Classifications

• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L1/00—Arrangements for detecting or preventing errors in the information received
  • H04L1/02—Arrangements for detecting or preventing errors in the information received by diversity reception
  • H04L1/06—Arrangements for detecting or preventing errors in the information received by diversity reception using space diversity
  • H04L1/0618—Space-time coding
  • H04L1/0625—Transmitter arrangements
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L1/00—Arrangements for detecting or preventing errors in the information received
  • H04L1/02—Arrangements for detecting or preventing errors in the information received by diversity reception
  • H04L1/06—Arrangements for detecting or preventing errors in the information received by diversity reception using space diversity
  • H04L1/0618—Space-time coding
  • H04L1/0637—Properties of the code
  • H04L1/0668—Orthogonal systems, e.g. using Alamouti codes
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L1/00—Arrangements for detecting or preventing errors in the information received
  • H04L1/02—Arrangements for detecting or preventing errors in the information received by diversity reception
  • H04L1/06—Arrangements for detecting or preventing errors in the information received by diversity reception using space diversity
  • H04L1/0618—Space-time coding
  • H04L1/0631—Receiver arrangements
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L5/00—Arrangements affording multiple use of the transmission path
  • H04L5/0001—Arrangements for dividing the transmission path
  • H04L5/0003—Two-dimensional division
  • H04L5/0005—Time-frequency
  • H04L5/0007—Time-frequency the frequencies being orthogonal, e.g. OFDM(A) or DMT

Definitions

• the present invention relates to the field of wireless communication
• the present invention relates to transmitting the same information
• the data for any particular cell user may be available to multiple base stations. Viewing each of the base stations with data for a particular user as an element of a virtual antenna array suggests using cooperative signal encoding schemes across these base stations to provide diversity benefits to the desired user.
• each active base station when transmitting the same data from two active base stations, each active base station must encode its data independently. Since the encoded signals, however, are transmitted by spatially dispersed base-stations, they arrive at the receiver with distinct relative delays with one another, i.e., asynchronously. That is, there can be a lack of time-synchronization between the transmissions from different base stations to the receiver.
• This asynchrony can arise due to the fact that the individual base stations may be operating independently, but also due to the fact that even if the signals the signals transmitted from spatially dispersed base stations to a receiver are transmitted synchronously, they may arrive asynchronously at the receiver. Although these relative delays can, in principle, be estimated at the receiver, they are not known (and thus cannot be adjusted for) at the transmitting base stations, unless there is relative-delay information feedback from the receiver to the transmitting base stations.
• STBCs space time block codes
• R which is equal to kit (i.e., the ratio of k over i).
• Full rate STBCs are STBCs whose rate R equals 1 symbol per channel use.
• Another important attribute of a STBC is its decoding complexity.
• Another class of approaches exploits space-time trellis code designs that are designed to provide full diversity regardless of the relative delay subject to maximum allowable relative delay offset, L.
• These codes exploit constructions based on shift-full rank matrices that guarantee that the matrices provide full diversity regardless of the relative delay set. Although they provide full diversity subject to a set of relative delays with only a small overhead in data rate, these designs have some important limitations.
• the decoding complexity of these designs is exponential both in the number of antennas, and the parameter L.
• the design is modulation scheme specific.
• they are guaranteed to provide full diversity subject to relative delay offsets that are integer multiples of the symbol duration. Strictly speaking, however, there are no guarantees for relative delays that are fractions of the symbol duration.
• STBCs is exponential in the number k of jointly encoded symbols, there exist designs with much lower complexity.
• One such attractive class of designs referred to as orthogonal space-time codes (OSTBCs)
• OSTBCs orthogonal space-time codes
• a class of schemes includes space-time bit-interleaved coded modulation systems with OFDM and can provide both spatial diversity and can cope with asynchronous transmission. Although these schemes can provide full diversity and very good data rates, they are disadvantageous because the decoder complexity of such schemes grows exponentially with the number of transmit antennas used over all the base stations, and the number of bits/per symbol in the employed modulation scheme.
• a method and apparatus for asynchronous space- time coded transmission from multiple stations, hi one embodiment, the method comprises one or more terminals and at least two base stations wirelessly communicating information-bearing signals to the one or more terminals using orthogonal space-time block codes.
• the proposed codes maintain "orthogonality" at the receiver even in the presence of asynchronous signal reception among signals transmitted from distinct transmit base stations.
• Figure 1 is a block diagram of one embodiment of a high-level encoder of the baseline orthogonal space time block codes (OSTBCs).
• OSTBCs baseline orthogonal space time block codes
• Figure 2 is a block diagram of one embodiment of a high-level encoder of the induced OSTBCs.
• Figure 3 illustrates the structure of one embodiment of a matrix V in
• Figure 4 illustrates a code construction for a two transmit-base station system with a single transmit antenna.
• Figure 5 illustrates a code construction for a four transmit-base station system in which each base station employs a single transmit antenna.
• Figure 6 is a block diagram of a data decoding structure for an induced
• Figure 7 illustrates an asynchronous communication system.
• Figure 8 is a block diagram of an encoding and transmission system of a sample antenna at a base station.
• Figures 9A and 9B illustrate embodiments of two different base station encoders for the OSTBC shown in Figure 4.
• Figure 10 is a block diagram of one embodiment of a front-end of a receiver that uses OFDM-based induced OSTBC.
• Methods and apparatuses for transmitting an information-bearing stream of symbols from multiple base stations to one or more designated mobile receivers are disclosed.
• the transmission is achieved by using a space-time block code across all transmitting base stations.
• a space-time block code is a block-by-block encoding and transmission method in which each block of information symbols to be transmitted by a base station is encoded into a base station specific block of samples for transmission.
• the space-time codes allow data decoding at any mobile, even in the case that the signals transmitted by distinct base stations are received at the mobile asynchronously, i.e., relatively delayed with respect to one another.
• Codes disclosed herein can be readily employed as an inner code component of a more elaborate encoding system, designed to harvest other available forms of diversity.
• Techniques described herein include, but are not limited to: (i) systematically constructing space-time block codes for the asynchronous setting via a set of systematic transformations of existing space-time block codes that are used for (single base-station) multiple transmit antenna systems; (ii) block-symbol encoding in which each block of input data symbols is mapped into a (base-station specific) block of data samples for transmission; (iii) data decoding at each mobile receiver based on an aggregate asynchronously received signal, estimates of the channel response coefficients and estimates of the relative signal transmission delays.
• these quantities take the form of electrical or magnetic signals capable of being stored, transferred, combined, compared, and otherwise manipulated. It has proven convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.
• This apparatus may be specially constructed for the required purposes, or it may comprise a general purpose computer selectively activated or reconfigured by a computer program stored in the computer.
• a computer program may be stored in a computer readable storage medium, such as, but is not limited to, any type of disk including floppy disks, optical disks, CD-ROMs, and magnetic-optical disks, read-only memories (ROMs), random access memories (RAMs), EPROMs, EEPROMs, magnetic or optical cards, or any type of media suitable for storing electronic instructions, and each coupled to a computer system bus.
• a machine-readable medium includes any mechanism for storing or transmitting information in a form readable by a machine (e.g., a computer).
• a machine-readable medium includes read only memory ("ROM”); random access memory (“RAM”); magnetic disk storage media; optical storage media; flash memory devices; electrical, optical, acoustical or other form of propagated signals (e.g., carrier waves, infrared signals, digital signals, etc.); etc. Overview
• a new space time code is constructed that provides full diversity, symbol-by-symbol decoding and good coding gains for asynchronous settings, assuming that the maximum of the relative delay offsets does not exceed a predetermined parameter L.
• data in blocks larger than the memory of the baseline orthogonal code are encoded at a lower rate, thereby increasing the dimensionality of the signal space and allowing diversity benefits to be realized for any set of pairwise relative delay offsets, where the maximum relative delay is at most L samples.
• asynchronous orthogonal space-time block codes are used in an encoding and transmission method.
• the inputs include a maximum relative delay (integer) parameter, L 0 , a blocking (integer) parameter, N, and an orthogonal space-time block code for a n transmit-antenna system, which encodes blocks of k information bearing symbols into blocks of t samples (per transmit antenna) for transmission.
• a class of space-time block codes may be used for an n transmit-base-station system.
• each such space-time block code are used to encode blocks of K symbols, where K equals, Nk (N times k) into blocks of T samples (per transmit base-station) for transmission, where T equals t(N+L) (i.e., t times the sum of N and L) and where the value of L is determined by the value of L 0 and other system parameters.
• K equals, Nk (N times k) into blocks of T samples (per transmit base-station) for transmission
• T equals t(N+L) (i.e., t times the sum of N and L) and where the value of L is determined by the value of L 0 and other system parameters.
• a class of space-time block codes is used with associated transceivers to enable reliable transmission of common information from a set of base stations to one or more receivers over wireless channels.
• the transceivers achieve reliable transmission of a common information signal by sending distinct encodings from a set of base stations without the need for synchronizing the transmissions, even with an arbitrary set of relative delays in the reception of the signals transmitted from different base stations, provided that the maximum relative delay between transmissions does not exceed an a priori determined value.
• the transceiver operates to provide asynchronous reliable communication of a symbol stream from n transmitting base stations to one or more receivers, where each transmitting base station has available the same information-bearing symbol stream that is to be communicated to the receiver(s).
• multiple base stations in a system use an orthogonal space-time block code, referred to herein as the "induced" code.
• the induced OSTBC is generated via systematic transformations of a multiple transmit-antenna orthogonal space-time block code, referred to herein as the "baseline" code.
• an (induced) n transmitting base station space-time block code is generated by applying systematic transformations on an existing (baseline) n transmit-antenna OSTBC. More specifically, in one embodiment, to design a space-time block code for an n transmit base-station system in which each base station has a single transmit antenna, a baseline k-by-t-by-n orthogonal space- time block code (i.e., an OSTBC that encodes k information symbols at a time into a block of t time slots, over n (collocated) antennas) is used.
• a baseline k-by-t-by-n orthogonal space- time block code i.e., an OSTBC that encodes k information symbols at a time into a block of t time slots, over n (collocated) antennas
• a K-by-T-by-n induced OSTBC is generated that encodes K information symbols at a time over a block of T time slots, and over n single-antenna transmit base-stations.
• a code is the mapping that dictates the samples that are going to be transmitted (over each antenna and over time) given a sequence of data symbols.
• a block code does the coding in blocks of symbols at a time: give a (arbitrary) set of k symbols is to be transmitted the code dictates what gets transmitted (over time and antenna). In one embodiment, the choice of the block code (mapping) is performed once and prior to signal transmission.
• a new induced code is designed (a priori and once).
• the new induced code is as shown in Figure 4 and where, in addition, F is chosen to be the inverse DFT matrix (normalized so that it's unitary), and U is chosen to be a time-reversal matrix.
• the implementation of the encoder for the resulting selection of an induced OSTBC at each of the transmitting base-station is as shown in Figure 9.
• the values of the integers K and Tare determined as follows.
• the integer K equals iV times k, signifying that, for each scalar symbol encoded by the baseline code, the induced code encodes a vector of N symbols.
• the induced code encodes K symbols at a time into blocks of T samples/per transmit base-station, where T equals t(N+L) (i.e., t times the sum of N and L) and where L is chosen so as to satisfy the maximum allowable relative delay constraint.
• T equals t(N+L) (i.e., t times the sum of N and L) and where L is chosen so as to satisfy the maximum allowable relative delay constraint.
• the induced code is expected to allow reliable asynchronous communication whenever the relative reception delay between any two signals transmitted from distinct base stations does not exceed L 0 times the symbol signaling period (and where L 0 denotes a pre-determined design parameter), the value of L set to be equal L 0 plus a predetermined modem-dependent constant L ⁇ .
• an encoding scheme performed by an encoder provides the full transmit base-station diversity available in the set of n transmitting base stations, provided the maximum relative delay between all pairs of received signals from transmitting base stations is at most L 0 times the symbol signaling period.
• the receiver For decoding at a receiver, given knowledge of the set of relative sampling delays at the receiver, knowledge (estimates in practice) of the n fading channel coefficients, and the OSTBC structure, the receiver performs a linear transformation of the received signal followed by symbol-by-symbol decoding. Note that the decoding operation for the induced code base system (system of interest) is different that the one for the baseline code system.
• the decoder operates according to the specific code employed, but it also depends on the channel model (synchronous vs. asynchronous, flat fading vs. frequency selective fading, etc) and on what is known about the channel at the receiver. Note that for a general space-time code the "best" (in terms of reliability, i.e., probability of error) detector can in principle be determined, although it might be too complex to implement in practice.
• this detector is simple over flat fading synchronous channels: (i) a linear operation (multiplication by a matrix) of the received column, yielding a vector of the size of the information vector; (ii) symbol-by-symbol decoding: the ith symbol in the information vector that was transmitted by the code is directly decoded from the ith sample of the vector generated by the linear operation in (i).
• the best detector for an induced code is also of the form (i)/(ii) for synchronous channels, but also for asynchronous (flat fading) channels as well (up to a maximum relative delay).
• the linear operation in (i) is different for the induced code than for the baseline code.
• the induced code does not only depend on the code but also on the channel characteristics as well as the actual relative delays (see figure 10).
• the induced codes presented above are for a setting involving n transmit base stations, each with a single antenna, the same induced codes can be applied in other settings where there is an arbitrary number of transmit antennas at each base station. In that case, " «" denotes the total number of antennas in the system. For example, if the information bearing signal is to be transmitted from three base stations, where base station 1 and 2 have two transmit antennas each, and base station 3 has four transmit antennas, "n" equals 8.
• Figures 1 and 2 show typical high-level encoder implementations of the baseline and induced OSTBCs, respectively.
• x(l), x(2), ... , x(k) are used to denote a typical block of k scalar, complex-valued information-bearing symbols that are inputs to the baseline OSTBC.
• Encoder 101 receives information-baring symbols x(k), . . ., JC(2), *(1). Given such a set of inputs, baseline encoder 101 generates a baseline code that is represented by a matrix B of dimensions t by n, represented below as:
• the (j,i)th element of B denotes the sample that is to be transmitted by the antenna i at timey, in the context of an n transmit-antenna system utilizing the baseline code.
• encoder 101 outputs bj(n), which represents the nth output sample of the baseline code over the /th transmit antenna.
• encoder 101 generates an output to a transmit antenna 103 1 (via modem 102i 1), that is represented as b ⁇ (t), . . ., b ⁇ (2), b ⁇ (l). The same can be said for each of the other output samples of baseline code from encoder 101 for transmission over the second through nth transmit antenna.
• s(l), s(2), ..., s(k) denote a typical block of k information-bearing symbol input vectors of dimension N that are inputs to the induced OSTBC according to one embodiment of the present invention.
• the Mh vector s(i ' ) is a vector (or block) of N scalar complex-valued, information-bearing symbols in the induced code (where N denotes the blocking factor in the construction).
• induced encoder 201 Given such a set of input vectors, induced encoder 201 generates an induced code that is represented by a matrix B with Trows and n columns, where T equals t times the sum of N and L.
• the output matrix B of induced code of dimension "T" x "n" may be represented as follows:
• the (/,0th element of B denotes the sample that is to be transmitted by the i-th base station at timey, in the context of an n transmit base-station system utilizing the induced code.
• the output vector of induced code (of dimension "T") associated with the ith transmit base station is as follows:
• encoder 201 in response to information-bearing symbol vectors s(k), . . . , s(2), s(l), encoder 201 outputs bfT), . . ., bj(2), b,(l).
• the output of encoder 201 is received by modem 202, which causes the symbols to be transmitted via transmit antenna 203 at the /th base station.
• the induced orthogonal space-time block code B is generated by systematic transformations of the elements of the associated baseline code B. More specifically, in one embodiment, each entry of the baseline OSTBC equals one of the following: (i) the value zero; (ii) one of the k input symbols; (iii) one of the k input symbols negated; (iv) the complex conjugate of an input symbol; (v) the negative of the complex conjugate of an input symbol. Therefore, to generate the induced OSTBC, initially the mth vector input to B, s(m), is paired with the mth scalar input to B (i.e., x(m)). Then B is generated from B by replacing each entry of B with a vector of dimension N+L according to the rules specified in Table 1 below:
• each information-bearing symbol x(m) in the baseline OSTBC matrix B is associated with a vector of information bearing symbols s(m), of dimension N, in the induced code B.
• each scalar entry of the original OSTBC matrix B is replaced by a vector entry of dimension N+L, according to the table above.
• the matrix F is a unitary matrix with N rows and N columns.
• the matrix G has N+L rows and N columns.
• the transformation process takes as an input a vector of dimension N and produces a vector of dimension N+L, in which the last N entries of the output vector are the entries of the input vector (in the same order) and in which the L first entries of the output vector are the last L entries of the input vector (in the same order).
• the superscripts " * " and " H denote element- wise conjugation and the Hermitian (conjugate transpose) operation, respectively. For more information on these operations, see R. A. Horn and C. R. Johnson, Matrix Analysis. New York: Cambridge Univ. Press, 1994.
• this entry is replaced by a vector v(m), of dimension N+L generated by: 1) generating the intermediate vector z(m), of dimension N, by multiplying the vector s(m) with the matrix F; 2) adding an L-sample circular prefix to z(m) to produce a vector v(m) (i.e., generating a vector v(/n) of dimension N+L where the first L entries of v(m) are the last L entries of z(m) and the last N entries of v(m) are the entries of ⁇ (m) (in the same order)); and 3) replacing the (/J)th entry of B (equal to x(m) by assumption) with v(/n).
• this entry is replaced by a vector u(m), of dimension N+L generated by: 1) generating an intermediate vector z(m), of dimension N, by multiplying the element-wise complex conjugate of the vector s(m) with the matrix F * , which is the element-wise complex conjugate of the matrix F; 2) generating an intermediate vector d(/n), of dimension N, by multiplying the vector z(m) with a matrix U * , which is the element-wise complex conjugate of the matrix U; 3) adding an L-sample circular prefix to d(m) to produce a vector u(/n), i.e., generate a vector u(m) of dimension N+L where the first L entries of u(m) are the last L entries of d(m) and the last N entries of
• Figure 4 shows the special case of generating a code to be used with a two-transmit base station system, where each base station has a single transmit antenna per base station using the Alamouti code.
• the induced code is shown with the matrix B above.
• the samples transmitted by the first antenna at times L+N+l through 2x(L+N) are generated as follows: (i) apply a unitary transformation F on the element-wise conjugate of the vector s(2); (ii) apply transformation U on the resulting vector; (iii) prepend the resulting vector of dimension N with its L-sample circular prefix.
• Figure 5 illustrates code construction for a four transmit base-station system, in which each base station employs a single transmit antenna.
• the baseline code is shown in Figure 5 by the matrix B and is the maximum-rate four transmit antenna OSTBC.
• the associated induced space-time code is constructed according to Table 1 and is depicted by the matrix B.
• the baseline OST ⁇ C, matrix B encodes 3 symbols over four samples per antenna at a rate of 3 A symbols/channel use
• the associated induced OSTBC, B encodes 3-times-N symbols at a time over 4-times-(N+L) samples/per base-station, at a rate % times N/(N+L) symbols/channel use.
• This code also provides a systematic induced OSTBC for a three transmit-base-station system (e.g., by dropping one of the columns of B).
• the code is used in the context of a two-base station system in which each base station has two transmit antennas. In this case, columns 1-2 of B are associated with the two transmit antennas at the one of the base stations and columns 3-4 are associated with the two transmit antennas at the other base station.
• encoding at the /th base station is performed as follows. First, the encoder takes as an input the symbol vector s of size K (where K equals N times k), where the symbol vector represents the information to be transmitted and is assumed to be generated at earlier stages.
• the vector s that is employed is assumed to be known at the receiver, while in the data transmission phase, the vector s represents a set of K symbols to be transmitted to the receiver and is thus unknown to the receiver.
• the vector s of dimension K is partitioned into k vectors s(l), s(2), ... , s(k), each of dimension N.
• the /th base station transmits a vector b, of dimension T (and where T equals t times the sum of N and L), generated according to the induced OSTBC B, where b, denotes the /th column of the matrix B that is generated with input s(l), s(2), ... , s(*).
• the encoding process is performed using a normalized DFT matrix the unitary matrix F, with U being a time reversal matrix (i.e., a matrix such that Ux produces an order-reversed version of x).
• transmitting the vector GFs(m) is equivalent to an OFDM transmission with input s(m).
• the transmitted vector, b, is a collection of t vectors (transmitted sequentially), each of dimension N+L, in which each vector is constructed via OFDM-type operations on a particular input vector (of dimension N) according to Table 1 above.
• Figure 6 illustrates one embodiment of a symbol detection algorithm performed at a receiver.
• This algorithm can be readily applied for detection of the information bearing vectors s(l), s(2), ... , s(k), in the case of data transmission over slowly varying flat fading channels, but can also be applied for symbol detection over slowly-time varying frequency selective channels.
• the channel is varying slowly enough so that accurate estimates of the channel fading coefficients can be obtained (via the pilot estimation phase) and that the receiver employs a standard front-end which consists of a linear filter (e.g., a square-root raised cosine pulse matched to the transmitter pulse-shaping filter) followed by a baud-rate sampler (with adjustable sampling times).
• a linear filter e.g., a square-root raised cosine pulse matched to the transmitter pulse-shaping filter
• baud-rate sampler with adjustable sampling times.
• the effective impulse response of a channel from a transmitting base station to a receiver is at most equal to Lg samples long.
• the constant L ⁇ is determined by the "effective" duration (in symbol periods) of the response of the pulse-shaping waveform (used at each encoder) through the receiver front-end filter.
• the value of Ls depends on the roll-off factor employed and is typically between four and six.
• the first nonzero samples in the impulse responses of any two such channel responses are at most L 0 samples apart (due to the assumed L 0 maximum relative delay constraint).
• estimates of the relative delays in the transmissions are first obtained, together with estimates of the channel fading coefficients.
• these operations are performed at the receiver: they are part of the timing
• the transmitting antennas transmit space-time coded sequences
• the receiver has an
• Divider 601 partitions r into t vectors r(l), r(2), ..., r ⁇ t), each of dimension
• prefix removal unit 602 constructs a vector y(m) of dimension N, as the last N entries
• Combiner 604 constructs a real-valued vector Y of dimension 2Nt (i.e., 2 times N times t) using the output of combiner 603. In one embodiment, combiner 604 concatenates the (element- wise) real part of y with the (element-wise) imaginary part of y.
• Writing the received signal in this form (and thus the associated representation, including the creation of the matrix H) are straightforward operations that are well known in the art.
• Slicer 610 performs symbol-by-symbol (element-wise) detection on each of the scalars extracted from the vector x in order to determine the (element-wise) estimate of communicated symbol vector s, thereby inputting symbol estimates 611.
• the estimates of the relative delays at the receiver are used to adjust the sampling timing of the baud-rate sampler at the receiver front-end.
• Such adjustments can have significant benefits from a receiver performance point of view. For instance, consider the case where a two base-station system is employed and where the relative delay between the received signals equals "half a symbol period.
• the receiver performance is improved, and potentially optimized, when the baud-rate sampler is a quarter-sample "behind” one of the transmission and a quarter-sample “ahead” of the other, and takes its worse value when the baud-rate sampler is in-sync with one of the transmissions and half-a-sample out of sync with the other.
• the baud-rate sequence of samples (that is partitioned into r-type vectors of the form shown in Figure 6) is generated by passing the (down converted continuous-time) received signal through the following cascade of systems that includes: a linear front-end filter (e.g., a square-root raised cosine filter); an oversample-by-M system where M is an integer greater than 1 (this system keeps M samples per symbol period); a discrete-time delay-by-n o -samples system (the output of this system is its input delayed by n o -samples); and a decimate-by-M system (this system generates an output by discarding all by every Mth sample of the input sequence).
• the estimates of the relative sampling times can be employed to adjust the (integer) delay parameter n 0 in the delay-by-n 0 system prior to the decimator in order to optimize the decoder performance.
• Figure 7 illustrates an asynchronous wireless transmission from multiple base stations to mobile receivers.
• multiple base stations 702 ⁇ . n are shown, and each of these base stations has multiple antennas for communicating with mobile receivers, such as mobile receiver 703.
• Central control unit 701 is communicably coupled to base stations 702 ⁇ . n to control base stations 702 1 . n-
• Figure 8 is a block diagram of one embodiment of an OFDM-based
• the OSTBC encoder used by one of base stations 702i - ⁇ for encoding for transmission on a typical antenna element.
• the unitary matrix F is a (normalized) inverse DFT matrix.
• the encoder takes as input the information- bearing symbols, or samples generated by an outer code in the system. Encoding is performed in blocks of size K, where K equals k times N.
• Serial-to-parallel converter 801 splits each block of K symbols into k sub-blocks, information symbol vectors s(l), s(2), ..., s(k), each of which has N symbols.
• Each block is then individually processed as shown in Figure 8, with the switches 82Ui.*, 8211.*, 822 1- ⁇ , and 823 1- * set according to the OSTBC column associated with the given antenna element.
• the encoder applies an N-pt IDFT to k sub- block s(l).
• the encoder projects a matrix U 803 1 onto the transformed data.
• matrix U 803 1 is a time reversal matrix.
• the encoder performs an element-wise complex conjugate 804i to the results of projecting the transformed data onto matrix U 803 1 .
• Switch 821 1 is set to the same path (lower or upper) as switch 820 1 .
• the results from the selected (upper or lower) path are then directed through, either the upper path of switch 822 1 (and the upper path of 823 1 , which is always set to the same path as switch 8220 to circular prefix unit 806 1 , or through the lower path of switch 822] to "negate" block 805 1 and through the lower path of switch 823 1 to circular prefix unit 806 1 .
• the remainder of the information symbol vectors s(2), ..., s(k) are processed in the similar fashion.
• Circular prefix units 8O6 1 .* insert an L-sample circular prefix to create an N+L sized processed block.
• reorder unit 807 reorders the processed blocks and inserts blocks of zero vectors (each of dimension N+L) where appropriate, and as specified by the column of the induced space-time block code that is associated with the given antenna.
• the reorder operation for the antenna associated with the 2 nd column of the matrix would place first the output of prefix unit 806 2 followed by the output of prefix unit 806 1 followed by a vector of N+L zeros, followed by the output of prefix unit 806 3 .
• the position of switches in Figure 8 and the reordering operation are dictated by the column of the OFDM-based OSTBC that is associated with the given antenna element.
• the T-sample vector generated by the reordering/zero insertion operation is passed through parallel-to-serial converter 808 to a modem 809, which performs pulse-shaping, amplification, and modulation to radio-frequency (RF) in a manner well-known in the art. Afterwards, modem 809 sends the data for transmission over antenna element 810.
• RF radio-frequency
• Figures 9A and 9B illustrate encoders at two base stations.
• Figures 9A and 9B support an OFDM-based OSTBC implementation in each base station in a scheme involving transmission from two base stations, each with a single antenna element for transmission.
• the system takes as input two blocks of N (complex-valued scalar) symbols at a time.
• the encoder includes serial-to parallel converter
• N-point IDFT 902 transforms the information symbol vector s(l) into N symbols.
• Circular prefex unit 903 adds a prefex of L bits to the N symbols, thereby creating an N+L set of data that is sent to parallel-to serial- converter 908, where it is converted to a serial steam.
• N-point EDFT 904 transforms the information symbol vector s(2) into N symbols. Then projection unit 905 projects the matrix output from iV-point IDFT transform 904 onto matrix U.
• matrix U is a time-reversal matrix.
• Conjugate unit 906 performs the conjugate operation on the output of projection unit 905.
• Circular prefetch unit 907 adds a prefex of L symbols onto the conjugated output of conjugate unit 906.
• the output of circular prefex unit 907 is input to parallel-to- serial converter 908, which converts data on its parallel inputs into serial form.
• the output of parallel-to-serial converter 908 is sent to modem 909, which causes the data to be transmitted via antenna 910.
• the encoder includes serial-to parallel converter 911 that receives a set of information symbol vectors of dimension "N" and converts them into parallel form.
• N-point IDFT 912 transforms the information symbol vector s(2) into N symbols.
• Circular prefex unit 913 adds a prefex of L bits to the N symbols, thereby creating an N+L set of data that is sent to parallel-to serial-converter 918, where it is converted to a serial steam.
• N-point IDFT 914 transforms the information symbol vector s(l) into N symbols. Then projection unit 915 projects the matrix output from N-point IDFT transform 914 onto matrix U.
• matrix U is a time-reversal matrix.
• Conjugate unit 916 performs the conjugate operation on the output of projection unit 915.
• Circular prefetch unit 917 adds a prefix of L symbols onto the conjugated output of conjugate unit 916.
• the output of circular prefix unit 917 is input to parallel-to- serial converter 918, which converts data on its parallel inputs into serial form.
• the output of parallel-to-serial converter 918 is sent to modem 919, which causes the data to be transmitted via antenna 920.
• FIG 10 is a block diagram of one embodiment of a front-end of a receiver that uses OFDM-based induced OSTBC. The operation of the receiver is straightforward (i.e., there are no timing/synchronization issues).
• splitter 1001 receives a T-sample block of T consecutive samples associated with transmission of an information-bearing vector s of dimension K and splits it into 2 vectors, r(l) and r(2), of dimension N+L.
• the 7-sample block is output from a baud-rate sampler.
• Prefix discarding units 1002 and 1002 remove the L-sample prefix from each of vectors r(l) and r(2), respectively, resulting in vectors y(l) and y(2), which are of dimension N. Note that the ith entry of y(m) equals the "L+/"th entry of r(m).
• N-pt DFT unit 1005 applies the N-pt DFT to vector y(l) and sends the result to multipliers 1015 and 1016.
• Conjugate unit 1004 performs a conjugate on vector y(2). The conjugate of the vector y(2) is projected onto a matrix l ⁇ by unit 1006.
• N-pt DFT unit 1007 applies the N-pt DFT to vector result from unit 1006 and sends the result to multipliers 1014 and 1018.
• the relative delay (estimate) associated with the signal from antenna 1 is used to determine the samples (a vector) in the impulse response of the relative-delay response filter 1008.
• the resulting vector signal is scaled element-wise by the scaling fading coefficient (estimate) / ⁇ (l) associated with the first channel.
• An N-pt DFT is applied by unit 1009 to the result.
• the output of N-pt DFT unit 1009 is an input to conjugate unit 1030 and multiplier 1014.
• Conjugate unit 1030 determines the conjugate and outputs that to multiplier 1015.
• the relative delay (estimate) associated with the signal from antenna 2 is used to determine the samples (a vector) in the impulse response of the relative- delay response filter 1011.
• the resulting vector signal is scaled element- wise by the scaling fading coefficient (estimate) h(2) associated with the second channel.
• An N-pt DFT is applied by unit 1012 to the result.
• the output of N-pt DFT unit 1012 is an input to conjugate unit 1013 and multiplier 1016.
• Conjugate unit 1013 determines the conjugate and outputs that to multiplier 1018.
• the outputs of multiple 1016 and 1018 are sent to one input of adders 1017 and 1019.
• Adder 1017 subtracts the output of multiplier 1016 from the output of multiplier 1014.
• Adder 1019 adds the output of multiplier 1015 to the output of multiplier 1018.
• each adder module 1017 and 1019 are vectors of dimension N.
• the output is the element-wise sum vector, i.e., a vector of dimension N, where each entry is the sum of the associated entries of the input vectors.
• the (two) inputs to each multiplies are also vectors of dimension N.
• the output is the element-wise product vector, i.e., a vector of dimension N, where each entry is the product of the associated entries of the input vectors.
• each scalar symbol-by- symbol detection module 1019 and 1020 takes as input a vector of dimension N and produces a vector of detected symbols of dimension N (either of s(l) or of s(2), depending on the module #).
• symbol-by-symbol detection modules 1019 and 1020 detects the /th element of the detected vector in each case (either s(l) or s(2)) is detected based on only the /th element of the input vector (input to the module).
• the schemes provide full "transmit base-station" diversity, for any set of relative delays between transmitting base stations, provided that none of the relative delays of arrival exceed the maximum allowable value of L 0 times the symbol signaling period.
• these techniques allow for high-reliability symbol-by-symbol decoding. Specifically, given that a symbol blocking factor N is employed, the decoding complexity is at most quadratic in iV and k (where k denotes the number of transmit base stations), and independent of the maximum relative delay factor, L 0 . Also, unlike other existing designs that are tailored to specific modulation schemes, the proposed techniques are applicable to any real- valued or complex-valued modulation scheme, including, for example, but not limited to, BPSK, QPSK, QAM, and M-PSK.

Landscapes

• Engineering & Computer Science (AREA)
• Signal Processing (AREA)
• Computer Networks & Wireless Communication (AREA)
• Radio Transmission System (AREA)
• Mobile Radio Communication Systems (AREA)

Priority Applications (2)

Applications Claiming Priority (4)

Publications (1)

Family

ID=39226902

Family Applications (1)

Country Status (5)

Cited By (2)

Families Citing this family (23)

Citations (3)

Family Cites Families (67)

• 2006
  • 2006-12-21 US US11/644,638 patent/US8027407B2/en active Active
• 2007
  • 2007-11-02 KR KR1020087025123A patent/KR20080104375A/ko not_active Ceased
  • 2007-11-02 WO PCT/US2007/023207 patent/WO2008057452A1/en not_active Ceased
  • 2007-11-02 EP EP07861675A patent/EP2080305A1/en not_active Withdrawn
  • 2007-11-02 JP JP2009535338A patent/JP5097210B2/ja active Active

Patent Citations (3)

Also Published As

Similar Documents

Legal Events