RU2700568C2 - Wireless communication system and method with distributed inputs and distributed outputs - Google Patents

Abstract

FIELD: wireless communication.

SUBSTANCE: invention relates to communication and can be used in wireless communication systems in which space-time coding techniques are used. Method for dynamic adaptation of characteristics of multi-antenna system (MAS) with multi-user (MU) transmissions ("MU-MAS") consists in the fact that it comprises stages, transmitting a training signal between each base station antenna and each of the plurality of wireless client devices analyzing each training signal for generating channel characteristic data; determining instant or statistical channel quality (link quality indicator) for wireless client devices using channel characteristic data; user subset and MU-MAS transmission mode are determined based on link quality indicator; calculating multiple weights of MU-MAS preliminary coder based on channel characteristic data; data are pre-coded with the help of MU-MAS preliminary coder weights for generation of pre-coded data signals for each base station antenna; and transmitting pre-coded data signals through each antenna of the base station to each corresponding client device in the selected subset.

EFFECT: technical result is compensation of frequency and time shift in a multi-antenna system.

27 cl, 52 dwg

Description

FIELD OF THE INVENTION

This invention relates generally to the field of communication systems. More specifically, the invention relates to a system and method for wireless communication with distributed inputs and distributed outputs, which use methods of space-time coding.

State of the art

Space-time coding of communication signals

A relatively new development in wireless technology known as spatial multiplexing and space-time coding is known. One particular type of space-time coding is called MIMO (with multiple inputs and multiple outputs), because several antennas are used at each end. By using multiple antennas for sending and receiving, multiple independent radio waves can be transmitted at the same time in the same frequency range. The following articles give an overview of MIMO.

IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 21, NO. 3, APRIL 2003: "From Theory to Practice: An Overview of MIMO Space-Time Coded Wireless Systems" (By theory to practice: Overview of MIMO space-time coding wireless systems), by David Gesbert, Member, IEEE, Mansoor Shafi, Fellow, IEEE, Dashan Shiu, Member, IEEE, Peter J. Smith, Member, IEEE, and Ayman Naguib, Senior Member, IEEE.

, Member, IEEE, Dhananjay A. Gore, and Arogyaswami J. Paulraj, Fellow, IEEE.IEEE TRANSACTIONS ON COMMUNICATIONS, VOL. 50, NO. 12, DECEMBER 2002: "Outdoor MIMO Wireless Channels: Models and Performance Prediction", David Gesbert, Member, IEEE, Helmut

, Member, IEEE, Dhananjay A. Gore, and Arogyaswami J. Paulraj, Fellow, IEEE.

Essentially, MIMO technology is based on the use of spatially distributed antennas to create parallel spatial data streams in a common frequency range. Radio waves are transmitted so that individual signals can be separated at the receiver and demodulated, despite the fact that they are transmitted in the same frequency range, which can lead to many statistically independent (i.e., effectively separated) communication channels. Thus, in contrast to standard wireless communication systems that attempt to suppress multipath signals (i.e., multiple signals at the same frequency, delayed in time and modified in amplitude and phase), MIMO can be based on uncorrelated or weakly correlated multipath signals to achieve higher bandwidth and improved signal to noise ratio in a given frequency range. As an example, MIMO technology achieves a much higher throughput under comparable conditions in terms of power and signal to noise ratio (S / N) (SNR), when a traditional non-MIMO system can only achieve lower throughput. This ability is described on the Qualcomm Incorporated website (Qualcomm is one of the largest wireless technology providers) on the page entitled "What MIMO Delivers" at http://www.cdmatech.com/products/what_mimo_delivers.jsp: MIMO is the only multi-antenna technique that increases spectral capacity by providing two or more times the peak data rate in the system per channel or per MHz spectrum. More specifically, for wireless LAN or Wi-Fi® applications, the fourth-generation QUALCOMM MIMO technology delivers 315 Mbps in the 36 MHz or 8.8 Mbps spectrum. Compare this to the peak bandwidth of the 802.11a / g standard (even with beamforming or diversity), which provides only 54 Mbps in the 17 MHz or 3.18 Mbps spectrum. ”

MIMO systems, as a rule, face a practical limitation of less than 10 antennas per device (and therefore less than 10x improvement in network bandwidth) for several reasons:

1. Physical limitations: MIMO antennas on a given device must be sufficiently spaced apart so that each receives a statistically independent signal. Although improvements in MIMO bandwidth can be seen even when the antennas are separated by a few fractions of the wavelength, the efficiency quickly deteriorates as the antennas approach, which leads to lower coefficients of increased MIMO bandwidth.

See, for example, the following links:

[1] D.-S. Shiu, G.J. Foschini, M.J. Gans, and J.M. Kahn, "Fading correlation and its effect on the capacity of multielement antenna systems," IEEE Trans. Comm., Vol. 48, no. 3, pp. 502-513, Mar. 2000.

[2] V. Pohl, V. Jungnickel, T. Haustein, and C. von Helmolt, "Antenna spacing in MIMO indoor channels," Proc. IEEE Veh. Technol. Conf., Vol. 2, pp. 749-753, May 2002.

[3] M. Stoytchev, H. Safar, A.L. Moustakas, and S. Simon, "Compact antenna arrays for MIMO applications," Proc. IEEE Antennas and Prop. Symp., Vol. 3, pp. 708-711, July 2001.

[4] A. Forenza and R.W. Heath Jr., "Impact of antenna geometry on MIMO communication in indoor clustered channels," (Proc. IEEE Antennas and Prop. Symp., Vol. 2, pp. 1700-1703, June 2004.

In addition, for small antenna spacing, the mutual coupling effects may degrade the performance of MIMO systems.

See, for example, the following links:

[5] M.J. Fakhereddin and K.R. Dandekar, "Combined effect of polarization diversity and mutual coupling on MIMO capacity," Proc. IEEE Antennas and Prop. Symp., Vol. 2, pp. 495-498, June 2003.

[7] P.N. Fletcher, M. Dean, and A.R. Nix, "Mutual coupling in multielement array antennas and its influence on MIMO channel capacity," (IEEE Electronics Letters, vol. 39, pp. 342-344, Feb. 2003.

[8] V. Jungnickel, V. Pohl, and C. Von Helmolt, "Capacity of MIMO systems with closely spaced antennas," IEEE Comm. Lett., Vol. 7, pp. 361-363, Aug. 2003.

[10] J.W. Wallace and M.A. Jensen, "Termination-dependent diversity performance of coupled antennas: Network theory analysis," (Terminal-dependent coupled antenna diversity characteristic: Theoretical network analysis) IEEE Trans. Antennas Propagat, vol. 52, pp. 98-105, Jan. 2004.

[13] C. Waldschmidt, S. Schulteis, and W. Wiesbeck, "Complete RF system model for analysis of compact MIMO arrays," IEEE Trans, on Veh. Technol., Vol. 53, pp. 579-586, May 2004.

[14] M.L. Morris and M.A. Jensen, "Network model for MIMO systems with coupled antennas and noisy amplifiers," IEEE Trans. Antennas Propagat., Vol. 53, pp. 545-552, Jan. 2005.

In addition, as the antennas get closer, these antennas should generally become smaller, which can also affect the performance of the antennas. See, for example, the following links:

[15] N.A. Wheeler, "Small antennas," IEEE Trans. Antennas Propagat., Vol. AP-23, n. 4, pp. 462-469, July 1975.

[16] J.S. McLean, "A re-examination of the fundamental limits on the radiation Q of electrically small antennas," (Review of the fundamental restrictions on the emissivity of an electrically small antenna) IEEE Trans. Antennas Propagat., Vol. 44, n. 5, pp. 672-676, May 1996.

Finally, at lower frequencies and longer wavelengths, the physical size of a single MIMO device can become cumbersome. The limiting case may be in the HF range, where it may be necessary to separate the antennas of MIMO devices from each other by 10 meters or more.

2. Limitations imposed by noise. Each subsystem of the transmitter-receiver MIMO creates a certain level of noise. As more and more subsystems are placed in close proximity to each other, the noise level increases. This requires an increasingly lower noise level, as an increasing number of different signals must be distinguished from one another in a multi-antenna MIMO system.

3. Limitations on cost and power. Although there are MIMO applications in which cost and energy consumption are not a problem, in a conventional wireless product, both cost and energy costs are critical constraints in developing a successful product. Each MIMO antenna requires a separate RF subsystem, including analog-to-digital and digital-to-analog converters (ADC and DAC). Unlike many aspects of digital systems, the scale of which varies according to Moore's law (an empirical observation made by Intel co-founder Gordon Moore, according to which the number of transistors in an integrated circuit for a minimum cost of components doubles approximately every 24 months; see http: // www .intel.com / technology / mooreslaw /), such subsystems saturated with analog components, as a rule, have specific requirements for physical design size and power and linear dependence of cost and power on scale. Thus, a multi-antenna MIMO device would become excessively expensive and energy intensive compared to a single-antenna device.

As a result of the above, the majority of currently designed MIMO systems have about 2-4 antennas, which leads to a two to four-fold increase in throughput and to some increase in the S / N ratio due to the benefits of diversity multi-antenna system. MIMO systems have been developed with up to 10 antennas (especially at higher microwave frequencies due to shorter wavelengths and shorter distance between antennas), but a further increase in the number of antennas is impractical, except for very specialized and expensive applications.

Virtual antenna arrays

One particular application of technology such as MIMO is a virtual antenna array. Such a system was proposed in a research report presented at the European Organization for Scientific and Technical Research EURO-COST, Barcelona, Spain, Jan 15-17, 2003: Center for Telecommunications Research, King's College London, UK: "A step towards MIMO: Virtual Antenna Arrays "(Step to MIMO: Virtual Antenna Array), Mischa Dohler & Hamid Aghvami.

Virtual antenna arrays, as presented in this document, are systems of collaborative wireless devices (such as cell phones) that communicate with each other (if and when they are close enough to each other) on a dedicated communication channel other than their the primary communication channel with their base station in order to work together (for example, if they are GSM cell phones in the ultra-high frequency range (UHF) (UHF), it can be industrial, scientific and medical (ISM) wireless apazone at 5 MHz). This provides a MIMO-like increase in the bandwidth of single-antenna devices by transmitting information between several devices in each other's coverage area (in addition to being in the coverage area of the base station), so that they work as if they were physically one device with many antennas.

In practice, however, such a system is extremely difficult to implement, and it has limited applicability. First of all, now there are at least two separate communication paths for each device, which must be supported in order to achieve improved throughput, and the second broadcast line often has unstable access. In addition, these devices are more expensive, physically larger, and consume more power because they have at least a second communication subsystem and higher computing needs. In addition, this system is more dependent on the very complicated coordination of all devices in real time, potentially through many different communication lines. Finally, as the simultaneous use of channels increases (for example, the transmission of simultaneous telephone calls using MIMO methods), the computational load on each device increases (potentially exponentially when channel use grows linearly), which can be very impractical for portable devices with severe restrictions on power and size.

SUMMARY OF THE INVENTION

A system and method for compensating frequency and phase shifts in a multi-antenna system (MAC) (MAS) with multi-user (MP) (MU) transmissions ("MP-MAC") ("MU-MAS") are described. For example, the method according to one embodiment of the invention comprises the steps of: transmitting a training signal from each antenna of the base station to one or to each of a plurality of wireless client devices, wherein one or each of the client devices analyzes each training signal to generate frequency shift compensation data , and receive frequency shift compensation data at the base station; calculating the weights of the MP-MAC precoder based on the frequency shift compensation data to pre-compensate the frequency shift at the transmitter; pre-encode the training signal using the weights of the pre-encoder MP-MAC to generate pre-encoded training signals for each antenna of the base station; transmitting a precoded training signal from each antenna of the base station to each of a plurality of wireless client devices, each client device analyzing each training signal to generate channel characteristic data, and receiving channel characteristic data at the base station; calculating a plurality of weights of the MP-MAC preliminary encoder based on the channel characteristic data, and these weights of the MP-MAC preliminary encoder are calculated for the pre-compensated frequency and phase shift and / or inter-user interference; pre-encoding data using these weights of the pre-encoder MP-MAC to generate signals of pre-encoded data for each antenna of the base station; and transmitting precoded data signals through each antenna of the base station to each corresponding client device.

Brief Description of the Drawings

The following description, together with the drawings, is for a better understanding of the present invention.

FIG. 1 illustrates an existing MIMO system.

FIG. 2 illustrates an N-antenna base station communicating with a plurality of single-antenna client devices.

FIG. 3 illustrates a three-antenna base station communicating with a plurality of single-antenna client devices.

FIG. 4 illustrates training signal techniques applied in one embodiment of the invention.

FIG. 5 illustrates channel characteristic data transmitted from a client device to a base station, according to one embodiment of the invention.

FIG. 6 illustrates a multi-input, distributed-output (“MIDO”) downstream transmission according to one embodiment of the invention.

FIG. 7 illustrates an upstream multiple-input multiple-output (“MIDO”) transmission according to one embodiment of the invention.

FIG. 8 illustrates a base station cycling across different groups of clients for bandwidth allocation according to one embodiment of the invention.

FIG. 9 illustrates proximity-based customer grouping according to one “embodiment of the invention.

FIG. 10 illustrates an embodiment of the invention applied to a near vertical spatial wave (NVIS) system.

FIG. 11 illustrates an embodiment of a DIDO transmitter with I / Q imbalance compensation functional units.

FIG. 12 illustrates an embodiment of a DIDO receiver with I / Q imbalance compensation functional units.

FIG. 13 illustrates one embodiment of DIDO-OFDM systems with I / Q imbalance compensation.

FIG. 14 illustrates one embodiment of a 2 × 2 DIDO mode with and without I / Q imbalance compensation.

FIG. 15 illustrates one embodiment of a 2 × 2 DIDO mode with and without I / Q imbalance compensation.

FIG. 16 illustrates one embodiment of an error symbol occurrence rate (SER) with and without I / Q imbalance compensation for different QAM constellations.

FIG. 17 illustrates one embodiment of 2 × 2 DIDO modes with and without I / Q imbalance compensation at various locations of user devices.

FIG. 18 illustrates one embodiment of a SER with and without I / Q imbalance compensation in ideal (independent and identically distributed (i.i.d.)) channels.

FIG. 19 illustrates one embodiment of a transmitter block diagram in adaptive DIDO systems.

FIG. 20 illustrates one embodiment of a receiver block diagram in adaptive DIDO systems.

FIG. 21 illustrates one embodiment of an adaptive DIDO-OFDM method.

FIG. 22 illustrates one embodiment of an antenna arrangement for DIDO measurements.

FIG. 23 illustrates one embodiment of lattice configurations for DIDO systems of various orders.

FIG. 24 illustrates the operation mode of DIDO systems of various orders.

FIG. 25 illustrates one embodiment of an antenna arrangement for DIDO measurements.

FIG. 26 illustrates one embodiment of 2 × 2 DIDO modes with 4-QAM and 1/2 forward error correction rate as a function of the location of the user device.

FIG. 27 illustrates one embodiment of an antenna arrangement for DIDO measurements.

FIG. 28 illustrates how, in one embodiment, an 8 × 8 DIDO provides more SE than a 2 × 2 DIDO to require lower transmitter power.

FIG. 29 illustrates one embodiment of a DIDO 2 × 2 antenna selection mode.

FIG. 30 illustrates an average bit error rate (BER) mode of various DIDO precoding schemes in i.i.d. channels.

FIG. 31 illustrates the signal-to-noise (SNR) gain for ASel as a function of the number of additional transmit antennas in i.i.d. channels.

FIG. 32 illustrates the signal-to-noise (SNR) thresholds as a function of the number (M) of users for block diagonalization (BD) and ASel with 1 and 2 additional antennas in i.i.d. channels.

FIG. 33 illustrates the dependence of BER on the average for SNR users for two users located in the same angular direction with different angular distance (AS) values.

FIG. 34 illustrates results similar to FIG. 33, but with a greater angular separation between users.

FIG. 35 is a graph of SNR thresholds as a function of AS for various mean angles of arrival (AOA) of users.

FIG. 36 illustrates an SNR threshold for an exemplary case of five users.

FIG. 37 provides a comparison of the SNR threshold for BD and ASel with 1 and 2 additional antennas for the case of two users.

FIG. 38 illustrates results similar to FIG. 37, but for the case of five users.

FIG. 39 illustrates SNR thresholds for a BD scheme with various AS values.

FIG. 40 illustrates SNR thresholds in spatially correlated channels at AS = 0.1 ° for BD and ASel with 1 and 2 additional antennas.

FIG. 41 illustrates the calculation of SNR thresholds for two scenarios of a larger number of channels at AS = 5 °.

FIG. 42 illustrates the calculation of SNR thresholds for two scenarios of a larger number of channels at AS = 10 °.

FIG. 43-44 illustrate signal-to-noise (SNR) thresholds as a function of number (M) of users and angular distance (AS) for BD and ASel circuits with 1 and 2 additional antennas, respectively.

FIG. 45 illustrates a receiver provided with a frequency shift estimator / compensation unit.

FIG. 46 illustrates a model of a 2 × 2 DIDO system according to one embodiment of the invention.

FIG. 47 illustrates a method according to one embodiment of the invention.

FIG. 48 illustrates SER results in DIDO 2 × 2 systems with and without frequency shift.

FIG. 49 compares the performance of different DIDO schemes in terms of SNR thresholds.

FIG. 50 compares the amount of overhead required for various embodiments of the methods.

FIG. 51 illustrates a simulation with a small frequency shift, f _max = 2 Hz, and without any shift correction.

FIG. 52 illustrates the results when the integer shift estimator is turned off.

Detailed Description of Preferred Embodiments

In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced without some of these specific details. In other instances, well-known structures and devices are shown in block diagram form in order to avoid obscuring the principles underlying the invention.

FIG. 1 shows a known MIMO system with transmit antennas 104 and receive antennas 106. Such systems can provide up to three times the bandwidth that is usually achievable on an available channel. There are several approaches in which the details of such a MIMO system are described in the published literature on this subject, and the following explanation describes one such approach.

Before transmitting in the MIMO system of FIG. 1, the channel is "characterized." This is achieved by initially transmitting the training signal from each of the transmitting antennas 104 to each of the receivers 105. The training signal is generated by the coding and modulation subsystem 102, converted into analog form by a digital-to-analog converter (not shown), and then converted from the main frequency band to a high frequency (HF) ) by each transmitter 103 in turn. Each receive antenna 105, coupled to its RF receiver 106, receives each training signal and converts it to the main frequency band. The baseband signal is converted to digital form by an analog-to-digital converter (not shown), and the signal processing subsystem 107 characterizes the training signal. The characteristic data of each signal can include many factors, including, for example, phase and amplitude with respect to the reference signal integrated in the receiver, the absolute reference signal, relative reference signal, characteristic noise, or other factors. The characteristic data of each signal is typically defined as a vector that characterizes changes in phase and amplitude of several aspects of the signal when it is transmitted over a channel. For example, in a signal with quadrature amplitude modulation (QAM) (QAM), the characteristic data can be a vector of phase and amplitude shifts for several multipath signal patterns. As another example, in a signal modulated by orthogonal frequency division multiplexing (“OFDM”), this may be a phase and amplitude shift vector for several or all of the individual subchannels in the OFDM spectrum.

The signal processing subsystem 107 stores the channel characteristic data received by each receiving antenna 105 and the corresponding receiver 106. After all three transmit antennas 104 have completed their training signal transmissions, the signal processing subsystem 107 will store three sets of channel characteristic data for each of the three receiving antennas 105, which gives a matrix 108 of size 3 × 3, designated as a characteristic matrix of channels "H". Each individual matrix element H _{i, j} represents channel characteristic data (which is usually a vector, as described above) when transmitting the training signal of the transmit antenna 104 i received by the receive antenna 105 j.

Here, the signal processing subsystem 107 inverts the matrix H 108 to obtain the matrix H ^-1 , and expects the transmission of real data from the transmitting antennas 104. Note that a variety of known MIMO methods described in the available literature can be used to provide inversion of the matrix H 108.

In the process, the payable data to be transmitted is supplied to the data input subsystem 100. Then they are divided into three parts by a divider 101 before entering the coding and modulation subsystem 102. For example, if the useful information is ASCII bits for “abcdef”, it can be divided by a divisor 101 into three subgroups of ASCII bits for “ad”, “be”, and “cf”. Then, each of these three subgroups is supplied separately to the coding and modulation subsystem 102.

Each of the subgroups of useful information is encoded individually using an encoding system that provides both statistical independence of each signal and the possibility of error correction. The coding includes, in particular, Reed-Solomon coding, Viterbi coding and turbo codes. Finally, each of the three subgroups of useful information is modeled using a modulation scheme suitable for the channel. Examples of modulation schemes are relative phase modulation (“DPSK”), 64-QAM modulation, and OFDM. It should be noted that the diversity gain obtained from MIMO provides modulation constellations of a higher order than would otherwise be possible in a single input and single output (SISO) system using the same channel. Each encoded and modulated signal is then transmitted through its own antenna 104 after the digital-to-analog conversion unit (not shown) and the generation of an RF by each transmitter 103.

Assume that there is sufficient spatial diversity between the transmitting and receiving antennas, with each of the receiving antennas 105 receiving an excellent combination of the three transmitted signals from the antennas 104. Each signal is received and downconverted to the main frequency band by each RF receiver 106 and digitized by analog a digital converter (not shown). If y _n is the signal received by the nth receiving antenna 105, and x _n is the signal transmitted by the nth transmitting antenna 104, and N is noise, this can be described by the following three equations:

y ₁ = x ₁ H ₁₁ + x ₂ H ₁₂ + x ₃ H ₁₃ + N

y ₂ = x ₁ H ₂₁ + x ₂ H ₂₂ + x ₃ H ₂₃ + N

y ₃ = x ₁ H ₃₁ + x ₂ H ₃₂ + x ₃ H ₃₃ + N.

If this is a system of three equations with three unknowns, then the signal-processing subsystem 107 uses linear algebra to obtain x ₁ , x _2, and x ₃ (assuming that N is low enough to allow signal decoding):

.

.

When the three transmitted signals x _{n are} obtained in this way, they are then demodulated, decoded, and corrected for errors in the signal processing system 107 to recover the three bit streams that were originally separated by the divider 101. These bit streams are combined in combiner 108 and output as a single data stream output 109 data. Assuming that the security of the system is such that it is able to overcome noise distortion, the data output 109 will produce the same data stream that was input at the data input 100.

Although the well-known system just described is usually feasible in the configuration of up to four antennas, and maybe even up to 10, but for the reasons described in the "Background" section of this description, it becomes impractical with a large number of antennas (for example, 25, 100 or 1000).

Typically, such a known system is bidirectional, and the return path is implemented in exactly the same way, but in the opposite direction, with each side of the communication channels having both a transmitting and a receiving subsystem.

FIG. 2 illustrates one embodiment of the invention in which the base station 200 is configured with a wide area network (WAN) network 201 (for example, to the Internet via a T1 or other high-speed connection) and is equipped with multiple (N) antennas 202. For now, we use the expression “ base station ”to refer to any wireless station that wirelessly communicates with a set of clients from a fixed location. Examples of base stations are access points in wireless local area networks (WLANs) or an antenna tower or WAN antenna array. There are several client devices 203-207, each with a single antenna, which are served wirelessly from base station 200. Although for the purposes of this example, such a base station is primarily thought of as being located in the office where it serves client devices 203-207, which are personal computers equipped with a wireless network, this architecture can be applied to a large number of applications, both indoors and outdoors, in which the base station is serving Vaeth wireless clients. For example, a base station may be based in a cell phone tower or in a television broadcast tower. In one embodiment, the base station 200 is located on the ground and is configured to transmit upward at high frequencies (for example, at frequencies up to 24 MHz) to deliver signals to the ionosphere, as described in the jointly pending application entitled "SYSTEM AND METHOD, IMPROVEMENT OF COMMUNICATION CLOSE TO A VERTICAL SPATIAL WAVE (NVIS) BY SPATIAL-TEMPORARY CODING ”, No. 10/817,731, filed on April 20, 2004, the rights to which belong to the same applicant as in this application and which is incorporated here by links.

Some of the details associated with the above base station 200 and client devices 203-207 are provided for illustrative purposes only and are not required to implement the fundamental principles of this invention. For example, a base station can connect to many different types of wide area networks through the 201 WAN interface, including application-specific wide area networks such as those used for digital video distribution. Similarly, client devices can be any range of wireless data processing and / or data transfer devices, including but not limited to cell phones, personal digital assistants (“PDAs”), receivers, and wireless cameras.

In one embodiment, the base station antennas 202 are spatially spaced so that each transmits and receives signals that do not have spatial correlation, as if the base station were a MIMO transceiver known in the art. As described in the BACKGROUND section, experiments were conducted in which antennas placed at λ / 6 spacing (i.e., 1/6 wavelength) successfully provided an increase in throughput from MIMO, but, generally speaking, the farther apart these antennas are from the base station, the better the system performance, and λ / 2 is a desirable minimum. Of course, the fundamental principles of the invention are not limited to any particular spacing between antennas.

Note that a single base station 200 may well have its own antennas located very far from each other. For example, in the RF spectrum, antennas can be spaced 10 meters or more (for example, in the NVIS implementation mentioned above). If 100 such antennas are used, the antenna array of the base station may well occupy several square kilometers.

In addition to spatial diversity techniques, one embodiment of the invention imparts polarization to the signal to increase the effective system throughput. Increasing channel capacity through polarization is a well-known method that has been used by satellite television providers for many years. Using polarization, you can have many (for example, three) antennas of a base station or user antennas very close to each other and still not have spatial correlation. Although conventional RF systems are usually used only for diversity in two dimensions (for example, x and y) of polarization, the architecture described here can further benefit from three dimensions of polarization (x, y and z).

In addition to spatial and polarization diversity, one embodiment of the present invention provides antennas with nearly orthogonal radiation patterns to improve link characteristics through beam diversity. Diversity diversity can improve the throughput and noise immunity of MIMO systems and their advantage over other antenna diversity methods shown in the following documents:

[17] L. Dong, N. Ling, and R.W. Heath Jr., "Multiple-input multiple-output wireless communication systems using antenna pattern diversity," (Proc.) Multiple-input multiple-output wireless communication systems using antenna diversity. IEEE Glob. Telecom Conf., Vol. 1, pp. 997-1001, Nov. 2002.

[18] R. Vaughan, "Switched parasitic elements for antenna diversity," IEEE Trans. Antennas Propagat., Vol. 47, pp. 399-405, Feb. 1999.

[19] P. Mattheijssen, MHAJ Herben, G. Dolmans, and L. Leyten, "Antenna-pattern diversity versus space diversity for use at handhelds," (IEEE Trans, Diversity Antenna Patterns) on Veh. Technol, vol. 53, pp. 1035-1042, July 2004.

[20] C. B. Dietrich Jr, K. Dietze, J.R. Nealy, and W.L. Stutzman, "Spatial, polarization, and pattern diversity for wireless handheld terminals," (Provision of space, polarization, and radiation pattern for wireless handheld terminals) Proc. IEEE Antennas and Prop. Symp., Vol. 49, pp. 1271-1281, Sep. 2001.

[21] A. Forenza and R.W. Heath, Jr., "Benefit of Pattern Diversity Via 2-element Array of Circular Patch Antennas in Indoor Clustered MIMO Channels" (IOE Trans, on Communications, IEEE Trans, on Communications , vol. 54, no. 5, pp. 943-954, May 2006.

By using beam diversity, one can have many base station antennas or user antennas very close to each other and still not have spatial correlation.

FIG. 3 provides further details of one embodiment of a base station 200 and client devices 203-207 shown in FIG. 2. For simplicity, the base station 300 is shown with only three antennas 305 and only three client devices 306-308. Note, however, that the embodiments described herein can be implemented with a virtually unlimited number of antennas 305 (i.e., limited only by available space and noise) and client devices 306-308.

FIG. 3 is similar to the known MIMO architecture shown in FIG. 1, in that both have three antennas on each side of the communication channel. A significant difference is that in the known MIMO system, three antennas 105 on the right side of FIG. 1 are all at a fixed distance from one another (for example, integrated into one device), and the received signals from each of the antennas 105 are processed together in the signal processing subsystem 107. In contrast, in FIG. 3, three antennas 309 on the right side of the circuit are each connected to an excellent client device 306-308, each of which can be distributed anywhere in the range of the base station 305. In fact, the signal that each client device receives is processed independently of the two other received signals in its coding, modulating, and signal processing subsystem 311. Thus, the opposite is the case of a MIMO system with multiple inputs (i.e., antenna 105) and multiple outputs (i.e., antenna 104), FIG. 3 illustrates a system with multiple inputs (ie, antenna 305) and distributed outputs (ie, antenna 305), hereinafter referred to as the “MIDO” system.

Note that this application uses excellent terminology than previous applications in order to more closely match those accepted in scientific industrial circles. In the previously cited co-parsed application, “SYSTEM AND METHOD FOR IMPROVING A SPACE-VERTICAL WAVE (NVIS) RELATED TO VERTICAL CIRCUIT”, No. 10/817,731, filed April 20, 2004, and application No. 10/30,90978 years, for which this application is a partial continuation, the meaning of "input" and "output" (in the context of SIMO, MISO, DIMO and MIDO) are inverse with respect to how these expressions are used in this application. In these previous applications, “input” refers to wireless signals when they arrive at receive antennas (for example, antennas 309 in FIG. 3), and “output” refers to wireless signals when they are output from transmit antennas (eg, antennas 305). In science and the wireless industry, the inverse of “input” and “output” is usually used, in which “input” refers to wireless signals when they are inserted into a channel (i.e., transmitted wireless signals from antennas 305), and “output” refers to wireless signals when they are output from the channel (i.e., wireless signals received by antennas 309). This application accepts this terminology, which is the reverse of the applications mentioned earlier in this paragraph. Thus, between applications, the following terminological equivalents should be considered:

The MIDO architecture shown in FIG. 3 achieves the same increase in throughput as MIMO over the SISO system for a given number of transmit antennas. However, one difference between MIMO and the particular MIDO embodiment illustrated in FIG. 3, is that in order to achieve the increase in throughput provided by the multiple antennas of the base station, each client device 306-308 requires only one receive antenna, while in MIMO each client device requires at least as many receive antennas as many times it is desirable to achieve an increase in throughput. Provided that there is usually a practical limit to how many antennas can be placed on a client device (as explained in the Background section), this typically limits MIMO systems to between four and ten antennas (and from four to ten times the bandwidth). Since base station 300 typically serves many client devices from a fixed and energized location, it is practical to expand it to a much larger number of antennas than ten and separate the antennas with suitable distances to achieve spatial diversity. As illustrated, each antenna is provided with a transceiver 304 and a portion of a coding, modulating, and signal processing section 303. In this embodiment, it is significant that regardless of the expansion of the base station 300, each client device 306-308 requires only one antenna 309, so that the cost of a separate user client device 306-308 will be low, and the cost of the base station 300 can be divided among large number of users.

An example of how MIDO transmission from base station 300 to client devices 306-308 can be provided is shown in FIG. 4-6.

In one embodiment of the invention, before the MIDO transmission begins, the channel is characterized. As in the MIMO system, a training signal (in the embodiment described here) is transmitted to each of the antennas 405 one after another. FIG. 4 illustrates only the first transmission of the training signal, but with three antennas 405 there are only three separate transmissions. Each training signal is generated by a signal encoding, modulating and processing subsystem 403, converted to analog form by means of a digital-to-analog converter and transmitted as an RF signal through each RF transceiver 404. Various methods of encoding, modulating and processing signals can be applied, including but not limited to - those described above (for example, Reed-Solomon codes, Viterbi coding; modulation of QAM, DPSK, QPSK, ... etc.).

Each client device 406-408 receives a training signal through its antenna 409 and converts this training signal to the main frequency band by a transceiver 410. An analog-to-digital converter (not shown) converts this signal to digital form, where it is processed by each coding, modulation subsystem 411 and signal processing. Then, the signal characterizing logic 420 characterizes the resulting signal (for example, identifying phase and amplitude distortions as described above) and stores this characteristic in memory. This characterization process is similar to how it is done in the well-known MIMO systems, with the significant difference being that each client device calculates only the characteristic vector for its one antenna, and not for n antennas. For example, the coding, modulation, and signal processing subsystem 420 of the client device 406 is initialized by a known training signal sequence (either during manufacture by receiving it in a transmitted message, or through another initialization process). When the antenna 405 transmits a training signal with this known sequence, the coding, modulation and signal processing subsystem 420 uses correlation methods to find the strongest received sequence of the training signal, it stores a phase and amplitude shift, then subtracts this sequence from the received signal. Following this, she finds the second strongest received sequence that correlates with the training signal, she keeps the phase and amplitude shift, then subtracts this second strongest sequence from the received signal. This process continues until either a fixed number of phase and amplitude shifts (for example, eight) are memorized, or the detectable sequence of the training signal does not fall below a given noise level. This vector of phase and amplitude shifts becomes element H _{11 of} vector 413. Similarly, the coding, modulation and signal processing subsystems for client devices 407 and 408 implement the same processing to obtain their own vector elements H ₂₁ and H ₃₁ .

The memory in which the characteristic is stored may be non-volatile memory, such as flash memory or a hard disk, and / or non-volatile memory, such as random access memory (e.g., SDRAM, DOSE (SDRAM, DRAM)). In addition, different client devices can simultaneously use different types of memory to store characteristic information (for example, PCA can use flash memory, while laptop computers can use hard drives). The fundamental principles of the invention are not limited to any particular type of storage mechanism on various client devices or base stations.

As mentioned above, depending on the scheme used, since each client device 406-408 has only one antenna, each stores only 1 × 3 rows 413-415 from the matrix H. FIG. 4 illustrates the stage after the first transmission of the training signal, where the first column of 1 × 3 rows 413-415 is stored with channel characteristic information of three base station antennas 405. The remaining two columns are stored after the channel characterization of the next two transmissions of the training signal from the other two antennas of the base station. Note that for the sake of illustration, three training signals are transmitted at different times. If three sequences of training signals are chosen so that they are not correlated with each other, they can be transmitted simultaneously, thereby reducing training time.

As indicated in FIG. 5, after all three training signal transmissions are completed, each client device 506-508 transmits back to the base station 500 a stored row 513-515 of size 1 × 3 of the matrix N. For the sake of simplicity, only one client device 506 is shown transmitting its characteristic information in FIG. 5. You can use a suitable modulation scheme (for example, DPSK, 64QAM, OFDM) for the channel, combined with appropriate error correction coding (for example, Reed-Solomon, Viterbi and / or turbo codes) to ensure that the base station 500 will accept this data in lines 513-515 for sure.

Although in FIG. 5, all three antennas 505 are shown receiving a signal, it is sufficient that one antenna and a transceiver of the base station 500 receive each transmission of lines 513-515 of size 1 × 3. However, under certain conditions, using a larger number or all antennas 505 and transceivers 504 for receiving each transmission (i.e., using known processing methods with a single input and multiple outputs (SIMO) in the coding, modulating and signal processing subsystem 503), one can obtain better signal to noise ratio (“SNR”) than using one antenna 505 and one transceiver 504.

When the coding, modulating, and signal processing subsystem 503 receives a 1 × 3 row 513-515 from each client device 507-508, it stores it in a 3 × 3 matrix H 516. As with client devices, the base station can use various storage technologies, including - but not limited to - non-volatile mass storage devices (e.g. hard drives) and (or) non-volatile devices (e.g. SDRAM) for storage matrices 516. FIG. 5 illustrates the stage at which the base station 500 received and stored a 1 × 3 row 513 from the client device 509. 1 × 3 row 514 and 515 can be transmitted and stored in matrix H 516 when they are received from the rest of the client devices, until all matrix H 516 will not be saved.

One embodiment of transmitting MIDO from base station 600 to client devices 606-608 will now be described with reference to FIG. 6. Since the client device 606-608 is an independent device, usually each device receives excellent data transmission. In fact, one embodiment of the base station 600 includes a router 602 located in terms of communicating between the WAN interface 601 and the coding, modulating, and signal processing subsystem 603 that is the source of a plurality of data streams (formatted into bit streams) from the WAN interface 601 , and routes them as separate bit streams u ₁ -u ₃ intended, respectively, for each client device 606-608. For this purpose, various known routing methods via router 602 can be applied.

. Эти три предварительно кодированных битовых потока преобразуются затем к аналоговому виду цифроаналоговыми преобразователями (не показаны) и передаются в качестве ВЧ-сигнала приемопередатчиками 604 и антеннами 605.The three bitstreams u ₁ -u ₃ shown in FIG. 6 are then sent to the coding, modulation and processing subsystem 603, encoded into statistically different error-correcting streams (e.g., using Reed-Solomon, Viterbi or turbo codes) and modulated using a channel modulation scheme (such as DPSK, 64QAM or OFDM). In addition, the embodiment illustrated in FIG. 6 includes a signal precoding logic 605 for uniquely encoding the signals transmitted from each of the antennas 605 based on the signal characteristic matrix 616. More specifically, instead of routing each of the three encoded and modulated bit streams to separate antennas (as this is done in Fig. 1), in one embodiment, the precoding logic 630 multiplies these three bit streams u ₁ -u ₃ in FIG. 6 per matrix inverse to H 616, receiving three new bitstreams

. These three precoded bitstreams are then converted to analog form by digital-to-analog converters (not shown) and transmitted as RF signals by transceivers 604 and antennas 605.

Before explaining how these bitstreams are received by client devices 606-608, the operations performed by the precoding unit 630 will be described. Similar to the above MIMO example in FIG. 1, the encoded and modulated signal for each of the three source bitstreams will be denoted by u _n . _. In the embodiment illustrated in FIG. 6, each u _i contains data from one of three bitstreams routed by router 602, and each such bitstream is for one of three client devices 606-608.

However, unlike the MIMO example of FIG. 1, where each x _i is transmitted by each antenna 104, in the embodiment of the invention illustrated in FIG. 6, each u _{i is} received at the antenna 609 of each client device (plus the noise N present in the channel). To achieve this result, the output of each of the three antennas 605 (each of which will be denoted as ν _i ) is a function of u _i and the matrix H, which characterizes the channel for each client device. In one embodiment, each ν _{i is} computed by precoding logic 630 in a coding, modulation and signal processing subsystem 630 by implementing the following formulas:

.

.

Thus, unlike MIMO, where each x _{i is} computed in the receiver after the signals are converted by the channel, the embodiments described here decide for each ν _i in the transmitter before the signals are converted by the channel. Each antenna 609 receives u _i already separated from other bit streams u _n-1 for other antennas 609. Each transceiver 610 converts each received signal to the main frequency band, where it is digitized by an analog-to-digital converter (not shown), and each coding subsystem 611 modulating and processing the signals demodulates and decodes the bit stream x _i intended for it, and sends its bit stream to the data interface 612 for use by the client device (for example, an application on the client device).

Embodiments of the invention described herein may be implemented using various coding and modulating schemes. For example, in an OFDM embodiment, where a frequency spectrum is divided into a plurality of subbands, the techniques described herein can be used to characterize each individual subband. As, however, mentioned above, the fundamental principles of the invention are not limited to any particular modulation scheme.

If client devices 606-608 are portable data processing devices, such as PDAs, laptops, and / or cordless phones, channel characterization can often change because client devices can move from one location to another. In fact, in one embodiment of the invention, the channel characteristic matrix 616 at the base station is continuously updated. In one embodiment, the base station 600 periodically (for example, every 250 milliseconds) sends a new training signal to each client device, and each client device continuously transmits its channel characteristic vector back to the base station 600 to maintain channel accuracy (for example if the environment changes, so that it affects the channel, or if the client device moves). In one embodiment, the training signal is interleaved in a real data signal sent to each client device. As a rule, training signals require much less bandwidth than data signals, so this will have little effect on the bandwidth of the system as a whole. Accordingly, in this embodiment, the channel characteristic matrix 616 can be updated continuously as the base station actively communicates with each client device, thereby maintaining accurate channel characterization when the client devices move from one location to the next or if the environment changes so that it affects the channel.

In one embodiment of the invention illustrated in FIG. 7, MIMO methods are applied to improve the uplink channel (i.e., the channel from client devices 706-708 to base station 700). In this embodiment, the channel from each of the client devices is continuously analyzed and characterized by an uplink channel characterization logic 741 in the base station. More specifically, each of the client devices 706-708 transmits a training signal to the base station 700, which is analyzed by a channel characterization logic 741 (for example, as in a conventional MIMO system) to generate a characteristic matrix of 741 channels of size N × M, where N is the number client devices, and M is the number of antennas used by the base station. Illustrated in FIG. 7, an embodiment uses three antennas 705 at the base station and three client devices 706-708, which gives a 3 × 3 channel characteristic matrix 741 stored in the base station 700. Illustrated in FIG. 7, uplink MIMO transmission can be used by client devices to both transmit data back to base station 700 and to transmit channel characteristic vectors back to base station 700, as illustrated in FIG. 5. But unlike the embodiment illustrated in FIG. 5, in which the channel characteristic vector of each client device is transmitted at a separate point in time, the method shown in FIG. 7 allows simultaneous transmission of channel characteristic vectors from a plurality of client devices back to base station 700, thereby significantly reducing the influence of channel characteristic vectors on the reverse channel throughput.

As mentioned above, each signal characterization can include many factors, including, for example, phase and amplitude with respect to a reference signal internal to the receiver, absolute reference signal, relative reference signal, characteristic noise, or other factors. For example, in a signal modulated by quadratic amplitude modulation (“QAM”), the characterization may be a phase and amplitude shift vector for several multipath copies of the signal. As another example, in a signal modulated by orthogonal frequency division multiplexing (“OFDM”), this may be a phase and amplitude shift vector for several or all of the individual sub-signals in the OFDM spectrum. The training signal can be generated by each client device's coding and modulation subsystem 711, converted to analog form by a digital-to-analog converter (not shown), and then converted from the main frequency band to the RF transmitter 709 of each client device. In one embodiment, to ensure synchronization of training signals, client devices transmit training signals at the request of the base station (e.g., in a circular system). In addition, training signals can be interleaved in a real data signal or transmitted simultaneously with this signal sent from each client device. Thus, even if the client devices 706-708 are mobile, training signals can be continuously transmitted and analyzed by the uplink channel characterization logic 741, thereby ensuring the relevance of the channel characteristic matrix 741.

The total channel capacity supported by the above embodiments of the invention can be defined as min (N, M), where M is the number of client devices and N is the number of antennas of the base station. That is, throughput is limited by the number of antennas either on the base station side or on the client side. In fact, in one embodiment of the invention, synchronization techniques are used to ensure that no more than min (N, M) antennas are transmitted / received at a given time.

In a typical scenario, the number of antennas 705 at the base station 700 is less than the number of client devices 706-708. An exemplary scenario is illustrated in FIG. 8, which shows five client devices 804-808 communicating with a base station having three antennas 802. In this embodiment, after determining the total number of client devices 804-808 and collecting the necessary information to characterize the channels (for example, as described above) base station 800 selects the first group of three clients 810 with which communication will be carried out (three clients in the example because min (N, M) = 3). After communicating with the first group of clients 810 for a predetermined period of time, the base station then selects another group of three clients 811 with which it will communicate. For uniform distribution of the communication channel, the base station 800 selects two client devices 807, 808, which were not included in the first group. In addition, since an additional antenna is available, base station 800 selects an additional client device 806 included in the first group. In one embodiment, base station 800 cycles through groups of clients in such a way that each client is effectively allocated the same amount of throughput over time. For example, to evenly allocate bandwidth, the base station can then select any combination of three client devices that excludes the client device 806 (i.e., because the client device 806 communicated with the base station 806 during the first two cycles) .

In one embodiment, in addition to transmitting standard data, the base station may use the methods described above to transmit training signals to each of the client devices and receive training signals and characteristic signal data from each of the client devices.

In one embodiment, various levels of bandwidth may be allocated to some client devices or groups of client devices. For example, priority may be assigned to client devices, so that client cycles with a relatively higher priority can be guaranteed more communication cycles (i.e., more bandwidth) than client devices with a relatively lower priority. The “priority” of the client device can be selected based on the number of variables, including, for example, the assigned subscription level of the user for the wireless service (for example, the user may wish to pay more for additional bandwidth) and (or) the type of data that is transmitted to to or from a client device (for example, real-time communications such as audio and video telephony can take precedence over non-real-time communications such as email).

In one embodiment of the invention, the base station dynamically allocates bandwidth based on the current load required by each client device. For example, if client device 804 transmits real-time streaming video, and other devices 805-808 perform non-real-time functions, such as email, then base station 800 may allocate relatively more bandwidth to this client 804. It should be noted however, that the fundamental principles of the invention are not limited to any particular method of bandwidth allocation.

As illustrated in FIG. 9, the two client devices 907, 908 can be so close to each other that the channelization for these clients is essentially the same. As a result, the base station receives and stores almost equivalent channel characteristic vectors for two client devices 907, 908 and therefore will not be able to create unique spatially distributed signals for each client device. Accordingly, in one embodiment, the base station ensures that any two or more client devices that are in close proximity belong to different groups. In FIG. 9, for example, base station 900 first communicates with a first group 910 of client devices 904, 905, and 908; and then with the second group of 911 client devices 905, 906, 907, ensuring that the client devices 907 and 908 fall into different groups.

Alternatively, in one embodiment, the base station 900 communicates with both client devices 907 and 908 simultaneously, but multiplexes the communication channel using known channel multiplexing techniques. For example, a base station may employ time division multiplexing (“TDM”), frequency division multiplexing (“FDM”), or code division multiple access (“CDMA”) techniques to split a single spatially correlated signal between client devices 907 and 908.

Although each client device described above is equipped with a single antenna, the fundamental principles of the invention can be applied when using client devices with multiple antennas to increase throughput. For example, when using the wireless systems described above, a client with 2 antennas realizes a two-fold increase in bandwidth, a client with 3 antennas realizes a three-fold increase in bandwidth, etc. (i.e. assuming that spatial and angular separation between the antennas is sufficient). A base station can apply the same general rules when cycling around client devices with multiple antennas. For example, each antenna can be considered as a separate client, and bandwidth can be allocated to this “client”, as would be done by any other client (for example, providing each client with an appropriate or equivalent communication period).

As mentioned above, in one embodiment of the invention, the methods of transmitting MIDO and / or MIMO signals described above are used to increase the signal-to-noise ratio and throughput in a near vertical spatial wave ("NVIS") system. In FIG. 10, in one embodiment, the first NVIS station 1001, equipped with an array of N antennas 1002, is configured to communicate with M client devices 1004. The NVIS antennas 1002 and antennas of various client devices 1004 transmit signals up to about 15 degrees from the vertical to reach desired NVIS and minimize the interference effects of a surface wave. In one embodiment, antennas 1002 and client devices 1004 support multiple independent data streams 1006 using the various MIDO and MIMO methods described above at a dedicated frequency in the NVIS spectrum (for example, at a carrier frequency of 23 MHz or lower, but typically below 10 MHz), due to which the throughput at the allocated frequency is significantly increased (i.e., with a coefficient proportional to the number of statistically independent data streams).

The NVIS antennas serving this station can be physically very distant from each other. Given long wavelengths below 10 MHz and long distances traveled by signals (such as 300 miles round trip), physically separating the antennas hundreds of yards or even miles can provide the benefits of diversity. In such situations, the signals of individual antennas can be sent back using traditional wired or wireless communication systems to a central location for processing. Alternatively, each antenna may have local means for processing its signals, and then use traditional wired or wireless communication systems to transfer data back to a central location. In one embodiment, the NVIS station 1001 has a broadband line 1015 to the Internet 1010 (or other wide area network), thereby providing the client devices 1003 with remote high-speed wireless network access.

In one embodiment, the base station and / or users can use the polarization or radiation pattern methods described above to reduce grating size and / or distance between users while providing diversity and increased throughput. As an example, in MIMO systems with RF transmissions, users can be in the same location, and yet their signals will not be correlated due to polarization or radiation pattern diversity. In particular, by using beam diversity, one user can communicate with a base station via a surface wave, while another user can communicate via NVIS.

Additional embodiments of the invention

I. DIDO-OFDM precoding for I / Q imbalance

In one embodiment of the invention, a system and method is used to compensate for the imbalance in phase and quadrature (I / Q) components in systems with distributed inputs and distributed outputs (DIDO) when multiplexing with orthogonal frequency division multiplexing (OFDM). Briefly, according to this embodiment, user devices evaluate the channel and return this information to the base station; the base station, on the other hand, calculates a precoding matrix to suppress interference between carriers and between users caused by an imbalance in I / Q; and parallel data streams are transmitted to a plurality of user devices through DIDO precoding; user devices demodulate data through a receiver with vanishing negligible coefficients (ZF), minimum mean square error (MMSE) or maximum likelihood (ML) to suppress residual interference.

As described in detail below, some of the essential features of this embodiment include, but are not limited to:

Pre-coding to suppress inter-carrier interference (ICI) from mirror tones (due to I / Q mismatch) in OFDM systems;

Pre-coding to suppress interference between users and ICI (due to I / Q mismatch) in DIDO-OFDM systems;

Methods for suppressing ICI (due to I / Q mismatch) through a ZF receiver in DIDO-OFDM systems employing a block diagonalization pre-encoder (BD);

Methods for suppressing interference between users and ICI (due to I / Q mismatch) through precoding (at the transmitter) and ZF or MMSE filter (at the receiver) in DIDO-OFDM systems;

Methods for suppressing interference between users and ICI (due to I / Q mismatch) through precoding (at the transmitter) and a non-linear detector such as maximum likelihood detector (ML) (at the receiver) in DIDO-OFDM systems;

Using precoding based on channel status information to suppress inter-carrier interference (ICI) from mirror tones (due to I / Q mismatch) in OFDM systems;

Using precoding based on channel status information to suppress inter-carrier interference (ICI) from mirror tones (due to I / Q mismatch) in DIDO-OFDM systems;

Using an I / Q Mismatch information pre-encoder DIDO at the station and having IQ Mismatch information the DIDO receiver at the user terminal;

Using an I / Q mismatch information DIDO precoder at the station, having IID mismatch information of a DIDO receiver at a user terminal and having IQ mismatch information of a channel estimator;

Using an I / Q mismatch information DIDO precoder at a station having IID mismatch information of a DIDO receiver at a user terminal, having IQ mismatch information of a channel estimator, and having IQ mismatch information of a DIDO feedback generator that sends channel status information from user terminal to station;

Using an I / Q mismatch information DIDO precoder at a station and having an IQ mismatch information DIDO configurator that uses I / Q channel information to perform functions including user selection, adaptive coding and modulation, space-time-frequency mapping or selection of a preliminary encoder;

Using an IID mismatch information DIDO receiver that suppresses ICI (due to I / Q mismatch) by the ZF receiver in DIDO-OFDM systems employing a block diagonalization pre-encoder (BD);

The use of an IID mismatch information DIDO receiver that suppresses ICI (due to I / Q mismatch) by precoding (in the transmitter) and a non-linear detector such as maximum likelihood detector (in the receiver) in DIDO-OFDM systems; and

Use of an IID Mismatch-aware DIDO receiver that suppresses ICI (due to I / Q mismatch) through a ZF or MMSE filter in DIDO-OFDM systems.

but. State of the art

Transmitted and received signals of typical wireless communication systems consist of in-phase and quadrature (I / Q) components. In practical systems, in-phase and quadrature components can be distorted due to defects during mixing operations and in the main frequency band. These distortions manifest themselves as phase mismatch, gain, and I / Q delay. Phase imbalance is caused by the fact that the sine and cosine components in the modulator / demodulator are not completely orthogonal. The gain imbalance is caused by various amplifications between in-phase and quadrature components. There may be additional distortion called delay imbalance due to the difference in delays between the I and Q paths in the analog circuit.

In orthogonal frequency division multiplexing (OFDM) multiplexing systems, an I / Q imbalance causes inter-carrier interference (ICI) from mirror tones. This effect has been studied in the literature, and methods for compensation for I / Q mismatch in systems with a single input and a single output of SISO-OFDM are proposed in the works: M.D. Benedetto and P. Mandarini, "Analysis of the effect of the I / Q baseband filter mismatch in an OFDM modem," (Analysis of the I / Q baseband filter mismatch effect in the OFDM modem) Wireless personal communications, pp. 175-186, 2000; S. Schuchert and R. Hasholzner, "A novel I / Q imbalance compensation scheme for the reception of OFDM signals," (New I / Q Imbalance Compensation Scheme for Receiving OFDM Signals) IEEE Transaction on Consumer Electronics, Aug. 2001; M. Valkama, M. Renfors, and V. Koivunen, "Advanced methods for I / Q imbalance compensation in communication receivers," IEEE Trans. Sig. Proc, Oct. 2001; R. Rao and B. Daneshrad, "Analysis of I / Q mismatch and a cancellation scheme for OFDM systems," (I / Q Mismatch Analysis and Suppression Scheme for OFDM Systems) IST Mobile Communication Summit, June 2004; A. Tarighat, R. Bagheri, and A.H. Sayed, "Compensation schemes and performance analysis of IQ imbalances in OFDM receivers," Signal Processing, IEEE Transactions on [see also Acoustics, Speech, and Signal Processing, IEEE Transactions on ], vol. 53, pp. 3257-3268, Aug. 2005.

Extending this work to multi-input and multi-output MIMO-OFDM systems was presented by R. Rao and B. Daneshrad, "I / Q mismatch cancellation for MIMO OFDM systems," (I / Q Mismatch Suppression for MIMO OFDM Systems) in Personal, Indoor and Mobile Radio Communications, 2004; PIMRC 2004.15th IEEE International Symposium on, vol. 4, 2004, pp. 2710-2714. R.M. Rao, W. Zhu, S. Lang, C. Oberli, D. Browne, J. Bhatia, J.F. Frigon, J. Wang, P; Gupta, H. Lee, D.N. Liu, S.G. Wong, M. Fitz, B. Daneshrad, and O. Takeshita, "Multiantenna testbeds for research and education in wireless communications," IEEE Communications Magazine, vol. 42, no. 12, pp. 72-81, Dec. 2004; S. Lang, M.R. Rao, and B. Daneshrad, "Design and development of a 5.25 GHz software defined wireless OFDM communication platform," (Design and Development of the OFDM Wireless Communication Platform with Specific 5.25 GHz Software) IEEE Communications Magazine, vol. 42, no. 6, pp. 6-12, June 2004, for spatial multiplexing (SM) and in A. Tarighat and A.H. Sayed, "MIMO OFDM receivers for systems with IQ imbalances," (MIMO OFDM Receivers for I / Q Imbalance Systems) IEEE Trans. Sig. Proc, vol. 53, pp. 3583-3596, Sep. 2005, for orthogonal space-time block codes (OSTBC).

Unfortunately, there is currently no literature on how to correct errors due to an imbalance in gain and phase in a communication system with distributed inputs and distributed outputs. Embodiments of the invention described below offer a solution to these problems.

DIDO systems consist of a single base station with distributed antennas that transmit parallel data streams (through precoding) to multiple users to improve downlink throughput using the same wireless resources (i.e., the same code segment length and frequency range) ), as well as traditional SISO systems. A detailed description of DIDO systems was provided in S.G. Perlman and T. Cotter “System and method for wireless communications with distributed inputs and distributed outputs” No. 10/902,978, filed July 30, 2004 (initial application), the rights to which belong to the same applicant as in this application, and which is included here by reference.

There are many ways to implement DIDO precoders. One solution is block diagonalization (BD), described in: Q.N. Spencer, A.L. Swindlehurst, and M. Haardt, "Zero forcing methods for downlink spatial multiplexing in multiuser MIMO channels," IEEE Trans. Sig. Proc, vol. 52, pp. 461-471, Feb. 2004. K.K. Wong, R.D. Murch, and K.B. Letaief, "A joint channel diagonalization for multiuser MIMO antenna systems," IEEE Trans. Wireless Comm., Vol. 2, pp. 773-786, Jul 2003; L.U. Choi and R.D. Murch, "A transmit preprocessing technique for multiuser MIMO systems using a decomposition approach," (IEEE Trans. Wireless Comm., Vol. 3, pp. 20-24, Jan 2004; Z. Shen, J.G. Andrews, R.W. Heath, and B.L. Evans, "Low complexity user selection algorithms for multiuser MIMO systems with block diagonalization," (accepted for publication in IEEE Trans. Sig. Proc, Sep. 2005; Z. Shen, R. Chen, J.G. Andrews, R.W. Heath, and B.L. Evans, "Sum capacity of multiuser MIMO broadcast channels with block diagonalization," submitted to IEEE Trans. Wireless Comm., Oct. 2005; R. Chen, R.W. Heath, and J.G. Andrews, "Transmit selection diversity for unitary precoded multiuser spatial multiplexing systems with linear receivers," accepted to IEEE Trans, on Signal Processing, 2005. Methods for The I / Q compensations presented in this document assume the BD precoder, but can be extended to any type of DIDO precoder.

In DIDO-OFDM systems, I / Q mismatch has two effects: ICI and interference between users. The first is due to interference from mirror tones, as in SISO-OFDM systems. The latter is due to the fact that the I / Q mismatch violates the orthogonality of the pre-encoder DIDO, which causes interference between users. Both of these types of interference can be suppressed at the transmitter and receiver using the methods described here. Three methods are described for I / Q compensation in DIDO-OFDM systems, and their characteristics are compared with systems with and without I / Q mismatch. The results are presented on the basis of both modeling and practical measurements carried out on the prototype DIDO-OFDM.

Embodiments of the present invention are extensions of an initial application. In particular, these embodiments relate to the following features of the initial application:

The system, as described in the initial application, where the I / Q paths are affected by an imbalance in gain and phase;

The training signals used for channel estimation are used to calculate the DIDO precoder with I / Q compensation in the transmitter; and

The characteristic data of the signals is responsible for distortion due to I / Q imbalance and is used in the transmitter to calculate the DIDO precoder according to the method proposed in this document.

b. Embodiments of the invention

First, a mathematical model and structure of the invention will be described.

Before presenting a solution, it is useful to explain the basic mathematical idea. We explain it under the assumption that the imbalance in gain and phase I / Q (phase delay is not included in the description, but is automatically taken into account in the version of the DIDO-OFDM algorithm). To clarify the main idea, suppose that we want to multiply two complex numbers s = s _I + js _Q and h = h _I + jh _Q , and let x = h * s. We use subscripts to indicate in-phase and quadrature components. Recall that

x = s _I h _I -s _Q h _Q

and x _Q = s _I h _Q + s _Q h _I.

In matrix form, this can be rewritten as

.

.

Note the unitary transformation by the channel matrix (H). Now suppose s is the transmitted character, and h is the channel. The presence of an imbalance in gain and phase I / Q can be modeled by creating a non-unitary transformation as follows:

The trick is to recognize what can be rewritten

Now, rewrite (A):

Define:

.and

.

Both of these matrices have a unitary structure and, therefore, can be equivalently represented by complex scalars as

h _e = h ₁₁ + h ₂₂ + j (h ₂₁ -h ₁₂ )

and h _c = h ₁₁ -h ₂₂ + j (h ₂₁ + h ₁₂ ).

Using all these results, it is possible to present the calculation equation back in a scalar form with two channels: a matched channel h _e and a conjugate channel h _c . Then the calculated transformation in equation (5) will become like this:

x = h _e s + h _c s *.

We call the first channel the coordinated channel, and the second channel the conjugate channel. The matched channel is one that would be observed if there were no gain imbalance and I / Q phase.

Using similar arguments, it can be shown that the input-output ratio of the discrete-time MIMO N × M system with an imbalance in gain and phase I / Q is (using scalar equivalents to build their matrix copies):

where t is a discrete time indicator, h _e , h _c ∈ C ^{M × N} , s = [s ₁ , ... s _N ], x = [x ₁ , ..., x _M ], and L is the number of channel taps.

In DIDO-OFDM systems, the received signal is represented in the frequency domain. Recall from signals and systems that if FFTk {s [t]} - S [k], then FFT _K {s * [t]} = S * [(- k)] = S * [Kk] for k = 0 , 1, ... K-1.

With OFDM, the equivalent I / O ratio for the MIMO-OFDM system for subcarrier k is:

where k = 0, 1, ... K-1 is the OFDM subcarrier index, H _e and H _c denote matrices of matched and conjugate channels, respectively, defined as

and

The second contribution to equation (1) is the interference from the mirror tone. It can be handled by compiling the following stack matrix system (carefully note the pairings):

и

являются векторами символов, соответственно, передачи и приема в частотной области.Where

and

are symbol vectors, respectively, of transmission and reception in the frequency domain.

Using this approach, a computational matrix is constructed for use in the work of DIDO. For example, with DIDO 2 × 2, the input-output ratio (assuming that each user has a single receiving antenna) that the user of the first device sees (in the absence of noise):

while the second user is observing

означает m-ю строку матриц Н_е и Н_с, соответственно, a W ∈ С^4×4 является матрицей предварительного кодирования DIDO. Из (2) и (3) видно, что на принятый символ

пользователя m воздействуют два источника, вызванных дисбалансом I/Q: помеха между несущими от зеркального тона (т.е.

) и помеха между пользователями (т.е.

и

при р≠m). Матрица W предварительного кодирования в (3) предназначена для подавления этих двух помеховых членов.Where

means the mth row of the matrices H _e and H _c , respectively, a W ∈ C ^{4 × 4} is the DIDO precoding matrix. From (2) and (3) it can be seen that the received symbol

of user m are affected by two sources caused by an I / Q imbalance: interference between carriers from a mirror tone (i.e.

) and interference between users (i.e.

and

for p ≠ m). The precoding matrix W in (3) is intended to suppress these two interference terms.

(а не из

). Таким образом, текущая система DIDO выбирает предварительный кодер так, что:There are several different embodiments of the DIDO precoder that can be used here depending on the co-discovery used at the receiver. In one embodiment, block diagonalization (BD) is used (see, for example, Q. N. Spencer, AL Swindlehurst, and M. Haardt, "Zeroforcing methods for downlink spatial multiplexing in multiuser MIMO channels," coefficients for downlink spatial multiplexing in MIMO multi-user channels) IEEE Trans. Sig. Proc, vol. 52, pp. 461-471, Feb. 2004. KK Wong, RD Murch, and KB Letaief, "A joint channel diagonalization for multiuser MIMO antenna systems, "(Joint channel diagonalization for MIMO multi-user antenna systems) IEEE Trans. Wireless Comm., vol. 2, pp. 773-786, Jul 2003. LU Choi and RD Murch," A transmit preprocessing techniq ue for multiuser MIMO systems using a decomposition approach, "(Transmission preprocessing method for multi-user MIMO systems using the decomposition approach) IEEE Trans. Wireless Comm., vol. 3, pp. 20-24, Jan 2004. Z. Shen, JG Andrews, RW Heath, and BL Evans, "Low complexity user selection algorithms for multiuser MIMO systems with block diagonalization," (accepted for publication in IEEE Trans. Sig. Proc, Sep. 2005. Z. Shen, R. Chen, JG Andrews, RW Heath, and BL Evans, "Sum capacity of multiuser MIMO broadcast channels with block diagonalization," submitted to IEEE Trans. Wireless Comm., Oct. 2005) computed from a composite channel

(not from

) Thus, the current DIDO system selects the precoder so that:

. Этот способ является преимущественным, потому что с помощью данного предварительного кодера возможно сохранять прочие аспекты предварительного кодера DIDO такими же, как раньше, поскольку эффекты дисбаланса усиления и фазы I/Q полностью подавляются в передатчике.where α _{i, j} are constant, and

. This method is advantageous because with this precoder it is possible to keep other aspects of the DIDO precoder the same as before, since the effects of gain imbalance and I / Q phases are completely suppressed in the transmitter.

It is also possible to design DIDO precoders that pre-suppress interference between users without pre-suppressing ICI due to IQ imbalance. With this approach, the receiver (instead of the transmitter) compensates for the IQ imbalance by applying one of the receive filters described above. Then, the precoding design criterion in (4) can be modified as:

,

для m-го символа передачи, а

,

есть вектор символа приема для пользователя m.Where

,

for the mth character of the transmission, and

,

there is a reception symbol vector for user m.

символа передачи пользователь m применяет фильтр ZF, и оцененного вектор символа задается выражением:On the receiving side for vector estimation

the transmission character, user m applies the ZF filter, and the estimated character vector is given by:

Although the ZF filter is the easiest to understand, the receiver can apply any number of other filters known to those skilled in the art. One popular choice is the MMSE filter, where:

and ρ is the signal-to-noise ratio. Alternatively, the receiver may perform maximum likelihood symbol detection (either a spherical decoder or iterative variation). For example, the first user can use the ML receiver and decide the following optimization:

where S is the set of all possible vectors s and depends on the size of the constellation. The ML receiver provides higher quality due to the requirement of greater complexity in the receiver. A similar system of equations is applied to the second user.

и

в (6) и (7) предполагаются имеющими нулевые элементы. Это предположение справедливо, только если предварительный кодер передачи способен полностью подавлять помехи между пользователями в отношении критерия в (4). Аналогично, матрицы

и

являются диагональными матрицами, только если предварительный кодер передачи способен полностью подавлять помехи между несущими (т.е. от зеркальных тонов).Note that the matrices

and

in (6) and (7) are assumed to have zero elements. This assumption is valid only if the preliminary transmission encoder is able to completely suppress interference between users with respect to the criterion in (4). Similarly, matrices

and

are diagonal matrices only if the transmission precoder is able to completely suppress interference between carriers (i.e., from mirror tones).

и

посредством обучающих символов и отправляет по обратной связи эти оценки в предварительный кодер 1302 в АР. БС вычисляет веса предварительного кодера DIDO (матрицу W) для предварительного подавления помех из-за дисбаланса усиления и фазы I/Q, а также помех между пользователями и передает эти данные пользователям по беспроводному каналу 1304. Пользовательское устройство m применяет приемник 1308 ZF, MMSE или ML путем использования оценок канала, предоставляемых узлом 1304, для подавления остаточных помех и демодулирует данные.FIG. 13 illustrates one embodiment of a structure for I / Q compensation DIDO-OFDM systems including an IQ-OFDM precoder 1302 in a base station (BS), a transmission channel 1304, a channel estimation logic 1306 in a user device, and a receiver 1308 ZF, MMSE or ML. Channel estimation logic 1306 estimates the channels

and

through training symbols and sends feedback these estimates to the preliminary encoder 1302 in the AP. The BS calculates the weights of the pre-encoder DIDO (matrix W) to pre-suppress interference due to gain imbalance and I / Q phase, as well as interference between users, and transmits this data to users via wireless channel 1304. User device m uses a receiver 1308 ZF, MMSE or ML by using the channel estimates provided by the node 1304 to suppress residual interference and demodulates the data.

The three following embodiments can be used to implement the I / Q compensation algorithm.

и

предполагаются диагональными матрицами. Следовательно, уравнение (8) упрощается как:Method 1 - compensation in the transmitter. In this embodiment, the transmitter calculates a precoding matrix according to the criterion in (4). In the receiver, user devices use the “simplified” ZF receiver, where

and

are assumed to be diagonal matrices. Therefore, equation (8) is simplified as:

Method 2 - compensation in the receiver. In this embodiment, the transmitter calculates a precoding matrix based on the traditional BD method described in: R. Chen, R.W. Heath, and J.G. Andrews, "Transmit selection diversity for unitary precoded multiuser spatial multiplexing systems with linear receivers," accepted to IEEE Trans, on Signal Processing, 2005, non-suppressive (single selection multi-user systems with spatial multiplexing and precoding with linear receivers) interference between carriers and between users with respect to the criterion in (4). With this method, the precoding matrix in (2) and (3) is simplified as:

для того, чтобы передатчик рассчитал предварительный кодер DIDO, в противоположность способу 1, который требует передавать по обратной связи оба

и

. Поэтому способ 2 пригоден, в частности для систем DIDO с низкоскоростными каналами обратной связи. С другой стороны, способ 2 требует несколько более высокой вычислительной сложности в пользовательском устройстве для вычисления приемника ZF в (8), а не (11).In the receiver, user devices use the ZF filter, as in (8). Note that this method does not previously suppress interference in the transmitter, as in the above method 1. Therefore, it suppresses interference between carriers in the receiver, but is not able to suppress interference between users. In addition, in method 2, users need to send feedback only vector

in order for the transmitter to calculate the pre-encoder DIDO, as opposed to method 1, which requires both feedback

and

. Therefore, method 2 is particularly suitable for DIDO systems with low speed feedback channels. On the other hand, method 2 requires slightly higher computational complexity in the user device for computing the ZF receiver in (8) rather than (11).

Method 3 - compensation at the transmitter and receiver. In one embodiment, the two methods described above are combined. The transmitter calculates the precoding matrix as in (4), and the receiver estimates the transmit symbols according to (8).

An I / Q imbalance — whether phase imbalance, gain imbalance, or delay imbalance — creates harmful signal degradation in wireless communication systems. For this reason, circuitry in the past has been designed to have a very low imbalance. However, as described above, this problem can be corrected using digital signal processing in the form of transmission pre-coding and (or) a special receiver. One embodiment of the invention comprises a system with several new functional units, each of which is important for implementing I / Q correction in an OFDM communication system or in a DIDO-OFDM communication system.

In one embodiment of the invention, precoding based on channel state information is used to suppress inter-carrier interference (ICI) from mirror tones (due to I / Q mismatch) in an OFDM system. As shown in FIG. 11, the DIDO transmitter according to this embodiment includes a user select node 1102, a plurality of coding modulating nodes 1104, a corresponding plurality of display nodes 1106, having IQ mismatch information DIDO precoding unit 1108, a plurality of RF transmitting nodes 1114, user feedback node 1112 and 1110 DIDO Configurator.

The user selection node 1102 selects data associated with a plurality of users U ₁ -U _M based on the feedback information received by the feedback node 1112, and provides this information to each of the plurality of coding modulating nodes 1104. Each coding modulating node 1104 encodes and modulates information bits of each user and sends them to the display node 1106. The display node 1106 maps the input bits to complex characters and sends the results to the IQ DIDO mismatch information node 1108 pre th coding. Having the IQ DIDO mismatch information, the precoding unit 1108 uses the channel state information obtained by the user feedback unit 1112 to calculate the precoding weights based on the IQ DIDO mismatch information and precodes the input symbols obtained from the mapping nodes 1106. Each of precoded data streams are sent having IQ DIDO mismatch information by precoding unit 1108 to OFDM unit 1115, which calculates IFFT (reverse The second fast Fourier transform) (IFFT) and adds a cyclic prefix. This information is sent to the digital-to-analog conversion unit 1116, which performs the digital-to-analog conversion and sends the result to the RF unit 1114. The RF unit 114 converts the baseband signal with increasing frequency to intermediate / high frequency and sends it to the transmitting antenna.

The precoder works on the primary and mirror tones together to compensate for I / Q imbalance. Any number of design criteria for the precoder can be used, including ZF, MMSE, or weighted MMSE designs. In a preferred embodiment, the precoder completely eliminates ICI due to I / Q mismatch, so that the receiver does not need to perform any additional compensation.

In one embodiment, the precoder uses a block diagonalization criterion to completely suppress interference between users while not fully suppressing ICI effects for each user, which requires additional receiver processing. In another embodiment, the precoder uses the criterion of nulling negligible coefficients to completely suppress interference between users and ICI due to I / Q imbalance. This embodiment may use a conventional DIDO-OFDM processor at the receiver.

In one embodiment of the invention, precoding based on channel state information is used to suppress inter-channel interference (ICI) from mirror tones (due to I / Q mismatch) in a DIDO-OFDM system, and each user uses an I / Q mismatch information receiver DIDO. As shown in FIG. 12, in one embodiment, a system including a receiver 1202 comprises a plurality of RF nodes 1208, a corresponding plurality of analog-to-digital conversion nodes 1210 having I / Q mismatch information, a channel estimator 1204, and a DIDO feedback generating node 1206.

The RF nodes 1208 receive the signals transmitted from the DIDO transmission nodes 1114, convert these signals with decreasing frequency to the main frequency band, and feed the signals converted with decreasing frequency to the analog-to-digital conversion nodes 1210. These analog-to-digital conversion nodes 1210 further convert the signal from analog to digital and send them to OFDM nodes 1213. OFDM nodes 1213 remove the cyclic prefix and perform FFT (Fast Fourier Transform) (FFT) to translate the signal into the frequency domain. During the training period, OFDM nodes 1213 send their output signal to the IQ mismatch information channel estimator 1214, which calculates channel estimates in the frequency domain. Alternatively, channel estimates can be calculated in the time domain. During the data transmission period, the OFDM nodes 1213 send their output signal to the I / Q mismatch information receiving node 1202. The IQ mismatch receiving node 1202 calculates an IQ receiver and demodulates / decodes the signal to obtain data 1214. IQ mismatch information the channel estimator 1204 sends channel estimates to a DIDO feedback generating unit 1206, which can quantize the channel estimates and send them back to the transmitter via the feedback control channel 1112.

The receiver 1202 shown in FIG. 12 may operate in accordance with any number of criteria known to those skilled in the art, including ZF, MMSE, maximum likelihood, or a MAP receiver. In one preferred embodiment, the receiver uses an MMSE filter to suppress ICI caused by IQ imbalance on mirror tones. In another preferred embodiment, the receiver uses a non-linear detector, such as maximum likelihood search, to jointly detect symbols on mirror tones. This method has improved quality due to higher complexity.

In one embodiment, having an IQ mismatch information, a channel estimator 1204 is used to determine receiver coefficients for eliminating ICI. Therefore, we are announcing a DIDO-OFDM system that uses channel state information precoding to suppress inter-carrier interference (ICI) from mirror tones (due to IQ mismatch), having DIDO IQ mismatch information and having mismatch information IQ channel estimation node. The channel estimator may use a conventional training signal or may use specially constructed training signals sent on in-phase and quadrature signals. Any number of scoring algorithms may be implemented, including least squares, MMSE, or maximum likelihood. The channel estimation node having the IQ mismatch information provides an input signal to the IQ mismatch receiver.

Information about the state of the channel can be provided to the station through the reversibility of the channel or through the feedback channel. One embodiment of the invention comprises a DIDO-OFDM system with an IQ mismatch information pre-encoder having an IQ mismatch information feedback channel for transmitting channel status information from user terminals to a station. The feedback channel may be a physical or logical control channel. It can be dedicated or shared, as in a random access channel. Feedback information can be generated using the DIDO feedback generator on the user terminal, which we also declare. The DIDO feedback generator takes as an input signal the output of the signal estimation node having IQ mismatch information. It can quantize channel coefficients or can use any number of known limited feedback algorithms.

The selection of users, the coding and modulation rate, and the mapping into space-time-frequency code segments can vary depending on the results of the DIDO feedback generator. Thus, one embodiment comprises an IQ mismatch information DIDO configurator that uses an IQ mismatch information channel estimate from one or more users to configure a precoder having IQ mismatch information, selects a modulation rate, a coding rate, a subset of users that transmission is allowed, and their mapping into space-time-frequency code segments.

To assess the quality of the proposed compensation method, we compare three DIDO 2 × 2 systems:

1. With I / Q mismatch: transmission in all tones (except for DC and edge tones) without I / Q mismatch compensation.

2. With I / Q compensation: all tone transmission and I / Q mismatch compensation using the “Method 1” described above.

3. Ideal: transmit data only on odd tones to avoid interference between users and between frequencies (ie, from mirror tones) caused by I / Q mismatch.

Hereinafter, the results obtained from measurements with the DIDO-OFDM prototype in real-world propagation scenarios are presented. FIG. 14 shows the 64-QAM constellations obtained from the three systems described above. These constellations were obtained for the same user locations and a fixed signal-to-noise ratio (~ 46 dB). The first constellation 1401 is very noisy due to interference from the mirror tones caused by the I / Q imbalance. The second constellation 1402 shows some improvements due to I / Q compensation. Note that the second constellation 1402 is not as clean as the ideal case shown as constellation 1403, due to possible phase noise that produces inter-carrier interference (ICI).

In FIG. 15 shows the average frequency 1501 of erroneous symbols (SER) and the effective transmission characteristic 1502 per user for DIDO 2 × 2 systems with 64-QAM and 3/4 coding rate, with and without I / Q mismatch. The OFDM bandwidth is 250 kHz, with 64 tones and a cyclic prefix length of L _cp = 4. Since in the ideal case we only transmit data using a subset of tones, the SER and the useful transmission characteristic are evaluated as a function of the average transmit power per tone (rather than the total transmit power) to ensure an honest comparison for different cases. In addition, in the following results we use normalized transmit power values (expressed in decibels), since our goal here is to compare the relative (rather than absolute) quality of the various circuits. In FIG. Figure 15 shows that with an I / Q imbalance, the SER value enters saturation, not reaching the target SER value (~ 10 ^-2 ), which is consistent with the results reported by A. Tarighat and A.N. Sayed, "MIMO OFDM receivers for systems with IQ imbalances," (MIMO OFDM Receivers for I / Q Imbalance Systems) IEEE Trans. Sig. Proc, vol. 53, pp. 3583-3596, Sep. 2005. This saturation effect occurs due to the fact that the power of both the signal and the interference (from specular tones) increases as the transmitter power increases. By the proposed compensation method, however, it is possible to suppress interference and obtain a better SER performance. Note that a small increase in SER at high SNR occurs due to the effects of amplitude saturation in the digital-to-analog converter due to the higher transmit power required for 64-QAM modulations.

In addition, it is seen that the SER characteristic with I / Q compensation is very close to ideal. A gap of 2 dB in transmitter power between the two cases occurs due to possible phase noise that provides additional interference between adjacent OFDM tones. Finally, the useful transmission curves 1502 show that you can transmit twice as much data using the I / Q method as compared to the ideal case, since we use all data tones, not just odd tones (as for an ideal case).

In FIG. 16 shows the SER characteristic of various QAM constellations with and without I / Q compensation. It can be seen that in this embodiment, the proposed method is especially beneficial for 64-QAM constellations. For 4-QAM and 16-QAM, this I / Q compensation method yields worse performance than the I / Q mismatch case, possibly because the proposed method requires more power to provide both data transmission and suppression of noise from mirror tones. In addition, 4-QAM and 16-QAM are not as susceptible to I / Q mismatch as 64-QAM due to the larger minimum distance between the constellation points. See A. Tarighat, R. Bagheri, and A.N. Sayed, "Compensation schemes and performance analysis of IQ imbalances in OFDM receivers," Signal Processing, IEEE Transactions on [see also Acoustics, Speech, and Signal Processing, IEEE Transactions on ], vol. 53, pp. 3257-3268, Aug. 2005. This can be observed in FIG. 16 by comparing I / Q mismatch with the ideal case for 4-QAM and 16-QAM. Therefore, the additional power required by the DIDO pre-encoder with noise suppression (from mirror tones) does not justify the small gain from I / Q compensation in 4-QAM and 16-QAM cases. Note that this moment can be fixed by using methods 2 and 3 for the I / Q compensation described above.

Finally, the relative SER characteristics of the three methods described above are measured under different propagation conditions. For comparison, the SER characteristic in the presence of I / Q mismatch is also described. In FIG. 17 shows a SER measured for a DIDO 2 × 2 system with 64-QAM at a carrier frequency of 450.5 MHz and a bandwidth of 250 kHz at two different user locations. At location 1, users are located at a distance of ~ 6λ from the BS in different rooms and out of line of sight (NLOS) conditions. At location 2, users are located at a distance of ~ λ from the BS in line of sight (LOS).

In FIG. Figure 17 shows that all three compensation methods always outperform the case of no compensation. In addition, it should be noted that method 3 is superior to the other two compensation methods in any channel scenario. The relative characteristic of method 1 and 2 depends on the propagation conditions. In practical measurements, it has been observed that method 1 is usually superior to method 2 because it pre-suppresses (at the transmitter) interference between users caused by an I / Q imbalance. When interference between users is minimal, method 2 may be superior to method 1, as illustrated in graph 1702 in FIG. 17 because it does not suffer from power loss due to the I / Q pre-compensation encoder.

So far, various methods have been compared when considering only a limited set of distribution scenarios, as in FIG. 17. Hereinafter, the relative characteristics of these methods are measured in ideal independent and perfectly distributed (i.i.d.) channels. DIDO-OFDM systems are modeled with an I / Q imbalance in phase and gain on the transmit and receive sides. In FIG. Figure 18 shows the characteristics of the proposed methods only with an imbalance of gain on the transmitting side (i.e., with a gain of 0.8 in the I-path of the first transmission chain and with a gain of 1 in other paths). It can be seen that method 3 is superior to all other methods. In addition, method 1 gives better results than method 2 in i.i.d. channels, in contrast to the results obtained at location 2 on graph 1702 of FIG. 17.

Thus, three methods are presented for compensating for I / Q imbalance in the DIDO-OFDM systems described above. Method 3 is superior to other proposed methods. In systems with low speed feedback channels, method 2 can be used to reduce the amount of feedback required by the DIDO precoder due to the worse SER performance.

II. Adaptive DIDO Transfer Scheme

Now another embodiment of a system and method for improving the performance of systems with distributed inputs and distributed outputs will be described. This method dynamically allocates wireless resources to various user devices by tracking the changing conditions in the channels to increase throughput while observing a certain target error rate. User devices evaluate the quality of the channel and transmit it by feedback to the base station (BS) (BS); the base station processes the channel quality received from user devices to select the best set of user devices, DIDO scheme, modulation / coding scheme (MCS) and lattice configuration for the next transmission; the base station transmits parallel data to a plurality of user devices through precoding, and the signals are demodulated at the receiver.

Also described is a system that efficiently allocates resources for a DIDO wireless line. This system includes a DIDO base station with a DIDO configurator that processes feedback signals received from users to select the best set of users, a DIDO scheme, a modulation / coding scheme (MCS), and a lattice configuration for the next transmission; a receiver in a DIDO system that measures channel parameters and other relevant parameters to generate a DIDO feedback signal; and a DIDO feedback control channel for transmitting feedback information from users to the base station.

As described in detail below, some of the essential features of this embodiment include, but are not limited to:

Adaptive number of user selection methods, DIDO transmission schemes (i.e. antenna selection or multiplexing), modulation / coding (MCS) schemes and grating configurations based on channel quality information to minimize SER or maximize spectral efficiency per user or downlink;

Methods for defining DIDO transmission mode sets as combinations of DIDO and MCS circuits;

Methods for assigning different DIDO modes for different time segments, OFDM tones, and DIDO substreams depending on channel conditions;

Methods for dynamically assigning DIDO carved modes to different users based on the quality of their channel;

Criterion for ensuring adaptive DIDO switching based on line quality indicators calculated in the time, frequency and spatial domains;

Criterion for providing adaptive DIDO switching based on lookup tables.

A DIDO system with a DIDO configurator in a base station, as in FIG. 19, for adaptively selecting the number of users, DIDO transmission schemes (i.e., selecting or multiplexing antennas), modulation / coding (MCS) schemes and lattice configurations based on channel quality information to minimize SER or maximize spectral efficiency per user or downstream lines;

A DIDO system with a DIDO configurator in the base station and a DIDO feedback generator in each user device, as in FIG. 20, which uses the estimated channel state and / or other parameters, like the estimated SNR at the receiver, to generate a feedback message to be input to the DIDO configurator.

A DIDO system with a DIDO configurator in the base station, a DIDO feedback generator, and a DIDO feedback control channel for transmitting DIDO-specific configuration information from users to the base station.

but. State of the art

On systems with multiple inputs and multiple outputs, diversity schemes such as orthogonal space-time block codes (OSTBC) (See V. Tarokh, N. Jafarkhani, and AR Calderbank, "Spacetime block codes from orthogonal designs," block codes from orthogonal constructions) IEEE Trans. Info. Th., vol. 45, pp. 1456-467, Jul. 1999) or antenna selection (see RW Heath Jr., S. Sandhu, and AJ Paulraj, "Antenna selection for spatial multiplexing systems with linear receivers, "(Antenna Selection for Spatial Multiplexing Systems with Linear Receivers) IEEE Trans. Comm., vol. 5, pp. 142-144, Apr. 2001), conceived to combat channel fading Niemi, providing increased line resistance, which leads to a broader surface. Spatial Multiplexing (SM), on the other hand, enables the transmission of multiple parallel data streams as a means to improve system throughput. See G.J. Foschini, G. D. Golden, R.A. Valenzuela, and P.W. Wolniansky, "Simplified processing for high spectral efficiency wireless communication employing multielement arrays," (IEEE Jour. Simplified processing for high-spectral efficiency wireless communications using multi-element arrays). Select Areas in Comm., Vol. 17, no. 11, pp. 1841-1852, Nov. 1999. These benefits can be simultaneously achieved in MIMO systems according to the theoretical diversity-multiplexing tradeoffs of L. Zheng and D.N. C. Tse, "Diversity and multiplexing: a fundamental tradeoff in multiple antenna channels," (IEEE Trans. Info Th., Vol. 49, no. 5, pp. 1073-1096, May 2003. One practical implementation is adaptively switching between diversity and multiplex transmission schemes by tracking changing conditions in channels.

To date, several adaptive MIMO transmission techniques have been proposed. The method of switching diversity / multiplexing in the work of R.W. Heath and A.J. Paulraj, "Switching between diversity and multiplexing in MIMO systems," IEEE Trans. Comm., Vol. 53, no. 6, pp. 962-968, Jun. 2005, was designed to improve bit error rate (BER) for fixed-rate transmission based on instantaneous channel quality information. Alternatively, channel statistics can be used to provide adaptation, as in S. Catreux, V. Erceg, D. Gesbert, and R.W. Heath. Jr., "Adaptive modulation and MIMO coding for broadband wireless data networks," IEEE Comm. Mag., Vol. 2, pp. 108-115, June 2002 ("Catreux"), which leads to a decrease in overhead feedback signals and the number of control messages. The adaptive transmission algorithm in Catreux was designed to improve spectral efficiency for a given target error rate in systems with orthogonal frequency division multiplexing based on time / frequency selectivity indices. Similar adaptive low-feedback approaches have been proposed for narrow-band systems using spatial channel selectivity to switch between diversity schemes and spatial multiplexing. See, for example, A. Forenza, M.R. McKay, A. Pandharipande, R.W. Heath. Jr., and I.V. Collings, "Adaptive MIMO transmission for exploiting the capacity of spatially correlated channels," accepted to the IEEE Trans, on Veh. Tech., Mar. 2007; M.R. McKay, I.B. Collings, A. Forenza, and R.W. Heath. Jr., "Multiplexing / beamforming switching for coded MIMO in spatially correlated Rayleigh channels," (Accepted multiplexing-beamforming for coded MIMOs in spatially correlated Rayleigh channels) accepted to the IEEE Trans, on Veh. Tech., Dec. 2007; A. Forenza, M.R. McKay, R.W. Heath. Jr., and I.B. Collings, "Switching between OSTBC and spatial multiplexing with linear receivers in spatially correlated MIMO channels," (Switching between OSTBC and spatial multiplexing with linear receivers in spatially correlated MIMO channels) Proc. IEEE Veh. Technol. Conf., Vol. 3, pp. 1387-1391, May 2006; M.R. McKay, I.B. Collings, A. Forenza, and R.W. Heath Jr., "A throughput-based adaptive MIMO BICM approach for spatially correlated channels," (Appearing throughput-based BICM MIMO for spatially correlated channels) to appear in Proc. IEEE ICC, June 2006.

In this document, we expand the scope of work presented in various well-known publications on DIDO-OFDM systems; see, for example, R.W. Heath and A.J. Paulraj, "Switching between diversity and multiplexing in MIMO systems," IEEE Trans. Comm., Vol. 53, no. 6, pp. 962-968, Jun. 2005. S. Catreux, V. Erceg, D. Gesbert, and R.W. Heath Jr., "Adaptive modulation and MIMO coding for broadband wireless data networks," IEEE Comm. Mag., Vol. 2, pp. 108-115, June 2002; A. Forenza, M.R. McKay, A. Pandharipande, R.W. Heath Jr., and I.B. Collings, "Adaptive MIMO transmission for exploiting the capacity of spatially correlated channels," IEEE Trans, on Veh. Tech., Vol. 56, n. 2, pp. 619-630, Mar. 2007. M.R. McKay, I.B. Collings, A. Forenza, and R.W. Heath Jr., "Multiplexing / beamforming switching for coded MIMO in spatially correlated Rayleigh channels," (Accepted multiplexing / beamforming for coded MIMO in spatially correlated Rayleigh channels) accepted to the IEEE Trans, on Veh. Tech., Dec. 2007; A. Forenza, M.R. McKay, R.W. Heath Jr., and I.B. Collings, "Switching between OSTBC and spatial multiplexing with linear receivers in spatially correlated MIMO channels," (Switching between OSTBC and spatial multiplexing with linear receivers in spatially correlated MIMO channels) Proc. IEEE Veh. Technol. Conf., Vol. 3, pp. 1387-1391, May 2006. M.R. McKay, I.B. Collings, A. Forenza, and R.W. Heath Jr., "A throughput-based adaptive MIMO BICM approach for spatially correlated channels," (Appearing throughput-based BICM MIMO for spatially correlated channels) to appear in Proc. IEEE ICC, June 2006.

Here, a new DIDO adaptive transmission strategy is described that switches between a different number of users, a number of transmitting antennas and transmission schemes based on channel quality information as a means to improve system performance. Note that the schemes that adaptively select users in multi-user MIMO systems have already been proposed by M. Sharif and B. Hassibi, "On the capacity of MIMO broadcast channel with partial side information," (About the bandwidth of a MIMO broadcast channel with partial side information side information) IEEE Trans. Info Th., Vol. 51, p. 506522, Feb. 2005; and W. Choi, A. Forenza, J.G. Andrews, and R.W. Heath Jr., "Opportunistic space division multiple access with beam selection," to appear in IEEE Trans, on Communications. Spatial Adaptive Spatial Channel Multiple Access (OSDMA) schemes in these publications, however, are designed to maximize total throughput using multi-user diversity, and they only achieve a fraction of the theoretical throughput of dirty paper codes, since interference is not completely pre-suppressed in the transmitter. The DIDO transmission algorithm described here uses block diagonalization to pre-suppress interference between users. However, the proposed adaptive transmission strategy can be applied to any DIDO system, regardless of the type of precoding method.

This patent application describes an extension of the embodiments of the invention described above in the previous application, including but not limited to the following additional features:

1. The training symbols of the previous application for channel estimation can be used by wireless client devices to evaluate the quality indicators of communication lines in the adaptive DIDO scheme;

2. The base station receives signal characteristic data from client devices, as described in the previous application. In the present embodiment, the characteristic data of the signal is defined as an indicator of the quality of the communication line used to provide adaptation;

3. The previous application describes a mechanism for selecting the number of transmitting antennas and users, and also determines the allocation of bandwidth. In addition, different levels of bandwidth can be dynamically assigned to different clients, as in the previous application. The present embodiment defines new criteria related to this selection and bandwidth allocation.

b. Embodiments of the invention

The purpose of the proposed adaptive DIDO method is to improve the spectral efficiency per user or on the downlink by dynamically allocating a wireless resource in the time, frequency and spatial domains for various users in the system. A common adaptation criterion is to increase throughput while meeting the target error rate. Depending on the propagation conditions, this adaptive algorithm can be used to improve the quality of user communication lines (or coverage) through diversity schemes. Shown in FIG. 21, the flowchart describes the steps in this adaptive DIDO scheme.

The base station (BS) (BS) collects channel status information (CSI) from all users in step 2102. In step 2104, from the received CSI, the base station calculates link quality indicators in the time / frequency / spatial domains. At step 2106, these link quality indicators are used to select the users to be serviced in the next transmission, as well as the transmission mode for each of these users. Note that transmission modes consist of various modulation / coding combinations and DIDO schemes. Finally, the BS transmits data to users through DIDO precoding, as in block 2108.

At step 2102, the base station collects channel status information (LSI) from all user devices. At step 2104, the SUIT is used by the base station to determine the instantaneous or statistical quality of channels for all user devices. In DIDO-OFDM systems, channel quality (or an indicator of link quality) can be estimated in the time, frequency, and spatial domains. Then, at step 2106, the base station uses a link quality metric to determine the best subset of users and transmission mode for existing distribution conditions. A set of DIDO transmission modes is defined as a combination of DIDO schemes (i.e., antenna selection or multiplexing), modulation / coding schemes (MCS), and grating configurations. At 2108, data is transmitted to user devices based on the selected number of users and transmission modes.

The choice of mode is provided by look-up tables (LUTs), pre-computed based on the characteristic of the error rate in DIDO systems in various distribution media. These DTs establish a correspondence between channel quality information and error rate characteristics. To construct a DT, the characteristic of the error rate in DIDO systems is evaluated in various propagation scenarios as a function of SNR. From the error rate curves, the minimum SNR required to achieve some given target error rate can be calculated. We define this SNR requirement as an SNR threshold. The SNR thresholds are then evaluated in different propagation scenarios and for different DIDO propagation modes and stored in look-up tables. For example, the SER results in FIG. 24 and 26 can be used to build the PT. Then, from the base station, the base station selects transmission modes for active users that increase throughput while observing a predetermined target error rate. Finally, the base station transmits data to selected users through DIDO precoding. Note that different DIDO modes can be assigned to different time segments, OFDM tones, and DIDO substreams, so adaptation can occur in the time, frequency, and spatial domains.

One embodiment of a system employing DIDO adaptation is illustrated in FIG. 19-20. Several new functional units have been introduced to ensure the implementation of the proposed DIDO adaptation algorithms. In particular, in one embodiment, the DIDO configurator 1910 performs many functions, including selecting a number of users, DIDO transmission schemes (i.e., selecting antennas or multiplexing), modulation / coding (MCS) schemes, and grating configurations based on channel quality information 1912 provided by user devices.

The user selection node 1902 selects data associated with a plurality of users U ₁ -U _M based on the feedback information obtained by the DIDO configurator 1910 and provides this information to each of the plurality of coding and modulation nodes 1904. Each coding and modulation unit 1904 encodes and modulates the information bits of each user and sends them to the mapping unit 1906. This mapping unit 1906 maps the input bits to complex symbols and sends them to the precoding unit 1908. Both the coding and modulation units 1904 and the display unit 1906 use the information received from the DIDO configurator 1910 to select the type of modulation / coding scheme used for each user. This information is computed by the DIDO configurator 1910 by using channel quality information from each of the users provided by the feedback unit 1912. The DIDO precoding unit 1908 uses the information obtained by the DIDO configurator 1910 to calculate the DIDO precoding weights and precodes the input characters received from the mapping nodes 1906. Each of the precoded data streams is sent by the DIDO precoding unit 1908 to the OFDM unit 1915, which calculates OBPF and adds a cyclic prefix. This information is sent to the digital-to-analog conversion unit 1916, which performs the digital-to-analog conversion and sends the resulting analog signal to the high-frequency unit 1914. The high-frequency unit 1914 converts the baseband signal with increasing frequency to intermediate / high frequency and sends it to the transmitting antenna.

The RF nodes 2008 of each client device receive signals transmitted from the DIDO transmit nodes 1914, convert these signals with a lower frequency to the main frequency band, and feed the signals converted with a lower frequency to the analog-to-digital conversion nodes 2010. The A / D conversion nodes 2010 then convert this signal from analog to digital and send it to the OFDM nodes 2013. OFDM nodes 2013 remove the cyclic prefix and perform an FFT to obtain a signal for the frequency domain. During the training period, OFDM nodes 2013 send an output signal to a channel estimate node 2004, which counts the channel estimates in the frequency domain. Alternatively, channel estimates can be calculated in the time domain. During the data period, OFDM nodes 2013 send an output signal to a receiving node 2002, which demodulates / decodes the signal to obtain 2014 data. The channel estimate node 2004 sends channel estimates to a feedback generator 2006, which can quantize the channel estimates and send them back to the transmitter via feedback control channel 1912.

The DIDO configurator 1910 may use the information output at the base station or, in a preferred embodiment, additionally use the output signal of the DIDO feedback generator 2006 (see FIG. 20) operating in each user device. The DIDO feedback generator 2006 uses the estimated channel state 2004 and / or other parameters, such as the estimated SNR at the receiver, to generate a feedback message to be input to the DIDO configurator 1910. The DIDO feedback generator 2006 may compress information at the receiver, may quantize the information, and / or use any feedback limiting strategies known in the art.

DIDO configurator 1910 may use information extracted from the DIDO feedback control channel 1912. This DIDO feedback control channel 1912 is a logical or physical control channel that is used to send the output of the DIDO feedback generator 2006 from the user to the base station. The control channel 1912 may be implemented by any method known in the art and may be a logical or physical control channel. As a physical channel, it may contain a dedicated time / frequency segment assigned to the user. It can also be a random access channel shared by all users. The control channel may be assigned in advance or may be created by capturing bits in a predetermined manner from an existing control channel.

The following discussion describes the results obtained by measuring with the prototype DIDO-OFDM in real distribution environments. These results demonstrate the potential gains achievable in adaptive DIDO systems. Characteristics of DIDO systems of various orders are presented initially, demonstrating that it is possible to increase the number of antennas / users to achieve greater downstream throughput. The DIDO characteristic is then described as a function of the location of the user device, demonstrating the need to track changes in conditions in the channel. Finally, the characterization of DIDO systems employing diversity methods is described.

i. Characterization of DIDO systems of various orders

The performance of various DIDO systems is evaluated by increasing the number of transmit antennas N = M, where M is the number of users. The characteristics of the following systems are compared: SISO, DIDO 2 × 2, DIDO 4 × 4, DIDO 6 × 6 and DIDO 8 × 8. DIDO N × M refers to DIDO with N transmit antennas at the base station and M users.

FIG. 22 illustrates an arrangement of a transmit / receive antenna. Transmitting antennas 2201 are placed in a square array configuration, and users are located around the transmission array. In FIG. 22 T denotes transmit antennas, and U refers to user devices 2202.

In an eight-element transmission array, different subsets of antennas are active depending on the N value selected for different measurements. For each (N) DIDO order, a subset of antennas was selected that covers the largest room area to limit a fixed size superimposed on an 8-element array. This criterion is expected to improve spatial diversity for any given value of N.

FIG. 23 shows the lattice configuration for various DIDO orders that are appropriate for an existing room (i.e. a dashed line). The square dotted frame is 24 '' × 24 '' in size, corresponding to ~ λ × λ at a carrier frequency of 450 MHz.

Now, based on the comments related to FIG. 23 and with reference to FIG. 22, the characteristics of each of the following systems will be determined and compared:

SISO with T1 and U1 (2301)

DIDO 2 × 2 with T1, 2 and U1, 2 (2302)

DIDO 4 × 4 with T1, 2, 3, 4 and U1, 2, 3, 4 (2303)

DIDO 6 × 6 with T1, 2, 3, 4, 5, 6 and U1, 2, 3, 4, 5, 6 (2304)

DIDO 8 × 8 with T1, 2, 3, 4, 5, 6, 7, 8 and U1, 2, 3, 4, 5, 6, 7, 8 (2305)

In FIG. 24 shows SER, BER, SE (spectral efficiency) and payload throughput as a function of transmit power (TX) for the above-described DIDO systems with 4-QAM and FEC (forward error correction) rate of 1/2. It can be seen that the characteristics of SER and BER deteriorate for increasing values of N. This effect is caused by two phenomena: for a fixed TX power, the input power to the DIDO array is divided between an increasing number of users (or data streams); spatial diversity decreases as the number of users in real (spatially correlated) DIDO channels increases.

To compare the relative characteristics of various order DIDO systems, the target BER is fixed at 10 ^-4 (this value may vary depending on the system), which corresponds to approximately SER = 10 ^-2 , as shown in FIG. 24. We call TX power values corresponding to this target value TH TX power thresholds (TRT). For any N, if the TX power is lower than TPT, we consider it impossible to transmit with order N DIDO, and we need to switch to a lower order DIDO. In addition, in FIG. Figure 24 shows that SE and payload throughput saturate when TX power exceeds TPT for any value of N. It follows that it is possible to construct an adaptive transmission strategy that switches between different order DIDOs to improve SE or payload throughput fixed target error rate.

ii. User variable location feature

The purpose of this experiment is to evaluate the DIDO characteristics for different user locations by modeling in spatially correlated channels. DIDO 2 × 2 systems are considered with 4QAM and a FEC rate of 1/2. User 1 is in the transverse direction relative to the transmission array, while user 2 is changing from the transverse direction to the direction along the antennas, as shown in FIG. 25. Transmitting antennas are spaced apart by ~ λ / 2 and allocated from users by ~ 2.5λ.

In FIG. 26 shows the results of SER and SE per user for various locations of user device 2. The arrival angles (AOA) of the user device are in the range between 0 ° and 90 ° when measured from a direction transverse to the transmission grid. It can be seen that as the angular separation of user devices increases, the DIDO characteristic improves due to the increasing diversity available in the DIDO channel. In addition, with the target SER = 10 ^-2, there is a 10 dB gap between the cases of AOA2 = 0 ° and AOA2 = 90 °. This result is consistent with the simulation results obtained in FIG. 35 for an angular divergence of 10 °. In addition, we note that for the case of AOA1 = AOA2 = 0 °, interference effects between two users can occur (due to the proximity of their antennas), which may distinguish their characteristics from the simulation results in FIG. 35.

iii. Preferred scenario for DIDO 8 × 8

In FIG. 24 shows that DIDO 8 × 8 gives more SE than lower order DIDOs due to higher TX power requirements. The purpose of this analysis is to show that there are cases where DIDO 8 × 8 is superior to DIDO 2 × 2 not only in terms of peak spectral efficiency (SE), but also in terms of TX (or TRT) power requirements to achieve this peak SE.

Note that in i.i.d. (ideal) channels have a gap of ~ 6 dB in TX power between SEs for DIDO 8 × 8 and DIDO 2 × 2. This gap is caused by the fact that DIDO 8 × 8 divides TX power between eight data streams, and DIDO 2 × 2 only between two streams. This result is shown through the simulation in FIG. 32.

In spatially correlated channels, however, TPT is a function of the characteristics of the propagation medium (for example, lattice orientation, user location, angular divergence). For example, FIG. 35 shows a span of ~ 15 dB for a lower angular difference between two different locations of user devices. Similar results are presented in FIG. 26 of this application.

Similar to MIMO systems, the performance of DIDO systems deteriorates when users are located in directions along the antennas relative to the TX array (due to lack of diversity). This effect was observed in measurements with the current DIDO prototype. Therefore, one way to show that DIDO 8 × 8 is superior to DIDO 2 × 2 is to place users in directions along the antennas with respect to the DIDO 2 × 2 arrays. In this scenario, DIDO 8 × 8 is superior to DIDO 2 × 2 due to the higher diversity provided by the eight-antenna array.

The following systems are considered in this analysis.

System 1: DIDO 8 × 8 with 4-QAM (transmission of 8 parallel data streams in each time segment);

System 2: DIDO 2 × 2 with 64-QAM (transmission to users X and Y in every 4 time segments). For this system, we consider four combinations of the locations of the transmitting and receiving antennas: a) T1, T2 U1, 2 (direction along the antennas); b) T3, T4 U3, 4 (direction along the antennas); c) T5, T6 U5, 6 (~ 30 ° from the direction along the antennas); d) T7, T8 U7, 8 (NLOS out of line of sight));

System 3: DIDO 8 × 8 with 64-QAM; and

System 4: MISO 8 × 1 with 64-QAM (transmission to user X in every 8 time segments).

For these cases, a FEC rate of 3/4 was used.

User locations are shown in FIG. 27.

In FIG. 28 SER results show a ~ 15 dB gap between systems 2a and 2c due to different lattice orientations and user locations (similar to the simulation results in FIG. 35). The first subgraph in the second row shows the TX power values for which the SE curves are saturated (i.e. correspond to BER 1e-4). We see that system 1 gives a higher SE per user for a lower TX power requirement (~ 5 dB less) than system 2. In addition, the advantages of DIDO 8 × 8 versus DIDO 2 × 2 are more obvious for the downstream (DL ) SE and downlink payload throughput due to gain from DIDO 8 × 8 multiplexing over DIDO 2 × 2. System 4 has a lower TX power requirement (8 dB less) than system 1 due to grating gain from beam formation (i.e., MRC with 8 × 1 MISO). But system 4 gives only 1/3 of the SE per user compared to system 1. System 2 has worse performance than system 1 (that is, it gives a lower SE for a larger TX power requirement). Finally, system 3 gives a much larger SE (due to higher order modulations) than system 1 with higher power requirements (~ 15 dB).

The following conclusions can be drawn from these results:

One channel scenario was identified for which DIDO 8 × 8 is superior to DIDO 2 × 2 (that is, it provides a higher SE for lower TX power requirements);

In this channel scenario, DIDO 8 × 8 gives a higher SE per user and SE on the downlink than DIDO 2 × 2 and MISO 8 × 1; and

The DIDO 8 × 8 performance can be further improved by using higher-order modulations (i.e. 64-QAM rather than 4-QAM) due to higher TX power requirements (~ 15 dB more).

iv. DIDO with antenna selection

Hereinafter, we evaluate the advantages of the antenna selection algorithm described in R. Chen, R.W. Heath, and J.G. Andrews, "Transmit selection diversity for unitary precoded multiuser spatial multiplexing systems with linear receivers," accepted to IEEE Trans, on Signal Processing, 2005. We present the results for one particular system DIDO with two users, 4-QAM and a FEC rate of 1/2. In FIG. 27, the following systems are compared:

DIDO 2 × 2 with T1, 2 and U1, 2; and

DIDO 3 × 2, using antenna selection, with T1, 2, 3 and U1, 2.

The transmit antennas and locations of user devices are the same as in FIG. 27.

In FIG. Figure 29 shows that DIDO 3 × 2 with antenna selection can provide a gain of ~ 5 dB compared to DIDO 2 × 2 systems (in the absence of selection). Note that the channel is almost static (i.e., there is no Doppler shift), so the selection algorithm adapts to the path loss and spatial channel correlation, and not to fast fading. In high Doppler shift scenarios, excellent wins should be expected. In addition, in this particular experiment, it was observed that the antenna selection algorithm selects antennas 2 and 3 for transmission.

v. SNR thresholds for lookup tables

Above, when choosing method 2 (with compensation in the receiver, we noted that the mode selection is provided by look-up tables (PTs). These CTs can be calculated in advance by evaluating the SNR thresholds to achieve some predetermined target error rate characteristic for DIDO transmission modes in different propagation media. Below we provide a characterization of DIDO systems with and without choice of antennas and a variable number of users, which can be used as a guide for constructing DTs. rakticheskih measurements prototype DIDO, the following drawings are obtained by modeling. The following results BER involve no FEC.

FIG. 30 shows a characteristic of the average BER of various DIDO precoding schemes in iid channels. A curve labeled “no choice” refers to the case where block diagonalization (DB) is applied. In the same drawing, the antenna selection characteristic (ASel) is shown for a different number of additional antennas (relative to the number of users). You can see that as the number of additional antennas increases, the ASel characteristic provides a higher diversity gain (characterized by the slope of the BER curve in high SNR mode), resulting in better coverage. For example, if we fix the target BER at the level of 10 ^-2 (practical value for non-encoded systems), then the SNR increase coefficient provided by ASel increases with the number of antennas.

FIG. 31 shows an ASel SNR gain as a function of the number of additional transmit antennas in iid channels for different target BERs. It can be seen that even with the addition of 1 or 2 antennas, ASel provides significant SNR increase coefficients compared to the DB. In the following scenarios, we evaluate the ASel response only for cases of 1 or 2 additional antennas and when the target BER is fixed at 10 ^-2 (for non-encoded systems).

In FIG. Figure 32 shows the SNR thresholds as a function of the number of users (M) for the database and ASel with 1 or 2 additional antennas in i.i.d. channels. We see that the SNR thresholds increase with increasing M due to the greater requirements for the receiving SNR for a larger number of users. Note that we assume a fixed total transmit power (with a variable number of transmit antennas) for any number of users. In addition, in FIG. 32 shows that the magnification factor from the choice of antennas is constant for any number of users in i.i.d. channels.

Next, we show the characteristics of DIDO systems in spatially correlated channels. We model each user channel using the COST259 spatial channel model described by X. Zhuang, FW Vook, KL Baum, T.A. Thomas, and M. Cudak, "Channel models for link and system level simulations," IEEE 802.16 Broadband Wireless Access Working Group, Sep. 2004. We generate a single cluster for each user. As a case study, we assume NLOS channels, a uniform linear array (RLR) (ULA) in a transmitter with an element spacing of 0.5 lambda. For the case of a system with two users, we model clusters with average angles of arrival AOA1 and AOA2, respectively, for the first and second users. AOA are measured in relation to the direction away from the HRD. When there are more than two users in the system, we generate user clusters with uniformly spaced average AOA in the range [-φ _m , φ _m ], where we determine:

where M is the number of users, and Δφ is the angular separation between the average AOA of users. Note that the angular range [-φ _m , φ _m ] is centered at an angle of 0 °, corresponding to the direction transverse to the radar. Next, we consider the BER characteristic of DIDO systems as a function of the angular divergence (AS) of the channels and the angular separation between users with DB and ASel transmission schemes and a different number of users.

In FIG. 33 shows the BER versus the average SNR per user for two users located in the same angular direction (i.e., AOA1 = AOA2 = 0 ° with respect to the direction transverse to the HRD) with different AS values. It can be seen that as the AS increases, the BER characteristic improves and reaches the case of i.i.d. In fact, a higher AS gives a statistically less overlap between the native modes of two users and a better characteristic of the database precoder.

In FIG. 34 shows results similar to FIG. 33, but with a higher angular diversity of users. We consider AOA1 = 0 ° and AOA2 = 90 ° (i.e. angular separation of 90 °). The best performance is now achieved in the case of low AS. In fact, for the case of higher angular diversity, there is less overlap between the user's own modes when the angular divergence is low. It is noteworthy that, as can be seen, the BER characteristic at low AS is better than i.i.d. channels for the reasons just mentioned.

Next, we calculate the SNR thresholds for the target BER of 10 ^-2 in scenarios with different correlations. In FIG. 35, SNR thresholds are plotted as a function of AS for various mean AOA values of users. For low angular diversity of users, reliable transmissions with a reasonable SNR requirement (i.e. 18 dB) are only possible for channels characterized by high AS. On the other hand, when users are spatially separated, a lower SNR is required to comply with the same target BER.

In FIG. 36 shows the SNR threshold for the case of five users. Average AOA values for users as defined in (13) for various values of angular diversity Δφ. We see that for Δφ = 0 ° and AS <15 °, block diagonalization is not performed enough due to the small angular separation between users, and the target BER is not observed. As the AS grows, the SNR requirement for meeting a fixed target BER decreases. On the other hand, for Δφ = 30 °, the least SNR requirement is obtained at low AS, which is consistent with the results in FIG. 35. As AS increases, the SNR thresholds become saturated for one of i.i.d. channels. Note that Δφ = 30 ° with 5 users corresponds to the AOA range [-60 °, 60 °], which is common for base stations in cellular systems with cell division into sectors at angles of 120 °.

Next, we consider a characteristic of the ASel transmission scheme in spatially correlated channels. In FIG. 37 compares the SNR threshold for the database and ASel with 1 and 2 additional antennas for the case of two users. We consider two different cases of angular separation between users: [AOA1 = 0 °, AOA2 = 0 °] and [AOA1 = 0 °, AOA2 = 90 °]. The curves for the OBD circuit (i.e., without antenna selection) are the same as in FIG. 35. We see that ASel gives SNR increase factors of 8 dB and 10 dB, respectively, with 1 and 2 additional antennas for high AS. As AS decreases, the coefficient of increase in ASel over the database becomes smaller due to the reduced number of degrees of freedom in the MIMO broadcast channel. It is noteworthy that for AS = 0 ° (that is, next to the LOS channels) and the case [AOA1 = 0 °, AOA2 = 90 °] ASel does not provide any magnification factor due to the lack of diversity in the spatial domain. In FIG. 38 shows the same results as in FIG. 37, but for the case of five users.

We calculate the SNR thresholds (assuming the usual target BER at the level of 10 ^-2 ) as a function of the number of users in the system (M) for transmission schemes of both the database and ASel. The SNR thresholds correspond to the average SNR, so that the total transmit power is constant for any M. We assume the maximum separation between the average AOA of each user cluster in the azimuth range [-φ _m , φ _m ] = [- 60 °, 60 °]. Then the angular separation between users is Δφ = 120 ° / (M-1).

In FIG. 39 shows SNR thresholds for a database schema with different AS values. We see that the requirement of the lowest SNR is achieved for AS = 0.1 ° (i.e., low angular divergence) with a relatively small number of users (i.e., K≤20) due to the large angular separation between users. However, for M> 50, the SNR requirement is well over 40 dB, since Δφ is very small and the database is impractical. In addition, for AS> 10 °, the SNR thresholds remain almost constant for any M, and the DIDO system in spatially correlated channels reaches the characteristic i.i.d. channels.

To reduce the SNR thresholds and improve the performance of the DIDO system, we use the ASel transmission scheme. In FIG. 40 shows SNR thresholds in spatially correlated channels at AS = 0.1 ° for a DB and ASel with 1 and 2 additional antennas. For comparison, we also give curves for the case i.i.d. shown in FIG. 32. It can be seen that for a small number of users (ie, M≤10), the choice of antennas does not help reduce the SNR requirement due to the lack of diversity in the DIDO broadcast channel. As the number of users grows, ASel gains an advantage due to multi-user diversity giving an increase in SNR (namely, 4 dB for M = 20). In addition, for M≤20, the ASel characteristic with 1 and 2 additional antennas in channels with high spatial correlation is the same.

Next, we calculate the SNR thresholds for two more channel scenarios: AS = 5 ° in FIG. 41 and AS = 10 ° in FIG. 42. In FIG. 41 shows that ASel also provides an SNR increase factor for a relatively small number of users (i.e., M 10 10), in contrast to FIG. 40, due to greater angular divergence. For AS = 10 °, the SNR thresholds are further reduced, and the increase factors due to ASel become higher, as can be seen in FIG. 42.

Finally, we summarize the results presented to date for correlated channels. In FIG. 43 and FIG. Figure 44 shows SNR thresholds as a function of the number of users (M) and angular divergence (AS) for DB and ASel circuits with 1 and 2 antennas, respectively. Note that the case AS = 30 ° corresponds actually to i.i.d. channels, and we used this AS value on the graph for graphical representation only. We see that although the spatial correlation of channels affects the database, ASel gives an almost identical characteristic for any AS. In addition, at AS = 0.1 °, ASel is performed similarly to the database for small M, while surpassing the database for large M (i.e., M≥20) due to multi-user diversity.

In FIG. 49 compares the performance of various DIDO schemes in terms of SNR thresholds. The considered DIDO schemes are: database, ASel, database with its own selection mode (database-ESel) and the method of optimal addition (MRC). Note that the MRC does not pre-suppress interference in the transmitter (unlike other methods), but provides a larger magnification factor when users are spatially separated. In FIG. 49, we plot the SNR threshold at the target BER = 10 ^-2 for DIDO N × 2 systems, when two users are located respectively at –30 ° and 30 ° relative to the transverse direction to the transmission grid. We see that for low AS, the MRC scheme provides a 3 dB gain compared to other schemes, since the user spatial channels are well separated and the interference effect between users is low. Note that the gain of MRC over DIDO N × 2 is obtained due to the grating gain. For ASs over 20 °, the QR-ASel scheme is superior to the others and gives a gain of about 10 dB compared to a 2 × 2 DB without a choice. QR-ASel and DB-ESel provide approximately the same characteristic for any AS value.

The above is a new adaptive transmission method for DIDO systems. This method dynamically switches between DIDO transmission modes to different users to improve throughput at a fixed target error rate. Characteristics of various order DIDO systems were measured under different propagation conditions, and it was found that significant gains in throughput can be achieved by dynamically selecting DIDO modes and the number of users as a function of distribution conditions.

III. Preliminary compensation of frequency and phase shift

but. State of the art

As described above, wireless communication systems use carrier waves to transmit information. These carrier oscillations are usually sinusoids that are modulated in amplitude and / or phase depending on the information to be transmitted. The nominal frequency of this sine wave is called the carrier frequency. To create this oscillation, the transmitter synthesizes one or more sinusoids and uses a frequency upconversion to create a simulated signal "sitting" on a sinusoid with a predetermined carrier frequency. This can be done by direct conversion, where the signal is directly modulated on the carrier, or by multi-stage conversion with increasing frequency. To handle this oscillation, the receiver must demodulate the received RF signal and effectively remove the modulation carrier. This requires the receiver to synthesize one or more sinusoidal signals to invert the modulation process in the transmitter, which is known as down-conversion. Unfortunately, the sinusoidal signals generated in the transmitter and receiver are obtained from different reference generators. No reference oscillator creates an ideal frequency reference; in practice, there is always some deviation from the true frequency.

In wireless communication systems, differences in the output signals of the reference oscillators in the transmitter and receivers create a phenomenon known as a carrier frequency shift or simply a frequency shift in the receiver. It is significant that in the received signal there is some residual modulation (corresponding to the carrier difference during transmission and reception), which appears after conversion with decreasing frequency. This creates distortion in the received signal, resulting in a higher error rate and lower bandwidth.

There are various methods to combat carrier frequency shift. Most approaches evaluate the carrier frequency shift at the receiver, and then apply the carrier frequency shift correction algorithm. The carrier frequency shift estimation algorithm can be performed blindly using a QAM shift (T. Fusco and M. Tanda, "Blind Frequency-offset Estimation for OFDM / OQAM Systems," Signal Processing, IEEE Transactions on [see also Acoustics, Speech, and Signal Processing, IEEE Transactions on], vol. 55, pp. 1828-1838, 2007); periodic properties (E. Serpedin, A. Chevreuil, GB Giannakis, and P. Loubaton, "Blind channel and carrier frequency offset estimation using periodic modulation precoders," Signal Blind channel and carrier frequency offset using periodic modulating precoders) Signal Processing, IEEE Transactions on [see also Acoustics, Speech, and Signal Processing, IEEE Transactions on], vol. 48, no. 8, pp. 2389-2405, Aug. 2000); or cyclic prefix in structural approaches of orthogonal frequency division multiplexing (OFDM)) (JJ van de Veek, M. Sandell, and PO Borjesson, "ML estimation of time and frequency offset in OFDM systems," (ML estimation of time and frequency shift in OFDM systems) Signal Processing, IEEE Transactions on [see also Acoustics, Speech, and Signal Processing, IEEE Transactions on], vol. 45, no. 7, pp. 1800-1805, July 1997; U. Tureli, H. Liu, and MD Zoltowski, "OFDM blind carrier offset estimation: ESPRIT," (OFDM blind carrier offset estimation: ESPRIT) IEEE Trans. Commun., Vol. 48, no. 9, pp. 1459-1461, Sept. 2000; M . Luise, M. Marselli, and R. Reggiannini, "Low-complexity blind carrier frequency recovery for OFDM signals over frequency-selective radio channels," formation of a carrier frequency of low complexity for OFDM signals in frequency selective radio channels (IEEE Trans. Commun., vol. 50, no. 7, pp. 1182-1188, July 2002).

Alternatively, you can use special training signals that include a repeatable data symbol (P.N. Moose, "A technique for orthogonal frequency division multiplexing frequency offset correction," IEEE Trans. Commun ., vol. 42, no. 10, pp. 2908-2914, Oct. 1994); two different symbols (T.M. Schmidl and D.C. Cox, "Robust frequency and timing synchronization for OFDM," (IEEE Trans. Commun., vol. 45, no. 12, pp. 1613-1621, Dec. 1997); or periodically inserted known symbol sequences (M. Luise and R. Reggiannini, "Carrier frequency acquisition and tracking for OFDM systems," IEEE Trans. Commun., vol. 44, no. 11, pp. 1590-1598, Nov. 1996). Correction can occur in analog or digital form. The receiver may also use a carrier frequency offset estimate to pre-correct the transmitted signal to eliminate the offset. Carrier frequency shift correction has been extensively studied for multi-frequency and OFDM systems due to their sensitivity to frequency shift (JJ van de Beek, M. Sandell, and P. O. Borjesson, "ML estimation of time and frequency offset in OFDM systems," time and frequency shift in OFDM systems) Signal Processing, IEEE Transactions on [see also Acoustics, Speech, and Signal Processing, IEEE Transactions on], vol. 45, no. 7, pp. 1800-1805, July 1997; U. Tureli, H. Liu, and MD Zoltowski, "OFDM blind carrier offset estimation: ESPRIT," (OFDM blind carrier offset estimate: ESPRIT) IEEE Trans. Commun., Vol. 48, no. 9, pp. 1459-1461, Sept . 2000; T.M. Schmidl and DC Cox, "Robust frequency and timing synchronization for OFDM," (Reliable frequency and time synchronization radonization for OFDM) IEEE Trans. Commun., vol. 45, no. 12, pp. 1613-1621, Dec. 1997; M. Luise, M. Marselli, and R. Reggiannini, "Low-complexity blind carrier frequency recovery for OFDM signals over frequency-selective radio channels, "(Blind recovery of carrier frequency of low complexity for OFDM signals in frequency selective radio channels) IEEE Trans. Commun., Vol. 50, no. 7, pp. 1182-1188, July 2002).

Evaluation and correction of the frequency shift is an important issue for multi-antenna communication systems or, in general, for MIMO systems (with multiple inputs and multiple outputs). In MIMO systems, where transmit antennas are tied to one reference frequency and receivers are tied to another reference frequency, there is only one shift between the transmitter and receiver. Several algorithms have been proposed to solve this problem using training signals (K. Lee and J. Chun, "Frequency-offset estimation for MIMO and OFDM systems using orthogonal training sequences," (Estimating the frequency shift for MIMO and OFDM systems using orthogonal training sequences ) IEEE Trans. Veh. Technol., Vol. 56, no. 1, pp. 146-156, Jan. 2007; M. Ghogho and A. Swami, "Training design for multipath channel and frequency offset calculation in MIMO systems," (A training scheme for evaluating multipath channels and frequency shift in MIMO systems) Signal Processing, IEEE Transactions on [see also Acoustics, Speech, and Signal Processing, IEEE Transactions on], vol. 54, no. 10, pp. 3957-3965 , Oct. 2006), and adaptive tracking I. (C. Oberli and B. Daneshrad, "Maximum likelihood tracking algorithms for MIMOOFDM," (Maximum Likelihood Tracking Algorithms for MIMOOFDM) in Communications, 2004 IEEE International Conference on, vol. 4, June 20-24, 2004, pp. 2468 -2472). A more complex problem is encountered in MIMO systems, where transmitting antennas are not tied to the same reference frequency, and receiving antennas are tied to each other. This practically happens on the uplink of a space division multiple access (SDMA) system, which can be thought of as a MIMO system in which different users correspond to different transmit antennas. In this case, the compensation of the frequency shift is much more complicated. In particular, the frequency shift interferes between different transmitted MIMO streams. They can be corrected using complex joint estimation and alignment algorithms (A. Kannan, T.P. Krauss, and M.D. Zoltowski, "Separation of cochannel signals under imperfect timing and carrier synchronization", (Separation of signals in joint channels with imperfect time and frequency synchronization) IEEE Trans. Veh. Technol., vol. 50, no. 1, pp. 79-96, Jan. 2001), and alignment followed by frequency shift estimation (T. Tang and RW Heath, "Joint frequency offset estimation and interference cancellation for MIMO-OFDM systems [mobile radio], "(Joint frequency shift estimation and interference suppression for MIMO-OFDM systems [mobile radio]) 2004. VTC2004-Fall. 2004 IEEE 60 ^th Vehicular Technology Confe rence, vol. 3, pp. 1553-1557, Sept. 26-29, 2004; X. Dai, "Carrier frequency offset estimation for OFDM / SDMA systems using consecutive pilots," (Carrier frequency offset estimation for OFDM / SDMA systems with using sequential training signals) IEEE Proceedings-Communications, vol. 152, pp. 624-632, Oct. 7, 2005). Some works solve the related problem of residual phase shift and error tracking, where residual phase shifts are estimated and compensated after estimating the frequency shift, but these works consider only the uplink of the SDMA OFDMA system (L. Haring, S. Bieder, and A. Czylwik, "Residual carrier and sampling frequency synchronization in multiuser OFDM systems," 2006. VTC 2006-Spring. IEEE 63rd Vehicular Technology Conference, vol. 4, pp. 1937-1941, 2006). The most difficult case in MIMO systems occurs when all transmit and receive antennas have different reference frequencies. The only available work on this topic provides only an asymptotic analysis of the estimation error in channels with smooth fading (O. Besson and P. Stoica, "On parameter estimation of MIMO flat-fading channels with frequency offsets," (On the parametric estimation of channels with smooth fading OFDM with frequency shifts) Signal Processing, IEEE Transactions on [see also Acoustics, Speech, and Signal Processing, IEEE Transactions on], vol. 51, no. 3, pp. 602-613, Mar. 2003).

A case that has not been studied properly is when the different transmit antennas of the MIMO system do not have the same reference frequency, and the receive antennas process the signals independently. This takes place in the so-called communication system with distributed inputs and distributed outputs (DIDO), also referred to in the literature as the MIMO broadcast channel. DIDO systems consist of a single access point with distributed antennas that transmit parallel data streams (through precoding) to multiple users to improve downlink throughput using the same wireless resources (i.e., the same segment length and frequency lane), as well as traditional SISO systems. A detailed description of DIDO systems has been presented in US Patent Application No. 2006/0023803 to S.G. Perlman and T. Cotter, "System and method for distributed input-distributed output wireless communications," (July 2004. There are many ways to implement DIDO precoders. One solution is to block diagonalization, described, for example, in the works of Q.N. Spencer, A.L. Swindlehurst, and M. Haardt, "Zero-forcing methods for downlink spatial multiplexing in multiuser MIMO channels," (IEEE Trans I / O methods for downlink spatial multiplexing in multi-user MIMO channels). Sig. Proc, vol. 52, pp. 461-471, Feb. 2004; K.K. Wong, R.D. Murch, and K.B. Letaief, "A joint-channel diagonalization for multiuser MIMO antenna systems," IEEE Trans. Wireless Comm., Vol. 2, pp. 773-786, Jul 2003; L.U. Choi and R.D. Murch, "A transmit preprocessing technique for multiuser MIMO systems using a decomposition approach," (IEEE Trans. Transmission preprocessing method for multi-user MIMO systems using a decomposition approach) Wireless Comm., Vol. 3, pp. 20-24, Jan 2004; Z. Shen, J.G. Andrews, R.W. Heath, and B.L. Evans, "Low complexity user selection algorithms for multiuser MIMO systems with block diagonalization," (Accepted low complexity user selection algorithms for multi-user MIMO systems with block diagonalization) accepted for publication in IEEE Trans. Sig. Proc, Sep. 2005; Z. Shen, R. Chen, J.G. Andrews, R.W. Heath, and B.L. Evans, "Sum capacity of multiuser MIMO broadcast channels with block diagonalization," submitted to IEEE Trans. Wireless Comm., Oct. 2005; R. Chen, R.W. Heath, and J.G. Andrews, "Transmit selection diversity for unitary precoded multiuser spatial multiplexing systems with linear receivers," accepted to IEEE Trans, on Signal Processing, 2005.

In DIDO systems, transmission precoding is used to separate data streams intended for different users. Shifting the carrier frequency causes several problems associated with the implementation of the system when the high-frequency chains of transmitting antennas do not share the same reference frequency. When this happens, each antenna actually transmits at a slightly different carrier frequency. This violates the error-free operation of the pre-encoder DIDO, causing each user the appearance of additional interference. Below are some solutions to this problem. In one embodiment of this solution, DIDO transmit antennas share a reference frequency via a wired, optical, or wireless network. In another embodiment of this solution, one or more users evaluate frequency shift differences (relative differences in shifts between pairs of antennas) and send this information back to the transmitter. The transmitter then pre-adjusts the frequency shifts and proceeds to the training and evaluation phase of the pre-encoder for DIDO. There is a problem with this embodiment when there are delays in the feedback channel. The reason is that there may be residual phase errors created by the correction process, which are not taken into account in the subsequent channel estimation. To solve this problem, in one additional embodiment, a new frequency shift and phase estimation unit is used that can correct the problem by estimating the delay. The results are presented both in modeling and in practical measurements carried out with the prototype DIDO-OFDM.

The method for compensating for frequency and phase shift proposed in this document may be sensitive to estimation errors due to noise at the receiver. Therefore, in one additional embodiment, methods for estimating the time and frequency offset that are stable even in low SNR conditions are provided.

There are different approaches for performing time and frequency offset estimates. Because of their sensitivity to timing errors, many of these approaches have been proposed specifically for OFDM oscillations.

Algorithms, as a rule, do not use the OFDM oscillation structure, thus they are common enough for oscillations with a single carrier as well as with multiple carriers. The algorithm described below belongs to the class of methods that use known reference symbols, for example, training data, to help synchronize. Most of these methods are extensions of the Muse frequency shift estimation node (see P. N. Moose, "A technique for orthogonal frequency division multiplexing frequency offset correction," IEEE Trans. Commun. , vol. 42, no. 10, pp. 2908-2914, Oct. 1994.). Moose suggested using two repetitive training signals and highlighted the frequency shift using the phase difference between the two received signals. The Muse method can only correct a fractional frequency shift. An extension of the Muse method was proposed by Schmidl and Sox (T.M. Schmidl and D.C. Sox, "Robust frequency and timing synchronization for OFDM," IEEE Trans. Commun., Vol. 45, no. 12, pp. 1613-1621, Dec. 1997.). A key innovation was the use of a single periodic OFDM symbol together with an additional relative coding training symbol. Relative coding in the second character provides an integer offset correction. Coulson considered the same setup as described by T.M. Schmidl and D.C. Cox, "Robust frequency and timing synchronization for OFDM," IEEE Trans. Commun., Vol. 45, no. 12, pp. 1613-1621, Dec. 1997, and proposed a detailed review of algorithms and analysis, as described in: A.J. Coulson, "Maximum likelihood synchronization for OFDM using a pilot symbol: analysis," (IEEE J. Select. Areas Commun., Vol. 19, no. 12, pp. 2495-2503, Dec. 2001 .; A.J. Coulson, "Maximum likelihood synchronization for OFDM using a pilot symbol: algorithms," (Maximum likelihood synchronization for OFDM using a pilot symbol: algorithms) IEEE J. Select. Areas Commun., Vol. 19, no. 12, pp. 2486-2494, Dec. 2001. One major difference is that Coulson uses maximum length repeating sequences to provide good correlation properties. He also suggests using linear frequency modulation signals due to their envelope properties in the time and frequency domains. Coulson considers a few practical details, but does not include integer valuations. Many repetitive training signals were reviewed by N. Minn, V.K. Bhargava, and K.V. Letaief, "A robust timing and frequency synchronization for OFDM systems," IEEE Trans. Wireless Commun., Vol. 2, no. 4, pp. 822-839, July 2003, but the learning structure has not been optimized. Shi and Serpedin show that the training structure has some optimal shape for the perspective of frame synchronization (K. Shi and E. Serpedin, "Coarse frame and carrier synchronization of OFDM systems: a new metric and comparison," (Rough frame and carrier synchronization in systems OFDM: A New Metric and Comparison) IEEE Trans. Wireless Commun., Vol. 3, no. 4, pp. 1271-1284, July 2004). One embodiment of the invention uses the Shi and Serpedin approach to perform frame synchronization and estimate fractional frequency shift.

Many approaches described in the literature focus on frame synchronization and fractional frequency shift correction. The correction of the integer shift is solved using an additional training symbol, as in the work of T.M. Schmidl and D.C. Cox, "Robust frequency and timing synchronization for OFDM," IEEE Trans. Commun., Vol. 45, no. 12, pp. 1613-1621, Dec. 1997. For example, Morellli et al. Received an improved version of T.M. Schmidl and D.C. Cox, "Robust frequency and timing synchronization for OFDM," IEEE Trans. Commun., Vol. 45, no. 12, pp. 1613-1621, Dec. 1997, by M. Morelli, A.N. D'Andrea, and U. Mengali, "Frequency ambiguity resolution in OFDM systems," IEEE Commun. Lett., Vol. 4, no. 4, pp. 134-136, Apr. 2000. An alternative approach using an excellent preamble structure was proposed by Morelli and Mengali (M. Morelli and U. Mengali, "An improved frequency offset estimator for OFDM applications," IEEE Commun. Lett. , vol. 3, no. 3, pp. 75-77, Mar. 1999). This approach uses the correlation between M repeating identical training symbols to increase the range of the fractional frequency shift estimator by M times. This is the best node of a linear unbiased estimate, and it perceives a large shift (with proper implementation), but does not provide good time synchronization.

System description

One embodiment of the invention uses precoding based on channel state information to suppress frequency and phase shifts in DIDO systems. See FIG. 11 and the related description above to describe this embodiment.

In one embodiment of the invention, each user employs a receiver equipped with a frequency shift estimation / compensation unit. As shown in FIG. 45, in one embodiment of the invention, the receiver-containing system includes a plurality of RF nodes 4508, a corresponding plurality of analog-to-digital conversion nodes 4510, a receiver provided with a frequency shift estimator / compensation node 4512, and a DIDO feedback generator 4506.

The RF nodes 4508 receive the signals transmitted from the transmitting DIDO nodes, convert these signals with decreasing frequency to the main frequency band and supply the signals converted with decreasing frequency to the nodes 4510 analog-to-digital conversion. The analog-to-digital conversion units 4510 convert the signal from analog to digital and send it to the frequency shift estimation / compensation units 4512. These frequency shift estimation / compensation nodes 4512 evaluate and compensate for the frequency shift, as described herein, and then send the compensated signal to OFDM nodes 4513. OFDM nodes 4513 remove the cyclic prefix and perform fast Fourier transform (FFT) (FFT) to translate the signal into the frequency domain. During the training period, OFDM nodes 4513 send an output signal to a channel estimate node 4504, which calculates channel estimates in the frequency domain. Alternatively, channel estimates can be calculated in the time domain. During the data transmission period, the OFDM nodes 4513 send an output signal to the OFDM receiving node 4502, which demodulates / decodes this signal to obtain data. The channel estimator 4504 sprinkles the channel estimates into a DIDO feedback generator 4506, which can quantize the channel estimates and send them back to the transmitter via the feedback control channel, as shown in the drawing.

Description of one embodiment of the algorithm for the DIDO 2 × 2 scenario

Embodiments of an algorithm for compensating for frequency / phase shift in DIDO systems are described below. The DIDO system model is first described with and without frequency / phase shifts. For the sake of simplicity, a specific implementation of the DIDO 2 × 2 system is presented. However, the fundamental principles of the invention can also be implemented on higher order DIDO systems.

DIDO system model without frequency and phase shift

Received DIDO 2 × 2 signals can be recorded for the first user as:

and for the second user as:

where t is the discrete time index, h _mn and w _mn are the channel and DIDO precoding weights between the mth user and the nth transmit antenna, respectively, and x _m is the signal transmitted to user m. Note that h _mn and w _mn are not functions of t, since we assume that the channel is constant over the period of time between the transmission of training and actual data.

In the presence of a frequency and phase shift, the received signals are expressed as:

and

where T _s is the symbol period, ω _Tn = 2Пƒ _Tn for the nth transmit antenna, ω _Um = 2Пƒ _Um for the mth user, and ƒ _Tn and ƒ _Un are real carrier frequencies (subject to shift), respectively, for n- nth transmit antenna and mth user. The values of t _mn mean random delays that cause a phase shift on the channel h _mn . In FIG. 46 shows a model of a DIDO 2 × 2 system.

For now, we use the following definitions:

to indicate the frequency shift between the mth user and the nth transmit antenna.

Description of one embodiment of the invention

The method according to one embodiment of the invention is illustrated in FIG. 47. This method includes the following general steps (which include sub-steps, as illustrated): a training period 4701 for estimating the frequency shift; training period 4702 for channel assessment; transmitting 4703 data through compensated DIDO precoding. These steps are described in detail below.

(a) Training period for estimating the frequency shift (4701)

During the first training period, the base station sends one or more training sequences from each transmitting antenna to one of the users (4701a). As described here, “users” are wireless client devices. For the case of DIDO 2 × 2, the signal received by the mth user is given by the equation

where p ₁ and p ₂ are training sequences transmitted respectively from the first and second antennas,

The mth user can apply any type of frequency shift estimation node (i.e., convolution by training sequences) and estimates the shifts Δω _m1 and Δω _m2 . Then, from these values, the user calculates the frequency shift between the two transmitting antennas as:

Finally, the value (7) is fed back to the base station (4701b).

Note that p ₁ and p ₂ in (6) are made orthogonal, so that users can evaluate Δω _m1 and Δω _m2 . Alternatively, in one embodiment, the same training sequence is used in two consecutive time segments, and the user estimates the shift from them. In addition, to improve the estimate of the shift in (7), the same calculations described above can be performed for all users in DIDO systems (not only for the nth user), and the final estimate can be a (weighted) average of the values obtained from all users. This option, however, requires more computation time and feedback volume. Finally, updates to the frequency shift estimate are only required if the frequency shift changes over time. Therefore, depending on the stability of the clock in the transmitter, this algorithm step 4701 may be carried out on a long-term basis (i.e., not for each data transmission), resulting in reduced feedback costs.

(b) Training period for channel assessment (4702)

During the second training period, the base station first receives feedback information regarding the frequency shift with a value in (7) from the mth user or from a plurality of users. The value in (7) is used for preliminary compensation of the frequency shift on the transmitting side. The base station then sends training data to all users for channel estimation (4702a).

For DIDO 2 × 2 systems, the signal received by the first user is given by the equation:

and

, и Δt есть случайная или известная задержка между первой и второй передачами базовой станции. Кроме того, p₁ и p₂ являются обучающими последовательностями, переданными от первой и второй антенн, соответственно, для оценки частотного сдвига и канала.Where

, and Δt is a random or known delay between the first and second transmissions of the base station. In addition, p ₁ and p ₂ are training sequences transmitted from the first and second antennas, respectively, for estimating the frequency shift and the channel.

Note that pre-compensation only applies to the second antennas in this embodiment. Making a substitution in (8), we obtain:

and

.Where

.

At the receiving side, users compensate for the residual frequency shift using training sequences p ₁ and p ₂ . Users then evaluate the vector channels through training (4702b):

These channels in (12) or channel status information (LST) are fed back to the base station (4702b), which calculates the preliminary DIDO encoder, as described in the next subsection.

(from). DIDO precoding with precompensation (4703)

The base station receives information about the channel status (ISK) in (12) from users and calculates the precoding weights by means of block diagonalization (DB) (4703a), so that:

where the vectors h _{1 are} defined in (12), w _m = [w _m1 , w _m2 ]. Note that the invention presented in this description can be applied to any other method of precoding DIDO in addition to the database. The base station also pre-compensates for the frequency shift by applying the estimate in (7) and the phase shift by estimating the delay (Δt _o ) between the second training transmission and the current transmission (4703a). Finally, the base station sends data to users through the DIDO precoder (4703b).

After processing the transmission, the signal received by user 1 is set as:

. Используя свойство (13), получим:Where

. Using property (13), we obtain:

Similarly for user (2) we have:

and, making a replacement in (16),

.Where

.

Finally, users calculate the residual frequency offset and channel estimate to demodulate the data streams x ₁ [t] and x ₂ [t] (4703 s).

Generalization to DIDO N × M

In this section, the previously described methods are generalized to DIDO systems with N transmit antennas and M users.

i. Training period for estimating frequency shift

During the first training period, the signal received by the mth user as a result of training sequences sent from N antennas is defined as:

where p _n are training sequences transmitted from the nth antenna.

After estimating the shifts Δω _mn , ∀n = 1, ..., N, the m-th user calculates the frequency shift between the first and nth transmit antennas as:

Finally, the values in (19) are fed back to the base station.

ii. Channel Assessment Training Period

During the second training period, the base station first receives feedback information regarding the frequency shift with the value in (19) from the mth user or from a plurality of users. The value in (19) is used for preliminary compensation of the frequency shift on the transmitting side. The base station then sends training data to all users for channel estimation.

For DIDO N × M systems, the signal received by the mth user is specified as:

,

, а Δt есть случайная или известная задержка между первой и второй передачами базовой станции. Кроме того, p_n есть обучающая последовательность, передаваемая из n-й антенны для оценки частотного сдвига и канала.Where

,

, and Δt is a random or known delay between the first and second transmissions of the base station. In addition, p _n is a training sequence transmitted from the nth antenna to estimate the frequency shift and channel.

At the receiving side, users compensate for the residual frequency shift using training sequences p _n . Then, each of m users evaluates the vector channel through training:

and feeds it back to the base station, which calculates the preliminary DIDO encoder, as described in the following section.

iii. DIDO precoding with precompensation

The base station receives the channel status information (ISK) in (12) from users and calculates the precoding weights by means of block diagonalization (DB), so that:

where the vectors h _{m are} defined in (21), w _m = [w _m1 , w _m2 , ..., w _mN ]. The base station also pre-compensates for the frequency shift by applying the estimate in (19) and the phase shift by estimating the delay (Δt _o ) between the second training transmission and the current transmission (4703a). Finally, the base station sends data to users through the DIDO precoder.

After processing the transmission, the signal received by user i is set as:

. Используя свойство (22), получим:Where

. Using property (22), we obtain:

Finally, users calculate the residual frequency offset and channel estimate for demodulating the flows x _i [t].

results

In FIG. 48 shows SER results for DIDO 2 × 2 systems with and without frequency shift. It is seen that the proposed method completely suppresses frequency / phase shifts, giving the same SER as systems without shifts.

Further, we evaluate the sensitivity of the proposed compensation method to errors in the estimation of the frequency shift and / or time fluctuations of the shift. Therefore, we rewrite (14) as:

where ∈ indicates the estimation error and (or) a change in the frequency shift between the training and data transfers. We note that the effect of e consists in violating the orthogonality property in (13), so that the interference terms in (14) and (16) are not previously completely suppressed in the transmitter. As a result, the SER characteristic degrades for increasing values of ∈.

In FIG. 48 shows the SER characteristic in a method of compensating for a frequency shift for different values of ∈. These results assume T _s = 0.3 ms (i.e., a signal with a frequency band of 3 kHz). We observe that for ∈ = 0.001 Hz (or less) the characteristic SER is similar to the case of no shift.

f. Description of one embodiment of an algorithm for estimating the time and frequency shift

Next, we describe additional embodiments for estimating the time and frequency offset (4701b in FIG. 47). The signal transmission structure under consideration is illustrated by N. Minn, VK Bhargava, and K.V. Letaief, "A robust timing and frequency synchronization for OFDM systems," IEEE Trans. Wireless Commun., Vol. 2, no. 4, pp. 822-839, July 2003, and more fully explored by K. Shi and E. Serpedin, "Coarse frame and carrier synchronization of OFDM systems: a new metric and comparison," (Rough Frame and Carrier Synchronization in OFDM Systems: New metric and comparison) IEEE Trans. Wireless Commun., Vol. 3, no. 4, pp. 1271-1284, July 2004. General sequences with good correlation properties are used for training. For example, our system uses Chu sequences, which are obtained as described in D. Chu, "Polyphase codes with good periodic correlation properties (correp.)," (IEEE Trans. Multiphase codes with good periodic correlation properties (reported)). Inform. Theory, vol. 18, no. 4, pp. 531-532, July 1972. These sequences have an interesting property in that they have perfect circular correlations. Let L _cp denote the length of the cyclic prefix and let N _t denote the length of the component training sequences. Let N _t = M _t , where M _t is the length of the training sequence. Under these assumptions, the transmitted character sequence for the preamble can be written as:

s [n] = t [nN _t ] for n = -1, ..., -L _cf s [n] = t [n] for n = 0, ..., N _t -1 s [n] = t [nN _t ] for n = N _t , ..., 2N _t -1 s [n] = - t [n-2N _t ] for n = 2N _t , ..., 3N _t -1 s [n] = t [n-3N _t ] for n = 3N _t , ..., 4N _t -1.

Note that the structure of this training signal can be extended to other lengths, but with a repeating block structure. For example, to use 16 training signals, we consider the following structure:

[CP, B, B, -B, B, B, B, -B, B, -B, -B, B, -B, B, B, -B, B].

Using this structure and assuming N _t = 4M _t , all the algorithms to be described can be applied without modification. In fact, we are repeating the training sequence. This is especially useful in cases where a suitable training signal may not be available.

Consider the following received signal after matched filtering and downsampling to the symbol frequency:

where ε is an unknown discrete-time frequency shift, Δ is an unknown frame shift, h [l] are unknown discrete-time channel coefficients, and v [n] is additive noise. In order to clarify key ideas in the following section, the presence of additive noise is ignored.

i. Rough frame sync

The goal of coarse frame synchronization is to find an unknown frame shift Δ. We introduce the following definitions:

,

,

,

,

,

,

,

,

,

,

,

,

,

,

.

.

The proposed coarse frame synchronization algorithm is derived from the algorithm by K. Shi and E. Serpedin, "Coarse frame and carrier synchronization of OFDM systems: a new metric and comparison," (Coarse frame and carrier synchronization in OFDM systems: a new metric and comparison ) IEEE Trans. Wireless Commun., Vol. 3, no. 4, pp. 1271-1284, July 2004, derived from the maximum likelihood criterion.

Method 1 - advanced coarse frame synchronization: the coarse frame synchronization evaluation unit solves the following optimization:

Where

Let the correction signal be defined as:

.

.

An additional correction term is used to compensate for small initial taps in the channel and can be adjusted based on the application. This additional delay may be further included in the channel.

ii. Partial frequency shift correction

A partial correction of the frequency shift follows the coarse frame synchronization block.

Method 2 - advanced correction of fractional frequency shift: fractional frequency shift is a solution for:

.

.

This is known as fractional frequency shift, because the algorithm allows correction only for shifts:

.

.

This problem will be resolved in the next section. Let the signal of the exact correction of the frequency shift be defined as:

.

.

, как описано выше. Использование

улучшает известный узел оценки, потому что он игнорирует отсчеты, которые могут быть засорены из-за межсимвольной помехи.Note that methods 1 and 2 represent an improvement for the method of operation by K. Shi and E. Serpedin, "Coarse frame and carrier synchronization of OFDM systems: a new metric and comparison," (Rough frame and carrier synchronization in OFDM systems: new metric and comparison) IEEE Trans. Wireless Commun., Vol. 3, no. 4, pp. 1271-1284, July 2004, which works best in frequency selective channels. One specific innovation here is the use of both r and

as described above. Using

improves the well-known evaluation node because it ignores samples that may be clogged due to intersymbol interference.

iii. Integer frequency shift correction

To correct an integer frequency shift, it is necessary to write an equivalent system model for the received signal after accurate correction of the frequency shift. Absorbing the remaining temporary errors in the channel, the received signal in the absence of noise has the following structure:

for n = 0. 1, ... 4N _t -1. The integer frequency shift is k, while the unknown equivalent channel is g [l].

Method 3 - advanced correction of integer partial shift: integer frequency shift is a solution for:

where r = D [k] Sg

This gives an estimate of the total frequency shift as:

.

.

In practice, method 3 has a rather high complexity. To reduce complexity, the following remarks can be made. First of all, the product SS (S * S) ^-1 S * can be calculated in advance. Unfortunately, after this, it is still necessary to perform a fairly large matrix multiplication. An alternative is to use the observation that with the proposed training sequences S * S≈I. This leads to the following method with reduced complexity.

Method 4 - improved correction of an integer frequency shift of low complexity: the evaluation unit of an integer frequency shift of low complexity decides:

.

.

iv. results

In this section, we compare the characteristics of the various proposed evaluation nodes.

First in FIG. 50 we compare the amount of costs required for each method. Note that both new methods reduce the required costs by 10 × to 20 ×. To compare the characteristics of the various evaluation nodes, experiments were performed using the Monte Carlo method. The layout under consideration is our usual transmitted NVIS oscillation, constructed from linear modulation with a symbol rate of 3K symbols per second, corresponding to a bandwidth of 3 kHz, and with the shape of an increasing cosine pulse. For each Monte Carlo implementation, a frequency shift is generated from a uniform distribution in the interval [-f _max , f _max ].

Modeling with a small frequency shift of f _max = 2 Hz and in the absence of integer-shift correction is illustrated in FIG. 51. From this comparison of characteristics, it can be seen that the characteristic at N _t / M _t = 1 is slightly worse compared to the initial evaluation node, although it still significantly reduces costs. The characteristic at N _t / M _t = 4 is much better, almost 10 dB. All curves have kinks at low SNR points due to errors in the integer shift estimate. A small error in the integer shift can create a large frequency error, and thus a large mean square error. Integer shift correction can be turned off at small offsets to improve performance.

In the presence of multipath channels, the characteristic of frequency shift estimation nodes is usually degraded. However, turning off the integer-shift estimation unit shows a very good characteristic in FIG. 52. Thus, in multipath channels, it is even more important to perform robust coarse correction, followed by an advanced algorithm for accurate correction. Note that the shear characteristic at N _t / M _t = 4 is much better in the multipath case.

Embodiments of the invention may include the various steps set forth above. These steps can be implemented in computer-executable instructions that cause the universal or specialized processor to perform certain steps. For example, the various components described above in base stations / APs and client devices can be implemented as software in a universal or specialized processor. In order to avoid misunderstanding of relevant aspects of the invention, various well-known components of personal computers, such as computer memory, hard disk drive, input devices, and the like, are omitted from the drawings.

Alternatively, in one embodiment, the various function modules illustrated here and the associated steps may be performed by specific hardware components that contain wired logic to complete the steps, such as a specialized integrated circuit (ASIC), or any combination of programmable computer components and custom hardware components.

In one embodiment, some modules, such as the coding, modulation, and signal processing logic 903 described above, may be implemented as a programmable digital signal processor (DSP) (or DSP group), such as a DSP using the Texas Instruments TMS320x architecture (e.g. , TMS320C6000, TMS320C5000, ... etc.). The DSP in this embodiment may be integrated in an additional board to a personal computer, such as, for example, a PCI board. Of course, many different DSP architectures can be used without violating the fundamental principles of the invention.

Elements of the present invention may also be presented as computer-readable media for storing computer-executable instructions. This computer-readable medium may include, but is not limited to, flash memory, optical disks, CD-ROMs, DVD ROMs, RAMs, EPROMs, EEPROMs, magnetic or optical cards, distribution media, or other types of computer-readable media suitable for storing electronic commands . For example, the present invention can be downloaded as a computer program that can be transferred from a remote computer (for example, a server) to the requesting computer (for example, a client) using data signals embedded in carrier wave or other propagation media via a communication line (for example, modem or network connection).

Throughout the foregoing description, for purposes of explanation, numerous specific details have been set forth in order to provide a thorough understanding of the present system and method. The specialist, however, it is obvious that these system and method can be implemented without some of these specific details. Accordingly, the scope and spirit of the present invention should be judged on the basis of the following claims.

In addition, numerous publications have been cited throughout the foregoing description to provide a thorough understanding of the present invention. All of these cited sources are incorporated herein by reference.

Claims (44)

1. A method for dynamically adapting the communication characteristics of a multi-antenna system (MAC) with multi-user (MP) transmissions (MP-MAC), comprising the steps of: transmitting a training signal between each antenna of the base station and each of a plurality of wireless client devices, analyzing each training signal to generate channel characteristic data; determine the instantaneous or statistical quality of the channel (an indicator of the quality of the communication line) for wireless client devices using the characteristic data of the channel; determining a subset of users and an MP-MAC transmission mode based on a link quality indicator; calculating a plurality of weights of the pre-encoder MP-MAC based on the characteristic data of the channels; pre-encode the data using the weights of the pre-encoder MP-MAC for generating signals of pre-encoded data for each antenna of the base station; and transmit precoded data signals through each antenna of the base station to each respective client device in the selected subset. 2. The method of claim 1, wherein the MP-MAC transmission modes include various combinations of antenna selection / diversity or multiplexing, modulation / coding (MCS) schemes, and lattice configurations / geometries. 3. The method according to p. 1, in which the quality index of the communication line is evaluated in the time, frequency and / or spatial areas. 4. The method according to claim 1, in which the indicator of the quality of the communication line includes a signal-to-noise ratio (SNR) for signals received at client devices. 5. The method of claim 1, wherein the MP-MAC system is a distributed antenna communication system, wherein the MP-MAC transmission mode is a transmission mode based on a link quality indicator, and the weights of the MP-MAC precoder are the weights of the pre encoder distributed antennas. 6. A system for dynamically adapting communication characteristics of an MP-MAC communication system, comprising: one or more encoding and modulation nodes for encoding and modulating information bits for each of a plurality of wireless client devices to obtain encoded and modulated information bits; one or more display nodes for mapping encoded and modulated information bits into complex symbols; and the MP-MAC configurator for determining a subset of users and the transmission mode of the MP-MAC based on the characteristic data of the channels and for controlling, in response to this, coding and modulation nodes and display nodes. 7. The system of claim 6, further comprising: an MP-MAC precoding unit operating under the control of the MP-MAC configurator for calculating precoding weights for precoding data signals before transmission to client devices. 8. The system of claim 7, further comprising: one or more orthogonal frequency division multiplexing (OFDM) nodes for receiving precoded signals from the precoded and modulated precoded signals node according to the OFDM standard. 9. The system of claim 8, wherein the OFDM standard comprises computing an inverse fast Fourier transform (IFFT) and adding a cyclic prefix. 10. The system of claim 8, further comprising one or more digital-to-analog conversion nodes for performing digital-to-analog conversion on the output signals of the OFDM nodes to generate an analog signal of the main frequency band; and one or more high-frequency (HF) nodes for conversion with increasing frequency of the signal of the main frequency band to a high frequency and transmission of these signals using the corresponding one or more transmitting antennas. 11. The system of claim 7, wherein the MP-MAC precoding unit is implemented as a preliminary encoder with a minimum mean square error (MMSE) weighted by MMSE, forced coefficient zeroing (ZF), or block diagonalization (DB). 12. The system of claim 6, wherein the MP-MAC system is a distributed antenna communication system, wherein the MP-MAC configurator is a configurator for determining a subset of users and a transmission mode based on channel characteristic data and for controlling, in response, encoding nodes and modulation and display nodes. 13. The system of claim 6, further comprising a base station configured to adaptively switch between different lattice geometries and various antenna diversity methods based on feedback from channel quality information from clients as a means to increase communication line spacing and downlink throughput. 14. The system of claim 6, further comprising a base station that is configured to determine user groups and schedule transmission for different sets of users based on relative priority and / or channel status. 15. The system of claim 6, wherein the MP-MAC configurator or channel estimator is configured to use polarization diversity and / or beam diversity methods as a means to reduce grating size when receiving diversity on a wireless communication line. 16. The system of claim 6, wherein the communication is effected by oscillation of the NVIS and / or surface wave as a means of increasing diversity and downstream throughput. 17. The system of claim 6, wherein the pattern diversity is used to communicate through a surface wave with some users and through NVIS oscillation with other users. 18. The system of claim 17, wherein each client is configured to use the spatial separation of the surface wave and the NVIS communication lines as a means to increase the spatial diversity of the communication line. 19. A wireless client device for use in a system for dynamically adapting communication characteristics of an MP-MAC communication system, comprising: one or more RF nodes for receiving signals transmitted from one or more transmitting nodes of the MP-MAC, and converting these signals with decreasing frequency to the main frequency band; one or more nodes of analog-to-digital conversion for receiving signals converted with lowering frequency and converting said signals from analog signals to digital signals; one or more OFDM nodes to remove a cyclic prefix and perform fast Fourier transform (FFT) on digital signals to represent signals in the frequency domain; a channel estimation unit, configured to receive signals issued from one or more OFDM nodes during the training period, and to calculate, in response thereto, communication line quality indicators; and feedback generator for transmitting quality indicators of the communication line or training signals to the base station for use in modulation / coding, precoding of signals and user selection before transmission to the wireless client device. 20. The wireless client device of claim 19, wherein the channel estimates are calculated in the time domain using input signals to OFDM nodes. 21. The wireless client device according to claim 19, wherein the feedback generator further comprises a logic circuit for quantizing channel estimates and / or quality indicators of the communication lines before transmitting to the base station. 22. The wireless client device according to claim 19, further comprising a receiving node configured to receive output signals from the OFDM nodes and demodulate / decode the signal in response thereto to obtain an estimate of the transmitted data. 23. The wireless client device of claim 21, wherein the receiving node is a minimum mean square error (MMSE) receiver, a coefficient zeroing (ZF) receiver, a maximum likelihood (ML) receiver, or a MAP receiver. 24. The wireless client device of claim 20, wherein the MP-MAC configurator or channel estimator is configured to use polarization diversity and / or radiation diversity techniques as a means to reduce grating size when receiving diversity on a wireless communication line. 25. The wireless client device according to claim 19, in which the communication is carried out through the oscillation of the NVIS and / or surface waves as a means to increase diversity and downstream bandwidth. 26. The wireless client device according to claim 19, in which the diversity in the radiation pattern is used for communication through a surface wave with some users and through the oscillation of NVIS with other users. 27. The wireless client device according to claim 26, wherein each client is configured to use spatial separation of the surface wave and the NVIS communication lines as a means to increase the spatial diversity of the communication line.

Images