RU2350029C2 - Transmission with increasing redundancy for numerous parallel channels in communication system with many inputs and many outputs (mimo) - Google Patents

Abstract

FIELD: information technologies.

SUBSTANCE: transmitter handles (for example, codes, parts, alternates and modulates) each package of data for each parallel channel, being grounded on the velocity chosen for the parallel channel, and gains numerous blocks of figures for a package. For each package of data the transmitter transmits one block of figures for once on the parallel channel until the receiver will not recover a package, or all blocks will not be transmitted yet. The receiver carries out detection and gains the blocks of figures transmitted on parallel channels. The receiver recovers the packages of data transmitted on parallel channels is independent or in the appointed order. The receiver handles (for example, demodulates, deinterlaces, carries out the recurring integration and decodes) all blocks of figures gained for each package of data, and gives out the decoded package. The receiver can estimate and kill the noises caused by recovered packages of data so packages of data recovered later can reach higher signal-noise-jamming ratio (SNJR).

EFFECT: maintenance of transmission with accruing redundancy on numerous parallel channels.

63 cl, 17 dwg, 2 tbl

Description

This application claims priority to provisional application US No. 60/501 776, entitled "H-ARQ for MIMO Systems with Successive Interference Cancellation" (Hybrid Automatic Repeat Request (HARQ) for systems with MIMO and sequential interference suppression), filed September 9, 2003 , and US application No. 60/531 393, entitled "Incremental Redundancy Transmission for Multiple Parallel Channels in a MIMO Communication System", filed December 19, 2003, the assignee of which is the assignee of this applications, and which are included in this application by reference in its entirety for all purposes.

FIELD OF THE INVENTION

The present invention relates mainly to communication and, more specifically, to methods for transmitting data on multiple parallel channels in a communication system with many inputs and many outputs (MIMO).

State of the art

A MIMO system uses multiple ( N _T ) transmit antennas and multiple ( N _R ) receive antennas for data transmission, and it is referred to as an ( N _T , N _R ) system. A MIMO channel formed by N _T transmit and N _R receive antennas can be decomposed into N _S spatial channels, where N _S ≤ min { N _T , N _R }, as described below. N _S data streams may be transmitted on N _S spatial channels. A MIMO system can provide increased throughput if N _S spatial channels created by multiple transmit and receive antennas are used for data transmission.

The main problem in a system with MIMO is the selection of suitable speeds for transmitting data on a channel with MIMO, based on the channel mode. A “rate” may indicate a particular data rate or information bit rate, a particular coding scheme, a particular modulation scheme, a specific data packet size, etc. The goal of selecting a rate is to maximize the overall throughput over N _S spatial channels, while at the same time satisfying certain quality requirements that can be measured by the target probability of error per packet (VOP) (e.g. 1% VOP) or some other measures.

The bandwidth of each spatial channel depends on the ratio of the signal to the sum of noise and interference (SINR) achieved by this spatial channel. The SINR for N _S spatial channels depends on the channel mode and may additionally depend on how the data streams are restored at the receiver. In one conventional MIMO system, a transmitter encodes, modulates, and transmits each data stream in accordance with a speed selected based on a static channel model with a MIMO. Good performance can be achieved if the model is accurate and if the MIMO channel is relatively static (i.e., does not change much over time). In another conventional MIMO system, the receiver estimates the MIMO channel, selects the appropriate speed for each spatial channel based on the channel estimates, and sends N _S selected speeds for N _S spatial channels to the transmitter. The transmitter then processes the N _S data streams in accordance with the selected rates and transmits these streams over the N _S spatial channels. The performance of this system depends on the nature of the MIMO channel and the accuracy of the channel estimates.

For both the conventional MIMO systems described above, the transmitter processes and transmits each data packet for each spatial channel at the rate selected for that spatial channel. The receiver decodes each data packet received on each spatial channel and determines whether the packet is decoded correctly or in error. The receiver may send back an acknowledgment (PP) if the packet is decoded correctly, or a non-acknowledgment (NPP) if the packet is decoded in error. The transmitter can retransmit each data packet decoded with an error by the receiver, completely when receiving the receiver for the packet.

The performance of both MIMO systems described above is highly dependent on the accuracy of the speed selection. If the selected speeds for spatial channels are too cautious (for example, since the actual SINRs are much higher than the estimates of SINRs), then excess system resources are spent on transmitting data packets and channel bandwidth is underutilized. Conversely, if the selected speeds for spatial channels are too aggressive, then data packets may be decoded with an error by the receiver, and system resources may be spent on retransmission of these packets. The choice of speed for a system with MIMO is a problem due to (1) greater difficulty in estimating a channel for a channel with MIMO, (2) the time-dependent and independent nature of spatial channels, and (3) the relationship of multiple data streams transmitted over spatial channels.

There is, therefore, a need for methods for efficiently transmitting data over multiple spatial channels in a MIMO system that do not require accurate speed selection to achieve good performance.

Disclosure of invention

This document presents methods for transmitting with increasing redundancy (NI) over multiple ( N _D ) parallel channels in a MIMO system. These parallel channels can be formed (1) by means of numerous spatial channels in a system with MIMO, (2) in such a way that they achieve similar UWB, or (3) in some other way. Initially, the receiver or transmitter in the MIMO system estimates the SINR for N _D parallel channels and selects N _D speeds for these parallel channels. The SINR can depend on various factors, such as the transmission scheme used by the transmitter, the processing performed by the receiver, etc. The transmitter is provided with selected speeds if the receiver makes a speed selection.

The transmitter processes (e.g., encodes, splits, interleaves, and modulates) each data packet for each parallel channel based on the speed selected for that channel and obtains multiple ( N _B ) data symbol blocks for the packet. The first block of data symbols usually contains sufficient information for the receiver to recover the data packet in a favorable channel mode. Each of the remaining blocks of data symbols contains additional redundancy, allowing the receiver to restore the data packet at a less favorable channel mode. For each data packet, the transmitter transmits one block of data symbols at a time until all blocks for the packet are transmitted. The transmitter completes the transmission of the data packet in advance if the packet is restored (i.e., successfully decoded) by the receiver with a smaller quantity than all blocks of data symbols.

The receiver detects N _R sequences of received symbols and obtains a block of detected symbols for each block of data symbols transmitted by the transmitter. Subsequent processing depends on whether the parallel channels are independent or interdependent.

N _D parallel channels are independent if the data transmission on each parallel channel is independent of data transmission on other parallel channels. In this case, for each data packet on each parallel channel, the receiver processes (for example, demodulates, deinterleaves, reassembles, and decodes) all the blocks of detected symbols received for the data packet and produces a decoded packet. The receiver may send back the PP if the decoded packet is good, and the NRI if the decoded packet is in error. The receiver completes processing for each data packet that is recovered, or if all blocks of data symbols have been received for the packet.

N _D parallel channels are interdependent if data transmission on each parallel channel depends on data transmission on other parallel channels. This is the case if the receiver uses a “sequential interference cancellation” (“interference suppression”) processing technique to obtain blocks of detected symbols. Using the RFP, whenever a data packet is restored on a parallel channel, the interference that this packet causes for data packets not yet received on other parallel channels is evaluated and suppressed before detection is performed in order to obtain blocks of detected symbols for these other data packets. The SINR for later data packets is thus higher, and higher rates can be selected for these packets. The data packets are then restored by the receiver in a specific order determined on the basis of their selected speeds, so that the UWB necessary to recover these data packets can be obtained.

For a “streamlined” transmission scheme with IFR, if the data packet on a given parallel channel x is restored earlier than the expected time, then one of several options is available. Firstly, the transmitter can transmit nothing on the parallel channel x and use most or all of the radiation power for data packets that have not yet been restored. Secondly, the transmitter can transmit a new “short” data packet on parallel channel x . The short packet is expected to be restored the moment the next data packet is to be restored, or before it. Thirdly, the transmitter can transmit a new “long” data packet on parallel channel x . A long packet is expected to be restored after the next data packet is to be restored. One of these options can be selected based on a metric that compares throughputs with packet transmission and without packet transmission on parallel channel x after premature termination.

For the IFR and cyclic repeat transmission scheme, whenever a data packet is recovered for a parallel channel, the transmitter transmits a new data packet on this parallel channel, and the receiver cycles through to the next parallel channel and attempts to recover the data packet on this next parallel channel.

IFR and other transmission schemes are described below. Various aspects and embodiments of the invention are also described in more detail below.

Brief Description of the Drawings

The distinguishing features and essence of the present invention will become more apparent from the detailed description set forth below, taken in conjunction with the drawings, in which like numbers are defined in the same way in all the drawings and in which:

figure 1 depicts a block diagram of a transmitter and a receiver in a system with MIMO, which implements transmission from NI for multiple ( N _D ) parallel channels;

figure 2 depicts the process of performing transmissions with NI for parallel channels;

figure 3 illustrates the transmission with NI for one data stream on one parallel channel;

4 depicts a transmission scheme for an orthogonal frequency division multiplexing (OFDM) system with MIMO;

5 illustrates NI transmission for N _D independent parallel channels;

6A-6C illustrate an ordered transmission scheme with IFR with three different options for premature termination of a data packet on one parallel channel;

Fig.7 depicts GPs for Packet 1b and Packet 2a, depending on the number of transmission cycles for Packet 2a;

Fig. 8 is a state diagram for an ordered transmission scheme with IFR;

figa and 9B illustrate the transmission scheme with IFR and cyclic repetition;

figure 10 depicts a data processor transmission (TX) at the transmitter;

11 illustrates the processing of one data packet by the transmitter;

Fig depicts the spatial processor TX and the transmitting node in the transmitter;

13 depicts one embodiment of a receiver;

FIG. 14 shows a reception data processor (RX) in the receiver of FIG. 13; and

Fig depicts a receiver that implements the IFR method.

DETAILED DESCRIPTION

The word “exemplary” in this document is used to mean “serving as an example, option, or illustration.” Any embodiment or development described herein as “exemplary” is not necessarily to be construed as preferred or advantageous with respect to other embodiments or developments.

For a system with MIMO, N _D data streams can be transmitted simultaneously on N _D parallel channels, one data stream for each parallel channel, where N _D > 1. Each parallel channel may correspond to a spatial channel or may be formed in some other way, as described below. Each data stream can be processed independently, based on the speed selected for this data stream, and transmitted on its parallel channel.

Figure 1 depicts a block diagram of a transmitter 110 and a receiver 150 in an MIMO system 100, which implements NR transmission for multiple data streams over multiple parallel channels. At transmitter 110, TX data processor 120 receives N _D data streams from data source 112. TX data processor 120 processes (for example, formats, encodes, splits, interleaves, and modulates) each data packet in each data stream in accordance with the speed selected for that data stream and provides N _B data symbol blocks for the packet, where N _B > 1 and may depend on the selected speed. Each block of data symbols may be transmitted in one time slot (or simply “slot”), which is a predetermined time period for the MIMO system 100. The selected rate for each data stream may indicate the data rate, coding scheme or coding rate, modulation scheme, packet size, number of data symbol blocks, etc., which are indicated by various control signals provided by the controller 140. The selected speed for each data stream can be static or continuously updated (e.g. based on channel mode). In order to transmit a given data stream from an NI, the blocks of data symbols for each data packet of this data stream are transmitted one block at a time until the packet is recovered by receiver 150 or all blocks are transmitted.

TX spatial processor 130 receives N _D data symbol streams from TX data processor 120. Each data symbol stream includes a set of data symbol blocks for each data packet in the corresponding data stream. TX spatial processor 130 performs processing (e.g., demultiplexing, spatial processing, etc.) to transmit N _D data symbol streams from N _T transmit antennas. Various transmission schemes may be implemented as described below. Depending on the transmission scheme selected for use, up to N _D data symbol blocks for up to N _D data streams are transmitted simultaneously on up to N _D parallel channels in any given interval. TX spatial processor 130 also multiplexes pilot symbols, which are used to estimate the channel by receiver 150, and provides N _T transmit symbol streams to transmitting node 132 (TMTR).

The transmitting node 132 receives and brings to a certain state (for example, converts to analog form, converts with increasing frequency, filters and amplifies) N _T transmission symbol streams to receive N _T modulated signals. Each modulated signal is then transmitted from the corresponding transmitting antenna (not shown in FIG. 1) and along the channel from the MIMO to the receiver 150. The channel from the MIMO distortes N _T transmitted signals in accordance with the characteristic of the channel with the MIMO and further degrades the transmitted signals with additive white Gaussian noise and possibly interference from other transmitters.

At a receiver 150 N _T transmitted signals are received by each of the N _R receive antennas (not shown in FIG. 1), and N _R received signals from N _R receive antennas are fed to a receiving node (RCVR) 154. The receiving node 154 is brought into a certain state, digitizes and preprocesses each received signal to obtain a stream of received symbols. The receiving node 154 supplies N _R received symbol streams (for data) to the RX spatial processor 160 and the received pilot symbols (for pilot) to the channel estimator 172. The RX spatial processor 160 processes (e.g., detects, multiplexes, demultiplexes, etc.) N _R received symbol streams to obtain N _D detected symbol streams, which are estimates of N _D data symbol streams sent by transmitter 110.

RX data processor 170 receives and processes N _D detected symbol streams to obtain N _D decoded data streams, which are estimates of N _D data streams sent by transmitter 110. For each data packet of each data stream, RX data processor 170 processes (for example, demodulates, deinterleaves, reassembles, and decodes) all blocks of data symbols received for this data packet in accordance with the selected rate, and produces a decoded packet, which is an estimate of chum data. RX data processor 170 also provides a status for each decoded packet, which indicates whether the packet is decoded correctly or in error.

The channel estimator 172 processes the received pilot symbols and / or the received data symbols to obtain channel estimates (eg, channel gain estimates and SINR estimates) for N _D parallel channels. A speed selector 174 receives channel estimates and selects a rate for each of the N _D parallel channels. The controller 180 receives the N _D selected speeds from the speed selector 174 and the state of the packet from the RX data processor 170 and collects feedback information for the transmitter 110. The feedback information may include N _D selected speeds, PP, and NRP for decoded packets, etc. P. Feedback information is processed by TX spatial processor 190, is additionally brought into a certain state by the transmitting node 192, and transmitted via the feedback channel to the transmitter 110.

At transmitter 110, the signal (s) transmitted by receiver 150 is received and conditioned by receiver 146 and further processed by RX spatial data processor 148 to obtain feedback information sent by receiver 150. Controller 140 receives feedback information, uses PP / NR to control the transmission from NI of the current data packets sent over N _D parallel channels, and uses N _D selected speeds to process new data packets to be sent over N _D parallel channels.

Controllers 140 and 180 control operation at transmitter 110 and receiver 150, respectively. The nodes 142 and 182 memory provide storage of program codes and data used by the controllers 140 and 180, respectively. Memory nodes 142 and 182 may be internal to controllers 140 and 180, as shown in FIG. 1, or external to these controllers. The processing units shown in FIG. 1 are described in detail below.

Figure 2 depicts a flowchart of a process 200 for performing NI transmissions for N _D data streams over N _D parallel channels in a MIMO system. Initially, the receiver estimates N _D parallel channels based on the pilot and / or data symbols received from the transmitter (step 210). The receiver selects a speed for each of the N _D parallel channels based on the channel estimates and sends N _D selected speeds to the transmitter (step 212). The transmitter receives N _D selected speeds and processes the data packets for N _D data streams in accordance with N _D selected speeds to obtain N _D data symbol streams (step 220). The transmitter can format, encode, split, interleave, and modulate each data packet of each data stream in accordance with the speed selected for that data stream to obtain N _B data symbol blocks for the data packet. The transmitter then transmits N _D data symbol streams over N _D parallel channels (block 222). For example, a transmitter may transmit one block of data symbols at a time for each data packet of each data stream until all blocks of data symbol blocks are transmitted or the PP for the data packet is received. Various transmission schemes may be used for transmission NO N _D data streams, as described below.

The receiver receives N _D data symbol streams from the transmitter using N _R receive antennas and processes N _R received symbol streams to obtain N _D detected symbol streams (step 230). The receiver then processes the N _D detected symbol streams and restores the data packets transmitted by the transmitter (step 232). For each interval, the receiver may attempt to restore the current data packet transmitted for each of the N _D data streams. For example, whenever a new block of detected symbols is obtained for a data packet, the receiver can demodulate, de-interleave, reassemble and decode all the blocks of detected symbols received for this packet to obtain a decoded packet. The receiver also checks each decoded packet to determine if the packet is correctly decoded (good) or erroneously (erased) (also step 232).

PP / NRE feedback can be achieved in various ways. In one embodiment, the receiver sends an LSC for each decoded packet that is erased, and the transmitter uses this feedback to transmit the next block of data symbols for the erased packet. In another embodiment, the transmitter sends one block of data symbols at a time for each data packet until a receiver for the packet from the receiver is received (the receiver may send back and may not send the NTP back). In any case, the receiver completes the processing for each data packet that has been recovered, or if all the data symbol blocks for the packet have been received (block 234).

Figure 2 depicts a typical embodiment of the transmission with NI for N _D data streams on N _D parallel channels. NI transmission for multiple parallel channels can also be performed in another way, and this is within the scope of the invention.

FIG. 3 illustrates an NI transmission for one data stream (designated as Stream i ) on one parallel channel (designated as Channel i ). The receiver estimates Channel i , selects the speed r _{i, 1} for Channel i , based on the channel estimates, and sends the selected speed to the transmitter in the interval 0. The transmitter receives the selected speed, processes the data packet (Packet 1) for Stream i in accordance with the selected speed and transmits the first block (Block 1) of the data symbols for Packet 1 in interval 1. The receiver receives and processes Block 1, determines that Packet 1 is decoded with an error, and sends back the NRT in interval 2. The transmitter receives the NRT and transmits the second block (Block 2 ) data characters for Packet 1 in interval 3. The receiver receives Block 2, processes Blocks 1 and 2, determines that Packet 1 is still decoded with an error, and sends back the CID in interval 4. The transmission of blocks and the response CID can be repeated any number of times. In the example shown in FIG. 3, the transmitter receives the NAK for a block of N _X -1 data symbols and transmits a block of N _X (Block N _X ) data symbols for Packet 1 in the interval m , where N _{X is} less than or equal to the total number of blocks for the Pack 1. The receiver receives a Block N _X , processes all N _X blocks of data symbols received for Packet 1, determines that the packet is decoded correctly, and sends back the PP in the interval m +1. The receiver also evaluates Channel i , selects the speed r _{i, 2} for the next data packet for Stream i, and sends the selected speed to the transmitter in the interval m +1. The transmitter receives the PP for Block N _X and completes the transmission of Packet 1. The transmitter also processes the next data packet (Packet 2) in accordance with the selected speed r _{i, 2} and transmits the first block of data symbols for Packet 2 in the interval m +2. The processing at the transmitter and receiver for Packet 2 continues in the same way as described for Packet 1.

For the embodiment shown in FIG. 3, there is a one-interval delay for the response PP / NAK from the receiver to transmit each block. To improve channel utilization, multiple data packets may be transmitted for each data stream in an alternating manner. For example, one data packet may be transmitted at odd-numbered intervals, and another data packet may be transmitted at even-numbered intervals. More than two data packets may also alternate if the PP / NAK delay is more than one interval.

N _D parallel channels in a system with MIMO can be formed in various ways, as described below. In addition, depending on the processing performed at the receiver, N _D parallel channels may be independent of one another or interdependent. For independent parallel channels, NR transmission for each data stream can be performed independently of or taking into account NR transmission for other data streams. For interdependent parallel channels, the transmission from the NI for each data stream depends on the transmission from the NI for other data streams.

1. NI transmission for multiple independent parallel channels

Various transmission schemes can be used to transmit N _D data streams simultaneously on N _D parallel channels, where N _D > 1. Some exemplary transmission schemes are described below. For simplicity, the following description assumes a full-rank channel with MIMO and N _D ≤ N _S = N _T ≤ N _R.

In the first transmission scheme, one data stream is transmitted from each of the N _T transmit antennas without any spatial processing at the transmitter. The model for this transmission scheme can be expressed as:

Lv. (1)

where s is a data vector { N _T × 1} with N _T elements for data symbols;

r _nsp is a receive vector { N _R × 1} with N _R elements for N _R received symbols obtained using N _R receive antennas;

H is a channel response matrix { N _R × N _T } for a channel with MIMO; and

n is the additive white Gaussian noise (ABGS) vector.

, где σ² представляет собой дисперсию шума, и I представляет собой единичную матрицу с единицами по диагонали и нулями во всех других местах.Vector s includes N _T elements for N _T transmit antennas, wherein N _D elements are set equal to N _D data symbols for N _D data streams, and the remaining N _T - N _D elements are set to zero. Vector n is assumed to have a zero mean and covariance matrix

, where σ ² is the variance of the noise, and I is the identity matrix with diagonal units and zeros in all other places.

Due to scattering in the channel with MIMO, N _D data streams transmitted from N _T transmit antennas interfere with each other at the receiver. The data stream transmitted from a given transmitting antenna can be received by all N _R receiving antennas with different amplitudes and phases. The received signal for each receiving antenna then includes a component of each of the N _D data streams.

The receiver can evaluate the data vector s based on various spatial and spatio-temporal processing schemes (ie, “detection”). For example, the receiver can estimate the data vector s using an optimal addition detector (OS), a minimum mean square error (ISCED) detector, a linear zero-force detector (FC) (which is also referred to as a channel correlation matrix inversion detector (CCMC)), ISCED linear corrector, critical feedback corrector, or some other detector / corrector. The spatial processing for some of these detectors is described below.

Spatial processing for an OS detector can be expressed as:

Ur (2)

where W _mrc is an OS detector characteristic, which is W _mrc = H ;

представляет собой вектор {N _T×1} для N _T детектированных символов от детектора ОС; и

represents the vector { N _T × 1} for N _T detected characters from the OS detector; and

“ ^H ” means the conjugate result of the transposition.

The spatial processing for an ISCED detector can be expressed as:

Ur (3)

where W _mmse = ( HH ^H + σ ² I ) ^-1 H for the ISCED detector.

Spatial processing for a zero-force detector can be expressed as:

Ur (four)

where W _zf = H ( H ^H H ) ^-1 for a zero-force detector. For the first transmission scheme, each spatial channel corresponds to a corresponding transmit antenna.

In the second transmission scheme, one data stream is transmitted over each “native mode” of the channel with the MIMO. The channel response matrix H can be decomposed using either decomposition by special values or decomposition by eigenvalues to obtain N _S eigenmodes of the channel with MIMO. N _S eigenmodes of the channel with MIMO are orthogonal to each other, and improved performance can be achieved by transmitting multiple data streams using these eigenmodes. The decomposition according to the special values of the matrix H of the channel characteristic can be expressed as:

H = UΣV ^H Ur (5)

where U is the identity matrix { N _R × N _R } of the left eigenvectors H ;

Σ is the diagonal matrix { N _R × N _T } of special values of H ; and

V is the identity matrix { N _T × N _T } of the right eigenvectors H.

The identity matrix is characterized by the property M ^H M = I. The unit matrices V and U are used for spatial processing by the transmitter and receiver, respectively, for transmitting N _D data streams over N _S eigenmodes of the channel with the MIMO.

The transmitter performs spatial processing using the matrix V as follows:

x _svd = Vs Ur (6)

where x _svd is a vector { N _T × 1} with N _T elements for N _T transmit symbols sent from N _T transmit antennas. The receive vector is then defined as r _svd = HVs + n . The receiver performs spatial processing using the matrix U as follows:

Ur (7)

For the second transmission scheme, each spatial channel corresponds to a corresponding eigenmode. N _S eigenmodes can be considered as orthogonal spatial channels obtained by decomposition.

For the first and second transmission schemes, N _D data streams can achieve various and possibly substantially varying SIRWs “after processing” or “after detection”, which are the SINRs achieved after linear detection by the receiver (for example, using the ISCED detector , with forcing zero or optimal addition). Then different speeds for data streams are required.

In the third transmission scheme, each of the N _D data streams from all N _T transmission symbols is transmitted, so that all data streams experience similar channel modes and achieve similar UWB after processing. Then the same or similar speeds can be used for N _D data streams. For this scheme, the transmitter multiplies the matrices of the data vector s with the base transfer matrix and the diagonal matrix as follows:

x _tbm = MΛs Ur (8)

where x _tbm is a vector { N _T × 1} with N _T transmit symbols for N _T transmit antennas;

M is a transmission base matrix { N _T × N _T }, which is a identity matrix; and

Λ is the diagonal matrix { N _T × N _T }.

, где E представляет собой матрицу Уолша-Адамара. Матрица M также может определяться как

, где F представляет собой матрицу дискретного преобразования Фурье (ДПФ) с (m,n)-элементом, определенным как

, где m представляет собой индекс строки, и n представляет собой индекс столбца для матрицы F , причем m=1…N _T и n=1…N _T. Диагональная матрица Λ включает в себя N _D ненулевых элементов по диагонали и нули во всех других местах. Эти N _D ненулевых элементов могут использоваться для назначения различной мощности излучения N _D потокам данных, в то же время соответствуя ограничению на полную мощность излучения P _ant для каждой передающей антенны.The base transmit matrix M makes it possible to send each data stream from all N _T transmit antennas and further makes it possible to use the total power P _{ant of} each transmit antenna for data transmission. Matrix M can be defined as

where E is the Walsh-Hadamard matrix. Matrix M can also be defined as

where F is the matrix of the discrete Fourier transform (DFT) with the ( m, n ) -element defined as

, where m is the row index and n is the column index for the matrix F , with m = 1 ... N _T and n = 1 ... N _T. The diagonal matrix Λ includes N _D non-zero elements along the diagonal and zeros in all other places. These N _D non-zero elements can be used to assign different radiation powers to N _D data streams, while at the same time meeting the limit on the total radiation power P _ant for each transmit antenna.

The “effective” channel response observed by the receiver for this transmission scheme is H _eff = HM . The receiver can evaluate the data vector s using the OS detector / corrector, according to ISCED, with zero force or some other, where the detector characteristic W (which may be W _mrc , W _mmse or W _zf ) is calculated using the matrix H _{eff of the} effective channel characteristic instead matrix H channel characteristics. The third transmission scheme is described in detail in US Patent Application No. 10/367 234, entitled “Sharing Adaptive Transmission Scheme for MIMO Systems”, filed February 14, 2003.

The third transmission scheme can transmit any number of data streams simultaneously from N _T transmit antennas (i.e., 1 ≤ N _D ≤ N _S ), allows N _D parallel channels to achieve similar SINRs after processing (which can simplify receiver operation with IFR) and additionally makes it possible to use the same or different radiation power for data streams.

The NI transmission methods described in this document can be implemented in a single-carrier MIMO system that uses one carrier for data transmission, and in a multi-carrier MIMO system that uses many carriers for data transmission. Many carriers can be provided with orthogonal frequency division multiplexing (OFDM), other modulation methods with many carriers, or some other design. OFDM effectively partitions the overall system bandwidth into multiple ( N _F ) orthogonal subbands, which are also commonly referred to as tones, bins, or frequency channels. In OFDM, each subband is associated with a corresponding carrier that can be modulated with data.

For a MIMO system that implements OFDM (i.e., an OFDM system with MIMO), N _D data streams can be transmitted over N _F subbands of N _T transmit antennas in various ways. For example, each data stream may be transmitted on N _F subbands of a respective transmit antenna. Alternatively, each data stream may be transmitted over multiple subbands and multiple transmit antennas to achieve frequency diversity and spatial diversity.

In the fourth transmission scheme, each data stream is transmitted diagonally over the N _F subbands and from all N _T transmit antennas. This scheme provides both frequency diversity and spatial diversity for all N _D data streams transmitted simultaneously, and additionally achieves similar SINRs after processing for N _D data streams after linear detection at the receiver.

FIG. 4 shows a fourth transmission scheme for a case in which two data streams ( N _D = 2) are transmitted in an exemplary OFDM system with MIMOs with four transmit antennas ( N _T = 4) and 16 subbands ( N _F = 16). For the first data stream, the first four symbols s _1,1 , s _1,2 , s _1,3 and s _{1,4 of} data are transmitted in subbands 1, 2, 3 and 4, respectively, of transmitting antennas 1, 2, 3 and 4, respectively. The following four symbols s _1,5 , s _1,6 , s _1,7 and s _{1,8 of} data perform a cyclic transition and are transmitted in subbands 5, 6, 7 and 8, respectively, of transmitting antennas 1, 2, 3 and 4, respectively. For the second data stream, the first four characters s _2,1 , s _2,2 , s _2,3 and s _{2,4 of the} data are transmitted in subbands 1, 2, 3 and 4, respectively, of the transmitting antennas 3, 4, 1 and 2, respectively. The following four data symbols s _2.5 , s _2.6 , s _2.7 and s _2.8 perform a cyclic transition and are transmitted in subbands 5, 6, 7 and 8, respectively, of transmit antennas 3, 4, 1 and 2, respectively. For the embodiment shown in FIG. 4, not all subbands are used for data transmission, and unused subbands are filled with zero signal values. Multiplexing / demultiplexing can also be performed in another way.

For an OFDM system with MIMO, the spatial processing described above for the transmitter and receiver can be performed for each subband k , for k = 1 ... N _F , based on the channel characteristic matrix H ( k ) for that subband.

For a MIMO system that implements orthogonal frequency division multiple access (OFDMA) (i.e., an MIMO system with MIMO), only a subset of N _F subbands may be available for transmitting data for each receiver. The processing described above for the MIMO OFDM system can also be used for the OFDM system with MIMO, although only on the subbands available for data transmission. For example, N _D data streams for a given receiver can be transmitted diagonally over available subbands (instead of all N _F subbands) and from N _T transmit antennas.

N _D parallel channels can be formed in various ways in systems with MIMO and in OFDM systems with MIMO. The four transmission schemes described above represent four exemplary methods for generating multiple parallel channels. Generally speaking, parallel channels can be formed using any combination of space, frequency and time.

In the following description, a “transmission cycle” (or simply a “cycle”) is a period of time that encompasses the transmission of a data symbol block by a transmitter and the transmission of a PP / NAK response signal for that block by a receiver. “F” denotes an unfavorable decoding outcome by the receiver, and “S” denotes a favorable decoding outcome. For simplicity, the interleaving of multiple data packets for each data stream is not shown in the following timing diagrams.

FIG. 5 illustrates ND transmission of N _D data streams over N _D independent parallel channels. Since these parallel channels are independent, the receiver can recover each data stream independently and provide a PP / NR feedback stream for this data stream. The transmitter sends a new block of data symbols for the current data packet of each data stream in each cycle.

In the example shown in FIG. 5, for data stream 1 transmitted on parallel channel 1 (Channel 1), the receiver encounters an unsuccessful decoding outcome (“F _1a ”) when it tries to recover data packet 1a (Packet 1a) with only a block 1 data symbols in cycle 1, with an unfavorable decoding outcome when trying to recover a Packet 1a with blocks 1 and 2 of data symbols in a cycle 2, with an unfavorable decoding outcome when trying to recover a Packet 1a with blocks 1, 2 and 3 data symbols in a cycle 3 and with a favorable outcome ( «S _1a») decodes Bani when trying to restore the package 1a with blocks 1-4 of the data symbols in the cycle 4. Then, the transmitter terminates the transmission of Packet 1a and begins transmitting blocks of data symbols for another 1b packet (1b) data. The receiver attempts to recover Packet 1b whenever a new block of data symbols is received for this packet, encounters an unfavorable decoding result in each of 5-8 cycles and can correctly decode Packet 1b with blocks of 1-5 data symbols in cycle 9. The receiver processes each from other data streams in a similar manner as shown in FIG.

2. NI transmission for multiple interdependent parallel channels

The receiver may process N _R received symbol streams using the IFR technique to obtain N _D detected symbol streams. For the SPT method, which is a nonlinear detection scheme, the receiver initially performs detection of N _R streams of received symbols (for example, using an OS detector, ISCED or zero-forcing) and receives one stream of detected symbols. The receiver further processes (for example, demodulates, de-interleaves, and decodes) this stream of detected symbols to obtain a stream of decoded data. The receiver then estimates the interference that this data stream causes for other N _D −1 data streams, and suppresses the estimated interference in the N _R received symbol streams to obtain N _R modified symbol streams. The receiver then repeats the same processing on N _R modified symbol streams to recover another data stream.

The receiver thus processes the N _R received symbol streams in N _D consecutive steps. For each stage, the receiver performs (1) detection of either N _R received symbol streams, or N _R modified symbol streams from the previous stage to obtain one detected symbol stream, (2) decodes this detected symbol stream to obtain the corresponding decoded data stream, and (3 ) estimates and suppresses interference caused by this stream to obtain N _R modified symbol streams for the next stage. If the interference caused by each data stream can be accurately estimated and suppressed, which requires recovery without errors or with small errors in the data stream, then later restored data streams experience less interference and can achieve large SINS after processing. The RFP method is described in detail in a US patent application with transfer of sharing rights No. 09/993 087, entitled “Multiple-Access Multiple-Input Multiple-Output (MIMO) Communication System” (MIMO) and with multiple access), filed November 6, 2001

After the processing of each data stream, for the SINR IFR method, it depends on (1) the SINR of this stream with linear detection and without suppressing interference, (2) the specific stage in which the data stream is restored, and (3) the interference caused by later restored data streams. Thus, even if N _D data streams can achieve similar SINRs after processing with linear detection (for example, using an ISCED detector, with zero force or OS), these streams usually reach different SINRs after processing with nonlinear detection using the IFR method. Generally speaking, after processing, the SINR is gradually improving for data streams restored at later stages, since interference from data streams restored at previous stages is suppressed. This, therefore, allows the use of higher speeds for later restored data streams.

The RFP method introduces the interdependence between data streams. In particular, the rates for N _D data streams are selected based on the UWB after processing achieved by these data streams, which, in turn, depends on the order in which the data streams are restored. After processing each data stream, the SINR assumes that all earlier data streams (that is, those intended to be restored to this data stream) were successfully decoded and suppressed. The receiver usually needs to restore N _D data streams in the assigned order, and it normally cannot restore this data stream until it is restored and suppressed in all earlier data streams.

Different transmission schemes can be used for a system with an MIMO with a receiver with a BCP. Several exemplary transmission schemes are described below. For simplicity, the following description assumes that two data streams ( N _D = 2) are transmitted on two parallel channels. However, the principles described below can be extended to any number of data streams.

A. Scheme of ordered transmission with IFR

In an ordered transmission scheme with IFR, N _D data streams are restored in the assigned order. For example, the receiver may first restore the data stream 1, then the next data stream 2, etc., and the last data stream N _D. The assigned order may depend on how the data streams are transmitted. For example, the received UWB for N _D data streams are likely to be similar for the third and fourth transmission schemes described above. In this case, the order in which N _D data streams are restored is minimally affected by the performance, and any order can be selected. The adopted SINRs for N _D data streams are likely to be different for the first transmission scheme described above. In this case, better performance can be achieved by first recovering the data stream with the highest received SINR, then the data stream with the next highest received SINR, etc. In any case, for an ordered transmission scheme from the IFR, the receiver tries to recover the data stream i only after of how interference from all earlier streams from 1 to i -1 of the data will be suppressed.

Initially, after-processing SINRs are evaluated for N _D data streams based on (1) received SINRs for data streams, for example, with equal radiation powers used for data streams, and (2) the designated recovery order of data streams. The SINR after processing the data stream restored at step l , SINR _pd ( l ), can be expressed as

Ur (9)

where w _l is the detector characteristic for the flow recovered at stage l , and σ ² is the noise variance in the receiver. The characteristic w _{l of the} detector is one column of the characteristic W _{l of the} detector (for example, OS, ISCED or zero-force) derived for stage l , based on the reduced matrix H _{l of the} channel characteristic for this stage. The matrix H _{l is} obtained by deleting ( l −1) columns in the original matrix H corresponding to data streams already restored to ( l −1) previous steps. The post-processing UWB calculation is described in detail in a U.S. patent application transferring the right to share No. entitled "Successive Interference Cancellation Receiver Processing with Selection Diversity" filed September 23, 2003.

The speed is selected for each data stream based on its SINR after processing. It is not necessary that the speed selection be accurate, since the data packet can be transmitted at a variable speed when transmitting from NI. The sizes of N _D data packets to be transmitted for N _D data streams are selected at given selected rates, so it is expected that all data packets will be recovered by the receiver with the same number of cycles ( N _est ), where N _est can be determined based on an estimate of a stock of OSPP after processing. The transmission of each data packet may terminate prematurely if the packet is restored before the N _est cycle, and may increase beyond the N _est cycle, if necessary, until the packet is restored.

, который находится перед циклом N _est, тогда целью этой схемы передачи является синхронизация обоих потоков данных как можно быстрее без потери спектральной эффективности. В таблице 1 показаны некоторые доступные варианты, если Пакет 1а восстанавливается перед циклом N _est.6A-6C illustrate an RFR ordered transmission scheme with three different transmission options for prematurely terminating a data packet in a single data stream. 6A-6C, two new data packets (Packets 1a and 2a) are transmitted, starting in cycle 1 for data streams 1 and 2, on parallel channels 1 and 2 (Channels 1 and 2), respectively. If Packet 1a for data stream 1 is restored in a loop

, which is before the N _est cycle, then the purpose of this transmission scheme is to synchronize both data streams as quickly as possible without loss of spectral efficiency. Table 1 shows some of the options available if Packet 1a is restored before the N _est cycle.

Table 1 Option Description one Do not transmit anything on Channel 1 and use all the radiation power for Packet 2a on Channel 2 after restoring Packet 1a, as shown in FIG. 6A. This increases the likelihood of Packet 2a recovering to the N _est cycle. 2

, где

и

. Скорость для короткого пакета выбирается на основе оценок канала, полученных в цикле

.Transmit a new “short” data packet on Channel 1, as shown in FIG. Short pack has a length

where

and

. The speed for the short burst is selected based on the channel estimates obtained in the loop

. 3 Transmit a new “long” data packet over Channel 1, as shown in FIG. The long packet has a length

where

. This may delay the recovery of Package 2a until the cycle.

, which is a cycle in which long packet recovery is expected.

и

(аналогично N _est) представляют количество циклов, в течение которых ожидается восстановление коротких и длинных пакетов данных, основываясь на оценке с запасом ОСШП после обработки.Table 1

and

(similar to N _est ) represent the number of cycles during which recovery of short and long data packets is expected, based on the estimate with a stock of SINR after processing.

A metric can be used to select one of the three options shown in Table 1 whenever premature termination occurs. This metric can be determined based on the total throughput, and is determined as follows:

Ur (10)

обозначает суммарную пропускную способность для Канала 2, если передача для Пакета 2а расширяется до цикла

из-за передачи длинного пакета по Каналу 1. Разность между этими двумя членами представляет снижение суммарной пропускной способности для Канала 2. Новый длинный пакет, таким образом, может передаваться по Каналу 1, если выигрыш суммарной пропускной способности для Канала 1 больше снижения суммарной пропускной способности для Канала 2 (т. е. Вариант 3 в таблице 1 может выбираться, если справедливо уравнение (10)).where R _i ( j , n ) is the total throughput predicted in cycle j for data stream i after n cycles. The left side of the inequality in equation (10) represents the gain in total throughput (Δ R _{1, long} ) for Channel 1 with the new long packet transmitted on Channel 1. The right side of the inequality in equation (10) represents the decrease in total throughput (Δ R _{2 , long} ) for Channel 2 due to the transmission of a new long packet on Channel 1. The term R ₂ (0, N _est ) denotes the total throughput for Channel 2 if Packet 2a is restored in the N _est cycle, as predicted. Member

denotes the total throughput for Channel 2, if the transmission for Packet 2a extends to the cycle

due to the transmission of a long packet on Channel 1. The difference between the two members represents a decrease in the total throughput for Channel 2. A new long packet can thus be transmitted on Channel 1 if the gain in the total throughput for Channel 1 is greater than the decrease in the total throughput for Channel 2 (i.e., Option 3 in table 1 can be selected if equation (10) is true).

. Это пессимистическое предположение, так как вероятность восстановления Пакета 2а до цикла N _est повышается, когда большая мощность излучения используется для Пакета 2а после цикла

. Уравнение (10) может быть модифицировано следующим образом:Equation (10) assumes that N _est cycles are required to restore Package 2a, even if the total radiation power is used for Package 2a after Package 1a has been restored in the cycle

. This is a pessimistic assumption, since the probability of Packet 2a recovering to the N _est cycle increases when a large radiation power is used for Packet 2a after the cycle

. Equation (10) can be modified as follows:

Ur (eleven)

представляет собой число предсказанных циклов, необходимых для восстановления Пакета 2а со всей мощностью излучения, используемой для Пакета 2а после цикла

, где

.Where

represents the number of predicted cycles needed to restore Package 2a with all the radiation power used for Package 2a after the cycle

where

.

. Для каждого цикла приемник пытается восстановить Пакет 1а, основываясь на всех блоках символов данных, принятых для Пакета 1а, и не пытается восстановить Пакет 2а («Х_2а»). Приемник встречается с неблагоприятным исходом («F_1a») декодирования для Пакета 1а в каждом из циклов с 1 по

и с благоприятным исходом («S_1a») декодирования в цикле

, который является более ранним, чем цикл N _est. Приемник затем оценивает и подавляет помехи, обусловленные Пакетом 1а, пытается восстановить Пакет 2а и встречается с неблагоприятным исходом («F_2a») декодирования для Пакета 2а.Fig. 6A depicts an NI transmission with zero transmission for premature termination (Option 1 in Table 1). 6A, two new data blocks are transmitted for Packets 1a and 2a on Channels 1 and 2 in each of cycles 1 through

. For each cycle, the receiver attempts to recover Packet 1a based on all blocks of data symbols received for Packet 1a and does not attempt to recover Packet 2a (“X _2a ”). The receiver encounters an unfavorable outcome (“F _1a ”) of decoding for Packet 1a in each of the cycles 1 through

and with a favorable outcome ("S _1a ") decoding in the loop

which is earlier than the cycle N _est . The receiver then evaluates and suppresses the interference caused by Packet 1a, tries to recover Packet 2a, and encounters an unsuccessful decoding outcome (“F _2a ”) for Packet 2a.

по

приемник пытается восстановить Пакет 2а, основываясь на всех блоках символов данных, принятых для Пакета 2а, причем из блоков, принятых между циклами с 1 по

, были удалены помехи от Пакета 1а, и блоки, принятые между циклами с

по

, имеют более высокую мощность излучения. Приемник встречается с неблагоприятным исходом («F_2a») декодирования для Пакета 2а в каждом из циклов с

по

и благоприятным исходом («S_2a») декодирования в цикле

. В данном примере Пакет 2а также восстанавливается преждевременно, т. е. до цикла N _est, из-за большей мощности излучения, используемой для Пакета 2а с цикла

и вперед. Два новых пакета (Пакеты 1b и 2b) данных затем передаются по Каналам 1 и 2, начиная в цикле

. Процесс декодирования повторяется над этими пакетами.For Option 1, the transmitter uses all the radiation power for Packet 2a after Packet 1a has been restored. For each of the cycles with

by

the receiver attempts to recover Packet 2a based on all blocks of data symbols received for Packet 2a, moreover, from the blocks received between cycles 1 through

, interference from Packet 1a has been removed, and blocks received between loops with

by

have a higher radiation power. The receiver encounters an unfavorable outcome (“F _2a ”) of decoding for Packet 2a in each of the cycles with

by

and a favorable outcome ("S _2a ") of decoding in the loop

. In this example, Package 2a is also restored prematurely, i.e., before the N _est cycle, due to the higher radiation power used for Package 2a from the cycle

and forth. Two new packets (Packets 1b and 2b) of the data are then transmitted on Channels 1 and 2, starting in a loop

. The decoding process is repeated on these packets.

. Для каждого цикла приемник пытается восстановить Пакет 1а и не пытается восстановить Пакет 2а. Приемник встречается с благоприятным исходом («S_1a») декодирования для Пакета 1а в цикле

(который является более ранним, чем цикл N _est), оценивает и подавляет помехи, обусловленные Пакетом 1а, пытается восстановить Пакет 2а и встречается с неблагоприятным исходом («F_2a») декодирования для Пакета 2а. Новый короткий Пакет 1b с длиной

затем передается по Каналу 1, начиная в цикле

. Для каждого из циклов с

по

приемник пытается восстановить Пакет 1b, основываясь на всех блоках символов данных, принятых для Пакета 1b, и встречается с благоприятным исходом («S_1b») декодирования в цикле

. В данном примере Пакет 1b также восстанавливается до цикла N _est. Однако данные не передаются по Каналу 1 после цикла

, например, так как пакет с наименьшей длиной не может быть полностью передан по Каналу 1 до цикла N _est. Передатчик затем использует всю мощность излучения для Пакета 2а после того, как Пакет 1b был восстановлен.Fig. 6B depicts NR transmission with short packet transmission for premature termination (Option 2 in Table 1). 6B, two new data blocks are transmitted for Packets 1a and 2a on Channels 1 and 2 in each of cycles 1 through

. For each cycle, the receiver tries to restore Package 1a and does not try to restore Package 2a. The receiver encounters a favorable decoding outcome (“S _1a ”) for Packet 1a in the loop

(which is earlier than the N _est cycle) evaluates and suppresses interference caused by Packet 1a, attempts to recover Packet 2a, and encounters an unsuccessful decoding outcome (“F _2a ”) for Packet 2a. New short package 1b with length

then transmitted on Channel 1, starting in a loop

. For each of the cycles with

by

the receiver attempts to recover Packet 1b based on all blocks of data symbols received for Packet 1b and encounters a favorable decoding outcome (“S _1b ”) in the loop

. In this example, Packet 1b is also restored to the N _est cycle. However, data is not transmitted on Channel 1 after the cycle

, for example, since the packet with the smallest length cannot be completely transmitted on Channel 1 until the N _est cycle. The transmitter then uses all the radiation power for Packet 2a after Packet 1b has been restored.

по

приемник пытается восстановить Пакет 2а, основываясь на всех блоках символов данных, принятых для Пакета 2а, причем из блоков, принятых между циклами с 1 по

, были удалены помехи от Пакета 1а, из блоков, принятых между циклами с

по

, были удалены помехи от Пакета 1b, и блоки, принятые после цикла

, имеют более высокую мощность излучения. Приемник встречается с благоприятным исходом («S_2a») декодирования для Пакета 2а в цикле

, который в данном примере находится перед циклом N _est. Два новых пакета данных затем передаются по Каналам 1 и 2, начиная в цикле

.For each of the cycles with

by

the receiver attempts to recover Packet 2a based on all blocks of data symbols received for Packet 2a, moreover, from the blocks received between cycles 1 through

, interference from Packet 1a has been removed from blocks received between loops with

by

, interference from Packet 1b was removed, and blocks received after the loop

have a higher radiation power. The receiver encounters a favorable decoding outcome (“S _2a ”) for Packet 2a in the loop

, which in this example is before the cycle N _est . Two new data packets are then transmitted on Channels 1 and 2, starting in a loop

.

. При встрече с благоприятным исходом («S_1a») декодирования для Пакета 1а в цикле

новый длинный Пакет 1b с длиной

передается по Каналу 1, начиная в цикле

. Для каждого из циклов с

по

приемник пытается восстановить Пакет 1b, основываясь на всех блоках символов данных, принятых для Пакета 1b, и встречается с благоприятным исходом («S_1b») декодирования в цикле

, который находится после цикла N _est.Fig. 6C depicts NI transmission with long packet transmission for premature termination (Option 3 in Table 1). 6C, two new data blocks are transmitted for Packets 1a and 2a on Channels 1 and 2 in each of cycles 1 through

. When faced with a favorable outcome ("S _1a ") decoding for Packet 1a in the loop

new long Package 1b with length

transmitted on Channel 1, starting in a cycle

. For each of the cycles with

by

the receiver attempts to recover Packet 1b based on all blocks of data symbols received for Packet 1b and encounters a favorable decoding outcome (“S _1b ”) in the loop

which is after the cycle N _est .

приемник пытается восстановить Пакет 2а, основываясь на всех блоках символов данных, принятых для Пакета 2а, причем удалены помехи от Пакета 1а, и встречает неблагоприятный исход («F_2a») декодирования. В цикле

приемник пытается восстановить Пакет 2а, основываясь на всех блоках символов данных, принятых для Пакета 2а, причем из блоков, принятых между циклами с 1 по

, удалены помехи от Пакета 1а, и из блоков, принятых между циклами с

по

, удалены помехи от Пакета 1b. Приемник встречается с благоприятным исходом («S_2a») декодирования для Пакета 2а в цикле

. Два новых пакета данных затем передаются по Каналам 1 и 2, начиная в цикле

.In the loop

the receiver attempts to recover Packet 2a based on all blocks of data symbols received for Packet 2a, where interference from Packet 1a is removed, and encounters an unfavorable decoding outcome (“F _2a ”). In the loop

the receiver attempts to recover Packet 2a based on all blocks of data symbols received for Packet 2a, moreover, from the blocks received between cycles 1 through

, interference from Packet 1a is removed, and from blocks received between cycles with

by

, removed interference from Package 1b. The receiver encounters a favorable decoding outcome (“S _2a ”) for Packet 2a in the loop

. Two new data packets are then transmitted on Channels 1 and 2, starting in a loop

.

, который находится позже, чем цикл N _est, тогда (1) скорость, достигаемая для Канала 2, уменьшается с R ₂(0,N _est) до

, и (2) ВОП для Пакета 2а будет ниже, чем целевая ВОП, так как большая избыточность была передана для Пакета 2а. Улучшенные рабочие характеристики могут достигаться посредством завершения передачи для Пакета 2а после некоторого предопределенного количества циклов (

) и использования всей мощности излучения для Пакета 1b.The transmission of a new long Packet 1b over Channel 1 may have an effect on the actual speed and GP achieved for Channel 2. As noted above, N _est is the number of cycles predicted to recover Packet 2a over Channel 2 with suppressed interference from packet (s) from Channel 1 and for the target GP. If the long Packet 1b on Channel 1 is restored in a cycle

, which is later than the cycle N _est , then (1) the speed achieved for Channel 2 decreases from R ₂ (0, N _est ) to

, and (2) the GP for Package 2a will be lower than the target GP because the large redundancy has been transferred for Package 2a. Improved performance can be achieved by completing the transfer for Packet 2a after a predetermined number of cycles (

) and use of all radiation power for Package 1b.

) передачи для Пакета 2а. Целевая ВОП достигается для Пакета 2а, если он передается в течение N _est циклов (т. е.

=N _est), как указано точкой 720. ВОП для Пакета 2а постепенно снижается ниже целевой ВОП, когда более длинный Пакет 2а передается больше N _est циклов, как показано графиком 712. Целевая ВОП достигается для Пакета 1b, если он передается в течение

циклов, что происходит в цикле

, как указано точкой 722. Это предполагает, что Пакет 2а передается в течение всего этого времени. ВОП для Пакета 1b постепенно снижается ниже целевой ВОП, когда завершается более ранний Пакет 2а, и вся мощность излучения используется для Пакета 1b, как показано графиком 710. ВОП для Пакета 1b и 2а пересекаются в цикле

. Если передача для Пакета 2а завершается в цикле

, тогда может достигаться одинаковая надежность для обоих Пакетов 1b и 2а, и также повышается вероятность восстановления Пакета 1b до цикла

.7 depicts a GP schedule 710 for Package 1b and a GP schedule 712 for Package 2a depending on the number of cycles (

) transfers for Package 2a. Target GPs are achieved for Package 2a if it is transmitted during N _est cycles (i.e.

= N _est ), as indicated by point 720. The GP for Packet 2a gradually decreases below the target GP when a longer Packet 2a transmits more N _est cycles, as shown in graph 712. The target GP is reached for Packet 1b if it is transmitted within

cycles what happens in a cycle

as indicated by point 722. This assumes that Packet 2a is transmitted during all this time. The GP for Package 1b gradually decreases below the target GP when the earlier Package 2a is completed, and all the radiation power is used for Package 1b, as shown in graph 710. The GP for Pack 1b and 2a intersect in a cycle

. If the transfer for Packet 2a completes in a loop

then equal reliability can be achieved for both Packets 1b and 2a, and the likelihood of Packet 1b being restored to a cycle is also increased

.

могут использоваться различные мощности излучения для Пакетов 1b и 2а для достижения аналогичных результатов. Например,

может выбираться, основываясь на использовании большей мощности излучения для Пакета 1b и меньшей мощности излучения для Пакета 2а в течение продолжительности Пакета 1b (т. е. циклов с

по

), так что ВОП Пакетов 1b и 2а аналогичны в цикле

. В качестве другого примера, мощность излучения для Пакета 1b может постепенно увеличиваться, и мощность излучения для Пакета 2а может постепенно снижаться после цикла N _est. Различные мощности излучения могут использоваться для различных потоков данных, используя третью или четвертую схему передачи, описанные выше.Alternatively, instead of completing the transmission of Packet 2a in a loop

different radiation powers may be used for Packets 1b and 2a to achieve similar results. For example,

can be selected based on the use of a higher radiation power for Package 1b and a lower radiation power for Package 2a for the duration of Package 1b (i.e., cycles with

by

), so the GPs of Packets 1b and 2a are similar in the loop

. As another example, the radiation power for Package 1b may gradually increase, and the radiation power for Package 2a may gradually decrease after the N _est cycle. Different radiation powers can be used for different data streams using the third or fourth transmission scheme described above.

Table 2 shows some of the available options with the transmission of the long Packet 1b, which can be extended for the cycle N _est .

table 2 Option Description BUT If Package 1b is being restored to Package 2a, then any one of the options shown in Table 1 may be selected. AT

(например,

=

), ожидать восстановление Пакета 1b, затем предпринимать попытку восстановления Пакета 2а с подавленными Пакетами 1а и 1b.Complete the transfer of Packet 2a after a predetermined number of cycles

(eg,

=

), wait for the restoration of Package 1b, then attempt to restore Package 2a with suppressed Packets 1a and 1b.

FIG. 8 depicts an example state diagram 800 that may be supported by a transmitter and a receiver for an IFR ordered transmission scheme. The state diagram 800 includes a synchronization state 810, a new packet transmission state 820, and a zero transmission state 830. In synchronization state 810, two new packets (Packets 1a and 2a) of data are transmitted on Channels 1 and 2, starting in the same cycle. As expected, these two packets will be restored in N _est cycles if the speed selection is sufficiently accurate.

The state diagram transitions from synchronization state 810 to new packet transmission state 820 if Packet 1a on Channel 1 is restored earlier N _est cycles and a new short or long packet (Packet 1b) of data is transmitted on Channel 1. In state 820, the receiver attempts to recover Packet 1b on Channel 1 and does not try to restore Packet 2a on Channel 2 until Packet 1b is restored and the interference from Packet 1b is suppressed. The state diagram remains in state 820 if Packet 1b is not restored, or if Packet 1b is restored, and a new packet (Packet 1c) of data is transmitted on Channel 1. The state diagram passes from state 820 back to state 810 if the packets are on both Channels 1 and 2 restored.

The state diagram transitions from synchronization state 810 to the zero transmission state 830 if Packet 1a on Channel 1 is restored earlier than N _est cycles and nothing is transmitted on Channel 1. The state diagram also passes from state 820 to state 830 if the current packet is on Channel 1 restored and nothing is transmitted on Channel 1. In state 830, the receiver tries to restore Packet 2a on Channel 2, and interference from all packets restored on Channel 1 is suppressed. The state diagram remains in state 830 if Packet 2a on Channel 2 is not restored anovlen and transitions back to state 810, if the packet 2a restored.

An ordered transmission scheme with IFR can provide good performance if the speed selection is accurate enough so that the restoration of later data streams is not unduly delayed.

B. Transfer scheme with IFR and cyclic repeat

In the transmission scheme with IFR and cyclic repetition, N _D data streams are restored by cyclic repetition over the data streams, so that the data stream that is most likely to be decoded correctly is the first to be restored. Initially, N _D rates for N _D data streams are selected, and N _D data packets are transmitted on N _D parallel channels. The choice of speed can be rough, and the packet size can be chosen, so it is expected that all data packets will be restored in N _est cycles. Whenever a data packet for a data stream is restored, a new packet is transmitted for that data stream, and the receiver attempts to decode the data packet for the next data stream, as described below.

и с благоприятным исходом («S_1a») декодирования в цикле

. Новый Пакет 1b затем передается по Каналу 1, начиная в цикле

. Пакет 1b имеет длину N _est и обрабатывается, основываясь на более высокой скорости, которая оценивается в цикле

и при предположении, что будут подавлены помехи от Канала 2.Figa depicts a transmission with NI with a transmission scheme with IFR and cyclic repetition. In FIG. 9A, two new data blocks are transmitted starting in cycle 1 for Packets 1a and 2a on Channels 1 and 2. Packet 1a, as indicated, is restored first and processed based on a lower speed due to interference from Packet 2a. Packet 2a, as indicated, is restored later and processed based on the higher speed achieved by suppressing interference from Packet 1a. Packets 1a and 2a have a length of N _est (i.e., they are expected to recover in N _est cycles). For each cycle, the receiver attempts to recover Packet 1a based on all blocks of data symbols received for this packet, and does not attempt to recover Packet 2a (“X _2a ”). The receiver encounters an unfavorable outcome (“F _1a ”) of decoding for Packet 1a in each of the cycles 1 through

and with a favorable outcome ("S _1a ") decoding in the loop

. The new Packet 1b is then transmitted on Channel 1, starting in a loop

. The packet 1b has a length of N _est and is processed based on a higher speed, which is estimated in the cycle

and under the assumption that interference from Channel 2 will be suppressed.

приемник оценивает и подавляет помехи, обусловленные Пакетом 1а, пытается восстановить Пакет 2а и встречается с неблагоприятным исходом («F_2a») декодирования для Пакета 2а. Для каждого из циклов с

по

приемник пытается восстановить Пакет 2а, основываясь на всех блоках символов данных, принятых для этого пакета, причем из блоков, принятых в циклах с 1 по

, удалены помехи от Пакета 1а, и из блоков, принятых в циклах с

по

, удалены помехи от Пакета 1b. Приемник встречается с неблагоприятным исходом («F_2a») декодирования для Пакета 2а в каждом из циклов с

по

и с благоприятным исходом («S_2a») декодирования в цикле

. Новый Пакет 2b затем передается по Каналу 2, начиная в цикле

. Пакет 2b имеет длину N _est и обрабатывается, основываясь на большей скорости, которая оценивается в цикле

и при предположении, что будут подавлены помехи от Канала 1.In the loop

the receiver evaluates and suppresses interference caused by Packet 1a, attempts to recover Packet 2a, and encounters an unfavorable decoding outcome (“F _2a ”) for Packet 2a. For each of the cycles with

by

the receiver attempts to recover Packet 2a based on all blocks of data symbols received for this packet, and from the blocks received in cycles 1 through

, interference from Packet 1a is removed, and from blocks received in cycles with

by

, removed interference from Package 1b. The receiver encounters an unfavorable outcome (“F _2a ”) of decoding for Packet 2a in each of the cycles with

by

and with a favorable outcome ("S _2a ") decoding in the loop

. The new Packet 2b is then transmitted on Channel 2, starting in a loop

. The packet 2b has a length of N _est and is processed based on the higher speed, which is estimated in the cycle

and under the assumption that interference from Channel 1 will be suppressed.

приемник оценивает и подавляет помехи, обусловленные Пакетом 2а, пытается восстановить Пакет 1b и встречается с неблагоприятным исходом («F_1b») декодирования для Пакета 1b. Для каждого из циклов с

по

приемник пытается восстановить Пакет 1b, основываясь на всех блоках символов данных, принятых для этого пакета, причем из блоков, принятых в циклах с

по

, удалены помехи от Пакета 2а, и из блоков, принятых в циклах с

по

, удалены помехи от Пакета 2b. Приемник встречается с благоприятным исходом («S_1b») декодирования для Пакета 1b в цикле

. Приемник пытается восстановить последующие пакеты на Каналах 1 и 2 аналогичным образом.In the loop

the receiver evaluates and suppresses interference caused by Packet 2a, attempts to recover Packet 1b, and encounters an unfavorable decoding outcome (“F _1b ”) for Packet 1b. For each of the cycles with

by

the receiver attempts to recover Packet 1b based on all blocks of data symbols received for this packet, and from the blocks received in cycles with

by

, interference from Packet 2a is removed, and from blocks received in cycles with

by

, removed interference from Package 2b. The receiver encounters a favorable decoding outcome (“S _1b ”) for Packet 1b in the loop

. The receiver attempts to recover subsequent packets on Channels 1 and 2 in a similar fashion.

. При восстановлении Пакета 1а в цикле

приемник пытается восстановить Пакет 2а на Канале 2 в циклах с

по

. При восстановлении Пакета 2а в цикле

приемник пытается восстановить Пакет 1b на Канале 1 в циклах с

по

. Приемник пытается восстановить последующие пакеты на Каналах 1 и 2 аналогичным образом.Figv depicts the recovery order of data streams for transmission schemes with IFR and cyclic repetition. The receiver attempts to recover Packet 1a on Channel 1 in cycles 1 through

. When restoring Package 1a in a cycle

the receiver attempts to recover Packet 2a on Channel 2 in cycles with

by

. When restoring Package 2a in a cycle

the receiver attempts to recover Packet 1b on Channel 1 in cycles with

by

. The receiver attempts to recover subsequent packets on Channels 1 and 2 in a similar fashion.

Generally speaking, the receiver may attempt to recover packets sent over N _D parallel channels based on the probability of recovery of these packets. The likelihood of recovering a packet sent on each parallel channel depends on various factors, such as (1) UWB after processing, achieved for the parallel channel with linear detection, and (2) the number of data symbol blocks already received for the parallel channel. In each cycle, the receiver may attempt to recover only the packet sent on the parallel channel, which is most likely to be restored in this cycle. Alternatively, the receiver may attempt to recover packets on all N _D parallel channels, one packet at a time, starting with the parallel channel being restored with the highest probability and ending with the parallel channel being restored with the lowest probability. If multiple parallel channels have the same probability of recovery, then the receiver can select one parallel channel (for example, at a time, randomly) for recovery.

The receiver can cyclically repeat over N _D parallel channels if (1) these channels reach similar SINRs after processing with linear detection, and (2) packets for these channels have the same length. As an example, consider the case in which N _D = 4 and four new packets are transmitted on four parallel channels, starting in cycle 1. In each cycle, the receiver may attempt to recover a packet sent on each parallel channel, based on all blocks of data symbols, accepted for this channel. The receiver can recover the first packet transmitted on, for example, Channel 2, and then will evaluate and suppress the interference caused by this packet. After that, in each cycle, the receiver may attempt to recover the packet sent on each of Channels 1, 3, and 4, based on all blocks of data symbols received for this packet. The receiver can recover the next packet transmitted on, for example, channel 3, and then will evaluate and suppress the interference caused by this packet. After that, in each cycle, the receiver may attempt to recover the packet sent on each of Channels 1 and 4, based on all blocks of data symbols received for this packet. The receiver can recover the next packet transmitted on, for example, Channel 1, and then will evaluate and suppress the interference caused by this packet. After that, in each cycle, the receiver may attempt to recover the packet sent on Channel 4 based on all blocks of data symbols received for this packet. After that, the receiver can simply cycle through four parallel channels in a predetermined order, i.e., channels 2, 3, 1, 4, then back 2, etc. This predetermined order is selected based on the order in which packets are restored for four parallel channels. Whenever a data packet is restored on the current parallel channel (the channel for which recovery is first attempted in a loop), a new data packet is transmitted on this channel, and this packet is then restored last.

The IFR and cyclic repeat transmission scheme can provide good performance even with a rough selection of speed. This is because NR transmission is effectively achieved for each data stream, as shown in FIGS. 9A and 9B. IFR and cyclic repeat transmission can provide good performance even if the channel mode changes rapidly. In addition, the implementation of the transmission scheme with IFR and cyclic repetition is relatively simple, since (1) the transmitter and receiver do not need to store status information for what is currently being transmitted, and (2) there is no need to change the packet sizes to fit specific time windows , as is the case for an ordered transmission scheme with IFR.

на фиг.6В и 6С).Schemes of ordered transmission from IFR and transmission from IFR and cyclic repeat are two exemplary schemes. Other transmission schemes may also be implemented for interdependent parallel channels. As an example, in the “hybrid” IFR scheme, the receiver tries to recover each of the data packets currently being transmitted for N _D data streams based on all blocks of data symbols received for this packet (that is, the receiver does not miss decoding no packet). Each block of data symbols for each packet has (1) suppressed interference from the recovered packets and (2) interference from packets that are not yet recovered. The SINR for each data packet, therefore, may vary throughout the packet depending on the degree of interference suppression, if any, for the packet. The IFR hybrid transmission scheme may also be used in combination with the IFR ordered transmission scheme and the IFR and cyclic repeat transmission scheme. For example, the receiver may try to recover a data packet on Channel 2 in each cycle after the first data packet on Channel 1 has been received and canceled (for example, for each cycle after the cycle

on figv and 6C).

3. Transmitter

10 is a block diagram of an embodiment of TX data processor 120 at transmitter 110. TX data processor 120 includes N _D TX channel data processors 1010a through 1010n for N _D data streams. Each TX channel data processor 1010 receives a corresponding data stream, processes each data packet in the data stream based on the speed selected for the stream, and provides a set of data symbol blocks for the packet. 11 illustrates the processing of one data packet by one data processor 1010.

In each TX channel data processor 1010, a cyclic redundancy check (CRC) generator 1012 receives a data packet in a data stream processed by the data processor 1010, generates a CRC value for the data packet, and attaches the CRC value to the end of the data packet to form a formatted packet. The CEC value is used by the receiver to check whether the packet is decoded correctly or with an error. Instead of the CEC, other error detection codes may also be used. The forward error correction encoder 1014 then encodes the formatted packet in accordance with the coding scheme or coding rate indicated by the selected rate, and provides an encoded packet or “codeword”. Encoding improves the reliability of packet transmission. The UCO encoder 1014 may implement a block code, a convolutional code, a turbo code, some other code, or a combination thereof. 11, an encoded packet includes a first portion with systematic bits for a formatted packet, a second portion with pilot bits from a first constituent encoder of a turbo encoder, and a third portion with pilot bits from a second constituent encoder of a turbo encoder.

Separation unit 1016 receives and splits the encoded packet into N _B encoded subpackets, where N _B may depend on the selected rate and is indicated by a separation control signal from controller 180. The first encoded subpacket typically contains all systematic bits and zero or more control bits. This allows the receiver to recover the data packet with only the first encoded subpacket under favorable channel mode. Other N _B -1 encoded subpackets contain the remaining control bits, with each subpacket typically containing control bits taken over the entire data packet.

Channel interleaver 1020 includes N _B block interleavers 1022a-1022nb that receive N _B coded subpackets from a partitioning node 1016. Each block interleaver 1022 interleaves (i.e., reorders) the code bits for its subpacket in accordance with the interleaving scheme and provides a subpacket with alternation. Interleaving provides time diversity, frequency diversity and / or spatial diversity for code bits. Multiplexer 1024 is connected to all N _B block interleavers 1022a-1022nb and provides N _B interleaved subpackets, one at a time, and if controlled by a transmission control signal from NI from controller 180. Multiplexer 1024 first outputs a subpacket interleaved from block interleaver 1022a, then the next subpacket is interleaved from an interleaver 1022b of blocks, etc., and the last subpacket is interleaved from an interleaver 1022nb of blocks. The multiplexer 1024 issues the next subpacket with interleaving, if the received NPP for the data packet. All N _B block interleavers 1022a-1022nb may be reset whenever an AC is received.

The symbol mapper 1026 receives the interleaved subpackets from the channel interleaver 1020 and maps the interleaved data in each subpacket to modulation symbols. Character mapping is performed in accordance with the modulation scheme indicated by the selected rate. Character mapping can be performed by (1) grouping sets of B bits, forming B-bit binary values, where B≥1, and (2) mapping each B-bit binary value to a point in a signal constellation having 2 ^V points. This signal constellation corresponds to the selected modulation scheme, which can be binary phase shift keying (DPS), quadrature phase shift keying (QPS), 2 ^V phase shift keying (2 ^V- FM), 2 ^V phase quadrature amplitude modulation (2 ^V- QAM) etc. As used herein, a “data symbol” is a modulation symbol for data, and a “pilot symbol” is a modulation symbol for a pilot. The symbol mapper 1026 provides a data symbol block for each encoded subpacket, as shown in FIG. 11.

]. Каждый символ s _i данных, где i=1 … N _SYM, получается посредством отображения В кодовых битов следующим образом: s _i=map ( b _i), где b _i=[b _i,1 b _i,2 … b _i,B].For each data packet, TX channel data processor 1010 provides N _B data symbol blocks, which together include N _SYM data symbols and may be referred to as { s } = [ s ₁ s ₂ ...

]. Each data symbol s _i , where i = 1 ... N _SYM , is obtained by mapping B code bits as follows: s _i = map ( b _i ), where b _i = [ b _{i, 1} b _{i, 2} ... b _{i, B} ].

12 shows a block diagram of an embodiment of TX spatial processor 130 and transmitting node 132. TX spatial processor 130 receives and processes N _D data symbol streams from TX data processor 120 and provides N _T transmission symbol streams to transmitting node 132. Processing by spatial TX processor 130 depends on the particular transmission scheme selected for use.

In the TX spatial processor 130, the matrix multiplier 1220 receives up to N _D data symbol blocks (represented by the data vector s ) for each slot. The node 1220 performs the multiplication of the matrices of the data vector s by (1) the identity matrix V for the second transmission scheme and (2) by the base transmission matrix M for the third transmission scheme. Node 1220 simply skips the data vector s for other transmission schemes. A multiplexer / demultiplexer (MUX / DEMUX) 1222 receives symbols from a node 1220 and provides these symbols to appropriate transmit antennas and subbands (if OFDM is used). A multiplexer / demultiplexer 1222 also multiplexes pilot symbols (eg, such as time division multiplexed transmission (MBP)) and provides N _T transmission symbol sequences for N _T transmit antennas in each slot. Each transmission symbol sequence is designed to transmit one transmit antenna in one slot.

The transmitter unit 132 includes N _T OFDM modulators 1230a-1230t and N _T radio frequency (RF) TX nodes 1236a-1236t for N _T transmit antennas. For a single-carrier MIMO system, OFDM modulators 1230 are not required, and TX spatial processor 130 provides N _T transmission symbol sequences directly to TX RF nodes 1236a-1236t. For an OFDM system with an MIMO, TX spatial processor 130 provides N _T transmission symbol sequences to OFDM modulators 1230a-1230t. Each OFDM modulator 1230 includes an Inverse Fast Fourier Transform (FFT) node 1232 and a cyclic prefix generator 1234. Each OFDM modulator 1230 receives a corresponding sequence of transmit symbols from TX spatial processor 130 and groups each set of N _F transmit symbols and zero signal values for N _F subbands. (The subbands not used for data transmission are filled with zeros.) OBPF node 1232 converts each set of N _F transmission symbols and zeros to the time domain using an N _F -point inverse fast Fourier transform and outputs a corresponding transformed symbol that contains N _F chips. The cyclic prefix generator 1234 repeats a portion of each transformed symbol to obtain a corresponding OFDM symbol that contains N _F + N _cp chips. The repeated part is referred to as a cyclic prefix, and N _cp represents the number of repeated chips. The cyclic prefix ensures that the OFDM symbol retains its orthogonal properties in the presence of a delay spread in multipath propagation caused by frequency selective fading (i.e., a frequency response that is not flat). A cyclic prefix generator 1234 provides an OFDM symbol sequence for a transmission symbol sequence.

The RF nodes 1236a-1236t TX receive and bring to a certain state N _T OFDM / transmit symbol sequences to generate N _T modulated signals that are transmitted from N _T transmit antennas 1240a-1240t, respectively.

4. Receiver

FIG. 13 is a block diagram of a receiver 150a, which is one embodiment of a receiver 150 of FIG. At a receiver 150a, the N _R receive antennas 1310a-1310r receive N _T modulated signals transmitted by the transmitter 110 and provide N _R received signals to N _R RF nodes 1312a-1312r RX, respectively, at a receiving node 154. Each RF node 1312 RX leads to a certain state and digitizes its received signal and produces a stream of characters / chips. An OFDM demodulator 1314a-1314r is not required for a system with an MIMO and a single carrier, and each RF node 1312 RX provides a symbol stream directly to the corresponding demultiplexer 1316. For an OFDM system with an MIMO and one carrier, each RF node 1312 RX provides a chip stream to the corresponding demodulator 1314 OFDM . Each OFDM demodulator 1314 demodulates the OFDM of its chip stream by (1) removing a cyclic prefix in each received OFDM symbol to obtain a received transformed symbol and (2) converting each received transformed symbol to the frequency domain using a fast Fourier transform (FFT) to obtain N _F received symbols for N _F subbands. For both systems, demultiplexers 1316a-1316r receive N _R symbol streams from RF RX nodes 1312 or OFDM demodulators 1314, provide N _R received symbol sequences (for data) for each slot to RX spatial processor 160a, and receive received pilot symbols to node 172 channel ratings.

The RX spatial processor 160a includes a detector 1320 and a multiplexer / demultiplexer 1322. The detector 1320 performs spatial or space-time processing (or “detection”) of N _R received symbol sequences to obtain N _T detected symbol sequences for each interval. Each detected symbol is an estimate of the data symbol transmitted by the transmitter. The detector 1320 may implement the OS detector shown in equation (2), the ISCED detector shown in equation (3), the zero-force linear detector shown in equation (4), the ISCED linear corrector, the decision feedback corrector, or some another detector / corrector. Detection may be performed based on an estimate of the channel response matrix H or the effective channel response matrix H _eff = HM depending on whether or not the data symbols are pre-multiplied by the base transmission matrix M at the transmitter. For an OFDM system with an MIMO, the receiver detects separately for each of the subbands used for data transmission.

. Мультиплексор/демультиплексор 1322 принимает N _T последовательностей детектированных символов и обеспечивает детектированные символы для N _D блоков детектированных символов для N _D потоков данных. Каждый блок детектированных символов представляет собой оценку блока символов данных, переданного передатчиком.For each interval, detector 1320 provides N _T sequences of detected symbols that correspond to N _T elements

. A multiplexer / demultiplexer 1322 receives N _T detected symbol sequences and provides detected symbols for N _D detected symbol blocks for N _D data streams. Each detected symbol block is an estimate of a data symbol block transmitted by the transmitter.

детектированных символов. Селектор 174 скорости (который реализуется контроллером 180 для варианта осуществления, показанного на фиг.13) выполняет выбор скорости, основываясь на оценках канала. Таблица 184 соответствия (ТС) хранит набор скоростей, поддерживаемых системой с МВМВ, и набор значений параметров для каждой скорости (например, скорость передачи данных, размер пакета, схема кодирования или скорость кодирования, схема модуляции и т. д. для каждой скорости). Селектор 174 скорости обращается к ТС 184 за информацией, используемой для выбора скорости.The channel estimator 172 estimates the channel response matrix H for the MIMO channel and the noise floor at the receiver (for example, based on the received pilot symbols) and provides channel estimates to the controller 180. In the controller 180, the matrix calculator 176 outputs the detector characteristic W (which can be W _mrc , W _mmse , W _zf or Σ ^-1 U ^H ), based on the estimated channel response matrix as described above, and provides the detector response to detector 1320. Detector 1320 pre-multiplies the vector r of received symbols by the character the logic of the W detector to obtain a vector

detected characters. A speed selector 174 (which is implemented by the controller 180 for the embodiment shown in FIG. 13) performs speed selection based on channel estimates. Correspondence table (TS) 184 stores a set of speeds supported by the MIMO system and a set of parameter values for each speed (for example, data transfer rate, packet size, coding scheme or coding rate, modulation scheme, etc. for each speed). The speed selector 174 accesses the vehicle 184 for information used to select a speed.

FIG. 14 is a block diagram of an RX data processor 170a, which is one embodiment of an RX data processor 170 of FIGS. 1 and 13. RX data processor 170a includes N _D RX channel data processors 1410a-1410n for N _D data streams . Each RX channel data processor 1410 receives and processes a corresponding detected symbol stream and provides a decoded data stream.

In each RX channel data processor 1410, the symbol demapper 1430 receives blocks of detected symbols from the RX spatial processor 160a, one block at a time. For each block of detected symbols, the symbol demapper 1430 demodulates the detected symbols in accordance with the modulation scheme used for this block (as indicated by the demodulation control signal from controller 180), and provides a block of demodulated data to the channel de-interleaver 1440. The channel de-interleaver 1440 includes a demultiplexer 1442 and N _B block de-interleavers 1444a-1444nb. Before receiving a new data packet, the de-interleavers 1444a-1444nb of the blocks are initialized with erasures. Erasure is a value that replaces a missing code bit (that is, one that has not yet been received) and is given the corresponding weight in the decoding process. A multiplexer 1442 receives blocks of demodulated data from a symbol demapper 1430 and provides each block of demodulated data to a proper block de-interleaver 1444. Each block de-interleaver 1444 de-interleaves the demodulated data in its block in a manner that is complementary to the interleaving performed on the transmitter for that block.

For independent parallel channels, whenever a new block of data symbols is received from the transmitter for a data packet on a parallel channel, decoding can be performed again on all blocks received for this packet. The reassembly node 1448 generates a data packet with deinterleaving for subsequent decoding. A de-interleaved data packet contains (1) de-interleaved data blocks for all data symbol blocks received for the current packet to be decoded, and (2) erasing for data symbol blocks not received for the current packet. The reassembly unit 1448 performs reassembly in a complementary manner to the separation performed by the transmitter, as indicated by the reassembly control signal from the controller 180. The UCO decoder 1450 decodes the data packet with deinterleaving in addition to the UCO encoding performed on the transmitter, as indicated by the decoding control signal from controller 180. For example, a Viterbi turbo decoder or decoder can be used for the UCO decoder 1450 if the transmitter is turbo encoded or convolutional encoded sc, respectively. The UCO decoder 1450 issues a decoded packet for the current packet. The CEC verification section 1452 checks the decoded packet to determine if the packet is decoded correctly or in error, and provides the status of the decoded packet.

FIG. 15 is a block diagram of a receiver 150b that implements the RFP technique and is another embodiment of receiver 150 of FIG. Receiver 150b includes an RX spatial processor 160b and an RX data processor 170b, which together implement N _D sequential (i.e., cascaded) receiver processing steps. Each of steps 1 through N _D −1 includes a detector 1510, an interference suppressor 1520, an RX channel data processor 1530, and a TX channel data processor 1540. The last stage N _D includes only the detector 1510n and the RX channel data processor 1530n.

For stage 1, the detector 1510a performs detection on N _R sequences of received symbols for each interval and provides a block of detected symbols for a data packet (Packet x ) in the data stream restored to stage 1. The RX channel data processor 1530a demodulates, deinterleaves, and decodes all blocks of the detected characters accepted for Package x . If Packet x is decoded correctly, then TX channel data processor 1540a encodes, interleaves, and modulates Packet x to obtain a sequence of remodulated symbols, which is an estimate of the data symbol sequence for Packet x . TX channel data processor 1540a performs the same processing as that performed by the transmitter for Packet x . The jammer 1520a receives and spatially processes the sequence of remodulated symbols in the same manner as those performed by the transmitter 110 for Packet x to obtain N _T transmission symbol sequences that contain only character components for Packet x . The jammer 1520a further processes the transmission symbol sequences using a channel response matrix to obtain interference components caused by Packet x . Interference components are then subtracted from N _R received symbol sequences to obtain N _R modified symbol sequences that are fed to stage 2.

Each of steps 2 through N _D -1 performs the same processing as step 1, although in N _R sequences of modified symbols from the previous step instead of N _R sequences of received symbols. Stage N _D performs detection and decoding on N _R sequences of modified symbols from stage N _D -1 and does not evaluate and suppress interference.

Each of the detectors 1510a-1510n may implement an OS detector, an ISCED detector, a zero-force linear detector, an ISCED linear corrector, a critical feedback corrector, or some other detector / corrector. Each RX channel data processor 1530 may be implemented as shown in FIG. 14, and each TX channel data processor 1540 may be implemented as shown in FIG. 10. As described above, the receiver may attempt to restore the data packet for a later stage only after the data packets for the earlier stages are restored. The buffers (not shown in FIG. 15) will then store characters from each step until they are ready for processing by later steps.

For both systems with MIMO and one carrier, and OFDM systems with MIMO, the receiver and / or transmitter can evaluate the received SINR or SINR after processing (depending on whether or not the IFR is used) for N _D parallel channels and select the appropriate speed for data transmission on each parallel channel. The choice of speed can be performed in various ways. In one speed selection scheme, the speed for each parallel channel is selected based on the SIRB necessary for an equivalent system with a channel with an ABHS (i.e., with a flat frequency response) to support the average spectral efficiency calculated for the parallel channel. This rate selection scheme is described in detail in US Patent Application No. 10/176 567, entitled “Rate Control for Multi-Channel Communication Systems,” filed June 20, 2002. In another speed selection scheme, the speed for each parallel channel is selected based on the operating SIRB calculated for the parallel channel, based on the average SINR after processing for the parallel channel and the SINR offset. The highest speed with the required SINR (in the channel with ABGS), which is less than or equal to the working SINR, is selected for the parallel channel. This rate selection scheme is described in detail in US Patent Application No. 10/394 529, entitled “Transmission Mode Selection for Data Transmission in a Multi-Channel Communication System”. ), filed March 20, 2003

The IFR transmission methods described in this document can be implemented in a system with full duplex frequency division (DLC) and in a system with full duplex division of time (AED). For a system with HDF, the forward channel with MIMO and the feedback channel use different frequency bands and are likely to detect different channel modes. In this case, the receiver can estimate N _D parallel channels, select speeds for parallel channels, and send back the selected speeds, as shown in FIGS. 1-3. For a system with DRV, the forward channel with MIMO and the feedback channel share the same frequency band and are likely to detect similar channel modes. In this case, the transmitter can estimate N _D parallel channels based on the pilot signal sent by the receiver and selects the rates for parallel channels. Channel estimation and speed selection can thus be performed by a receiver, transmitter, or both.

The transmission methods with NI described in this document can be implemented by various means. For example, these methods may be implemented in hardware, software, or a combination thereof. For hardware implementation, the processing nodes used in the transmitter for transmission from NI can be implemented in one or more specialized integrated circuits (specialized ICs), digital signal processors (DSP), digital signal processing devices (DSP), programmable logic devices (PLU) , programmable gate arrays (FDAs), processors, controllers, microcontrollers, microprocessors, other electronic components designed to perform the functions described in this document, or combinations. The processing nodes used in the receiver to receive transmission from NI can also be implemented in one or more specialized ISs, DSPs, UTsOS, PLU, FDA, processors, controllers, etc.

For software implementation, transmission methods from NI can be implemented using modules (for example, procedures, functions, etc.) that perform the functions described in this document. Software codes can be stored in a memory node (for example, memory nodes 142 and 182 in FIG. 1) and executed by a processor (for example, controllers 140 and 180). The memory node can be implemented inside the processor or outside the processor, in which case it can be connected with the possibility of organizing communication with the processor using various means that are known in the art.

Headings are included in this document for reference and assistance in locating specific sections. These headings are not intended to limit the scope of the principles described herein, and these principles may be applicable in other sections throughout the specification.

The foregoing description of the disclosed embodiments is provided so that any person skilled in the art can make or use the present invention. Various modifications to these embodiments are readily apparent to those skilled in the art, and the general principles defined herein can be applied to all embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited by the embodiments shown herein, but should be within the broadest scope consistent with the principles and new features described herein.

Claims (63)

1. A method of performing transmission with increasing redundancy (NI) in a wireless communication system with many inputs and many outputs (MVMV), containing
processing the first data packet to obtain a first plurality of symbol blocks;
processing a second data packet to obtain a second plurality of symbol blocks;
transmitting the first plurality of symbol blocks, one symbol block at a time, on a first parallel channel to a receiver;
transmitting a second plurality of symbol blocks, one symbol block at a time, on a second parallel channel to a receiver;
prematurely terminating the transmission of the first plurality of symbol blocks if the first data packet is restored by the receiver with a smaller amount than all of the first plurality of symbol blocks; and prematurely terminating the transmission of the second plurality of symbol blocks if the second data packet is restored by the receiver with a smaller number than all of the second plurality of symbol blocks. 2. The method according to claim 1, additionally containing
processing a third data packet to obtain a third plurality of symbol blocks;
transmitting a third plurality of symbol blocks, one symbol block at a time, on a third parallel channel to a receiver; and prematurely terminating the transmission of the third plurality of symbol blocks if the third data packet is restored by the receiver with less than all of the third plurality of symbol blocks. 3. The method according to claim 1, additionally containing
receiving indications that the first data packet has been restored;
bandwidth estimation for the first and second parallel channels with no transmission on the first parallel channel until the second data packet is restored;
estimating the throughput for the first and second parallel channels with the transmission of a new data packet on the first parallel channel after the first data packet; and
transmission of a new data packet on the first parallel channel if the throughput with transmission on the first parallel channel is greater than throughput with no transmission on the first parallel channel. 4. The method according to claim 1, additionally containing
receiving indications that the first data packet has been restored; and
lack of transmission of data packets on the first parallel channel until a second data packet is restored. 5. The method according to claim 4, in which the symbol blocks for the second data packet are transmitted with full radiation power after completion of the transmission of the first plurality of symbol blocks for the first data packet. 6. The method according to claim 1, additionally containing
receiving indications that the first data packet has been restored;
processing a third data packet to obtain a set of at least one character block for the third data packet; and transmitting a set of at least one character block, one character block at a time, on the first parallel channel. 7. The method according to claim 6, in which it is expected that the third data packet will be restored by the receiver at or before the time when the second data packet is expected to be restored. 8. The method according to claim 6, in which it is expected that the third data packet will be restored by the receiver after the point in time when the restoration of the second data packet is expected. 9. The method of claim 8, further comprising completing the transmission of the second plurality of symbol blocks after a predetermined number of symbol blocks. 10. The method according to claim 6, further comprising increasing the radiation power for the third packet and reducing the radiation power for the second packet at or after the time when the second data packet is expected to be restored. 11. The method according to claim 1, further comprising receiving an indication that the first data packet has been restored;
processing a third data packet to obtain a third plurality of symbol blocks for a third data packet;
transmitting a third plurality of symbol blocks, one symbol block at a time, on a first parallel channel after the first data packet;
receiving indications that the second data packet has been restored;
processing a fourth data packet to obtain a fourth plurality of symbol blocks; and
transmitting a fourth plurality of symbol blocks, one symbol block at a time, on a second parallel channel after the second data packet. 12. The method according to claim 1, additionally containing
receiving a first speed for the first parallel channel and a second speed for the second parallel channel, and in which the first and second data packets are processed in accordance with the first and second speeds, respectively. 13. The method according to item 12, in which the processing of the first data packet includes
encoding the first data packet in accordance with the coding scheme indicated by the first rate to obtain an encoded packet,
dividing an encoded packet into a plurality of encoded subpackets, and
modulating a plurality of encoded subpackets in accordance with a modulation scheme indicated by a first rate to obtain a first plurality of symbol blocks. 14. The method according to claim 1, in which one block of characters in the first set of blocks of characters includes all systematic bits for the first data packet and transmitted first for the first data packet. 15. The method according to claim 1, further comprising receiving at least one block of symbols selected from the first and second sets of symbol blocks for transmission in one time slot on the first and second parallel channels; and spatial processing of at least one block of symbols by means of a basic transmission matrix to obtain a plurality of transmission symbol sequences for a plurality of transmit antennas. 16. The method according to claim 1, in which the first and second parallel channels are formed so that a similar signal-to-noise-noise ratio (SNR) is achieved after linear detection at the receiver. 17. The method according to claim 1, in which the first and second parallel channels correspond to the first and second transmitting antennas in the transmitter in the system with MIMO. 18. The method according to claim 1, in which the first and second parallel channels correspond to the first and second spatial channels in the system with MIMO. 19. The method according to claim 1, in which the MIMO system implements orthogonal frequency division multiplexing (OFDM), and in which each of the first and second parallel channels is formed with multiple subbands and multiple transmit antennas. 20. The method of claim 19, wherein the plurality of parallel channels are formed by cyclically repeating diagonally across the plurality of subbands of the plurality of transmit antennas, the plurality of parallel channels including a first and a second parallel channel. 21. The method according to claim 1, in which the system with MIMO implements multiple access with orthogonal frequency division multiplexing (OFDM), and in which each packet is transmitted over a set of subbands available for data transmission. 22. The method according to claim 1, in which a plurality of data packets are processed and transmitted simultaneously on a plurality of parallel channels, in which the transmission of symbol blocks for each data packet terminates prematurely if the data packet is restored by the receiver with a smaller number than all symbol blocks generated for a data packet, and in which the total radiation power is distributed over data packets that have not yet been completed. 23. A transmitter that serves to perform transmission with increasing redundancy (NI) in a wireless communication system with many inputs and multiple outputs (MIMO), containing
a data processor configured to process a first data packet to obtain a first plurality of symbol blocks and to process a second data packet to obtain a second plurality of symbol blocks; and
a controller configured to
initiating the transmission of the first plurality of symbol blocks, one symbol block at a time, on the first parallel channel to the receiver, initiating the transmission of the second plurality of symbol blocks, one symbol block at a time, on the second parallel channel to the receiver, prematurely completing the transmission of the first plurality of symbol blocks, if the first data packet is restored by the receiver with a smaller number than all of the first plurality of symbol blocks, and the transmission of the second plurality of symbol blocks is completed prematurely If the second data packet by the receiver is recovered from fewer than all of the second plurality of symbol blocks. 24. The transmitter of claim 23, wherein the controller is further configured to
receiving an indication that the first data packet has been restored;
bandwidth estimates for the first and second parallel channels with no transmission on the first parallel channel until the second data packet is restored;
bandwidth estimates for the first and second parallel channels with the transmission of a new data packet on the first parallel channel after the first data packet; and
initiating the transmission of a new data packet on the first parallel channel, if the bandwidth with transmission on the first parallel channel is greater than the bandwidth with no transmission on the first parallel channel. 25. The transmitter of claim 23, wherein the data processor is further configured to process a third data packet to obtain a third plurality of symbol blocks, and wherein the controller is further configured to initiate transmission of a third plurality of symbol blocks, one symbol block at a time, the first parallel channel upon receipt of an indication that the first data packet has been restored. 26. The transmitter of claim 23, wherein the data processor is configured to
encoding the first data packet in accordance with the encoding scheme indicated by the speed selected for the first parallel
channel to receive an encoded packet,
dividing the encoded packet into a plurality of encoded subpackets, and
modulating a plurality of coded subpackets in accordance with a modulation scheme indicated by a rate to obtain a first plurality of symbol blocks. 27. The transmitter according to item 23, further comprising a spatial processor configured to receive at least one block of characters selected from the first and second sets of blocks of characters for transmission in one time interval on the first and second parallel channels and with the possibility of spatial processing at least one block of symbols by means of a basic transmission matrix to obtain a plurality of transmission symbol sequences for a plurality of transmit antennas. 28. A device configured to perform transmission with increasing redundancy (NI) in a wireless communication system with many inputs and multiple outputs (MIMO), containing
means for processing the first data packet to obtain a first plurality of symbol blocks;
means for processing the second data packet to obtain a second plurality of symbol blocks;
means for transmitting a first plurality of symbol blocks, one symbol block at a time, on a first parallel channel to a receiver;
means for transmitting a second plurality of symbol blocks, one symbol block at a time, on a second parallel channel to a receiver;
means for prematurely completing the transmission of the first plurality of symbol blocks if the first data packet is restored by the receiver with a smaller number than all of the first plurality of symbol blocks; and
means for prematurely completing the transmission of the second plurality of symbol blocks if the second data packet is restored by the receiver with a smaller number than all of the second plurality of symbol blocks. 29. The device according to p, optionally containing
means for processing the third data packet to obtain a set of at least one character block for the third data packet; and
means for transmitting a set of at least one character block, one character block at a time, on the first parallel channel upon receipt of an indication that the first data packet has been restored. 30. The device according to p, optionally containing
means for processing the third data packet to obtain a third plurality of symbol blocks for the third data packet;
means for transmitting a third plurality of symbol blocks, one symbol block at a time, on a first parallel channel upon receipt of an indication that the first data packet has been restored;
means for processing the fourth data packet to obtain a fourth plurality of symbol blocks; and
means for transmitting a fourth plurality of symbol blocks, one symbol block at a time, on a second parallel channel upon receipt of an indication that the second data packet has been recovered. 31. A method of receiving transmission with increasing redundancy (NI) on the first and second parallel channels in a wireless communication system with many inputs and multiple outputs (MIMO), containing
receiving a symbol block for a first data packet transmitted by a first parallel channel, in which a first plurality of symbol blocks are generated for the first data packet and transmitted one symbol block at a time on the first parallel channel;
decoding all symbol blocks received for the first data packet to obtain a first decoded packet;
determining whether the first data packet is restored based on the first decoded packet;
completion of reception, decoding and determination for the first data packet if the first data packet is restored, or if all of the first plurality of symbol blocks have been received;
receiving a symbol block for a second data packet transmitted by a second parallel channel in which a second plurality of symbol blocks is generated for the second data packet and transmitted by one symbol block at a time on the second parallel channel;
decoding all symbol blocks received for the second data packet to obtain a second decoded packet;
determining whether the second data packet is restored based on the second decoded packet; and
completion of reception, decoding, and determination for the second data packet if the second decoded packet is recovered, or if all of the second plurality of symbol blocks have been received. 32. The method according to p. 31, in which decoding, determination and completion for the first data packet is performed whenever a block of characters is received for the first data packet, and in which decoding, determination and completion for the second data packet is performed whenever character block for the second data packet. 33. The method of claim 31, further comprising detecting a plurality of received symbol sequences for a plurality of receiving antennas to obtain a symbol block for the first data packet and a symbol block for the second data packet. 34. The method of claim 33, wherein the detection is based on a minimum mean square error (ISCED) detector, an optimal addition detector (OS), or a zero-force linear detector (FN). 35. The method according to p, in which the reception, decoding, determination and completion for the first data packet are performed regardless of the reception, decoding, determination and completion for the second data packet. 36. The method according to p, in which the restoration of the first data packet is assigned to the second data packet, and in which decoding, determination and completion for the second data packet are not performed until the first data packet is restored. 37. The method according to p, optionally containing
if the first data packet is restored,
estimating interference caused by the first data packet to the second data packet, and
suppressing interference caused by the first data packet in blocks of symbols received for the second data packet, and in which all symbol blocks received for the second data packet with the suppressed interference from the first data packet are decoded to obtain a second decoded packet. 38. The method according to p, in which the first data packet is restored before the second data packet, and the new data packet is not transmitted on the first parallel channel until the second data packet is restored. 39. The method according to clause 37, further comprising
if the first data packet is restored,
receiving a symbol block for a third data packet transmitted by the first parallel channel, in which a set of at least one symbol block is generated for the third data packet and transmitted by one symbol block at a time on the first parallel channel after the first data packet,
decoding all symbol blocks received for the third data packet to obtain a third decoded packet,
determining whether a third data packet is restored based on the third decoded packet, and
completion of reception, decoding and determination for the third data packet if the third data packet is recovered, or if all of the set of at least one block of characters have been received. 40. The method according to § 39, further comprising
if the third data packet is restored,
assessing interference from the third data packet to the second data packet, and
suppressing interference caused by the third data packet in blocks of symbols received for the second data packet, and in which all symbol blocks received for the second data packet with the suppressed interference from the first and third data packets are decoded to obtain a second decoded packet. 41. The method according to § 39, in which it is expected that the third data packet will be restored at or before the time when it is expected to restore the second data packet. 42. The method according to § 39, in which it is expected that the third data packet will be restored after the point in time when it is expected to restore the second data packet. 43. The method according to clause 37, further comprising
if the first data packet is restored,
receiving a symbol block for a third data packet transmitted by a first parallel channel in which a third plurality of symbol blocks are generated for the third data packet and transmitted one symbol block at a time on the first parallel channel after the first data packet,
decoding all symbol blocks received for the third data packet to obtain a third decoded packet,
determining whether a third data packet is restored based on the third decoded packet, and
completion of reception, decoding and determination for the third data packet if the third data packet is restored, or if all of the third plurality of symbol blocks have been received; and
if the second decoded packet is restored,
assessing interference caused by the second data packet to the third data packet, and
suppressing interference caused by the second data packet in the symbol blocks received for the third data packet, and in which all the symbol blocks received for the third data packet with the interference cancellation from the second data packet are decoded to obtain a third decoded packet. 44. The method according to p. 31, further comprising obtaining estimates of the ratio of the signal to the sum of noise and interference (SINR) for the first and second parallel channels; and selecting a first speed for the first parallel channel and a second speed for the second parallel channel based on the SINR estimates, and in which the first and second data packets are decoded in accordance with the first and second speeds, respectively. 45. The method of claim 31, further comprising sending an acknowledgment (PP) if the first data packet is recovered, or non-acknowledged (NPP) if the first data packet is not recovered. 46. A method for receiving transmission with increasing redundancy (NI) over multiple parallel channels in a wireless communication system with many inputs and many outputs (MIMO), comprising
receiving a plurality of symbol blocks for a plurality of data packets transmitted over a plurality of parallel channels in the current period, one symbol block for each data packet and one data packet for each parallel channel in which multiple symbol blocks are generated for each data packet and transmitted one symbol block per once through a parallel channel associated with them;
the choice of one of many parallel channels for recovery;
decoding all symbol blocks received for a data packet transmitted over a selected parallel channel to obtain a decoded packet;
determining whether the data packet transmitted on the selected parallel channel is restored based on the decoded packet;
completion of the acquisition, decoding, and determination for the data packet transmitted over the selected parallel channel if the data packet is recovered, or if all of the multiple symbol blocks have been received for the data packet; and
Evaluation and suppression of interference caused by the data packet transmitted on the selected parallel channel, if the data packet is restored. 47. The method according to item 46. in which a parallel channel with the highest probability of recovery from a plurality of parallel channels is selected for recovery. 48. The method according to item 46, in which to restore the selected parallel channel, which is restored by the last, latest in time from the current period, from a variety of parallel channels. 49. The method according to item 46, in which to restore the selected parallel channel with the largest number of blocks of data symbols in the current period from the set of parallel channels. 50. The method according to item 46, in which the selection, decoding, determination, completion and evaluation and suppression are performed for each of the many parallel channels in the current period. 51. The method according to item 46, in which the selection, decoding, determination, completion and evaluation and suppression are performed for many parallel channels, one parallel channel at a time and in the order with cyclic repetition, and the order with cyclic repetition is determined so that one or several parallel channels restored by the latest are placed last and restored, subsequently, by the last. 52. The method according to item 46, in which the selection, decoding, determination, completion and evaluation and suppression are performed for many parallel channels, one parallel channel at a time and in a predetermined order in the current period. 53. The method according to paragraph 52, in which a predetermined order is selected based on the probability of recovering a data packet for each of the many parallel channels. 54. The method of claim 52, wherein the predetermined order is selected based on the order in which data packets previously transmitted over a plurality of parallel channels are restored. 55. The method of claim 46, wherein the plurality of parallel channels have similar signal-to-noise-to-noise (SNR) ratios after linear detection at the receiver. 56. The method of claim 46, wherein a plurality of parallel channels are formed by diagonally transmitting across a plurality of subbands of a plurality of transmit antennas. 57. A receiver for receiving transmission with increasing redundancy (NI) on the first and second parallel channels in a wireless communication system with many inputs and multiple outputs (MIMO), containing
data processor configured
receiving a symbol block for the first data packet through a first parallel channel in which a first plurality of symbol blocks is generated for the first data packet and transmitted one symbol block at a time on the first parallel channel,
decoding all symbol blocks received for the first data packet to obtain a first decoded packet,
determining whether the first data packet is restored based on the first decoded packet,
receiving a symbol block for the second data packet using a second parallel channel, in which a second plurality of symbol blocks are generated for the second data packet and transmitted by one symbol block at a time on the second parallel channel,
decoding all symbol blocks received for the second data packet to obtain a second decoded packet, and
determining whether the second data packet is restored based on the second decoded packet; and
a controller configured to
completing the processing by the data processor for the first data packet if the first data packet is restored, or if all of the first plurality of symbol blocks have been received, and
completing the processing by the data processor for the second data packet,
if the second decoded data packet is restored, or if all of the second plurality of symbol blocks have been received. 58. The receiver of claim 57, further comprising a spatial processor configured to receive a plurality of symbol sequences for a plurality of receiving antennas and capable of detecting a plurality of received symbol sequences to obtain a symbol block for a first data packet and a symbol block for a second data packet. 59. The receiver of claim 58, wherein the spatial processor is configured to, if the first data packet is recovered, evaluate interference caused by the first data packet to the second data packet, and suppress interference caused by the first data packet in blocks of symbols received for the second a data packet, and in which the data processor is configured to decode all the blocks of symbols received for the second data packet with suppressed interference from the first data packet to obtain a second decoded packet. 60. The receiver of claim 57, further comprising
a channel estimator, configured to obtain estimates of the signal-to-noise-interference sum (SNR) for the first and second parallel channels; and
a speed selector configured to select a first speed for the first parallel channel and a second speed for the second parallel channel based on estimates of the SINR, and
in which the data processor is configured to decode the first and second data packets in accordance with the first and second speeds, respectively. 61. A device for receiving transmission with increasing redundancy (NI) on the first and second parallel channels in a wireless communication system with many inputs and multiple outputs (MIMO), containing
means for receiving a symbol block for the first data packet through a first parallel channel in which a first plurality of symbol blocks is generated for the first data packet and transmitted one symbol block at a time on the first parallel channel;
means for decoding all symbol blocks received for the first data packet to obtain a first decoded packet;
means for determining whether the first data packet is restored based on the first decoded packet;
means for completing reception, decoding, and determining for the first data packet if the first data packet is recovered, or if all of the first plurality of symbol blocks have been received;
means for receiving a symbol block for the second data packet through a second parallel channel, in which a second plurality of symbol blocks are generated for the second data packet and transmitted by one symbol block at a time on the second parallel channel;
means for decoding all symbol blocks received for the second data packet to obtain a second decoded packet;
means for determining whether the second data packet is restored based on the second decoded packet; and means for completing reception, decoding, and determining for the second data packet if the second decoded packet is recovered, or if all of the second plurality of symbol blocks have been received. 62. The device according to p, optionally containing
means for receiving multiple symbol sequences for multiple receiving antennas; and
means for performing detection of a plurality of received symbol sequences to obtain a symbol block for the first data packet and a symbol block for the second data packet. 63. The device according to p, optionally containing means for assessing interference caused by the first data packet to the second data packet, if the first data packet is restored; and means for suppressing interference caused by the first data packet in symbol blocks received for the second data packet, and in which all symbol blocks received for the second noise suppressed data packet from the first data packet are decoded to obtain a second decoded packet.

Images