SYSTEMS AND METHODS FOR OBTAINING LOW COMPLEXITY AND HIGH CORRECTIVE CAPACITY DECODERS

Description

The radio channel is a very hostile medium for mobile communications, mainly due to signal fading derived from multipath propagation and interference caused by neighboring cells. All this causes that the sequence of bits detected in the receiver contains numerous errors. In this context, turbo codes are today presented as a fundamental pillar of many mobile communication standards due to their enormous correction capacity.
Turbo decoding consists of an iterative process in which a soft measure decoder is executed at each iteration. These devices are capable not only of obtaining the decoded bit sequence, but of associating a reliability about the decoding of each bit. The information provided by these reliabilities improves with the passing of the iterations and is what allows correcting new errors with each iteration.
There are two families of soft measure decoders, those based on the SOVA (soft-output Viterbi algorithm) algorithm, and those based on the MAP (maximum a-posteriori) algorithm. The MAP algorithm offers greater corrective capabilities than SOVA, but at the cost of greater complexity that translates into higher chip area consumption and higher latency. As decoding is an iterative process, higher latency translates into lower binary rate.
The present invention is based on the fact that executing the SOVA algorithm in two different directions (forward and backward), although it produces the same sequence of decoded bits, gives different information about the reliabilities of each bit. This information produced in two different directions can be used to improve turbo decoding. However, generating this information in the same iteration would require using two SOVA decoders (one forwards and one backwards), thereby doubling the area consumption. The present invention proposes to alternate the directions in which the SOVA algorithm runs intelligently to take advantage of this different information that is generated throughout the iterations. This allows the turbo decoder to be implemented using a single special SOVA decoder that is capable of forward and backward decoding and whose area and latency consumption is similar to that of a conventional SOVA decoder.

Advantages

The present invention makes it possible to obtain turbo decoder implementations with similar corrective capabilities as those based on the MAP algorithm, but with lower consumption of chip area and higher bit rate. If we compare the present invention with solutions based on the conventional SOVA algorithm, the area consumption and binary rate are similar, but the corrective capabilities associated with the invention are higher.

Uses and Applications

Turbo codes are used in many wireless communication systems, such as modern mobile communication standards (UMTS, HSPA, LTE, WiMAX), satellite communication systems, underwater communications, unguided optical communications, etc.