpose the signal energy estimates generated at several previous periods of fast signal processing determine the best direction for signal transmission as dominant the direction of signal propagation, which corresponds to the maximum estimated value of the signal energy, and the number of this direction in an ordered M th sequence of dominant directions of propagation of the signal estimates are transmitted on a reverse channel to the base station as a short-term control information at the base station receiving the sequence number of the best directions for transmitting a signal and transmitting the subscriber station information signal, forming a directivity pattern with a maximum in this direction. 2. The method according to claim 1, characterized in that the bulk of the power of the information signal of the subscriber station is transmitted, forming a radiation pattern with a maximum in this direction, and the rest of the power is evenly distributed for transmission in the remaining M minus 1 dominant directions. 3. The method according to claim 1, characterized in that the assessment of M dominant directions of signal propagation relative to the antenna array of the base station is performed at the subscriber station based on the reception of pilot signals transmitted from various antenna elements of the antenna array of the base station and the known configuration of the antenna array. 4. The method according to claim 1, characterized in that the assessment of M dominant directions of signal propagation relative to the antenna array of the base station is performed based on the reception of K pilot signals transmitted from the base station with different radiation patterns. 5. The method according to claim 1, characterized in that when M is equal to 1 at the mobile station, the obtained one estimate of the dominant direction of signal propagation is quantized and transmitted along the reverse channel from the subscriber to the base station as control information, and at the base station, they are received via the reverse channel from subscriber station estimates the dominant direction of signal propagation and transmit the information signal to the subscriber station, forming a radiation pattern with a maximum in this direction. 6. The method of directional feedback transmission, which consists in the fact that the information of the subscriber station is transmitted from the base station via the transmitter, forming a radiation pattern based on the control information received from the subscriber station via the feedback channel, while at the base and subscriber stations slow and fast signal processing, and slow signal processing is performed periodically with a period of T ₁ , and fast signal processing is performed periodically with a period of T ₂ , while m T ₂ less than T ₁ , at the subscriber station periodically with a period of slow signal processing T ₁ evaluate the spatial communication channel between the base and subscriber stations, and this assessment is carried out using pilot signals that are generated by the transmitter of the base station and transmit in addition to the information signal, characterized in that at each period T _{1 of} slow signal processing at the subscriber station, the spatial directions of signal propagation from the base station to the subscriber station are estimated and, M dominant signal propagation directions are selected from the indicated estimated spatial directions, which are characterized by the highest received signal energy, the obtained M estimates of the dominant signal propagation directions are ordered, quantized, subjected to error-correcting coding, thereby forming a sequence of ordered estimates of M dominant signal propagation directions, a sequence of ordered estimates of M dominant signal propagation directions are transmitted about the return channel from the subscriber station to the base station, as a long-term monitoring information, at the base station, a sequence of ordered estimates of the M dominant signal propagation directions is received from the subscriber station via the return channel, they are decoded and stored in the order in which they were received at each period T ₂ fast signal processing in the subscriber station is evaluated values of the ratios of the amplitudes and phase differences between the signal components taken from different directions dominant disintegrations signal elimination, the obtained estimates are quantized and transmitted along the reverse channel to the base station as short-term control information, estimates of the values of the amplitudes and phase differences ratios are received at the base station and the information signal of the subscriber station is transmitted in the M dominant directions of signal propagation, setting the values of the amplitudes and differences phases between signal components transmitted in different directions, in accordance with the estimates obtained in such a way as to maximize oschnost signal at the reception point. 7. The method according to claim 2, characterized in that at each period of fast processing of the T ₂ signal at the subscriber station, only the phase difference values between the signal components received from different dominant signal propagation directions are evaluated, the resulting estimates are quantized and transmitted along the reverse channel to the base station as a short-term control information, while at the base station receive phase difference estimates and transmit the information signal of the subscriber station in M dominant signal propagation directions, setting the values of the phase difference between the signal components transmitted in different directions, in accordance with the obtained estimates, so as to maximize the signal power at the receiving point. 8. The method according to claim 2, characterized in that the assessment of M dominant directions of signal propagation relative to the antenna array of the base station is performed at the subscriber station based on the reception of pilot signals transmitted from various antenna elements of the antenna array of the base station and the known configuration of the antenna array. 9. The method according to claim 2, characterized in that the assessment of M dominant signal propagation directions relative to the antenna array of the base station is performed based on the reception of K pilot signals transmitted from the base station with different radiation patterns. 10. The method according to claim 2, characterized in that when M is equal to 1 at the mobile station, the obtained one estimate of the dominant direction of signal propagation is quantized and transmitted via the return channel from the subscriber to the base station as control information, and at the base station, they are received via the return channel from subscriber station estimates the dominant direction of signal propagation and transmit the information signal to the subscriber station, forming a radiation pattern with a maximum in this direction.

Images