Abstract: In a TDS-OFDM receiver a device for frequency-domain compensation for channel estimation at an over-sampling rate is provided. The device includes a correlater for correlating a first signal and a second signal; a transformer for transforming the correlated signals into a frequency domain; and an interpolater and truncater block for interpolating and truncating the transformed signal to have the interpolated and truncated, transformed signal as a reference for future use.

Abstract: Method and system for determining the number of one or more of a sequence of M+1 consecutive OFDM frames from analysis of the designated preambles of two or more consecutive frames (m=0, 1, . . . , M; M?1). An overlap function OF(m;k) is formed for each frame with a sequence of selected reference signals indexed by k (k=1, 2, . . . , K), dependent upon the frame number m and the index k, and a phase (sequence location corresponding to largest amplitude of overlap function) is determined. An Mth-order phase difference is computed that corresponds to frame number of one of the M+1 frames. A consistency check is provided for the phase numbers.

Abstract: A method for constructing a simplified trellis diagram for BCH-encoded information is disclosed. BCH-encoded information is received, having a corresponding parity check matrix H. The parity check matrix H is expressed as an ordered sequence of columns of matrices. A sequence of sub-code words is provided, corresponding to one or more code words, each satisfying a given condition. A matrix Hcp, having columns that are generated as a selected permutation of the columns of the matrix H through a column-permutation-for-binary-matching process, is provided, and a sequence of sub-matrices and a corresponding sequence of permuted sub-code words is provided. A trellis diagram, representing an ordered sequence of code word transitions in the received information and symmetric about a central location, is provided for each code word c, connecting n+1 stages, numbered i=0, 1, . . . , n, in an ordered sequence.

Abstract: Method and system for determining whether one or more multipath signals is present in a received signal frame that includes a reference symbol (RS) sequence RS(t;m1) as part of the frame, and for iteratively evaluating a multipath signal that is present. The RS sequence index (m1) for the frame is identified, and first and second Fourier transforms ?(f;m1;ref) and ?(f;rec) for the received frame are provided for a selected time segment IS that includes the RS sequence of the received frame. An inverse Fourier transform ?(t;m1) of the ratio of the transforms is computed and analyzed. If at least one multipath signal is present, ?(t;m1) is nonlinearly filtered, and the filtered function is convolved with RS(t;m1). This convolution represents a received signal extension, due to multipath signal(s) presence. A refinement procedure analyzes a contribution to multipath signals from the preceding frame(s). The procedure is iteratively driven toward an optimal representation of any multipath signal(s) present.

Abstract: Method and system for adaptively estimating and compensating for phase noise error in a received signal s(t;data). A plurality of transmission parameter symbol (TPS) sets, each with a representative TPS group frequency, are incorporated in the signal before transmission. Received channel response amplitudes are analyzed to select a particular TPS frequency group, and inverse Fourier transforms, N(t) and D(t), are computed for these chosen signals. A complex ratio, r(t)=N(t)/D(t), is computed and used to determine a phase angle ?(t). A phase noise error compensated, time domain signal, equal to exp{?j?(t)}·s(t??t(delay);data) is provided.

Abstract: A method for efficient decoding of block product code format signals, using a (22p)QAM signal constellation for mapping of a received signal, with p=2, 3, 4, 5, . . . A received signal value rx, corresponding to an x=I or x=Q coordinate, is converted to a p-tuple (B(p-1)x, . . . , B0x) corresponding to a “closest” I-coordinate or Q-coordinate numerical value, and a p-stage algorithm is applied to the signal values rx and to the p-tuples (B(p-1)x, . . . , B0x) to determine a p-tuple (r(p-1)x, . . . , r0x) representing a decoded p-bit value for the received signal value rx. Depending upon a communication channel parameter Eb/N0 and the bit error ratio BER associated with each of the p bits, the received signal values rx may be suitable for some or all communications activities (e.g., HDTV, SDTV, mobile comm).