Abstract:
A method of estimating initial channel quality in a receiver when allocated a new channel in order to select an optimal codec mode in a multi-rate service is disclosed. One implementation initially fills the filter state with the first received channel quality measurement. Another embodiment proportionally fills the entire filter state with the calculated channel quality measurements. Yet another embodiment uses the hysteresis and threshold parameters in conjunction with the initial codec mode to calculate an initial fill value for the filter state.
Abstract:
A method of insuring the accuracy of transmitted or stored digital data involves the use of a cyclical redundancy check (CRC) code. The method is particularly useful for ensuring the accuracy of frames transmitted between multi-mode vocoders. The method allows a different CRC code to be used for each mode of a transmitting multi-mode vocoder. A receiving multi-mode vocoder checks the CRC code against the CRC coding formulas of the various modes. If the CRC code is satisfied under any one of the modes, the frame is labeled as “good”. If the CRC code fails under all the modes, the frame is labeled as “bad”. If the bit frame includes bits for indicating the mode of the transmitting multi-mode vocoder, the receiving multi-mode vocoder checks the CRC code against the CRC coding formula for the indicated mode only. If the CRC code passes for the indicated mode, the frame is labeled as “good”, otherwise, the frame is labeled as “bad”.
Abstract:
A received information signal is decoded to obtain the received information and to produce at least one feature of the received information signal. The received information signal is preliminarily classified as containing a normal burst or a truncated burst based upon the at least one feature, to obtain a preliminary classification. Cyclic redundancy checking of the received information that is decoded is performed. The received information signal is then further classified as containing a normal burst or a truncated burst based upon the preliminary classification and whether the cyclic redundancy checking is valid, to obtain a further classification. The received information signal may be still further classified as containing a normal burst or a truncated burst based upon the further classification and at least one transition rule for normal bursts and truncated bursts between the received information signal and a previously received information signal.
Abstract:
Novel techniques are disclosed for detecting discontinuous transmission (DTX) over a communication channel. A received data frame is characterized as one of a Good frame, Erasure, or DTX. If a Good frame is not initially detected, a multi-dimensional quality metric is used to characterize the received frame as either an Erasure or DTX. A two dimensional quality metric may be generated using energy per bit to noise power ratio as a first dimension and re-encoded symbol error count as a second dimension. Alternatively, re-encoded symbol energy may be used as the second dimension of the quality metric. The communication channel may be a CDMA wireless communication channel capable of DTX and the device may be a base station, base station controller, or mobile station.
Abstract:
A window error detector for a receiver capable of operating in a discontinuous transmit mode includes a soft decision decoder (210) producing soft output and generating window error signals. A detector (214) is coupled to the soft decision decoder for detecting a bad frame when the window error exceeds a bad frame threshold, wherein the threshold is altered based on the discontinuous transmit state. A turbo decoder (210) for a receiver includes a soft decision decoder and a window error detector (214) coupled to the soft decision decoder. The window error detector generates a bad frame indication. The turbo decoder stops iterative processing of the data associated with a window when the window error detector detects that the window does not result in a bad frame indication.
Abstract:
A communication system implements detection of bad frames of information by utilizing multiple bit correction thresholds. Equipment used within the communication system adapts to different signaling environments by dynamically altering the bit correction threshold based on a history of the number of consecutive bad frames of information that have been previously erased and the number of bits corrected by a channel decoder (202). By implementing this dynamic bit correction threshold, sufficient bad frame indication (BFI) detection and receiver sensitivity can be obtained simultaneously, which results in an improved perceived audio quality to the end user.
Abstract:
Method and apparatus for processing a data stream. Errors are concealed in the data stream by detecting loss or interruption of data delivery and signalling decoders to invoke error concealment.
Abstract:
A method for frame quality detection and a receiver for implementing the method. For frame quality detection, results A and B are compared with each other and with a predetermined threshold. The result A is formed as a sum/product of soft bit decisions or as a bit error rate from bursts comprising bits from successive frames N and N+1. The result B is formed in the same way from frames N-1 and N. When comparing the result A and the result B with each other, the frame N is determined to be bad if one of the result A and the result B is essentially smaller than another of the result A and the result B. The frame N is also determined to be bad if the both result A and the result B are smaller than the predetermined threshold value. The method improves detection of a bad frame and accordingly reduces interference sound received by a receiver and caused by erroneously interpreted bad frames, especially during periods of silence.
Abstract:
In a digital radio communication system in which information is transmitted and received in time slots and some bits of each received audio frame is interleaved over at least two time slots and other bits in each audio frame at are non-interleaved, bad received audio frames are detected by detecting possible bit errors in a channel encoded non-interleaved bit sequence in each received time slot. If bit errors have been detected in this bit sequence a bad received frame is indicated.
Abstract:
An error detector circuit (300) for a discrete receiver indicates bad frames of binary information signals which contain distorted bits of data in numbers so great as to prevent a convolutional decoder (738) from generating, accurately, a decoded signal. A variable threshold generator (440) generates a variable threshold level according to the signal quality estimate for the received signal. When bit errors are detected in numbers beyond the variable threshold or when a first preselected value of the signal quality of a received signal combined with the detected number of bit errors forms a signal beyond a second preselected value, a bad frame is indicated.