US20100226448A1 - Channel extrapolation from one frequency and time to another - Google Patents

Channel extrapolation from one frequency and time to another Download PDF

Info

Publication number
US20100226448A1
US20100226448A1 US12/478,564 US47856409A US2010226448A1 US 20100226448 A1 US20100226448 A1 US 20100226448A1 US 47856409 A US47856409 A US 47856409A US 2010226448 A1 US2010226448 A1 US 2010226448A1
Authority
US
United States
Prior art keywords
frequency
scattering
delay
extrapolated
scatterer parameters
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Abandoned
Application number
US12/478,564
Other languages
English (en)
Inventor
Paul Wilkinson Dent
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Telefonaktiebolaget LM Ericsson AB
Original Assignee
Individual
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to US12/478,564 priority Critical patent/US20100226448A1/en
Assigned to TELEFONAKTIEBOLAGET L M ERICSSON (PUBL) reassignment TELEFONAKTIEBOLAGET L M ERICSSON (PUBL) ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: DENT, PAUL WILKINSON
Priority to EP10714471A priority patent/EP2404419A2/fr
Priority to CN2010800114040A priority patent/CN102342075A/zh
Priority to PCT/IB2010/000438 priority patent/WO2010100551A2/fr
Publication of US20100226448A1 publication Critical patent/US20100226448A1/en
Abandoned legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/0202—Channel estimation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L25/00—Baseband systems
    • H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
    • H04L25/0202—Channel estimation
    • H04L25/024—Channel estimation channel estimation algorithms
    • H04L25/0242—Channel estimation channel estimation algorithms using matrix methods
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2626—Arrangements specific to the transmitter only
    • H04L27/26265—Arrangements for sidelobes suppression specially adapted to multicarrier systems, e.g. spectral precoding
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2647—Arrangements specific to the receiver only
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L27/00—Modulated-carrier systems
    • H04L27/26—Systems using multi-frequency codes
    • H04L27/2601—Multicarrier modulation systems
    • H04L27/2626—Arrangements specific to the transmitter only
    • H04L27/2627—Modulators
    • H04L27/2634—Inverse fast Fourier transform [IFFT] or inverse discrete Fourier transform [IDFT] modulators in combination with other circuits for modulation

Definitions

  • a transmitted signal reflects off objects (e.g. buildings, hills, etc.) in the environment, referred to herein as scattering objects.
  • the reflections arrive at a receiver from different directions and with different delays.
  • the reflections or multi-paths can be characterized by a path delay and a complex delay coefficient.
  • the complex delay coefficients show fast temporal variation due to the mobility of the vehicle while the path delays are relatively constant over a large number of OFDM symbol periods.
  • Channel estimation is the process of characterizing the effect of the radio channel on the transmitted signal.
  • Channel estimates approximating the effect of the channel on the transmitted signal may be used for interference cancellation, diversity combining, ML detection, and other purposes.
  • Many channel estimation techniques in common use do not produce sufficiently accurate estimates of the channel for use by higher order modulations. Further, it is difficult to predict how the channel will change due to the mobility of the vehicle. Therefore, there is a need for new channel estimation techniques that will produce more accurate channel estimates for higher order modulation and enable prediction of the channel from current channel estimates.
  • the present invention provides an improved channel estimation technique that determines accurate scatterer parameters for the scattering objects in the wireless channel, and extrapolates the scatterer parameters in both time and frequency to characterize the scattering objects for a different time and a different frequency.
  • the present invention may determine scatterer parameters that characterizes the scattering objects of a reception channel, and extrapolates the scatterer parameters in both time and frequency to predict the scatterer parameters for a future time and frequency, e.g., a future transmission time and frequency.
  • a future time and frequency e.g., a future transmission time and frequency.
  • a processor in a transceiver determines a first set of scatterer parameters based on signal samples received via a first communication channel associated with a first frequency and a first time.
  • the first set of scatterer parameters comprises a set of scattering coefficients for each of a plurality of scattering objects, where each scattering object has an associated path delay and a rate of change of path delay, or equivalently, Doppler shift.
  • the processor may determine the first set of scatterer parameters using the method provided by co-pending application Ser. No. 12/478473 and/or co-pending application Ser. No. 12/478520, which are both incorporated herein by reference.
  • the processor determines a second set of scatterer parameters for a second communication channel associated with a second frequency and a second time.
  • the processor determines the second set of scatterer parameters by calculating extrapolated scattering coefficients for the second set of scatterer parameters based on the first set of scatterer parameters, the first and second frequencies, and the first and second times.
  • the processor may use the resulting second set of scatterer parameters to characterize the second communication channel, e.g., to determine channel estimates for the second communication channel.
  • the processor calculates the extrapolated scattering coefficients by rotating the phase of the scattering coefficients in the first set of scatterer parameters based on a frequency difference between the first and second frequencies and a time difference between the first and second times.
  • each scattering coefficient in the first set of scatterer parameters is associated with a rate of change of delay.
  • the processor calculates the extrapolated scattering coefficients based on the second frequency and extrapolated path delays calculated for the second set of scatter parameters based on the rate of change of delay and the path delays of the first set of scatterer parameters.
  • FIG. 1 shows the relationship between different scattering objects and different path delays relative to a transmitter and receiver in a wireless system.
  • FIG. 2 shows exemplary Doppler frequency vectors associated with the scattering objects relative to the receiver at a particular instant of time.
  • FIG. 3 shows one exemplary process for calculating extrapolated scatterer parameters for a transmission channel.
  • FIG. 4 shows one exemplary Continuous Transform process for calculating extrapolated scatterer parameters for a transmission channel.
  • FIG. 5 shows one exemplary Kalman process for calculating extrapolated scatterer parameters for a transmission channel.
  • FIG. 6 shows an exemplary transmitter
  • FIG. 7 shows an exemplary receiver according to an exemplary embodiment of the present invention.
  • FIG. 8 shows measured channel impulse responses obtained from field test data.
  • FIG. 9 maps the intensity of signal rays in the Doppler-delay domain.
  • FIG. 10 shows an exemplary method for determining path delays and the corresponding Doppler parameters according to the present invention.
  • FIGS. 11A and 11B show details for implementing the transform methods of FIG. 7 .
  • FIG. 12 shows the change in ⁇ due to time delay differences.
  • FIG. 13 shows one integrated Doppler process for tracking scatterer parameters according to one exemplary embodiment of the present invention.
  • FIG. 14 shows extrapolating delay slopes from one frequency band to another.
  • FIG. 15 shows one exemplary process for calculating extrapolated scatterer parameters for a multiple antenna element device.
  • FIG. 16 shows another exemplary process for calculating extrapolated scatterer parameters for a multiple antenna element device.
  • Transmitted signals traveling through wireless communication channels typically encounter several scattering objects, e.g., buildings, mountains, trees, etc., before reaching a receiver.
  • scattering object refers to both single scattering objects and clusters of scattering objects that are too close to separate. Each scattering object reflects the transmitted signal along a different propagation path. The receiver relies on accurate channel estimates to process the received signal. Channel estimation is the process of characterizing the effect of these scattering objects on the transmitted signal.
  • Channel estimates typically used by receivers can include complex coefficients of the channel impulse response at a number of equally-spaced delays, e.g., equally-spaced chip delays in a CDMA (Code Division Multiple Access) system, as well as complex coefficients indicative of the channel frequency response at a number of equally-spaced frequencies, e.g., equally-spaced subcarrier frequencies in an OFDM (Orthogonal Frequency Division Multiplex) system.
  • the receiver may use channel estimates approximating the effect of the channel on the transmitted signal for interference cancellation, diversity combining, ML detection, and other purposes.
  • Accurate channel estimation relies on the accurate characterization of the scattering objects within the channel.
  • the receiver determines the path delays, complex delay coefficients, and Doppler or rate-of-change of delay parameters corresponding to the scattering objects.
  • each scattering object in the multi-path channel corresponds to a different path delay.
  • FIG. 1 illustrates the scenario where different scattering objects cause reflected signals to have the same path delay even though the reflected signals traverse different paths.
  • FIG. 1 shows a plurality of ellipses surrounding a transmitter 12 and a receiver 14 , where the transmitter 12 and receiver 14 mark the foci of the ellipses, where a scattering object 10 falls on one of the ellipses, and where different ellipses correspond to different path delays.
  • the paths of different scattering objects 10 positioned along the same ellipse have the same path delay, while the paths of different scattering objects 10 positioned along different ellipses have different path delays.
  • scattering objects 10 a , 10 b both fall on ellipse 2
  • scattering object 10 c falls on ellipse 1 .
  • the path delay associated with path 18 c differs from the path delay associated with path 18 a or path 18 b, where path 18 a and path 18 b both have the same path delay. Because scattering objects 10 a and 10 b apply the same path delay to a transmitted signal, receiver 14 cannot use path delay alone to distinguish scattering object 10 a associated with path 18 a from scattering object 10 b associated path 18 b.
  • the receiver 14 may use another characterizing parameter, such as a Doppler parameter, to further characterize and distinguish the different scattering objects 10 .
  • a Doppler parameter such as a Doppler parameter
  • objects 10 with the same delay at different angular positions on the same delay-ellipse exhibit different Doppler shifts for moving a transmitter 12 and/or receiver 14 .
  • receiver 14 may determine the Doppler shift (or the equivalent rate of change of delay) to further distinguish between scattering objects 10 having identical path delays.
  • FIG. 2 shows an example of the different Doppler frequency vectors associated with the scattering objects relative to a moving receiver 14 at a particular instant of time.
  • receiver 14 may distinguish different paths 18 having the same path delay but caused by different scattering objects 10 .
  • receivers rely on the channel estimates to accurately process (e.g., decode/demodulate) received signals. If the transmitter also had a reliable characterization of the transmission channel, the transmitter could use such channel information to improve the reliability of communications between the transmitter and receiver.
  • errors in the reception channel estimates coupled with the differences between transmission and reception channels e.g., different transmission and reception times, different transmission and reception frequencies, etc., generally cause transmission channel estimates derived from reception channel estimates to be prohibitively inaccurate, and therefore, useless.
  • the present invention provides a method for accurately characterizing a transmission channel based on information obtained for a reception channel.
  • FIG. 3 shows one exemplary method 50 for characterizing a transmission channel.
  • a wireless device determines a first set of scatterer parameters for a reception channel based on received signal samples associated with a reception frequency and time (block 52 ).
  • the first set of scatterer parameters comprises a set of scattering coefficients for each of a plurality of path delays, where each scattering coefficient corresponds to a scattering object.
  • the device determines a second set of scatterer parameters for a transmission channel associated with a transmission frequency and time by calculating extrapolated scattering coefficients for the second set of scatterer parameters based on the first set of scatterer parameters, the reception and transmission frequencies, and the reception and transmission times (block 54 ).
  • reception and transmission channels generally applies to communication channels associated with different communication times and frequencies. It will further be appreciated that the present invention does not require the use of all reception channel scattering coefficients. Instead, in determining the scatterer parameters for the transmission channel, the processor may only use some of the reception channel scattering coefficients. The remaining reception channel scattering coefficients may be discarded or used as transmission channel scattering coefficients.
  • a processor determines the scatterer parameters for the reception channel based on the received signal samples.
  • the present invention may, for example, use the teachings of the Ser. No. 12/478473 application and/or the teachings of U.S. application Ser. No. 12/478520 to determine and/or track the scatterer parameters for a reception channel associated with a reception time and a reception frequency.
  • the scatterer parameters comprise a plurality of scattering coefficients and associated Doppler frequencies for each of a plurality of path delays.
  • the set of scattering coefficients and associated Doppler frequencies corresponding to a given path delay characterize the different scattering objects in the reception channel associated with that path delay.
  • the determined scatterer parameters collectively correspond to the reception time and frequency.
  • FIG. 4 shows one exemplary Continuous Transform method 60 for calculating the extrapolated scattering coefficients for the transmission channel associated with a transmission time and transmission frequency.
  • the processor determines a frequency difference between the reception and transmission frequencies (block 62 ), and determines rotated scattering coefficients by rotating a phase of the reception channel scattering coefficients based on the frequency difference (block 64 ).
  • the processor also determines a time difference between reception and transmission times (block 66 ), and calculates the extrapolated scattering coefficients by rotating a phase of the rotated scattering coefficients based on the time difference (block 68 ). While not shown in FIG.
  • the processor may also calculate extrapolated Doppler frequencies by determining the ratio between the transmission channel frequency and the reception channel frequency, and using the scaling factor to scale the reception channel Doppler frequencies.
  • the processor uses the resulting scatterer parameters to characterize the transmission channel, e.g., to determine channel estimates for the transmission channel.
  • the processor determines the scatterer parameters for the reception channel based on the received signal samples.
  • the present invention may, e.g., use the teachings of the Ser. No. 12/478473 application and/or the Ser. No. 12/478520 application to determine and/or track the scatterer parameters for a reception channel associated with a reception time and a reception frequency.
  • the scatterer parameters comprise a plurality of scattering coefficients and associated rate of change of delay for each of a plurality of path delays.
  • the scattering coefficients comprise reflection coefficients, which are independent of time and frequency, rotated by the time and frequency of the corresponding channel, where each reflection coefficient corresponds to a different scattering object, and where each reflection coefficient has an associated rate of change of delay.
  • the determined scatterer parameters collectively correspond to the reception time and frequency.
  • FIG. 5 shows one exemplary Kalman method 70 for calculating the extrapolated scattering coefficients for the transmission channel associated with a transmission time and transmission frequency using a Kalman filter process.
  • the processor calculates extrapolated path delays for the transmission channel based on the reception channel path delays and the corresponding rates of change of delay (block 72 ).
  • the processor then calculates the extrapolated scattering coefficients based on the extrapolated path delays and the transmission frequency by rotating a phase of the reception channel reflection coefficients based on the extrapolated path delays and the transmission channel frequency (block 74 ).
  • the Ser. No. 12/478473 invention determines the actual path delay and Doppler parameter information for a plurality of scattering objects in a wireless channel. Based on the path delay and Doppler frequency information, the receiver determines channel estimates useful for higher order modulation processing, channel prediction, etc. More particularly, the receiver 14 applies a frequency-to-time transform to a plurality of OFDM subcarrier signal samples received over a plurality of OFDM symbol periods to determine a set of non-equally spaced path delays and a set of associated complex delay coefficients.
  • the receiver 14 applies a time-to-frequency transform to the complex delay coefficients determined for individual path delays over multiple OFDM symbol periods to determine a set of Doppler parameters comprising a plurality of non-equally spaced Doppler frequencies and their corresponding scattering coefficients for individual path delays.
  • the transform operations of the Ser. No. 12/478473 invention are not constrained to determining output values at equally spaced time or frequency intervals, the non-equally spaced path delays, the associated complex delay coefficients, and the associated Doppler parameter sets have fewer errors than those produced by conventional techniques.
  • the transform operations described therein therefore fully characterize the scattering objects 10 while avoiding the accuracy problems of the prior art.
  • the increased accuracy of the resulting scattering object characterizations enable the receiver 14 to better track the channel estimates.
  • FIGS. 6 and 7 first show simplified internal details of an exemplary OFDM transmitter 12 and OFDM receiver 14 , respectively.
  • Transmitter 12 comprises an antenna 13 , Inverse Fourier transform unit 20 , parallel-to-serial converter 22 , modulator 24 , and power amplifier 26 .
  • the transmitter uses an Inverse Discrete Fourier Transform (IDFT) to encode symbols
  • the receiver uses a Discrete Fourier Transform (DFT) to decode signals.
  • IDFT Inverse Discrete Fourier Transform
  • DFT Discrete Fourier Transform
  • the IDFT and DFT may be interchanged, and are so similar that they are simply referred to herein as Fourier Transform units 20 .
  • Signal values to be transmitted S 1 . . .
  • Sn are input to the Fourier transform unit 20 which may be a specialized, hardwired FFT (or IFFT) circuit or a DSP implementation.
  • Parallel-to-serial unit 22 converts the output values of the Fourier transform unit 20 to serial form by selecting them successively in a fixed order. Each value is complex, so the serial stream comprises a stream of real parts and a stream of imaginary parts, i.e., a stream of (I, Q) values.
  • the stream of I-values and the stream of Q-values are converted to continuous-time I and Q signals by digital-to-analog conversion and filtering within modulator 24 .
  • the filter frequency response is required to pass frequencies corresponding to the used spectral bins, e.g., the 700 bins exemplified above, while attenuating frequencies beyond the exemplary 1024 bins.
  • the modulator 24 further uses the continuous-time I and Q signals to modulate cosine and sine wave carrier frequency signals, respectively, to generate an OFDM modulated radio frequency signal, which is amplified to a transmit power level in amplifier 26 and transmitted via antenna 13 .
  • FIG. 7 shows a receiver 14 according to one exemplary embodiment of the present invention.
  • Receiver 14 comprises an antenna 15 , front-end receiver elements (e.g., amplifier 30 , down converter 32 , serial-to-parallel converter 34 , and Fourier transform unit 36 ), channel processor 38 , and signal processor 40 .
  • the front-end receiver elements generate a plurality of signal samples corresponding to a plurality of frequencies from a signal received via antenna 15 . More particularly, amplifier 30 amplifies an OFDM symbol received via antenna 15 , and down converter 32 down converts the amplified OFDM symbol to the complex digital baseband.
  • the down converter 32 may comprise any known down converter that has means to select an operating frequency, means to filter the received signal to select the signal bandwidth centered on the selected operating frequency, and means to sample and analog-to-digital convert the filtered signal to generate the complex digital I, Q signals.
  • down converter 32 may comprise a zero-IF or homodyne down converter, a low-IF down converter, or a conventional superheterodyne down converter in which the final IF signal is demodulated by mixing with cosine and sine reference signal waveforms in a quadrature mixer arrangement.
  • Exemplary down converters include those described by U.S. Pat. No. 5,048,059 (reissued as U.S. Pat. No. RE37,138), U.S. Pat. Nos. 5,084,669, and 5,070,303.
  • the digital I, Q samples from the I, Q downconverter are then assembled into a block by serial-to-parallel converter 34 , which can for example comprise a DSP memory.
  • the block is then Fourier Transformed by Fourier transform unit 36 which is the reverse or conjugate process to the transmit Fourier transform unit 20 .
  • the output of Fourier transform unit 36 comprises the same number of samples as in the input block, which, with oversampling, is greater than n. Only n samples are used however, and the rest, which correspond to out-of-band spectral components not completely suppressed by the signal selection filters, are discarded.
  • the output samples ⁇ tilde over (S) ⁇ 1 to ⁇ tilde over (S) ⁇ n correspond to the samples input to the transmitter 12 , with the addition of transmission noise and any distortion effects caused by the propagation channel.
  • Channel processor 38 processes samples ⁇ tilde over (S) ⁇ 1 to ⁇ tilde over (S) ⁇ n to determine the channel estimates.
  • Signal processor 40 uses the channel estimates to process (e.g., decode) samples ⁇ tilde over (S) ⁇ 1 to ⁇ tilde over (S) ⁇ n to recover the transmitted data symbols S 1 to Sn.
  • channel processor 38 applies a frequency-to-time transform to the pilot samples within samples ⁇ tilde over (S) ⁇ 1 to ⁇ tilde over (S) ⁇ n to determine a set of non-equally spaced path delays and the corresponding complex delay coefficients.
  • the frequency-to-time transform may be jointly applied to a matrix of pilot symbols obtained from multiple OFDM symbol periods to determine a matrix of complex delay coefficients, where a given row of the delay coefficient matrix corresponds to a given OFDM symbol period, and where a given column of the delay coefficient matrix corresponds to a given path delay within the set of non-equally spaced path delays.
  • the joint operation may alternately be replaced with an individual operation, where the frequency-to-time transform is individually applied to individual sets of pilot samples from individual OFDM symbol periods.
  • different rows of the delay coefficient matrix are produced by individual frequency-to-time transform operations applied to signal samples from individual OFDM symbols, where different rows of the matrix correspond to different OFDM symbol periods.
  • the complex delay coefficients in each column of the resulting matrix generally correspond to a common path delay, plus or minus a small path delay differential. It will be appreciated that while the operation of jointly determining a common set of non-equally spaced delays that apply over multiple different OFDM symbol periods represents a preferred implementation, other implementations of the invention may determine the delay values independently for each OFDM symbol period.
  • channel processor 38 applies a time-to-frequency transform to individual columns of the delay coefficient matrix to determine a Doppler spectrum for each path delay.
  • the determination of a Doppler Spectrum from a column of delay coefficients for a given path delay presumes that the path delay is common to all the OFDM symbol periods of the column, and thus is optimum when the receiver determines the path delays jointly over multiple different OFDM symbol periods.
  • individual determination of the path delays could be used for each OFDM symbol period, providing the delay coefficients in individual columns of the resulting matrix are conformed to the same path delay prior to Doppler analysis of the column of complex delay coefficients.
  • This conforming operation may be achieved by rotating each delay coefficient in phase angle by W o dT, where W o represents the center frequency of the signal, and dT represents the amount of delay change needed to conform the path delay for a complex delay coefficient in a particular column to a common delay for that column.
  • each Doppler spectrum comprises a set of determined Doppler parameters, which each comprise a plurality of non-equally spaced Doppler frequencies and their corresponding complex scattering coefficients.
  • determining the Doppler spectra is not a joint operation over different path delays. Instead, the channel processor 38 determines the Doppler spectrum individually for a given path delay, e.g., delay column.
  • Collecting different sets of Doppler parameters determined for different ones of the non-equally spaced path delays into a matrix produces a Doppler parameter matrix, where a given column of the Doppler parameter matrix provides a set of Doppler parameters for a given path delay from the set of non-equally spaced path delays, and where each entry in the Doppler parameter matrix comprises at least a Doppler frequency and an associated complex coefficient for a particular path delay.
  • Channel processor 38 uses the non-equally spaced path delays and corresponding Doppler parameters to characterize the channel, e.g., to determine the channel estimates as described herein or according to any known means. Because the path delays and Doppler parameters have significantly more accuracy than those obtained from conventional approaches, the resulting channel estimates are also significantly more accurate, as discussed above.
  • the simplified receiver 14 of FIG. 7 was deliberately illustrated in the same form as the simplified transmitter 12 of FIG. 6 to explain how the transmitter 12 and receiver 14 processes are essentially inverses of each other, with the result that n complex samples (S 1 ,S 2 , . . . ,Sn) input to the transmitter 12 appear at the receiver output, effectively establishing n parallel channels of communication.
  • n complex samples S 1 ,S 2 , . . . ,Sn
  • Sn complex samples
  • S 1 ,S 2 , . . . ,Sn input to the transmitter 12 appear at the receiver output, effectively establishing n parallel channels of communication.
  • These are normally employed to send digital information, using a suitable modulation constellation to map bit patterns to points in the complex I, Q plane.
  • a practical OFDM communication system comprises many more details than shown in FIGS.
  • a wideband signal is produced by modulating a carrier frequency with a time-waveform that changes rapidly, in a short period that may be termed a modulation interval, a chip period, or the like. This is the shortest time period involved.
  • An OFDM symbol comprises a large number of such modulation intervals—at least as many as there are subcarrier frequencies in the OFDM symbol.
  • the set of modulation samples, spaced in time by the modulation interval is computed by Inverse Fourier Transforming a set of phases and amplitudes, one per subcarrier frequency.
  • Data symbols are encoded into the choice of each phase and amplitude by some chosen modulation scheme, such as 256 QAM, so that every subcarrier frequency carries a data symbol.
  • the total duration of the time-waveform output by the IFT is equal to the reciprocal of the subcarrier frequency spacing, and is called the OFDM symbol period.
  • This may be extended by appending a so-called cyclic prefix, but some OFDM systems, known as Pulse-Shaped OFDM, do not need to extend the duration of the OFDM symbol to accommodate a cyclic prefix.
  • the cyclic repeats of the OFDM symbol in pulse shaped OFDM symbols are permitted to overlap adjacent symbols, and therefore do not add a time-overhead. Therefore the potential use of a cyclic prefix is ignored for the rest of the discussion.
  • a number of OFDM symbols may be collected together over a total analysis time interval, the total analysis time interval therefore being an integral number of OFDM symbol periods.
  • the frequency domain of the signal comprises the frequency span from the first to the last OFDM subcarrier frequency used.
  • the OFDM signal also exists as a time waveform in the signal time domain, which is related to the signal frequency domain by the Fourier Transform.
  • a second frequency domain arises when looking at variations in signals arriving via scattered rays that are received from different objects with different Doppler shifts, due to having different relative velocities to the communicating station. If data symbol modulation is removed, the signal on any subcarrier frequency would still therefore be perceived to vary with time, and therefore possess a spectrum of finite width.
  • This Doppler spectrum exists in the frequency domain also, but is very narrow even compared to a single OFDM subcarrier frequency spacing. For example, a typical subcarrier frequency spacing is 15 kHz, while a typical Doppler spectrum is only 100-200 Hz wide.
  • the signal time variation that gives rise to the Doppler spectrum is from one OFDM symbol period to the next, and a total analysis interval of many OFDM symbol periods is required to resolve the Doppler spectrum.
  • the value of the amplitude and phase of a given sub-carrier frequency in a given OFDM symbol, ignoring data symbol modulation, is the result of the sum of many scattered waves of different phase and amplitude, and these may add constructively or destructively in each sub-carrier frequency bin. If the resultant phase and amplitude is plotted versus sub-carrier frequency, it will exhibit a frequency variation which is the channel frequency response. If the channel frequency response is inverse frequency transformed, the channel impulse response will be obtained. The impulse response indicates very approximately that the composite signal comprises the sum of a number of relatively delayed rays, and is a plot of phase and amplitude versus delay. This is therefore referred to as the Delay Domain.
  • Orthogonal Frequency Division Multiplexing is one method of reducing the complexity of equalizing methods needed to communicate high data rates in a multi-path channel.
  • OFDM Orthogonal Frequency Division Multiplexing
  • signals received from an OFDM transmitter 12 are applied to the Fourier transform unit 36 to produce a complex numerical value in each of the plurality of subcarrier frequency bins for each OFDM block period.
  • the receiver 14 may process data comprising a 1296 sub-carrier OFDM system having 15 kHz subcarrier frequency spacing, each OFDM symbol period thus being approximately 66.7 ⁇ s in duration (the reciprocal of 15 kHz).
  • the total occupied bandwidth of such a signal is a little over 1296 ⁇ 15 kHz or 19.44 MHz, and therefore, the symbol period is 1/19.44 MHz, or 51.44 ns.
  • a 2048-point Fourier transform unit 36 may be used, leaving a margin for filtering as described above.
  • every fourth subcarrier frequency contained a known pilot symbol, meaning that the phase of the pilot sub-carrier frequencies were set to values that were pre-agreed between receiver 14 and transmitter 12 .
  • 324 pilot symbols are transmitted per OFDM symbol interval of 66.7 ⁇ s. The pilot symbols are used to estimate the phase in adjacent channels by interpolation in both the frequency and time domain.
  • an average along the time domain of samples corresponding to the same frequency in the frequency domain was employed. This is typical of conventional channel estimation methods, but not the preferred implementation according to this invention.
  • the result is an estimate of the transmission channel phase and attenuation at equally spaced points spaced 60 kHz apart along the frequency domain.
  • These may be inverse frequency transformed to produce complex delay coefficients.
  • these complex delay coefficients may represent a first estimate of the channel impulse response of the channel.
  • the pilot symbols are equi-spaced in both a first half of the subcarriers and in a second half of the subcarriers, but the spacing between the first half and the second half is non-commensurate.
  • This can be handled by treating the first half and second half symbols as two separate symbols with frequency displacement between the pilots of one half and the pilots of the other half.
  • the method described herein can handle an arbitrary frequency displacement between the pilots of one symbol, or part symbol, and the pilots of another symbol, or part symbol, and still process them jointly to uncover a common set of scatterer delays.
  • symbols that have already been decoded are included in the calculation, it is possible that all of the OFDM subcarrier frequencies can be used, and not just those bearing known pilot symbols.
  • FIG. 8 shows typical values of the impulse response magnitudes computed from field test data recorded while driving through Swedish, Sweden. Only the magnitudes of the value for each delay bin are shown in FIG. 8 , although the values are actually complex.
  • the channel impulse response can be used to compute channel phases and amplitudes for the other subcarrier frequencies lying in between the pilot channels. These values are then used as channel phase references for the decoding of data symbols carried in the data-modulated subcarrier frequencies. As discussed above, noise or other impairments in these phase references may hinder the use of higher order modulations such as 256 QAM and/or the prediction of channel estimates for future time intervals.
  • Complex delay coefficients from successive time intervals may be stored to form a two dimensional array. Applying an individual time-to-frequency transform to the set of delay coefficients associated with individual delays results in a Doppler spectrum for each delay. The different Doppler spectrums for different delays may be collected into a new 2-D array called the Doppler parameter matrix. Because signal components are now separated by both delay and their relative velocities of the receiver 14 (or transmitter 12 ) with respect to their scattering objects 10 , which is related to the bearing between the receiver 14 and the scattering object 10 , the scattering objects 10 are now separated in two spatial dimensions (distance and angle) with the expectation that individual scattering objects 10 will now become resolvable. This indeed appears to be so, as shown in the plot of FIG.
  • Prony's algorithm implements a time-to-frequency transform that may be used to generate the Doppler parameter matrix without the assumption of equally spaced path delays or Doppler frequencies.
  • Prony's algorithm has been traditionally applied to diverse fields such as Linear Predictive Speech Coding, direction finding using antenna arrays, and spectral analysis in Nuclear Magnetic Resonance Spectroscopy.
  • a version in finite field arithmetic, known as the Massey-Berlekamp algorithm is used for decoding Reed-Solomon error-correcting codes.
  • the Prony algorithm is basically a method of spectral analysis that does not assume the spectrum falls into integrally-related frequency bins.
  • the Prony algorithm is specifically formulated to spectrally analyze finite time segments of a signal, and therefore, gives precise results.
  • the Prony algorithm decomposes the signal segment into a sum of exponentially decaying, exponentially growing, or static sinusoids which are all described by the expression:
  • Equation (2) When the signal waveform is recorded at equally spaced intervals of time idt, Equation (2) becomes:
  • Equation (3) Equation (3)
  • the Ser. No. 12/478473 invention provides an inverse modified Prony algorithm for one exemplary frequency-to-time transform, where the inverse modified Prony algorithm is adapted to accept an input comprising channel values taken at equally spaced sample frequencies along the frequency domain, as in the OFDM test system described above, and to produce an output of delay domain parameters, comprising non-equally spaced delays and their associated complex delay coefficients (phase and amplitude of a delayed signal).
  • the resulting non-equally spaced path delays are not restricted to multiples of any particular time interval, e.g., a 51.44 ns signal sampling period.
  • Equation (2) The inverse modified Prony algorithm described herein is different than that normally expressed by Equation (2) above. Normally, if one had obtained frequencies Z k and their associated amplitude/phase coefficients C k using Prony's analysis method, then the inverse, namely determining the signal at desired times t (other than the given times idt) would involve substituting the determined frequencies and coefficients into Equation (2). Equation (2) therefore represents the conventional inverse of the Prony frequency analysis procedure.
  • Equation (2) may also be written in term of Z k as:
  • the frequency-to-time transform described herein referred to herein as an inverse modified Prony Algorithm, comprises an algorithm for determining the delays T i and their associated complex delay coefficients S i,q in the equation:
  • Equation (7) Equation (7)
  • Equation (8) does not yet represent the inverse modified Prony Algorithm described herein, but it does represent the inverse of the inverse modified Prony Algorithm.
  • the inverse modified Prony Algorithm is a method of solving (e.g., inverting) Equation (8) for Z i and A i,q given C k,q , and using Equation (9) to determine the path delays T i once Z i has been determined.
  • the coefficients of this polynomial (p 0 ,p 1 ,p 2 , . . . , p M ) may be found by multiplying all factors (T ⁇ T 1 )(T ⁇ T 2 ) . . . (T ⁇ T M ), given the T 1 .
  • T i can be found by root-finding programs, which are well developed, reliable, and fast.
  • Equation (11) By interchanging the order of summation, Equation (11) becomes:
  • the desired solution should minimize the length of the error vector ⁇ . Because it is the direction of the p vector, and not the length, that determines the roots of the polynomial, the search for the best p using Equation (14) should search only direction space while keeping the length of p unchanged, e.g., setting
  • 1.
  • Equation (16) represents the definition of an eigenvector of a matrix, namely that the product of the matrix with a vector yields that same vector with just a length scaling by ⁇ .
  • the maxima and minima of ⁇ # ⁇ occur when p is an eigenvector of [C]#[C], and the associated eigenvalue is the value of ⁇ # ⁇ at that point.
  • the absolute minimum is thus obtained by choosing p to be the eigenvector associated with the smallest eigenvalue of [C]#[C].
  • Equations (13) and (14) can first be simplified by requiring that all roots of P(z) have unit magnitude. This is physically justified because the effect of a delayed ray adding to an undelayed ray is to create a sinusoidal variation of the channel frequency response along the frequency domain, and this sinusoidal curve is of constant amplitude, i.e., undamped.
  • Equation (14) Exploiting the conjugate palindromic property of P(z) is best achieved by writing out Equation (14) in terms of its real and imaginary parts, respectively indicated by R and I, as:
  • Equation (17) has the above form if M is even. When M is odd, there is a slight modification due to the center coefficient of the conjugate palindromic polynomial being real. Equation (17) has twice as many equations in the same number of variables, as before, but all quantities are real. Calling the matrix in Equation (17) Q, and denoting by q the vector of p-values arranged as in Equation (17), the error ⁇ is now minimized when q is the eigenvector of Q#Q associated with the smallest eigenvalue.
  • Equation (17) may be rewritten by reverse-ordering PR 0 . . . PR M/2 and PI 0 . . . PI M/2 , which makes the four partitions of each matrix Toeplitz instead of Hankel.
  • Equation (8) Equation (8)
  • Equation (19) is solved for each OFDM symbol period (or every few symbol periods if the values C k were the average over a number of successive symbol periods) to give the amplitude/phase coefficients of signals with the determined delays for each symbol period.
  • the inverse modified Prony Algorithm shall produce the same path delay estimates for each time period, at least for a total evaluation period over which the delays can be assumed to change by negligible amounts. For example, a change in delay of 5 ns may occur due to a movement of 5 feet by the transmitter 12 or receiver 14, which movement would occur over the relatively long time interval of 56.8 ms at 60 mph.
  • the inverse modified Prony algorithm still allows the amplitudes and phases of the signal for each delay value to be determined independently for each successive time interval within the evaluation period. Therefore, we want to find P(z) according to Equations (13), (14), and (17) by including all of the pilot symbol amplitude/phase values C k for all OFDM symbols in the evaluation period, the resulting path delays being those path delays that best explain the signal in all of the OFDM symbols. This is done by adding blocks vertically to the matrices of Equations (13), (14), and (17) for each symbol period. Denoting the matrix in Equation (17) by Q 1 for symbol period 1, Q 2 for symbol period 2, etc., the solution q that we seek is the eigenvector associated with the smallest eigenvalue of:
  • Equation (19) may then be used with the common Z-roots but with their individual C-values to determine the amplitude/phase coefficients A separately for each symbol period.
  • the values of the original amplitude/phase values S i,q may be computed according to:
  • the Doppler parameters may be obtained. This is what is shown in FIG. 9 . It is undesirable to constrain the Doppler frequencies to discrete bins. Thus, an application of Prony's frequency analysis procedure provides a preferred method of Doppler analysis. Further modifications to both the inverse modified Prony Algorithm for estimating the delay profile and the Prony Algorithm for estimating the Doppler spectrum are possible.
  • Equation (8) the least square error obtained is given by C#[I ⁇ Z(Z#Z) ⁇ 1 Z#]C.
  • [I ⁇ Z(Z#Z) ⁇ 1 Z#] G(G#G) ⁇ 1 G#.
  • C#G may be replaced by g#[C]#, where [C] comprises an (N ⁇ M) ⁇ (M+1) matrix given by:
  • [ C ] [ C 1 C 2 C 3 ... C M + 1 C 2 C 3 C 4 C M + 2 C 3 C 4 C 5 C M + 3 ⁇ ⁇ ⁇ ⁇ C N - M ... ... C N ] , ( 23 )
  • C#[I ⁇ Z(Z#Z) ⁇ 1 Z#]C g#[C]#(G#G) ⁇ 1 [C]g.
  • Minimizing this expression means that g should be chosen to be the eigenvector associated with the smallest eigenvalue of [C]#(G#G) ⁇ 1 [C], as opposed to [C]#[C] in the above-mentioned improvement to Prony's algorithm proposed in the Howard J. Price reference.
  • G depends on g
  • the procedure is to find an initial approximation for g using, for example, Prony's method or Price's method, which is the equivalent to initializing G#G to I, and then using that value of g to calculate G, followed by iteration. It is also desirable to restrict P, and therefore, g, to being conjugate palindromes, e.g.,
  • Equation (17) will have one more value of PR to find than values of PI.
  • Equation (17) the matrix [C] in Equation (17) in order to solve for the real and imaginary parts of the palindromic polynomial coefficients, and then iterated. Furthermore, it is desired that the same delay profile should hold for an evaluation period of L successive OFDM symbol periods, similar to Equation (20). Namely, letting QL equal the matrix [C] obtained from symbol period L, then the expression from Equation (20) should be used to obtain the starting value of g as the eigenvector associated with its smallest eigenvalue. Then g should be successively refined to be the eigenvector associated with the smallest eigenvalue of:
  • FIG. 10 provides a process 100 for determining path delays and Doppler parameters according to one exemplary embodiment of the Ser. No. 12/478473 invention.
  • process 100 includes applying a frequency-to-time transform to a plurality of received signal samples corresponding to a plurality of frequencies (e.g., OFDM signal samples in an OFDM symbol) to determine a set of non-equally spaced path delays and a set of associated complex delay coefficients (block 110 ).
  • Each of the non-equally spaced path delays and its associated complex delay coefficient correspond to one or more scattering objects of the wireless communication channel.
  • the complex delay coefficients and path delays are determined for multiple OFDM symbols in an evaluation period to provide a matrix of complex delay coefficients.
  • the complex delay coefficients in a given column of the matrix correspond to a given path delay in the set of non-equally spaced path delays, and the complex delay coefficients in a given row correspond to a given OFDM symbol in the evaluation period.
  • process 100 applies a time-to-frequency transform to the complex delay coefficients in a column of the delay coefficient matrix to determine a set of Doppler parameters for that path delay (block 150 ).
  • the Doppler parameters comprise a plurality of non-equally spaced Doppler frequencies and their corresponding complex scattering coefficients, where each Doppler frequency/complex scattering coefficient pair corresponds to a scattering object.
  • the sets of Doppler parameters are collected into a matrix of Doppler parameters, where a given column of the Doppler parameter matrix corresponds to a given path delay in the set of non-equally spaced path delays.
  • FIGS. 11A and 11B show details for implementing the frequency-to-time transform (block 110 ) and the time-to-frequency transform (block 150 ), respectively, according to one exemplary embodiment.
  • the frequency-to-time transform comprises the inverse modified Prony algorithm discussed above.
  • the time-to-frequency transform comprises the Prony algorithm discussed above. It will be appreciated, however, that the present invention is not limited to the Prony algorithms discussed herein. Instead, any transforms that determine the actual time and frequency data without restricting the time and frequency data to equally spaced and predefined bins may be used.
  • FIG. 11A shows the inverse modified Prony process 110 of the Ser. No. 12/478473 invention.
  • the channel processor 38 initializes a counter (block 112 ) to track the OFDM symbols of the evaluation period.
  • Channel processor 38 arranges the N C-values for symbol period q as per the matrix of Equation (14) to obtain the matrix Q q (block 120 ).
  • the channel processor 38 After Q q has been formed for the L OFDM symbols of the evaluation period, the channel processor 38 forms the matrix sum
  • the channel processor 38 finds the initial values of the conjugate palindromic polynomial coefficients g as the eigenvector of the Toeplitz+Hankel matrix partitioning associated with the smallest eigenvalue (block 126 ), uses g to form the matrix G, and computes (G#G) ⁇ 1 (block 128 ). Channel processor 38 then forms the matrix sum
  • ⁇ q 1 L ⁇ Q q # ⁇ ( G # ⁇ G ) - 1 ⁇ Q q ,
  • ⁇ q 1 L ⁇ g # ⁇ Q q # ⁇ ( G # ⁇ G ) - 1 ⁇ Q q ⁇ g
  • the channel processor 38 finds the roots z k of the conjugate palindromic polynomial P(z) whose coefficients are given by the converged value of g from block 134 (block 136 ).
  • Channel processor 38 uses the matrix [Y] computed for block 138 to find N coefficients A i,q for symbol period q from the corresponding coefficients C k,q obtained for block 118 for symbol period q, using Equation (1 9) (block 140 ).
  • the channel processor 38 converts the coefficient vector values A i,q into corresponding complex delay coefficient values S i,q by using Equation (21) (block 142 ).
  • Blocks 112 - 142 of FIG. 11A detail the inverse modified Prony algorithm, which provides an improved method of finding path delays and the associated complex delay coefficients that explain the observed channel frequency responses for one or more OFDM symbol periods. Further improvements are possible by performing a Doppler frequency analysis of the complex delay coefficients for each path delay over many symbol periods, using a second transform, namely a time-to-frequency transform such as provided by Prony's algorithm.
  • FIG. 11B proves details for the exemplary Prony process 150 .
  • the channel processor 38 applies the Prony algorithm to the L complex delay coefficients S i,q corresponding to one path delay. More particularly, channel processor 38 arranges the L, S-values for path delay i into a matrix (block 152 ), as shown by Equation (27).
  • [ S ] [ S i ⁇ 1 S i ⁇ 2 ... S i ⁇ m + 1 S i ⁇ 2 S i ⁇ 3 ... S i ⁇ m + 2 ⁇ ⁇ ⁇ S i ⁇ L - m ... ... S i ⁇ L ] ( 27 )
  • the channel processor 38 finds the eigenvector h associated with the smallest eigenvalue of [S]#[S] (block 154 ), constructs an L ⁇ L matrix [H] by using h in place of g in Equation (22) (block 156 ), and finds an improved version of h as the eigenvector associated with the smallest eigenvalue of [S]#[H#H] ⁇ 1 [S] (block 158 ).
  • the matrix [H#H] is Hermitian-symmetric Toeplitz, having only 2(L ⁇ m) ⁇ 1 distinct element values, which represent the autocorrelation function of h* with h.
  • the autocorrelation may be rapidly calculated by padding out h with zeros and using an FFT, according to known theory.
  • the inversion of the Hermitian-symmetric Toeplitz matrix [H#H] can be carried out in O(N 2 ) operations using Trench's algorithm.
  • its inverse has to multiply another Toeplitz/Hankel matrix S, whose rows are shifts of each other.
  • This multiplication may also be performed using fewer operations with the help of an FFT, as is well known, and by using the Gohberg-Semencul form of the inverse.
  • the Fourier Transform of the Toeplitz generating sequence for S may be saved and used again when the result is pre-multiplied by S#.
  • Doppler-delay diagrams have been obtained from cellular telephone signals before, as reported in “Estimation of scatterer locations from urban array channel measurements at 1800 MHz”, Henrik Asplundh and Jan-Erik Berg, conference proceedings of Radio Vetenskap och Varunikation, (RVK99), pp. 136-140, Karlskrona, Sweden June 1999. These diagrams were obtained using more conventional transform operations.
  • the Delay-Doppler analysis using Prony frequency analysis and the inverse modified Prony Algorithm described herein to compute a delay profile common to several successive OFDM symbols, is expected to yield superior results to those previously obtained.
  • the path delays and Doppler parameters obtained by the Ser. No. 12/478473 invention may be used to determine channel estimates having the requisite accuracy to process higher order modulation (e.g., 256 QAM) signals and/or to accurately predict future channel estimates.
  • An objective of determining more accurate delay-Doppler analysis is to enable more precise channel estimation for decoding data. Having obtained the more accurate delay-Doppler profile, it is now filtered by one or other methods to reduce or eliminate noise. For example, elements corresponding to impossibly high Doppler values can be set to zero. Likewise, impossible delays, such as negative delays, may be eliminated, as may delays of unlikely high values. Also, some number of spectral coefficients may be selected based on predetermined criteria. For example, the N largest spectral coefficients may be selected, or a threshold may be used to select the spectral coefficients.
  • MMSE Minimum Mean Square Error
  • a noise level in the Delay-Doppler diagram can be estimated either by reprocessing recently decoded data to determine the noise level on data symbols, or by looking at the values eliminated on the grounds of impossible or unlikely Doppler and delay. In the remaining values, it is not known whether a value of the same order as the RMS value of the estimated noise is a true scatterer or noise. Suppose the value is d.
  • N/S may be identified as the reciprocal of the signal-to-noise ratio that would be present on a Doppler-delay value of magnitude
  • 1.
  • N/S may be identified as the reciprocal of the signal-to-noise ratio that would be present on a Doppler-delay value of magnitude
  • 1.
  • the Doppler delay values should be scaled down in dependence on their signal to noise ratios and their amplitudes, such that smaller values are reduced much more than larger values. This is a “softer” method than just clipping out values below a threshold.
  • Other variations of the MMSE method may be used taking account of the correlation matrices between error sources. When the delay-Doppler coefficients scaled by MMSE are employed to re-compute the channel estimate for any OFDM symbol period and subcarrier frequency, substantial reduction of noise on the channel estimate will be evident.
  • the value of a Delay coefficient at some desired time instant is computed by summing terms comprising the Doppler-Delay coefficient rotated in phase by the time difference between the reference time and the desired time instant multiplied by the Doppler frequency. Then to find the channel estimate for any OFDM subcarrier frequency, the sum is computed of terms, each of which is the just-found Delay coefficient for the desired time instant rotated in phase by the product of the subcarrier frequency multiplied by the associated path delay.
  • One application of the Ser. No. 12/478473 invention is thus the reduction of noise on channel estimates to facilitate the use of higher order modulations for increasing data rate.
  • the invention may also be used for other purposes, for example, to predict the frequency response of the channel for some future symbol, or to determine the location of a mobile transmitter by performing the inventive analysis of the signal received from the mobile transmitter at a network station to identify a pattern of scattering objects 10 , the pattern then being compared with a previously stored data base of scattering object patterns using a pattern recognition algorithm.
  • the above methods could be used on a block-by-block basis, e.g., by processing a block of previous OFDM symbol waveform samples jointly with the signal samples for the current OFDM symbol to determine a set of scattering objects 10 pertinent to that block.
  • each OFDM symbol would need to be processed along with previous symbols, resulting in previous symbol data being reprocessed multiple times.
  • CSIPA Continuous Sequential Inverse Prony Algorithm
  • At least two versions of CSIPA can be envisaged: Moving Window, and Exponential Forgetting.
  • Moving Window and Exponential Forgetting.
  • Equation (26) To obtain the CSIP algorithm with exponential forgetting, rewrite Equation (26) to be an infinite sum with exponential de-weighting of older values, as shown in Equation (29).
  • ⁇ is a factor less than unity which de-weights older errors by a factor that reduces by ⁇ for each successive symbol period further into the past.
  • the matrix G can also be assumed to change only marginally from one symbol period to the next. This is justified because, in the block formulation, G was assumed constant over the block, and delays were assumed not to drift over the block. Which assumption is more accurately representative of reality is therefore moot. Moreover, traditional Prony analysis succeeds even with (G#G) ⁇ 1 omitted. Thus, using for each symbol period the value of G indexed with the index q is thus an improvement over the traditional Prony method and close to optimum. We thus obtain:
  • ⁇ L ⁇ L ⁇ 1 +Q# L ( G# L G L ) ⁇ 1 Q L . (31)
  • the CSIP algorithm over a moving window can be obtained similarly by adding the contribution at one end of the block while subtracting the oldest contribution from the other end, obtaining:
  • ⁇ L ⁇ L ⁇ 1 +Q# L ( G# L G L ) ⁇ 1 Q L ⁇ Q# L ⁇ m ( G# L ⁇ m G L ⁇ m ) ⁇ 1 Q L ⁇ m (32)
  • each delay value comprises only one dominant scatterer and therefore only one Doppler shift
  • the delay of that scatterer for period L will be its delay determined for period L ⁇ 1 minus the distance moved in its direction during one symbol period, divided by the speed of light. The distance moved in units of wavelengths is simply the Doppler frequency times the elapsed time, e.g.,
  • T i ⁇ ( L ) T i ⁇ ( L - 1 ) - ⁇ i ⁇ T ⁇ o , ( 33 )
  • T i (L) represents the delay of the i th scatterer for OFDM symbol period L
  • T i (L ⁇ 1) represents the delay of the i th scatterer for OFDM symbol period L ⁇ 1
  • T represents the ODFM symbol period
  • ⁇ i represents the Doppler shift of the i th scatterer
  • ⁇ o represents the transmission frequency.
  • each delay value comprises for example two scattering objects 10 of different Doppler shifts ⁇ q1 , ⁇ q2 , each contributing a complex portion D 1,q1 and D 1,q2 to the total delay coefficient S i,L ⁇ 1 for time period L ⁇ 1, then their contributions for time period L may be updated to D 1,q1 e j ⁇ i,q1 ⁇ T and D 1,q2 e j ⁇ i,q2 ⁇ T for period L, ⁇ T later.
  • the above method gives a method of defining the mean delay change, at least so long as the delays of the scattering objects 10 do not diverge so far as to become considered different delays.
  • the latter may be detected by maintaining the updated delays of individual scattering objects 10 by means of Equation (33) and determining whether the updated delay of any scatterer of one group has become closer to the delays of scattering objects 10 in a second group, at which point its contribution may be removed from the first group and added to the second group using similar vector combination procedures.
  • modified inverse Prony process is merely providing a long term correction or “nudge” of the delays towards their correct positions, while the short term corrections to the delays are provided by integrating their Doppler shifts.
  • This is analogous to a navigation system in which short term position changes are calculated by dead-reckoning using velocity times time, while eliminating drift error from that open-loop process by means of an occasional absolute position fix.
  • a per-symbol-period modified inverse Prony algorithm may be used to estimate delays (or g-coefficients), and g-coefficients estimated using previous values updated by Doppler are then moved towards the Prony values.
  • a limitation of per-symbol period inverse modified Prony is that the number of estimated path delays cannot exceed half the number of subcarriers in the OFDM symbol (e.g., about 160 for the 324-subcarrier test system).
  • a hybrid method of using past history may be to perform CSIPA over a moving rectangular window, using the results as a means of drift-correcting values obtained just before the window that have been updated using integrated Doppler. If delays are updated using integrated Doppler, this takes the place of updating Doppler based on the symbol-to-symbol variation of phase, and an additional means of estimating Doppler from symbol-period to symbol period would then be required. This is best described in the context of using an integrated Doppler approach in the form of a Kalman process to track delays and their derivatives, which may be formally equated with Doppler shifts.
  • a Kalman tracking process may be constructed by assuming that the channel at frequency ⁇ is explained by a sum of scattered waves from individual scattering objects 10 of strength S i,q at time qT. Each scattered wave is delayed in phase by propagating through a delay T i which is assumed to change linearly with time due to the mobile station's resolved velocity in its direction, leading to the formula for delay at time qT as:
  • ⁇ T is the (OFDM) symbol period.
  • a Kalman tracking filter seeks to successively refine estimates of the scatterer parameters based on a channel frequency response vector C( ⁇ ) of channel frequency response observations at different frequencies at the current time q.
  • An exemplary integrated Doppler process 200 corresponding to the operations of a Kalman tracking filter is shown in FIG. 13 .
  • the channel processor 38 predicts the scatterer parameters and corresponding channel frequency responses for the new symbol period associated with time q (block 210 ). For example, the channel processor 38 predicts the scatterer parameters by predicting the values of S i,q and T* i,q , which can collectively be denoted by a vector U, from the values of U at time q ⁇ 1.
  • Channel processor 38 then calculates an observed channel frequency response vector C( ⁇ ) based on the OFDM symbol received at time q (block 220 ), and computes an error vector ⁇ between the predicted C( ⁇ ) and the observed C( ⁇ ) (block 230 ). The channel processor 38 subsequently corrects the scatterer parameter prediction (block 240 ) in such a way as to reduce the error vector ⁇ .
  • the scatterer parameter prediction of block 210 may be written in matrix form as:
  • Equation (36) is often denoted by ⁇ .
  • correction of the predictions made in block 210 comprises the use of the inverse of the covariance matrix, commonly denoted by P. Since P is also initially unknown, it is also predicted and corrected based on receipt of new observations.
  • P may be predicted using:
  • P may be predicted using:
  • P may be predicted using:
  • the GRADient vectors for each ⁇ for which C( ⁇ ) will be observed are stacked side by side, the result is a 3 m ⁇ N matrix where N represents the number of scattering objects 10 and m represents the number of frequencies at which C is observed.
  • the GRADient is not constant but is a function of the parameters (S and T)
  • the process is known as a “linearized” or “extended” Kalman filter process.
  • the correction implemented in block 240 comprises updating the vector of parameters denoted by U using:
  • Another variation of the extended Kalman filter may be derived by defining the error between predicted and observed C( ⁇ ) vectors to be the scalar ⁇ # ⁇ .
  • the new gradient vector would then be 2 ⁇ #GRAD(C( ⁇ )).
  • the benefit of this formulation is that the matrix inverse [R+GRAD#PGRAD] ⁇ 1 is avoided, being replaced by division by a scalar 1+ ⁇ GRAD#PGRAD ⁇ .
  • the above extended Kalman filter can be used to track the scatterer parameters, e.g., delay, Doppler (rate of change of delay), and signal strength of already identified scattering objects 10 , and functions even when the updated delays of different scattering objects 10 move through one another in value.
  • a tracked scatterer may become distant and weak, while a previously untracked scatterer gets nearer and stronger.
  • a procedure intended to operate for long periods of time e.g., for minutes or hours, should include ways to discard scattering objects 10 that have become insignificant and to detect the appearance of and then track new scattering objects 10 .
  • the channel estimator 38 may track known scattering objects 10 using the extended Kalman filter described above, subtract out their contributions to the received signal to leave a residual which would be expected to comprise noise and untracked scattering objects 10 , and process the residual using the Prony-based algorithms of FIGS. 11A and 11B to identify new scattering objects 10 . Periodically, any scattering objects 10 identified by the Prony algorithm would be compared in strength to those being tracked by the Kalman algorithm. The channel estimator 38 replaces a weak, Kalman-tracked scatterer with a Prony-detected scatterer in the event of the latter becoming stronger.
  • the addition or deletion of a tracked scatterer from the Kalman algorithm occurs through deletion or addition of appropriate rows and columns of the P, Q, and ⁇ matrices. For example, if 128 scattering objects 10 are being tracked, and it is desired to discard scattering object number 79 , then rows and columns numbered 3 ⁇ 78+1, 3 ⁇ 78+2, and 3 ⁇ 78+3 are deleted from the P, Q, and ⁇ matrices. Conversely, if 127 scattering objects 10 are being tracked and it is desired to add a 128 th , then rows and columns 1 , 2 , and 3 of the Q and ⁇ matrices are copied to rows and columns 3 ⁇ 128+1, 3 ⁇ 128+2 and 3 ⁇ 128+3, extending the dimensions by 3.
  • the P-matrix is also extended by three rows and columns from 3 ⁇ 127 to 3 ⁇ 128 square.
  • the additional rows and columns are initialized to zero everywhere except on the main diagonal.
  • the main diagonal is initialized to values inversely indicative of the confidence in the initial values of the parameters S, T, and T* of the 128 th scattering object 10 . If the values are obtained from Prony analysis of the residual signal mentioned above, and are considered reasonably accurate, the three new diagonal P-elements may be initialized to zero, or to the mean of corresponding P-matrix elements for scattering objects 1 - 127 .
  • the block formulation of the Prony procedures may be appropriate to use the block formulation of the Prony procedures, such that the above-defined signal residuals are collected over a number of OFDM symbol periods and processed en block to detect the appearance of new scattering objects 10 .
  • the time over which new scattering objects 10 are expected to appear or disappear is of the order of the time required for the mobile station to move a few meters, which, at 70 mph, would be on the order of 100 ms.
  • older OFDM symbols may have been error-correction decoded, making observations of C( ⁇ ) available not only at the subcarrier frequencies of pilot symbols, but also at the data symbol frequencies.
  • any new scattering objects 10 thereby identified are compared with scattering objects 10 being tracked by the extended Kalman filter, and transferred from the Prony results to the Kalman tracking procedure by addition of new scatterer parameters, with or without deletion of scatterer parameters corresponding to scattering objects 10 that have become weak.
  • the following shows a procedure for processing a signal received continuously based on a collection of the algorithms disclosed above.
  • Doppler shift of each scattered wave caused by motion of either the transmitter of the receiver relative to the scattering objects 10 was furthermore determined either as a frequency shift, or more precisely, as a rate-of-change of delay.
  • Doppler shift as a rate-of-change of delay is sometimes preferable, as it is a description that is independent of the actual frequency, whereas a Doppler frequency shift caused by a given relative motion is proportional to the frequency on which it is measured.
  • FIG. 14 illustrates a number of phase slopes as determined across a first frequency band and extrapolated to a second frequency band.
  • the upper line corresponds to a path when an excess delay over line-of-sight equal to 3 ⁇ s (an excess path length of 900 m), which causes an additional phase shift of 3 ⁇ 2000 ⁇ 2 ⁇ at a frequency of 2000 MHz (the lower edge of the first frequency band) increasing to 3 ⁇ 2010 ⁇ 2 ⁇ at 2010 MHz (the upper edge of the first frequency band).
  • the line can be determined by best-fitting a straight line to the determined phase shifts, which are marked as x.
  • the lower line corresponds to an excess path delay of 0.15 ⁇ s (an excess path length of 45 m), or 0.15 ⁇ 2000 ⁇ 2 ⁇ in excess phase shift at 2000 MHz.
  • Both the upper and lower lines are then extrapolated to the second frequency band, which may for example be 2100 MHz to 2110 MHz. Due to the 10:1 extension of the line, any error in matching the phase/frequency points within the 10 MHz first frequency band will be multiplied by 10 upon reaching the second frequency band. Thus, if an accuracy of say ⁇ 10° is required at the second frequency band, an accuracy of ⁇ 1° must be achieved in the first frequency band.
  • the Ser. Nos. 12/478473 and 12/478520 applications provide several improvements to the art which each allow significant advance in delay determination.
  • the disclosure of a modified Inverse Prony Algorithm in the Ser. No. 12/478473 application estimates the best set of delays that explain a frequency response given at equally spaced frequencies across the first frequency band, and the use of a Joint Inverse Prony Algorithm determines the best set of delays that jointly explain a plurality of frequency responses measured for the first frequency band at successive instants at which the delays are expected to be constant.
  • a single transmitter having a single antenna is able to predict the propagation channel frequency response to a receiver before transmitting, it may be able to optimize its transmission to provide improved communications.
  • the receiver would need to know what it did in order to take advantage of the improvement, but it is possible that such information could be included in the transmission as an overhead or deduced by the receiver, for example by attempting decoding in a plurality of ways corresponding to the different ways the transmitter may have adapted to its foreknowledge of the frequency response.
  • one simple adaptation would be for the transmitter to choose a higher order modulation constellation for OFDM sub-carrier frequencies located on transmission peaks and a lower order modulation constellation, or else placing no data, on sub-carrier frequencies corresponding to troughs in the frequency response.
  • Such adaptation in its optimum form is sometimes known as “Waterfilling”. Since the receiver can determine the actual downlink channel upon receipt of the signal, it can surmise therefrom what the transmitter may likely have done and decode accordingly. In borderline cases, there may be some uncertainty for some sub-carrier frequencies requiring the receiver to attempt decoding in a plurality of ways.
  • the number of different decodings may be limited by reducing the number of combinations that the transmitter can select. For example, in the limit, the transmitter can be restricted to selecting either high order modulation or low order modulation for all sub-carrier frequencies, but less restrictive choices can also be devised, with the aim of reducing the number of decodings the receiver shall attempt.
  • the scattering parameter estimation is now extended to encompass multi-antenna scatterer estimation, which adds a third dimension to the Doppler and delay dimensions.
  • a first case of multi-antenna scattering parameter estimation arises with close element spacing such that the array does not generate significant grating lobes.
  • the different path delays to different antenna elements may be treated as identical path delays but with a different per-element phase shift, without ambiguity.
  • Joint estimation over antennas is done in the same way as joint estimation over multiple OFDM symbol periods, namely, by summing matrices related to each and finding the smallest eigenvalue/vector of the sum. In this way a joint estimation of delays may be done over all antenna elements and over multiple symbol periods. Having obtained scatterer delays by joint estimation as above, the associated complex coefficients are determined for each symbol period and antenna element. The Doppler spectrum for each path delay is then determined by joint estimation over the antenna elements, which are assumed to be of identical directivity. Antennas of identical directivity should be receiving each scattering object wave at equal strength and thus the expectation is that the Doppler spectrum will be the same for each.
  • the set of complex coefficients for different antenna elements but the same Doppler/delay combination is submitted to another Prony-type analysis to partition scattered waves by direction-of-arrival (d.o.a). Only two possible d.o.a.'s are anticipated for each Doppler/delay combination, therefore the polynomials found by Prony analysis will simply be quadratics.
  • a three-dimensional array of complex values is obtained. Negligible or unlikely values in the three-dimensional array can then be eliminated and the values down-selected to those values corresponding to physically likely scattering objects.
  • the values can be used by the transmitter to optimize its transmission to the same station from which it just received.
  • FIG. 15 shows one exemplary procedure 250 for translating the retained values from the reception frequency to the transmission frequency for a multiple antenna element device.
  • the transmitter rotates the phase of each value though an angle equal to its associated delay (the phase/frequency slope) times the frequency difference between the reception and transmission frequencies (block 252 ).
  • the transmitter further rotates the phase of each value though an additional angle equal to the associated Doppler frequency times the time difference between the reception time and the future transmission time, times the ratio of the transmission frequency to the reception frequency (block 254 ).
  • the transmitter then calculates the phase progression required across the antenna elements to obtain the same transmission direction as the reception direction for each scattering object 10 , based on the transmission and reception frequencies (block 256 ).
  • the following procedure 260 shown in FIG. 16 can be used.
  • the transmitter interprets Doppler as rate-of-change of delay, then the delay for each scattering object 10 is updated for the new time period by adding or subtracting an amount proportional to the Doppler, thus obtaining a new phase slope (block 262 ).
  • the transmitter uses the updated phase slopes from block 262 to extrapolate the phase from the reception frequencies to the transmission frequencies by adding to the reception phase shifts the product of the updated phase slope and frequency difference (block 264 ).
  • the transmitter calculates the phase progression required across the antenna elements to obtain the same transmission direction as the reception direction for each scattering object 10 , based on the transmission and reception frequencies (block 266 ).
  • the delay T(i) may now be updated to T(i)+dT(i) for the new period, where dT(i) is determined from velocity multiplied by time divided by the speed of light, and where the velocity is given by the Doppler shift for the i th scattering object 10 .
  • dT(i) is determined from velocity multiplied by time divided by the speed of light, and where the velocity is given by the Doppler shift for the i th scattering object 10 .
  • the transmitter may perform more advantageous coding or weighting of the signals transmitted by each antenna in order to optimize reception at the receiver, including steps which result in cancellation of interference at unintended receivers, MISO and MIMO systems or beamforming.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Power Engineering (AREA)
  • Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Mathematical Physics (AREA)
  • Transmitters (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Radar Systems Or Details Thereof (AREA)
  • Monitoring And Testing Of Transmission In General (AREA)
US12/478,564 2009-03-05 2009-06-04 Channel extrapolation from one frequency and time to another Abandoned US20100226448A1 (en)

Priority Applications (4)

Application Number Priority Date Filing Date Title
US12/478,564 US20100226448A1 (en) 2009-03-05 2009-06-04 Channel extrapolation from one frequency and time to another
EP10714471A EP2404419A2 (fr) 2009-03-05 2010-03-03 Extrapolation d'un canal à un autre à partir d'une fréquence et d'un temps
CN2010800114040A CN102342075A (zh) 2009-03-05 2010-03-03 从一个频率和时间到另一个的信道推断
PCT/IB2010/000438 WO2010100551A2 (fr) 2009-03-05 2010-03-03 Extrapolation d'un canal à un autre à partir d'une fréquence et d'un temps

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US15761509P 2009-03-05 2009-03-05
US12/478,564 US20100226448A1 (en) 2009-03-05 2009-06-04 Channel extrapolation from one frequency and time to another

Publications (1)

Publication Number Publication Date
US20100226448A1 true US20100226448A1 (en) 2010-09-09

Family

ID=42678236

Family Applications (3)

Application Number Title Priority Date Filing Date
US12/478,564 Abandoned US20100226448A1 (en) 2009-03-05 2009-06-04 Channel extrapolation from one frequency and time to another
US12/541,426 Active 2031-06-02 US8315335B2 (en) 2009-03-05 2009-08-14 OFDM spectral control
US12/562,225 Expired - Fee Related US8160176B2 (en) 2009-03-05 2009-09-18 Compensating pre-filter for an OFDM transmitter

Family Applications After (2)

Application Number Title Priority Date Filing Date
US12/541,426 Active 2031-06-02 US8315335B2 (en) 2009-03-05 2009-08-14 OFDM spectral control
US12/562,225 Expired - Fee Related US8160176B2 (en) 2009-03-05 2009-09-18 Compensating pre-filter for an OFDM transmitter

Country Status (4)

Country Link
US (3) US20100226448A1 (fr)
EP (3) EP2404418B1 (fr)
CN (1) CN102342075A (fr)
WO (3) WO2010100061A1 (fr)

Cited By (56)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100278277A1 (en) * 2009-04-30 2010-11-04 Electronics And Telecommunications Research Institute Decoding apparatus and method using orthogonal space-time block codes robust timing errors
US20120127030A1 (en) * 2009-01-27 2012-05-24 Ohio University Tight optical integration (toi) of images with gps range measurements
US20120128042A1 (en) * 2010-11-23 2012-05-24 Ching-Kae Tzou Scattering-parameter estimation method and transceiver using the same
US8963764B1 (en) 2011-01-14 2015-02-24 Lockheed Martin Corporation Ship heading and pitch using satellite ephemerides and radar range measurement of satellite
WO2015073905A3 (fr) * 2013-11-14 2015-11-12 The Board Of Trustees Of The Leland Stanford Junior University Estimation de rétrodiffusion utilisant une annulation progressive d'auto-interférence
US9276682B2 (en) 2014-05-23 2016-03-01 Kumu Networks, Inc. Systems and methods for multi-rate digital self-interference cancellation
US9325432B2 (en) 2012-02-08 2016-04-26 The Board Of Trustees Of The Leland Stanford Junior University Systems and methods for full-duplex signal shaping
US9331737B2 (en) 2012-02-08 2016-05-03 The Board Of Trustees Of The Leland Stanford Junior University Systems and methods for cancelling interference using multiple attenuation delays
US9337885B2 (en) 2013-12-12 2016-05-10 Kumu Networks, Inc. Systems and methods for hybrid self-interference cancellation
US9455756B2 (en) 2013-08-09 2016-09-27 Kumu Networks, Inc. Systems and methods for frequency independent analog self-interference cancellation
US9459344B1 (en) * 2011-01-14 2016-10-04 Lockheed Martin Corporation Ship position and velocity using satellite ephemerides and radar range measurement of satellite
US9520983B2 (en) 2013-09-11 2016-12-13 Kumu Networks, Inc. Systems for delay-matched analog self-interference cancellation
US9521023B2 (en) 2014-10-17 2016-12-13 Kumu Networks, Inc. Systems for analog phase shifting
US9634823B1 (en) 2015-10-13 2017-04-25 Kumu Networks, Inc. Systems for integrated self-interference cancellation
US9667299B2 (en) 2013-08-09 2017-05-30 Kumu Networks, Inc. Systems and methods for non-linear digital self-interference cancellation
US9673854B2 (en) 2015-01-29 2017-06-06 Kumu Networks, Inc. Method for pilot signal based self-inteference cancellation tuning
US9698860B2 (en) 2013-08-09 2017-07-04 Kumu Networks, Inc. Systems and methods for self-interference canceller tuning
US9712313B2 (en) 2014-11-03 2017-07-18 Kumu Networks, Inc. Systems for multi-peak-filter-based analog self-interference cancellation
US9712312B2 (en) 2014-03-26 2017-07-18 Kumu Networks, Inc. Systems and methods for near band interference cancellation
US9742593B2 (en) 2015-12-16 2017-08-22 Kumu Networks, Inc. Systems and methods for adaptively-tuned digital self-interference cancellation
US9755692B2 (en) 2013-08-14 2017-09-05 Kumu Networks, Inc. Systems and methods for phase noise mitigation
US9774405B2 (en) 2013-12-12 2017-09-26 Kumu Networks, Inc. Systems and methods for frequency-isolated self-interference cancellation
US9819325B2 (en) 2015-12-16 2017-11-14 Kumu Networks, Inc. Time delay filters
US9887728B2 (en) 2011-02-03 2018-02-06 The Board Of Trustees Of The Leland Stanford Junior University Single channel full duplex wireless communications
US9979374B2 (en) 2016-04-25 2018-05-22 Kumu Networks, Inc. Integrated delay modules
US10103774B1 (en) 2017-03-27 2018-10-16 Kumu Networks, Inc. Systems and methods for intelligently-tuned digital self-interference cancellation
US10177836B2 (en) 2013-08-29 2019-01-08 Kumu Networks, Inc. Radio frequency self-interference-cancelled full-duplex relays
US10230422B2 (en) 2013-12-12 2019-03-12 Kumu Networks, Inc. Systems and methods for modified frequency-isolation self-interference cancellation
US10236922B2 (en) 2017-03-27 2019-03-19 Kumu Networks, Inc. Systems and methods for tunable out-of-band interference mitigation
US10243719B2 (en) 2011-11-09 2019-03-26 The Board Of Trustees Of The Leland Stanford Junior University Self-interference cancellation for MIMO radios
US10284356B2 (en) 2011-02-03 2019-05-07 The Board Of Trustees Of The Leland Stanford Junior University Self-interference cancellation
US10338205B2 (en) 2016-08-12 2019-07-02 The Board Of Trustees Of The Leland Stanford Junior University Backscatter communication among commodity WiFi radios
US10382085B2 (en) 2017-08-01 2019-08-13 Kumu Networks, Inc. Analog self-interference cancellation systems for CMTS
US10404297B2 (en) 2015-12-16 2019-09-03 Kumu Networks, Inc. Systems and methods for out-of-band interference mitigation
US10425115B2 (en) 2018-02-27 2019-09-24 Kumu Networks, Inc. Systems and methods for configurable hybrid self-interference cancellation
US10454444B2 (en) 2016-04-25 2019-10-22 Kumu Networks, Inc. Integrated delay modules
US20200112928A1 (en) * 2017-06-06 2020-04-09 Supply, Inc. Method and system for wireless power delivery
US10666305B2 (en) 2015-12-16 2020-05-26 Kumu Networks, Inc. Systems and methods for linearized-mixer out-of-band interference mitigation
US10673519B2 (en) 2013-08-29 2020-06-02 Kuma Networks, Inc. Optically enhanced self-interference cancellation
US10778044B2 (en) 2018-11-30 2020-09-15 Supply, Inc. Methods and systems for multi-objective optimization and/or wireless power delivery
US10811908B2 (en) 2014-09-25 2020-10-20 Supply, Inc. System and method for wireless power reception
US10827445B2 (en) 2017-06-06 2020-11-03 Supply, Inc. Method and system for wireless power delivery
US10868661B2 (en) 2019-03-14 2020-12-15 Kumu Networks, Inc. Systems and methods for efficiently-transformed digital self-interference cancellation
US20210028877A1 (en) * 2015-07-12 2021-01-28 Cohere Technologies, Inc. Orthogonal time frequency space modulation over a plurality of narrow band subcarriers
US10952163B2 (en) 2018-11-28 2021-03-16 Supply, Inc. System and method for wireless power delivery
CN112543074A (zh) * 2019-09-23 2021-03-23 清华大学深圳国际研究生院 一种非视距通信信道建模方法
US11159349B2 (en) * 2017-12-27 2021-10-26 Orange Method for estimating the channel between a transceiver and a mobile communicating object
US11163050B2 (en) * 2013-08-09 2021-11-02 The Board Of Trustees Of The Leland Stanford Junior University Backscatter estimation using progressive self interference cancellation
US11178625B2 (en) 2017-06-06 2021-11-16 Supply, Inc. Method and system for wireless power delivery
US11183886B2 (en) 2018-03-08 2021-11-23 Supply, Inc. Method and system for wireless power delivery
US11211969B2 (en) 2017-03-27 2021-12-28 Kumu Networks, Inc. Enhanced linearity mixer
US11209536B2 (en) * 2014-05-02 2021-12-28 The Board Of Trustees Of The Leland Stanford Junior University Method and apparatus for tracking motion using radio frequency signals
US11282531B2 (en) * 2020-02-03 2022-03-22 Bose Corporation Two-dimensional smoothing of post-filter masks
US11483836B2 (en) 2016-10-25 2022-10-25 The Board Of Trustees Of The Leland Stanford Junior University Backscattering ambient ism band signals
US11611242B2 (en) 2021-04-14 2023-03-21 Reach Power, Inc. System and method for wireless power networking
US11991034B1 (en) 2022-11-09 2024-05-21 Qualcomm Incorporated Sample-level error-correcting code

Families Citing this family (14)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2008054322A2 (fr) * 2006-11-02 2008-05-08 Telefonaktiebolaget Lm Ericsson (Publ) Système ofdm étalé de dft
TWI382683B (zh) * 2009-05-06 2013-01-11 Ind Tech Res Inst 預補償光纖色散所引起的延遲的方法、應用該方法的多子載波訊號產生器、以及應用該訊號產生器的光正交分頻多工系統之傳送器
US8995569B2 (en) * 2011-12-30 2015-03-31 Samsung Electronics Co., Ltd. Quadrature digital-IF transmitter without inter-stage SAW filter and devices using same
US8693561B2 (en) * 2012-03-16 2014-04-08 Posedge Inc. Receive signal detection of multi-carrier signals
US8873655B2 (en) * 2013-01-10 2014-10-28 Intel Corporation Sending information at a band edge within an orthogonal frequency-division multiplexing (OFDM) symbol
US9252823B2 (en) * 2013-08-06 2016-02-02 Purdue Research Foundation Phase compensation filtering for multipath wireless systems
US10797918B2 (en) * 2015-07-06 2020-10-06 Telefonaktiebolaget Lm Ericsson (Publ) Resource allocation for data transmission in wireless systems
DE102015115760B4 (de) * 2015-09-18 2019-06-06 Intel IP Corporation Vorrichtung und verfahren zum verringern einertiefpassfiltergruppenverzögerung
ES2909839T3 (es) * 2015-10-01 2022-05-10 Sony Group Corp Dispositivo, método y programa
EP3427520B1 (fr) * 2016-03-07 2020-02-19 Telefonaktiebolaget LM Ericsson (PUBL) Communication fiable à des récepteurs de détection d'énergie
US10601624B2 (en) * 2018-07-17 2020-03-24 Allen Le Roy Limberg COFDM DCM signaling that employs labeling diversity to minimize PAPR
CN112314007B (zh) * 2018-04-20 2024-04-23 瑞典爱立信有限公司 使用混叠来进行信号的节能传输和接收的方法和设备
CN111010249B (zh) * 2019-12-23 2021-03-02 华中科技大学 一种角度时延域信道预测方法、预测系统及应用
WO2021145956A1 (fr) * 2020-01-15 2021-07-22 Ast & Science, Llc Système à signal modulé permettant de compenser des erreurs de fréquence dans des signaux lte

Citations (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5959967A (en) * 1995-08-19 1999-09-28 Northern Telecom Limited Digital transmission system
US6175550B1 (en) * 1997-04-01 2001-01-16 Lucent Technologies, Inc. Orthogonal frequency division multiplexing system with dynamically scalable operating parameters and method thereof
US20020097686A1 (en) * 2000-11-20 2002-07-25 Qiu Robert C. Long-range prediction of fading signals for WCDMA high speed downlink packet access (HSDPA)
US6456653B1 (en) * 1999-08-25 2002-09-24 Lucent Technologies Inc. Fast and accurate signal-to-noise ratio estimation technique for OFDM systems
US20030103578A1 (en) * 2001-12-04 2003-06-05 Yeh Alex C. Method and system for determining tap gain values for a transmit frequency domain equalizer to achieve unity power gain
US6608529B2 (en) * 2001-06-28 2003-08-19 Intel Corporation Frequency synthesis apparatus, systems, and methods
US20030215007A1 (en) * 2002-05-02 2003-11-20 Mitsubishi Denki Kabushiki Kaisha Method for performing an equalisation per carrier in a MC-DMA receiver
US20040252632A1 (en) * 2002-08-22 2004-12-16 Andre Bourdoux Method and apparatus for multi-user multi-input multi-output transmission
US20050118977A1 (en) * 2003-12-02 2005-06-02 Drogi Serge F. Method, apparatus, and systems for digital radio communication systems
US20050163257A1 (en) * 2004-01-28 2005-07-28 Keerthi Arvind V. Channel estimation for a communication system using spectral estimation
US20060019601A1 (en) * 2004-07-26 2006-01-26 Ibiquity Digital Corporation Method and apparatus for blending an audio signal in an in-band on-channel radio system
US7023929B2 (en) * 2002-07-10 2006-04-04 Texas Instruments Incorporated Digital pre-compensation filter for DMT type transmitter
US20060250935A1 (en) * 2003-02-28 2006-11-09 Ntt Docomo, Inc Radio communication system and radio communication method
US20070211747A1 (en) * 2006-02-21 2007-09-13 Samsung Electronics Co., Ltd. Adaptive channel prediction apparatus and method for performing uplink pre-equalization depending on downlink channel variation in OFDM/TDD mobile communication system
US7339918B2 (en) * 2000-06-26 2008-03-04 Nokia Corporation Method for improving the quality of data transmission
US20080159422A1 (en) * 2007-01-03 2008-07-03 Freescale Semiconductor Inc. Reducing a peak-to-average ratio of a signal
US20080240265A1 (en) * 2007-04-02 2008-10-02 Infineon Technologies Ag System having an ofdm channel estimator
US7450532B2 (en) * 2003-12-05 2008-11-11 Samsung Electronics Co., Ltd Apparatus and method for transmitting data by selected eigenvector in closed loop MIMO mobile communication system
US20080318613A1 (en) * 2007-06-20 2008-12-25 Kumar Balachandran System and apparatus for interference suppression using macrodiversity in mobile wireless networks
US20090054012A1 (en) * 2007-08-22 2009-02-26 Realtek Semiconductor Corp. Transmitter and Transmission Method Thereof
US20090092193A1 (en) * 2005-11-16 2009-04-09 Matsushita Electric Industrial Co., Ltd. Multi carrier transmission device, multi carrier reception device, and communication method
US8130858B1 (en) * 2007-05-30 2012-03-06 Marvell International Ltd. Method and apparatus for implementing transmit diversity in OFDM systems

Family Cites Families (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US864743A (en) 1907-04-25 1907-08-27 Charles E Johnson Padlock.
GB2131785B (en) 1982-11-24 1986-09-03 Exxon Research Engineering Co Corrosion inhibitors
SE463540B (sv) 1988-09-19 1990-12-03 Ericsson Telefon Ab L M Saett foer att i ett radiokommunikationssystem digitalisera godtyckliga radiosignaler samt anordning foer utoevande av saettet
US5084669A (en) 1990-03-08 1992-01-28 Telefonaktiebolaget L M Ericsson Direct phase digitization
US5070303A (en) 1990-08-21 1991-12-03 Telefonaktiebolaget L M Ericsson Logarithmic amplifier/detector delay compensation
US6278867B1 (en) 1998-11-25 2001-08-21 Ericsson Inc. Methods and systems for frequency generation for wireless devices
US6996380B2 (en) 2001-07-26 2006-02-07 Ericsson Inc. Communication system employing transmit macro-diversity
US6996375B2 (en) 2001-07-26 2006-02-07 Ericsson Inc. Transmit diversity and separating multiple loopback signals
US6868276B2 (en) * 2003-06-17 2005-03-15 Nokia Corporation Method and apparatus for estimating carrier frequency offset and fading rate using autoregressive channel modeling
US7308286B2 (en) * 2003-11-20 2007-12-11 Telefonaktiebolaget Lm Ericsson (Publ) Multi-dimensional joint searcher and channel estimators
US7668265B2 (en) 2004-07-01 2010-02-23 Texas Instruments Incorporated Ultra wideband interference cancellation for orthogonal frequency division multiplex transmitters by protection-edge tones
WO2007092945A2 (fr) 2006-02-08 2007-08-16 Qualcomm Incorporated Mise en forme spectrale destinee a reduire le rapport valeur de crête sur valeur moyenne dans une communication sans fil
US7599418B2 (en) * 2006-02-16 2009-10-06 Pine Valley Investments, Inc. Method and apparatus for a frequency hopper
US7729602B2 (en) * 2007-03-09 2010-06-01 Eastman Kodak Company Camera using multiple lenses and image sensors operable in a default imaging mode
US8068797B2 (en) * 2007-09-28 2011-11-29 Freescale Semiconductor, Inc. Gain control methods for wireless devices and transmitters

Patent Citations (22)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5959967A (en) * 1995-08-19 1999-09-28 Northern Telecom Limited Digital transmission system
US6175550B1 (en) * 1997-04-01 2001-01-16 Lucent Technologies, Inc. Orthogonal frequency division multiplexing system with dynamically scalable operating parameters and method thereof
US6456653B1 (en) * 1999-08-25 2002-09-24 Lucent Technologies Inc. Fast and accurate signal-to-noise ratio estimation technique for OFDM systems
US7339918B2 (en) * 2000-06-26 2008-03-04 Nokia Corporation Method for improving the quality of data transmission
US20020097686A1 (en) * 2000-11-20 2002-07-25 Qiu Robert C. Long-range prediction of fading signals for WCDMA high speed downlink packet access (HSDPA)
US6608529B2 (en) * 2001-06-28 2003-08-19 Intel Corporation Frequency synthesis apparatus, systems, and methods
US20030103578A1 (en) * 2001-12-04 2003-06-05 Yeh Alex C. Method and system for determining tap gain values for a transmit frequency domain equalizer to achieve unity power gain
US20030215007A1 (en) * 2002-05-02 2003-11-20 Mitsubishi Denki Kabushiki Kaisha Method for performing an equalisation per carrier in a MC-DMA receiver
US7023929B2 (en) * 2002-07-10 2006-04-04 Texas Instruments Incorporated Digital pre-compensation filter for DMT type transmitter
US20040252632A1 (en) * 2002-08-22 2004-12-16 Andre Bourdoux Method and apparatus for multi-user multi-input multi-output transmission
US20060250935A1 (en) * 2003-02-28 2006-11-09 Ntt Docomo, Inc Radio communication system and radio communication method
US20050118977A1 (en) * 2003-12-02 2005-06-02 Drogi Serge F. Method, apparatus, and systems for digital radio communication systems
US7450532B2 (en) * 2003-12-05 2008-11-11 Samsung Electronics Co., Ltd Apparatus and method for transmitting data by selected eigenvector in closed loop MIMO mobile communication system
US20050163257A1 (en) * 2004-01-28 2005-07-28 Keerthi Arvind V. Channel estimation for a communication system using spectral estimation
US20060019601A1 (en) * 2004-07-26 2006-01-26 Ibiquity Digital Corporation Method and apparatus for blending an audio signal in an in-band on-channel radio system
US20090092193A1 (en) * 2005-11-16 2009-04-09 Matsushita Electric Industrial Co., Ltd. Multi carrier transmission device, multi carrier reception device, and communication method
US20070211747A1 (en) * 2006-02-21 2007-09-13 Samsung Electronics Co., Ltd. Adaptive channel prediction apparatus and method for performing uplink pre-equalization depending on downlink channel variation in OFDM/TDD mobile communication system
US20080159422A1 (en) * 2007-01-03 2008-07-03 Freescale Semiconductor Inc. Reducing a peak-to-average ratio of a signal
US20080240265A1 (en) * 2007-04-02 2008-10-02 Infineon Technologies Ag System having an ofdm channel estimator
US8130858B1 (en) * 2007-05-30 2012-03-06 Marvell International Ltd. Method and apparatus for implementing transmit diversity in OFDM systems
US20080318613A1 (en) * 2007-06-20 2008-12-25 Kumar Balachandran System and apparatus for interference suppression using macrodiversity in mobile wireless networks
US20090054012A1 (en) * 2007-08-22 2009-02-26 Realtek Semiconductor Corp. Transmitter and Transmission Method Thereof

Cited By (90)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9726765B2 (en) * 2009-01-27 2017-08-08 Ohio University Tight optical integration (TOI) of images with GPS range measurements
US20120127030A1 (en) * 2009-01-27 2012-05-24 Ohio University Tight optical integration (toi) of images with gps range measurements
US20100278277A1 (en) * 2009-04-30 2010-11-04 Electronics And Telecommunications Research Institute Decoding apparatus and method using orthogonal space-time block codes robust timing errors
US8189702B1 (en) * 2010-11-23 2012-05-29 Ralink Technology Corp. Scattering-parameter estimation method and transceiver using the same
US20120128042A1 (en) * 2010-11-23 2012-05-24 Ching-Kae Tzou Scattering-parameter estimation method and transceiver using the same
US8963764B1 (en) 2011-01-14 2015-02-24 Lockheed Martin Corporation Ship heading and pitch using satellite ephemerides and radar range measurement of satellite
US9459344B1 (en) * 2011-01-14 2016-10-04 Lockheed Martin Corporation Ship position and velocity using satellite ephemerides and radar range measurement of satellite
US10230419B2 (en) 2011-02-03 2019-03-12 The Board Of Trustees Of The Leland Stanford Junior University Adaptive techniques for full duplex communications
US9887728B2 (en) 2011-02-03 2018-02-06 The Board Of Trustees Of The Leland Stanford Junior University Single channel full duplex wireless communications
US10284356B2 (en) 2011-02-03 2019-05-07 The Board Of Trustees Of The Leland Stanford Junior University Self-interference cancellation
US10243719B2 (en) 2011-11-09 2019-03-26 The Board Of Trustees Of The Leland Stanford Junior University Self-interference cancellation for MIMO radios
US9325432B2 (en) 2012-02-08 2016-04-26 The Board Of Trustees Of The Leland Stanford Junior University Systems and methods for full-duplex signal shaping
US9331737B2 (en) 2012-02-08 2016-05-03 The Board Of Trustees Of The Leland Stanford Junior University Systems and methods for cancelling interference using multiple attenuation delays
US10243718B2 (en) 2012-02-08 2019-03-26 The Board Of Trustees Of The Leland Stanford Junior University Systems and methods for full-duplex signal shaping
US9667299B2 (en) 2013-08-09 2017-05-30 Kumu Networks, Inc. Systems and methods for non-linear digital self-interference cancellation
US9832003B2 (en) 2013-08-09 2017-11-28 Kumu Networks, Inc. Systems and methods for self-interference canceller tuning
US11163050B2 (en) * 2013-08-09 2021-11-02 The Board Of Trustees Of The Leland Stanford Junior University Backscatter estimation using progressive self interference cancellation
US10050659B2 (en) 2013-08-09 2018-08-14 Kumu Networks, Inc. Systems and methods for non-linear digital self-interference cancellation
US9698860B2 (en) 2013-08-09 2017-07-04 Kumu Networks, Inc. Systems and methods for self-interference canceller tuning
US9455756B2 (en) 2013-08-09 2016-09-27 Kumu Networks, Inc. Systems and methods for frequency independent analog self-interference cancellation
US9755692B2 (en) 2013-08-14 2017-09-05 Kumu Networks, Inc. Systems and methods for phase noise mitigation
US11637623B2 (en) 2013-08-29 2023-04-25 Kumu Networks, Inc. Optically enhanced self-interference cancellation
US10979131B2 (en) 2013-08-29 2021-04-13 Kumu Networks, Inc. Self-interference-cancelled full-duplex relays
US10673519B2 (en) 2013-08-29 2020-06-02 Kuma Networks, Inc. Optically enhanced self-interference cancellation
US10177836B2 (en) 2013-08-29 2019-01-08 Kumu Networks, Inc. Radio frequency self-interference-cancelled full-duplex relays
US9520983B2 (en) 2013-09-11 2016-12-13 Kumu Networks, Inc. Systems for delay-matched analog self-interference cancellation
WO2015073905A3 (fr) * 2013-11-14 2015-11-12 The Board Of Trustees Of The Leland Stanford Junior University Estimation de rétrodiffusion utilisant une annulation progressive d'auto-interférence
US9337885B2 (en) 2013-12-12 2016-05-10 Kumu Networks, Inc. Systems and methods for hybrid self-interference cancellation
US10230422B2 (en) 2013-12-12 2019-03-12 Kumu Networks, Inc. Systems and methods for modified frequency-isolation self-interference cancellation
US9774405B2 (en) 2013-12-12 2017-09-26 Kumu Networks, Inc. Systems and methods for frequency-isolated self-interference cancellation
US9712312B2 (en) 2014-03-26 2017-07-18 Kumu Networks, Inc. Systems and methods for near band interference cancellation
US11209536B2 (en) * 2014-05-02 2021-12-28 The Board Of Trustees Of The Leland Stanford Junior University Method and apparatus for tracking motion using radio frequency signals
US9455761B2 (en) 2014-05-23 2016-09-27 Kumu Networks, Inc. Systems and methods for multi-rate digital self-interference cancellation
US9276682B2 (en) 2014-05-23 2016-03-01 Kumu Networks, Inc. Systems and methods for multi-rate digital self-interference cancellation
US11211826B2 (en) 2014-09-25 2021-12-28 Supply, Inc. System and method for wireless power reception
US11742700B2 (en) 2014-09-25 2023-08-29 Reach Power, Inc. System and method for wireless power reception
US10811908B2 (en) 2014-09-25 2020-10-20 Supply, Inc. System and method for wireless power reception
US9521023B2 (en) 2014-10-17 2016-12-13 Kumu Networks, Inc. Systems for analog phase shifting
US9712313B2 (en) 2014-11-03 2017-07-18 Kumu Networks, Inc. Systems for multi-peak-filter-based analog self-interference cancellation
US9673854B2 (en) 2015-01-29 2017-06-06 Kumu Networks, Inc. Method for pilot signal based self-inteference cancellation tuning
US11601213B2 (en) * 2015-07-12 2023-03-07 Cohere Technologies, Inc. Orthogonal time frequency space modulation over a plurality of narrow band subcarriers
US20210028877A1 (en) * 2015-07-12 2021-01-28 Cohere Technologies, Inc. Orthogonal time frequency space modulation over a plurality of narrow band subcarriers
US10243598B2 (en) 2015-10-13 2019-03-26 Kumu Networks, Inc. Systems for integrated self-interference cancellation
US9634823B1 (en) 2015-10-13 2017-04-25 Kumu Networks, Inc. Systems for integrated self-interference cancellation
US11082074B2 (en) 2015-12-16 2021-08-03 Kumu Networks, Inc. Systems and methods for linearized-mixer out-of-band interference mitigation
US10541840B2 (en) 2015-12-16 2020-01-21 Kumu Networks, Inc. Systems and methods for adaptively-tuned digital self-interference cancellation
US9742593B2 (en) 2015-12-16 2017-08-22 Kumu Networks, Inc. Systems and methods for adaptively-tuned digital self-interference cancellation
US11671129B2 (en) 2015-12-16 2023-06-06 Kumu Networks, Inc. Systems and methods for linearized-mixer out-of-band interference mitigation
US9819325B2 (en) 2015-12-16 2017-11-14 Kumu Networks, Inc. Time delay filters
US10666305B2 (en) 2015-12-16 2020-05-26 Kumu Networks, Inc. Systems and methods for linearized-mixer out-of-band interference mitigation
US10404297B2 (en) 2015-12-16 2019-09-03 Kumu Networks, Inc. Systems and methods for out-of-band interference mitigation
US10050597B2 (en) 2015-12-16 2018-08-14 Kumu Networks, Inc. Time delay filters
US10200217B2 (en) 2015-12-16 2019-02-05 Kumu Networks, Inc. Systems and methods for adaptively-tuned digital self-interference cancellation
US10454444B2 (en) 2016-04-25 2019-10-22 Kumu Networks, Inc. Integrated delay modules
US9979374B2 (en) 2016-04-25 2018-05-22 Kumu Networks, Inc. Integrated delay modules
US10338205B2 (en) 2016-08-12 2019-07-02 The Board Of Trustees Of The Leland Stanford Junior University Backscatter communication among commodity WiFi radios
US11483836B2 (en) 2016-10-25 2022-10-25 The Board Of Trustees Of The Leland Stanford Junior University Backscattering ambient ism band signals
US10236922B2 (en) 2017-03-27 2019-03-19 Kumu Networks, Inc. Systems and methods for tunable out-of-band interference mitigation
US11121737B2 (en) 2017-03-27 2021-09-14 Kumu Networks, Inc. Systems and methods for intelligently-tuned digital self-interference cancellation
US10862528B2 (en) 2017-03-27 2020-12-08 Kumu Networks, Inc. Systems and methods for tunable out-of-band interference mitigation
US11764825B2 (en) 2017-03-27 2023-09-19 Kumu Networks, Inc. Systems and methods for tunable out-of-band interference mitigation
US10382089B2 (en) 2017-03-27 2019-08-13 Kumu Networks, Inc. Systems and methods for intelligently-tuned digital self-interference cancellation
US10547346B2 (en) 2017-03-27 2020-01-28 Kumu Networks, Inc. Systems and methods for intelligently-tuned digital self-interference cancellation
US10623047B2 (en) 2017-03-27 2020-04-14 Kumu Networks, Inc. Systems and methods for tunable out-of-band interference mitigation
US11515906B2 (en) 2017-03-27 2022-11-29 Kumu Networks, Inc. Systems and methods for tunable out-of-band interference mitigation
US11211969B2 (en) 2017-03-27 2021-12-28 Kumu Networks, Inc. Enhanced linearity mixer
US10840968B2 (en) 2017-03-27 2020-11-17 Kumu Networks, Inc. Systems and methods for intelligently-tuned digital self-interference cancellation
US10103774B1 (en) 2017-03-27 2018-10-16 Kumu Networks, Inc. Systems and methods for intelligently-tuned digital self-interference cancellation
US10827445B2 (en) 2017-06-06 2020-11-03 Supply, Inc. Method and system for wireless power delivery
US10798665B2 (en) * 2017-06-06 2020-10-06 Supply, Inc. Method and system for wireless power delivery
US11178625B2 (en) 2017-06-06 2021-11-16 Supply, Inc. Method and system for wireless power delivery
US11743841B2 (en) 2017-06-06 2023-08-29 Reach Power, Inc. Method and system for wireless power delivery
US20200112928A1 (en) * 2017-06-06 2020-04-09 Supply, Inc. Method and system for wireless power delivery
US10952162B2 (en) 2017-06-06 2021-03-16 Supply, Inc. Method and system for wireless power delivery
US10382085B2 (en) 2017-08-01 2019-08-13 Kumu Networks, Inc. Analog self-interference cancellation systems for CMTS
US11159349B2 (en) * 2017-12-27 2021-10-26 Orange Method for estimating the channel between a transceiver and a mobile communicating object
US10425115B2 (en) 2018-02-27 2019-09-24 Kumu Networks, Inc. Systems and methods for configurable hybrid self-interference cancellation
US11128329B2 (en) 2018-02-27 2021-09-21 Kumu Networks, Inc. Systems and methods for configurable hybrid self-interference cancellation
US10804943B2 (en) 2018-02-27 2020-10-13 Kumu Networks, Inc. Systems and methods for configurable hybrid self-interference cancellation
US11183886B2 (en) 2018-03-08 2021-11-23 Supply, Inc. Method and system for wireless power delivery
US10952163B2 (en) 2018-11-28 2021-03-16 Supply, Inc. System and method for wireless power delivery
US10778044B2 (en) 2018-11-30 2020-09-15 Supply, Inc. Methods and systems for multi-objective optimization and/or wireless power delivery
US10944299B2 (en) 2018-11-30 2021-03-09 Supply, Inc. Methods and systems for multi-objective optimization and/or wireless power delivery
US11562045B2 (en) 2019-03-14 2023-01-24 Kumu Networks, Inc. Systems and methods for efficiently-transformed digital self-interference cancellation
US10868661B2 (en) 2019-03-14 2020-12-15 Kumu Networks, Inc. Systems and methods for efficiently-transformed digital self-interference cancellation
CN112543074A (zh) * 2019-09-23 2021-03-23 清华大学深圳国际研究生院 一种非视距通信信道建模方法
US11282531B2 (en) * 2020-02-03 2022-03-22 Bose Corporation Two-dimensional smoothing of post-filter masks
US11611242B2 (en) 2021-04-14 2023-03-21 Reach Power, Inc. System and method for wireless power networking
US11955815B2 (en) 2021-04-14 2024-04-09 Reach Power, Inc. System and method for wireless power networking
US11991034B1 (en) 2022-11-09 2024-05-21 Qualcomm Incorporated Sample-level error-correcting code

Also Published As

Publication number Publication date
WO2010100551A2 (fr) 2010-09-10
US8315335B2 (en) 2012-11-20
EP2404418B1 (fr) 2015-01-28
US20100226416A1 (en) 2010-09-09
US20100226458A1 (en) 2010-09-09
EP2404417A1 (fr) 2012-01-11
WO2010100061A1 (fr) 2010-09-10
CN102342075A (zh) 2012-02-01
EP2404418A1 (fr) 2012-01-11
WO2010100076A1 (fr) 2010-09-10
WO2010100551A3 (fr) 2011-07-07
EP2404419A2 (fr) 2012-01-11
US8160176B2 (en) 2012-04-17

Similar Documents

Publication Publication Date Title
US20100226448A1 (en) Channel extrapolation from one frequency and time to another
US8116710B2 (en) Continuous sequential scatterer estimation
US8401487B2 (en) Radio channel analyzer to determine doppler shifts across multiple frequencies of a wideband signal
US6765969B1 (en) Method and device for multi-user channel estimation
US8260208B2 (en) Radio channel estimator
US9020050B2 (en) Accounting for inter-carrier interference in determining a response of an OFDM communication channel
US8422595B2 (en) Channel estimation for communication systems with multiple transmit antennas
US9313065B2 (en) Scattered pilot pattern and channel estimation method for MIMO-OFDM systems
Berger et al. Application of compressive sensing to sparse channel estimation
KR101647171B1 (ko) 스파스 피드백 데이터로부터의 채널 상태 정보의 재구성
US8379782B2 (en) Mobile radio channel estimation
US7385617B2 (en) Methods for multi-user broadband wireless channel estimation
US8023595B2 (en) Method and system of time-of-arrival estimation for ultra wideband multi-band orthogonal frequency division multiplexing signals
US8379706B2 (en) Signal and noise power estimation
US10270624B2 (en) Channel estimation method and apparatus for use in wireless communication system
CN104539562A (zh) 基于多输入多输出正交频分复用的宽带短波信道估计方法
KR20080087397A (ko) 광대역 무선통신 시스템에서 슬라이딩 윈도우 채널 추정장치 및 방법
US20080084938A1 (en) Intersymbol interference mitigation
US8335284B1 (en) System and method for inter-channel interference reduction
KR20050007432A (ko) 시변수 채널을 통해 전송된 데이터 신호의 등화 및 복조방법
Sun Integrated Sensing and Communications in Time-Varying Channels
Nisar Channel Estimation and Equalization for Evolved UTRA Uplink
Prasad et al. Fuzzy extended Kalman filter to estimate Rayleigh fading channel with PSAM for MIMO-OFDM
Burchill Performance and receiver structures for OFDM on Rayleigh fading channels
Valera Zorita et al. Space-frequency block coding for underwater acoustic communications

Legal Events

Date Code Title Description
AS Assignment

Owner name: TELEFONAKTIEBOLAGET L M ERICSSON (PUBL), SWEDEN

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:DENT, PAUL WILKINSON;REEL/FRAME:022787/0788

Effective date: 20090604

STCB Information on status: application discontinuation

Free format text: ABANDONED -- FAILURE TO PAY ISSUE FEE