Patent Application
20080069250
Many aspects of the disclosed systems and methods can be better understood with reference to the following drawings. The components in the drawings are not necessarily to scale, emphasis instead being placed upon clearly illustrating the principles of the disclosed systems and methods. Moreover, in the drawings, like reference numerals designate corresponding parts throughout the several views.
Disclosed herein are various embodiments of multipath (MP) processing systems and methods (herein, simply MP processing systems unless noted otherwise). Such embodiments implement a fast Fourier transform (FFT) on a frequency domain channel response to provide multipath information that can be used to avoid (e.g., all or some) of the multipath interference in a received signal. For instance, the multipath information can be used by sampling logic in a receiver to adjust a sampling window in a manner that avoids adjacent channel interference in the time domain.
The MP processing system 200 as described above and hereinafter can be implemented in hardware, software, firmware, or a combination thereof. In the disclosed embodiment(s), a combination of hardware and software or firmware is implemented. With regard to software or firmware, the MP processing system 200 comprises one or more modules (e.g., code) that are stored in a memory and executed by a suitable instruction execution system. The one or more modules of the MP processing system 200 may comprise a program, which comprises an ordered listing of executable instructions for implementing logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions.
With regard to hardware, the MP processing system 200 can be implemented with any or a combination of the following technologies, which are all well known in the art: a discrete logic circuit(s) having logic gates for implementing logic functions upon data signals, an application specific integrated circuit (ASIC) having appropriate combinational logic gates, a programmable gate array(s) (PGA), a field programmable gate array (FPGA), etc.
Additionally, the MP processing system 200 may be embodied in any wireless or wired communication device, including computers (desktop, portable, laptop, etc.), consumer electronic devices (e.g., multi-media players), compatible telecommunication devices, personal digital assistants (PDAs), or any other type of network devices, such as printers, fax machines, scanners, hubs, switches, routers, set-top boxes, televisions with communication capability, etc.
In view of the above description, a wireless orthogonal frequency domain multiplexing (OFDM) implementation for an embodiment of the MP processing systems 200 is described below to illustrate various aspects of the MP processing systems. Although described in the context of an OFDM system, such as found in current and future IEEE 802.11 standards, the scope of the various embodiments can be applied to, and thus include, other multiplexing techniques and/or wired or wireless standards. Additionally, although described in the context of a wireless multipath environment, one skilled in the art would understand within the context of this disclosure that the MP processing system 200 similarly applies to wired multipath environments.
The sampling logic 202a comprises a frequency recovery unit 318, timing recovery unit 320, timing alignment unit 306, frequency loop 322, and timing loop 324. The FDCR logic 204a comprises fast Fourier transform (FFT) logic 314, channel estimation unit 310, and channel compensation unit 312. The FFT logic 206 may be a device separate from the FFT logic 314 in some embodiments, or functionality of the FFT logic 314 and FFT logic 206 may be implemented in the same device in some embodiments. For instance, the FFT logic 314 and 206 may be implemented in a single device (as represented by the dashed boundary) using auxiliary digital-switching circuits, among other mechanisms. Additionally, one embodiment of the FFT logic 314 and/or 206 comprises an N-point FFT unit corresponding to N sub-carriers in an OFDM signal. In general, the receiver 308 comprises functionality responsible for filtering and demodulating the received signals, and generally, functionality responsible for performing processing that is complementary to the processing performed at a transmitting device. Additional processing, not shown, may include signal separation, in-phase/quadrature (I/Q) signal determination, cyclic extension (e.g., guard interval or GI) removal, among other well-known receive functionality.
In one exemplary operation, signals (e.g., OFDM multipath signals) are received at one or more antennas (e.g., antennas not shown, but similar to antennas 106 of
The sampling logic 202a identifies a symbol boundary and extracts a symbol from the processed (i.e., processed in the one or more initial stages of the receiver 308) signal(s) 301 (hereinafter, simply signal for brevity), as well as estimates any carrier-frequency offsets and provides timing alignment. The FDCR logic 204a receives signal 303 from the sampling logic 202a and performs an FFT operation on the signal 303 to transform the signal 303 from a time-domain to a frequency-domain, thereby generating frequency-domain signal 305. It would be understood by those having ordinary skill in the art that other mechanisms for providing a frequency domain signal may be used in some embodiments (e.g., not limited to using FFT logic). The channel estimation unit 310 samples a series of pilot signals in the frequency-domain signal from the FFT logic 314 and generates a multipath channel response. In some embodiments, alternative mechanisms for providing a reference signal (e.g., in lieu of or in addition to pilot signals) may be used to provide a multipath channel response. The channel-estimation unit 310 provides the multipath channel response to the channel compensation unit 312. The channel-compensation unit 312 compensates the frequency domain signal 305 received from the FFT logic 314, based on the multipath channel response, for provision to a demodulating unit (not shown in the figure), where it is further processed to extract information data.
The channel estimation unit 310 also provides the multipath channel response to the FFT logic 206. The FFT logic 206 performs an FFT operation on the multipath channel response and generates multipath channel information. Such multipath channel information may include information related to signal strength, phase and/or relative delays of multipath signals received at the receiver 308 (e.g., similar to direct signals 107 and delayed signals 109). In one embodiment, the multipath channel information is provided to the frequency loop 322 and the timing loop 324 of the sampling logic 202a. The frequency loop 322 and the timing loop 324 provides control signals to the frequency recovery unit 318 and the timing recovery unit 320, respectively, to synchronize bit timing and sample timing for extracting a symbol from the signal 301.
Now that a general description of operation is provided of the MP processing system 200a, attention is directed to various features of the MP processing system 200a in further detail. With regard to an embodiment of the sampling logic 202a, the description that follows presents further detail on how multipath channel information is utilized in implementing sampling functionality. The frequency loop unit 322 regulates bit-timing offset correction in the frequency recovery unit 318. In one embodiment, the frequency loop unit 322 identifies from the multipath channel information a multipath signal with the greatest signal strength, hereinafter referred to as the strongest multipath signal. The signal strength of the strongest multipath is denoted as |P(p)|.
The frequency loop unit 322 further identifies, from the multipath channel information, multipath signals that exceed a predefined ratio of the signal strength of the strongest path. The strongest signal and other multipath signals identified by the frequency loop unit 322 are hereinafter collectively referred to as significant multipath signals. In one embodiment, frequency loop unit 322 subsequently selects a predetermined, programmable number of significant multipath signals from the significant multipath signals identified, based on the relative signal strength of the identified significant multipath signals. If the number of significant multipath signals is less than the predetermined number, then all of the significant multipath signals are selected. The frequency loop unit 322 determines the locations of the significant multipath signals from the multipath channel information and generates an error metric, based on the correlation between the symbols of each significant multipath signal. In one embodiment, an energy metric ‘E’ is calculated for each path, where E is defined as:
E=Σ[|P(p−1)|−|P(p+1)]*|P(p)| Equation (I)
wherein P represents signal energy, p represents the strongest multipath signal, (p−1) represents a preceding multipath signal, and (p+1) represents a succeeding multipath signal.
An error metric for frequency loop unit 322 is calculated by the frequency loop unit 322, based on the difference between the value of the energy metric, E, of a given symbol and the value of E of the succeeding symbol of each significant multipath signal identified from the multipath channel information.
In one embodiment, based on the value of the error metric, the frequency loop unit 322 determines the positioning of a sampling window. Therefore, the frequency recovery unit 318 samples the signal 301 to extract a symbol by suitably positioning a sampling window under the control of the frequency loop unit 322. In one embodiment, the frequency loop unit 222 is implemented by using a phase-locked loop (PLL), although one skilled in the art would understand that other mechanisms with like-functionality to a PLL can similarly be employed.
The timing loop unit 324 regulates sample timing offset correction in the timing recovery unit 320 to accurately resolve the start of a symbol boundary in a suitable significant multipath signal. The start of a symbol boundary is also referred to as the leading end of the symbol. A suitable multipath signal is selected from the significant multipath signals by using the frequency loop unit 322. The timing loop unit 324 determines a start-point for recovery from the symbol provided by the frequency recovery unit 318. The timing recovery unit 320 starts the recovery of the symbol from the start-point indicated by the timing loop unit 324. The timing recovery unit 320 further buffers symbol information from a significant multipath signal with a signal strength immediately next to the strongest significant multipath signal. The buffering is performed from the beginning of the symbol, as extracted using the frequency recovery unit 318, until the start-point from the strongest significant multipath signal starts. The buffered portion is appended to the information extracted from the strongest significant multipath signal. In one embodiment, the timing loop unit 324 is implemented by using a phase locked loop (PLL), although one skilled in the art would understand that other mechanisms with like-functionality to the PLL can similarly be employed.
Note that the sampling logic 202a includes functionality to estimate and remove carrier-frequency offset as well as carrier-phase offset between a carrier of the signal 301 and a local oscillator (not shown). Frequency alignment is carried out by frequency loop unit 322, which estimates carrier-frequency offset to preserve mutual orthogonality between the sub-carriers of the OFDM signal. The OFDM signal received at receiver 308 may suffer from carrier-frequency offset. Carrier-frequency offset results in the sub-carriers of the OFDM signal being shifted in a frequency-domain, thereby compromising the mutual orthogonality between the sub-carriers of the OFDM signal. Frequency-loop unit 322 utilizes one of a plurality of carrier-frequency offset-estimation algorithms to estimate the carrier-frequency offset in the OFDM signal. Frequency-loop unit 322 provides control signals to frequency recovery unit 318. In one embodiment, frequency-loop unit 322 utilizes digital signal processing to correct the carrier-frequency offset in the OFDM signal.
Note that the timing alignment unit 306 comprises well-known timing alignment functionality, and thus the discussion of the same is omitted for brevity.
Having described features of an embodiment of the sampling logic 202a, attention is now directed to various features of an embodiment of the FDCR logic 204a. The channel estimation unit 310 performs a dynamic estimation of a multipath channel response, and therefore carries out initial training and dynamic tracking for the receiver 308. In one implementation, the receiver 308 is used in a frequency-selective and time-varying multipath channel. Therefore, the receiver 308 can dynamically estimate the multipath channel response in real time. In one embodiment, the channel estimation unit 310 estimates the multipath channel response in the frequency-domain for each of the sub-carriers, hence generating the frequency-domain multipath channel response. For instance, the channel estimation unit 310 generates an estimate of the multipath channel response using one of a plurality of training-based algorithms. Training-based algorithms typically generate an estimate of the multipath channel response based on a set of reference values included in the OFDM signal received at the receiver 308. The set of reference values may comprise pilot signals or other training signals. Examples of well-known training-based algorithms include a minimum mean square error (MMSE) algorithm, a least square (LS) algorithm, a maximum likelihood (ML) algorithm, among others.
The channel estimation unit 310, in one embodiment, generates a pilot interpolation curve, which represents an estimate of the multipath channel response estimate. In some embodiments, the channel estimation unit 310 generates an estimate of the multipath channel response using one of a plurality of blind algorithms. Blind algorithms typically generate an estimate of the multipath channel response using differential methods that exploit knowledge pertaining to one or more properties of the symbols containing the information data in the OFDM signal received at the receiver 308. In some embodiments, the channel estimation unit 310 generates an estimate of the multipath channel response using one of a plurality of semi-blind algorithms, where training signals as well as differential methods are employed to generate the multipath channel response estimate. One having ordinary skill in the art would understand that one or a combination of these and/or other mechanisms for generating an estimate of the frequency domain channel response can be used in some embodiments.
In one embodiment, the channel compensation unit 312 multiplies the OFDM symbol for each sub-carrier with the respective transfer functions of the multipath channel, as predicted by channel estimation unit 310, to compensate for the effects of the multipath propagation environment (e.g., environment 100a) on an OFDM signal.
S
(j)
=a
i
*e
jφ
i=1, 2, . . . N sub-carriers Equation (II)
wherein Si is the signal-strength vector of ith sub-carrier, ai is the magnitude of the signal-strength vector, and φ is the phase of the signal-strength vector.
In view of the above description, it will be appreciated that one MP processing method embodiment 200b for processing a signal in a multipath environment 100a, 100b, as illustrated in
With regard to the exemplary OFDM implementation described above,
At 602, an OFDM signal is received and pre-processed (e.g., down converted to a baseband range, converted from analog to digital and subjected to such initial processing) as may be necessary to make the OFDM signal suitable for being further processed.
At 604, an FFT operation is performed on the pre-processed signal (e.g., signal 303). For instance, the FFT logic 314 transforms the pre-processed signal from a time-domain to a frequency-domain, thereby generating a frequency-domain signal (e.g., signal 305). In one embodiment, the multipath signals are mathematically represented by the following equations:
y(n)=x(n)+a1x(n−τ1)+a2x(n−τ2)+a3x(n−τ3)+ . . . +a1x(n+τ−1)+a2x(n+τ−2).
Y(k)=FFT [y(n)]=X(k)[1+a1WNkτ1+a2WNkτ2+ . . . +a1WNkτ−1+a−1WNkτ−1 . . . ],
where WN=exp(−2Πj/N), and y(n) is the time-domain OFDM signal received at the receiver (e.g., receiver 308); x(n), x(n−τ1) and a−1x(n+τ−1) are a direct signal, post-echoes and pre-echoes, respectively; a is the attenuation effected on a multipath signal; τ1, τ2, and so on, are the respective time delays of the multiple path signals.
Similarly, Y(k) is the frequency-domain signal and, X(k) a frequency-domain direct path. N is the total number of sub-carriers in the received OFDM signal.
At 606, an estimate of the frequency-domain multipath channel response is generated, based on one of a plurality of channel response estimation algorithms. The estimate of the frequency-domain multipath channel response predicts the transfer function of each sub-carrier, as explained by the following mathematical equations:
Y(k)=X(k)H(k)
H(k)=1+a1WNkτ1+a2WNkτ2+ . . . +a−1WNkτ−1+a−1WNkτ−1 . . . ,
where H(k) defines the transfer function of the multipath channel for each sub-carrier in the received OFDM signal.
At 608, an FFT operation is performed on the frequency-domain multipath channel response (e.g, on the estimate) to generate multipath channel information (e.g., multipath channel profile), as described mathematically by the following equations:
P(n)=FFT[H(k)]=FFT[1+a1WNkτ1+a2WNτ2+ . . . +a−1WNkτ−1+a−1WNkτ−1 . . . ]
P(n)=1+a1δ(1+τ1)+a2δ(1+τ2)+a3δ(1+τ3)+ . . . +a−1δ(1+τ−1)+a−2δ(1+τ−2),
where δ(n)=0 for n=0 and δ(n)≠0 for n≠0.
Therefore, in one embodiment, the multipath channel profile P(n) is represented as a series of impulses with relative path locations.
At 610, a sampling window is synchronized based on the multipath channel information (e.g., multipath channel profile) generated at 608. Synchronizing the sampling window includes performing a first level of synchronization to position the sampling window and a second level of synchronization to determine a start-point within the sampling window. The start-point is the point from where the extraction is started for the symbol obtained by using the sampling window.
At 612, carrier synchronization is performed. Carrier synchronization includes removing carrier-frequency offset and carrier-phase offset from the OFDM signal. Carrier synchronization is performed using the multipath channel information (e.g., profile) as well as one of a plurality of carrier-frequency offset algorithms.
At 614, symbol data is extracted from the OFDM signal, based on the sampling window obtained at 610. The symbol data is extracted starting from the start-point determined at 610. The sampling window data located before the start-point is buffered and appended to the symbol extracted by using the start-point as a reference. The symbols are then demodulated in accordance with one of a plurality of demodulation techniques.
The flow diagrams of
Certain embodiments of the multipath processing systems 200 (e.g., 200a-200c) disclosed herein improve the processing of signals in various multipath environments. The multipath channel information comprises information related to the amplitude, phase and relative delay of one or more direct signals and one or more delayed signals, and provides an improved prediction of channel behavior. Such disclosed embodiments improve sampling-window synchronization, which may reduce inter-symbol interference (ISI). The multipath channel information also reduces the effect of phase noise in synchronization loops such as the PLL, thereby enabling improved carrier synchronization. Therefore, various embodiments of the MP processing systems 200 may improve the performance of the communication systems under adverse multipath channel conditions.
While various embodiments of MP processing systems 200 have been illustrated and described, it will be clear that these and other embodiments are not limited to the above description. Numerous modifications, changes, variations, substitutions and equivalents will be apparent to those skilled in the art without departing from the spirit and scope of the present disclosure as described in the claims.
