Efficient Rate Conversion Filtering on GPUs with Shared Memory Access Pattern Scrambling | IEEE Conference Publication | IEEE Xplore