Universal parallel processing decoder

An apparatus for universal decoding of data, the data having been encoded utilizing a plurality of encoder coefficients and transmitted as a signal from a transmitter to a receiver to form a received signal, the apparatus comprising:

• a memory;
• and
• a processor coupled to the memory, wherein the processor, when operative, is configured, for each current state of a plurality of states, to determine a most likely previous state to form a plurality of previous states, and from the plurality of previous states, to determine a terminating state and a penultimate terminating state, the penultimate terminating state having a most significant bit;
• the processor further configured, if the data has been feedforward encoded by a feedforward convolutional encoder, to determine a decoded least significant input bit as equal to the most significant bit of the penultimate terminating state, and if the data has been feedback encoded by a feedback convolutional encoder, to determine the decoded least significant input bit as equal to the most significant subset bit.

An apparatus for decoding of feedback encoded data utilizing parallel processing, the data having been feedback encoded by a feedback convolutional encoder utilizing a plurality of encoder coefficients and transmitted as a signal from a transmitter to a receiver to form a received signal, the apparatus comprising:

• a memory;
• and
• a processor coupled to the memory, wherein the processor, when operative, includes a plurality of parallel processing elements, each of the plurality of parallel processing elements configured to determine for for each current state of a respective plurality of current states, a most likely previous state to form a plurality of previous states;
• the processor further configured to determine a terminating state and a penultimate terminating state from previous states formed by the plurality of parallel processing elements, and from the terminating state and the penultimate terminating state, to determine a value of a plurality of subset bits, the plurality of subset bits having a most significant subset bit;
• and the processor further configured to determine a decoded least significant input bit as equal to the most significant subset bit.

An apparatus for universal decoding of data, the data having been encoded utilizing a plurality of encoder coefficients and transmitted as a signal from a transmitter to a receiver to form a received signal, the apparatus comprising:

• an equalizer to determine an equalized output from the received signal;
• a first memory;
• a second memory;
• a branch metric calculator coupled to the equalizer, wherein the branch metric calculator, when operative, is configured to determine, from the equalized output, a closest point of a signaling constellation subset, and to determine a corresponding branch metric as a distance between the equalized output and the closest point of the signaling constellation subset, for each combination of values of the plurality of subset bits;
• a plurality of processing elements coupled to the first memory and to the second memory, further coupled to the branch metric calculator, wherein the plurality of processing elements, when operative, are configured, for each current state of the plurality of states, to determine a plurality of potential previous states, and for each potential previous state of the plurality of potential previous states, to determine a corresponding value of the plurality of subset bits for a state transition from the potential previous state to the current state by re-encoding the potential previous state and the current state;
• the plurality of processing elements further configured to read a path metric associated with a previous state alternatingly from the first memory and from the second memory, and to determine a potential path metric for each current state of the plurality of states as an arithmetic sum of the corresponding branch metric and a path metric associated with the potential previous state to from a plurality of potential path metrics for each current state of the plurality of states;
• and for each current state of the plurality of states, to determine a path metric by selecting a comparatively smallest potential path metric of the plurality of potential path metrics;
• and
• a lookback processor couples to the plurality of processing elements, to the first memory and to the second memory, wherein the lookback processor, when operative, is configured to determine a terminating state and a penultimate terminating state, from the plurality of previous states, based upon a comparatively smallest path metric, the penultimate terminating state having a most significant bit;
• the lookback processor further configured to determine a specific value of a plurality of subset bits by re-encoding the terminating state and the penultimate terminating state, the plurality of subset bits having a most significant subset bit;
• and the lookback processor further configured, when the data has been feedforward encoded, to determine a decoded least significant input bit as equal to the most significant input bit as equal to the most significant bit of the penultimate terminating state, and when the data has been feedback encoded, to determine the decoded least significant input bit as equal to the most significant subset bit.

An apparatus for decoding of feedforward encoded data utilizing parallel processing, the data having been feedforward encoded by a feedforward convolutional encoder utilizing a plurality of encoder coefficients and transmitted as a signal from a transmitter to a receiver to form a received signal, the apparatus comprising:

• a memory;
• and
• a processor coupled to the memory, wherein the processor, when operative, includes a plurality of parallel processing elements, the plurality of processing elements configured to determine, for each of a plurality of current states, a most likely previous state to form a plurality of previous states;
• the processor further configured to determine a terminating state and a penultimate terminating state from previous states formed by the plurality of parallel processing elements, and from the terminating state and the penultimate terminating state, to determine a value of a plurality of subset bits;
• and the processor further configured to determine a decoded least significant input bit as equal to a most significant bit of the penultimate terminating state, the apparatus also for decoding of feedback encoded data, wherein the plurality of subset bits has a most significant subset bit, and wherein the processor is further configured, when the data has been feedback encoded by a feedback convolutional encoder, to determine a decoded least significant input bit as equal to the most significant subset bit.

A method for universal decoding of data, the method comprising:

• receiving an encoded signal having been encoded by a feedback convolutional encoder;
• using a feedforward convolutional decoder to decode the encoded signal;
• producing subset information and state information during decoding of the encoded signal by the feedforward convolutional decoder;
• determining that the encoded signal is feedback encoded;
• and
• setting, based on the determining, a value of a decoded bit produced by the feedforward convolutional decoder equal to a value of a bit of the subset information.