Abstract: Methods and apparatus for encoding codewords which are particularly well suited for use with low density parity check (LDPC) codes and long codewords are described. The described methods allow encoding graph structures which are largely comprised of multiple identical copies of a much smaller graph. Copies of the smaller graph are subject to a controlled permutation operation to create the larger graph structure. The same controlled permutations are directly implemented to support bit passing between the replicated copies of the small graph. Bits corresponding to individual copies of the graph are stored in a memory and accessed in sets, one from each copy of the graph, using a SIMD read or write instruction. The graph permutation operation may be implemented by simply reordering bits, e.g., using a cyclic permutation operation, in each set of bits read out of a bit memory so that the bits are passed to processing circuits corresponding to different copies of the small graph.

Abstract: A flexible and relatively hardware efficient LDPC decoder is described. The decoder can be implemented with a level of parallelism which is less than the full parallelism of the code structure used to control the decoding process. Each command of a relatively simple control code used to describe the code structure can be stored and executed multiple times to complete the decoding of a codeword. Different codeword lengths are supported using the same set of control code instructions but with the code being implemented a different number of times depending on the codeword length. The decoder can switch between decoding codewords of different lengths, without the need to change the stored code description information, by simply changing a code lifting factor that is indicative of codeword length and is used to control the decoding process. When decoding codewords shorter than the maximum supported codeword length some block storage locations may go unused.

Abstract: A flexible and relatively hardware efficient LDPC decoder is described. The decoder can be implemented with a level of parallelism which is less than the full parallelism of the code structure used to control the decoding process. Each command of a relatively simple control code used to describe the code structure can be stored and executed multiple times to complete the decoding of a codeword. Different codeword lengths are supported using the same set of control code instructions but with the code being implemented a different number of times depending on the codeword length. The decoder can switch between decoding codewords of different lengths, without the need to change the stored code description information, by simply changing a code lifting factor that is indicative of codeword length and is used to control the decoding process. When decoding codewords shorter than the maximum supported codeword length some block storage locations may go unused.

Abstract: In modern iterative coding systems such as LDPC decoder and turbo-convolutional decoder in which the invention may be used, the core computations can often be reduced to a sequence of additions and subtractions alternating between logarithm and linear domains A computationally efficient and robust approximation method for log and exp functions is described which involves using a simple bit mapping between fixed point fractional data format and floating point format. The method avoids costly lookup tables and complex computations and further reduces the core processing to a sequence of additions and subtractions using alternating fixed point and floating point processing units.

Abstract: A method, apparatus, and machine readable medium for processing a plurality of Z-vectors. Each Z-vector includes Z elements, and each element includes K bits. The Z-vectors correspond to a binary codeword, portions of which have a relationship to a plurality of transmission units. The Z-vectors are stored in a set of D memory arrays. Each memory array includes Z rows of memory locations. Each memory location corresponds to a different array column, and each array column corresponds to a different Z-vector. Each Z-vector identifies one column. A series of sets of control information is generated. Each set includes a transmission unit identifier, a Z-vector identifier, and a row identifier. For at least one set, P times K divided by D bits is read from each column identified by the Z-vector that is identified by the Z-vector identifier included in the set.

Abstract: Methods and apparatus for encoding codewords which are particularly well suited for use with low density parity check (LDPC) codes and long codewords are described. The described methods allow encoding graph structures which are largely comprised of multiple identical copies of a much smaller graph. Copies of the smaller graph are subject to a controlled permutation operation to create the larger graph structure. The same controlled permutations are directly implemented to support bit passing between the replicated copies of the small graph. Bits corresponding to individual copies of the graph are stored in a memory and accessed in sets, one from each copy of the graph, using a SIMD read or write instruction. The graph permutation operation may be implemented by simply reordering bits, e.g., using a cyclic permutation operation, in each set of bits read out of a bit memory so that the bits are passed to processing circuits corresponding to different copies of the small graph.

Abstract: Techniques for implementing message passing decoders, e.g., LDPC decoders, are described. To facilitate hardware implementation messages are quantized to integer multiples of ½ ln2. Messages are transformed between more compact variable and less compact constraint node message representation formats. The variable node message format allows variable node message operations to be performed through simple additions and subtractions while the constraint node representation allows constraint node message processing to be performed through simple additions and subtractions. Variable and constraint nodes are implemented using an accumulator module, subtractor module and delay pipeline. The accumulator module generates an accumulated message sum. The accumulated message sum for a node is stored and then delayed input messages from the delay pipeline are subtracted there from to generate output messages.

Abstract: Methods and apparatus for encoding codewords which are particularly well suited for use with low density parity check (LDPC) codes and long codewords are described. The described methods allow encoding graph structures which are largely comprised of multiple identical copies of a much smaller graph. Copies of the smaller graph are subject to a controlled permutation operation to create the larger graph structure. The same controlled permutations are directly implemented to support bit passing between the replicated copies of the small graph. Bits corresponding to individual copies of the graph are stored in a memory and accessed in sets, one from each copy of the graph, using a SIMD read or write instruction. The graph permutation operation may be implemented by simply reordering bits, e.g., using a cyclic permutation operation, in each set of bits read out of a bit memory so that the bits are passed to processing circuits corresponding to different copies of the small graph.

Abstract: High throughput parallel LDPC decoders are designed and implemented using hierarchical design and layout optimization. In a first level of hierarchy, the node processors are grouped on the LDPC decoder chip, physically co-locating the processing elements in a small area. In a second level of hierarchy, clusters, e.g., subsets, of the processing elements are grouped together and a pipeline stage including pipeline registers is introduced on the boundaries between clusters. Register to register path propagating signals are keep localized as much as possible. The switching fabric coupling the node processors with edge message memory is partitioned into separate switches. Each separate switch is split into combinational switching layers. Design hierarchies are created for each layer, localizing the area where the interconnect is dense and resulting in short interconnect paths thus limiting signal delays in routing.

Abstract: In modern iterative coding systems such as LDPC decoder and turbo-convolutional decoder in which the invention may be used, the core computations can often be reduced to a sequence of additions and subtractions alternating between logarithm and linear domains A computationally efficient and robust approximation method for log and exp functions is described which involves using a simple bit mapping between fixed point fractional data format and floating point format. The method avoids costly lookup tables and complex computations and further reduces the core processing to a sequence of additions and subtractions using alternating fixed point and floating point processing units.

Abstract: Methods and apparatus for decoding codewords using message passing decoding techniques which are particularly well suited for use with low density parity check (LDPC) codes and long codewords are described. The described methods allow decoding graph structures which are largely comprised of multiple identical copies of a much smaller graph. Copies of the smaller graph are subject to a controlled permutation operation to create the larger graph structure. The same controlled permutations are directly implemented to support message passing between the replicated copies of the small graph. Messages corresponding to individual copies of the graph are stored in a memory and accessed in sets, one from each copy of the graph, using a SIMD read or write instruction. The graph permutation operation may be implemented by simply reordering messages, e.g.

Abstract: A flexible and relatively hardware efficient LDPC decoder is described. The decoder can be implemented with a level of parallelism which is less than the full parallelism of the code structure used to control the decoding process. Each command of a relatively simple control code used to describe the code structure can be stored and executed multiple times to complete the decoding of a codeword. Different codeword lengths are supported using the same set of control code instructions but with the code being implemented a different number of times depending on the codeword length. The decoder can switch between decoding codewords of different lengths, without the need to change the stored code description information, by simply changing a code lifting factor that is indicative of codeword length and is used to control the decoding process. When decoding codewords shorter than the maximum supported codeword length some block storage locations may go unused.

Abstract: A flexible and relatively hardware efficient LDPC decoder is described. The decoder can be implemented with a level of parallelism which is less than the full parallelism of the code structure used to control the decoding process. Each command of a relatively simple control code used to describe the code structure can be stored and executed multiple times to complete the decoding of a codeword. Different codeword lengths are supported using the same set of control code instructions but with the code being implemented a different number of times depending on the codeword length. The decoder can switch between decoding codewords of different lengths, without the need to change the stored code description information, by simply changing a code lifting factor that is indicative of codeword length and is used to control the decoding process. When decoding codewords shorter than the maximum supported codeword length some block storage locations may go unused.

Abstract: In modern iterative coding systems such as LDPC decoder and turbo-convolutional decoder in which the invention may be used, the core computations can often be reduced to a sequence of additions and subtractions alternating between logarithm and linear domains A computationally efficient and robust approximation method for log and exp functions is described which involves using a simple bit mapping between fixed point fractional data format and floating point format. The method avoids costly lookup tables and complex computations and further reduces the core processing to a sequence of additions and subtractions using alternating fixed point and floating point processing units.

Abstract: Data communication over a block-coherent channel in a communication system is described. Low-complexity demodulation techniques that allow good performance are described. A dwell, e.g., a set of block coherent symbols transmitted including a known symbol, e.g., a pseudo pilot symbol, are received, demodulated and decoded by a joint decoder/demodulator employing soft inputs, soft outputs, and interleaving of messages. Low-complexity SISO demodulator is suitable for processing pseudo-pilot modulated information corresponding to each of one or more dwells. The low-complexity method achieves good performance when turbo equalization is used. Some decoding and demodulation embodiments include independent phase estimates and updated independent phase estimates following the extrinsic principle to generate soft symbol values and soft bits.

Abstract: A flexible and relatively hardware efficient LDPC decoder is described. The decoder can be implemented with a level of parallelism which is less than the full parallelism of the code structure used to control the decoding process. Each command of a relatively simple control code used to describe the code structure can be stored and executed multiple times to complete the decoding of a codeword. Different codeword lengths are supported using the same set of control code instructions but with the code being implemented a different number of times depending on the codeword length. The decoder can switch between decoding codewords of different lengths, without the need to change the stored code description information, by simply changing a code lifting factor that is indicative of codeword length and is used to control the decoding process. When decoding codewords shorter than the maximum supported codeword length some block storage locations may go unused.

Abstract: Methods and apparatus for implementing memory efficient LDPC decodes are described. In accordance with the invention message information is stored in a compressed state for check node processing operations. The state for a check node is fully updated and then subject to an extraction process to generate check node to variable node messages. The signs of messages received from variable nodes may be stored by the check node processor module of the invention for use in message extraction. The check node processor can process messages in variable node order thereby allowing the variable node processor and check node processor to operate on messages in the same order reducing or eliminating the need to buffer and/or reorder messages passed between check nodes and variable nodes. Graph structures which allow check node processing on one graph iteration to proceed before the previous graph iteration has been completed are also described.

Abstract: Methods and apparatus for communication over a block-coherent communication system are described. The present invention is directed to methods of interleaving coded bits that are encoded by codes, e.g., LDPC codes, having graph structures largely comprised, e.g., of multiple identical copies of a much smaller graph.

Abstract: Methods and apparatus for scaling soft values as part of an error correction decoding process are described. Accurate decoding depends on use of the appropriate scale factor. Selection and use of the scale factor to scale soft values is designed to improve and/or optimize decoder performance without the need for prior knowledge of the correct scale factor or the actual channel conditions at the time the signal from which the soft values were obtained was transmitted through a communications channel. The techniques of the present invention assume that the soft values to be processed were transmitted through a communications channel having a quality that can be accurately described by a channel quality value. A scale factor is determined from the distribution of soft values to be scaled and an assumption that the channel through which they were transmitted was of the quality corresponding to a preselected channel quality value.

Abstract: Methods and apparatus for implementing memory efficient LDPC decodes are described. In accordance with the invention message information is stored in a compressed state for check node processing operations. The state for a check node is fully updated and then subject to an extraction process to generate check node to variable node messages. The signs of messages received from variable nodes may be stored by the check node processor module of the invention for use in message extraction. The check node processor can process messages in variable node order thereby allowing the variable node processor and check node processor to operate on messages in the same order reducing or eliminating the need to buffer and/or reorder messages passed between check nodes and variable nodes. Graph structures which allow check node processing on one graph iteration to proceed before the previous graph iteration has been completed are also described.