Pctel, Inc.

General Datacomm, Inc.

Ic Tel Inc.

Yuri Goldstein

Yuri Okunev

Vitaly Drucker

Qin Wang

Paul D Cole

William Hanna

Oin Wang

Val Maizenberg

William E Abdelsayed

Yuri Okuney

William Eskander Abdelsayed

Paul Dana Cole

Methods, apparatus and systems for implementing an algorithm for N-symbol noncoherent processing of M-ary DPSK signals in a pilotless, wireless system is provided. The algorithm is carried out with (N&#x2212;1) recurrent steps (iterations) plus a decision step. Each iterative step includes simple trigonometrical transformation of quadrature components of the current symbol and summation of the transforms with results of the previous step. A final N-symbol decision regarding the current transmitted symbol corresponds to the vector of maximum length, calculated after the (N&#x2212;1)-th step of the iterative procedure. The general algorithm is optionally implemented with one or more intersymbol processors, one or more memory blocks for saving results of the intersymbol processors, and a decision block. In addition, shift registers for quadrature components of the received signal may be utilized.

Method and apparatus for noncoherent signal processing in pilotless wireless systems

Systems, methods and apparatus are provided for demapping a received wireless signal. The wireless signal includes signal-bearing data without accompanying pilot signals, and demapping is accomplished by extracting information from the signal-bearing data by reducing and averaging differential quadrature components or quadrature components of the received signal. Two embodiments are provided for the demapping procedure within a phase tracking loop. A first embodiment corrects the received signal, while a second embodiment corrects the constellation points

Pilotless, wireless, telecommunications apparatus, systems and methods

Systems, methods and apparatus are provided for equalizing a received multicarrier wireless telecommunications data signal. The wireless signal includes data without accompanying pilot signals, and equalization is accomplished by extracting information from the data, by processing the information in order to obtain equalization indications, and by equalizing the multicarrier wireless telecommunications data signal by modifying indications of the wireless telecommunications data signal using the equalization indications.

Pilotless, wireless, telecommunications apparatus, systems and methods having equalization

Methods, apparatus, and systems for mode assignment and mode adaptation to channel conditions are provided which are based on estimations of real signal-to-noise ratios (SNR) for each frequency carrier bearing information during a data transmission session. The invention utilizes two principal procedures: signal-to-noise ratio (SNR) estimation, and a corresponding mode assignment. SNR estimation is obtained by an averaging of squared Euclidean distances between normalized received signals and reference signals corresponding to either current hard decisions from the output of a demodulator or soft decisions provided by a decoder, for each frequency carrier bearing random information. Using the SNR estimations, various algorithms for determining mode assignment are provided.

Mode adaptation in wireless systems

A modem includes an LDPC encoder which utilizes a deterministic H-matrix, optionally via a generation matrix, to generate redundant parity bits for a bit block. Ones are placed into the H-matrix in a completely diagonal manner with diagonals subdivided into sets of diagonals. The first diagonal in each set i begins with coordinates H(1,k), where k=(1+(i*M)). The remaining diagonals in the sets are offset from the first diagonals so that the column distances between any two pairs of diagonals is unique. In another embodiment, the H-matrix is determined by assigning '1s' in a first column, and then assigning '1s' of subsequent columns deterministically by causing each &#x201c;1&#x201d; in a previous ancestor column to generate a &#x201c;1&#x201d; in the next descendant column based on the rule that a descendant is placed one position below an ancestor except where rectangles would be generated. Interrupted descending diagonals are generated.

Modems utilizing low density parity check codes

An easily implementable soft decision decoding (SDD) metric is provided for telecommunications systems and apparatus. The SDD metric is based on weighted average distances or weighted minimum distances between scaled received signals and all reference signals related to the corresponding binary symbol. An important property of the SDD metric of the invention is that the distance weight is completely defined by the received signal scaling factor which is readily available as a result of conventional frequency domain equalization procedures. The invention is particularly advantageous in wireless systems having variable parameters such as wireless OFDM systems with selective fading.

Methods, apparatus, and systems employing soft decision decoding

Distinguishing between final coding of received signals in a PCM modem

Handshake information for xDSL services is transmitted utilizing a spread spectrum modulated system where a plurality (n) of carrier tones (n>2) are summed and utilized as a spread spectrum carrier (SSC), and data is modulated onto the carrier (at all utilized frequencies). Preferably, phase shift keying (PSK) modulation or a variation thereof is used as the encoding/modulation technique.

Spread spectrum handshake for digital subscriber line telecommunications systems

A DSL modem with a receiver and a transmitter includes an LDPC encoder which utilizes a generation matrix G which is derived from a substantially deterministic H matrix in order to generate redundant parity bits for a block of bits. The H matrix is determined by assigning &ldquo;ones&rdquo; of a first column N randomly or deterministically. Then, beginning with the second column, assignment of &ldquo;ones&rdquo; is carried out deterministically with each &ldquo;1&rdquo; in a previous ancestor column generating a &ldquo;1&rdquo; in the next descendant column based on the rule that a descendant is placed one position below or above an ancestor. As a result, descending or ascending diagonals are generated. When distributing &ldquo;ones&rdquo; in any given column, care is taken to ensure that no rectangles are generated in conjunction with other &ldquo;ones&rdquo; in the current column and previous columns. By avoiding generation of rectangles, diagonals are interrupted.

DSL modem utilizing low density parity check codes

The receiver of a PCM modem utilizes the V.90 TRN training signal for detecting whether any of the six slots of the received signal may be subject to alternating robbed bit signaling (ARBS), and what the alternating robbed bit signaling pattern may be. This is accomplished by accumulating the received level over a plurality of frames for each slot and by comparing the received levels for each slot to an average level for that slot. If the difference between the received levels and the average levels exceeds a threshold, the slot is determined to be subject to ARBS. Where a slot is subject to ARBS, DIL sequence signals for frames of the alternating robbed bit signaling slot having LSB&equals;0 and LSB&equals;1 are accumulated separately (if available) in order to generate two translation tables (TRT , TRT ) for that slot. Where the DIL sequence is found only in the frames having the alternating robbed bit signaling slot having LSB&equals;0 or having LSB&equals;1, only one translation table can be generated. Different algorithms are provided for designing a constellation for the alternating robbed bit signaling slot depending on (a) whether two translation tables were generated for the slot, (b) whether the frame-to-TRT correspondence is known, and (c) where only one translation table is generated, whether the translation table generated was TRT or TRT

Translation table and constellation design for a PCM modem subjected to alternating robbed bit signaling

Methods are provided for determining whether any slots of a modem frame are subject to robbed bit signaling (RBS), and whether the robbed bit signaling is &bgr;-codec robbed bit signaling. From an obtained DIL sequence, an ordered table of level values for each slot is generated. In determining whether any slots are RBS slots, distances between adjacent levels of the table for each slot are obtained and are compared to a &ldquo;zero distance&rdquo; threshold value in order to determine the number of distances which exceed the &ldquo;zero distance&rdquo; value. For each slot, the number of zero distances are compared to a threshold value, and if the number of zero distances exceeds the threshold value, the slot is declared to be a RBS slot. Zero distance thresholds for each slot are generated by finding a function of a maximum distance for that slot among the adjacent levels. In determining whether any RBS slots are &bgr;-codec slots, distances are found between corresponding levels of the ordered tables. If substantially more than half of the distances from one slot to the others are found to be greater than a zero distance threshold value, the slot is determined to be a &bgr;-codec slot.

Robbed bit signaling identification in a PCM modem

A method of providing a high bit rate DMT signal includes providing information on a plurality of DMT tones, with at least two of the plurality of DMT tones sending information in parallel. Of the tones sending information in parallel, at least one has a SNR too low to be individually used for the number of bits it is transmitting. Accordingly, the tones are coherently &ldquo;combined&rdquo; at the receiver in order to generate a SNR sufficient together to carry, at the desired BER, the number of bits being transmitted in parallel. Two or more tones which alone cannot carry a single bit of information may be used together to provide a sufficient SNR at the receiver to transmit one or more bits. Similarly, two or more tones which alone cannot carry two bits of information, may be used together via parallel transmission to provide a sufficient SNR at the receiver. Further, the bit-carrying capacity of a first tone which can carry one or more bits may be increased by transmitting an increased number of bits in parallel with one or more additional tones which cannot carry any bits of information. According to another embodiment of the invention, rather than sending two or more DMT tones in parallel, one or more DMT tones are repeated over a plurality of DMT symbols, and &ldquo;combined&rdquo; coherently in order to generate a SNR sufficient together to carry the number of bits being carried by the repeated tone.

Method and apparatus for data transmission using discrete multitone technology

Methods are provided for generating at least one translation table for a modem having a frame having a plurality of slots. DIL sequence information from another modem is used to generate ordered table of levels, and distances between adjacent levels of the ordered table for found for each slot. The distances are used to distinguish between slots subject to robbed bit signaling (RBS) and slots not subject to RBS, and the ordered tables are used to make a determination as to whether the DIL sequence information is A-law or &mgr;-law information. Scaling factors are generated from the A-law/&mgr;-law determination as well as from a PAD estimation, and the scaling factors are utilized in generating translation tables. A PAD estimation function is utilized in finding the PAD estimation.

Translation table design for a PCM modem

A method in a modem having a frame having a plurality of slots, for determining whether incoming DIL sequence information contains A-law or &mgr;-law information is provided. From the DIL sequence information, an ordered table of levels for each slot is generated. The levels of all non-robbed-bit slots are averaged, and average levels are subjected to an A-law/&mgr;-law separation function. The separation function

Growing community of inventors

Southbury, CT, United States of America

Yuri Goldstein