研究了一种在并行Turbo译码器中同时进行存储器访问的新颖的存储方案.该方案采用了图论中的节点着色法,与其他也在存储器中采用的非规则方法相比,所需的存储块(RAM)要多2-5块,但当码长变化时,这种配置方法更简单,可以在片上实时实现.实验表明,对于中高速的译码器(40-100 Mb/s),其硬件开销对3GPP标准中的交织器依然是可以承受的.
This paper discussed a novel storage scheme for simultaneous memory access in parallel Turbo decoder. The new scheme employs vertex coloring in graph theory. Compared to similar methods which also use un-natural order in storage, the scheme requires more memory blocks but allows a simpler configuration method when code length changes, which can be implemented on-chip. The experiment shows, for a moderate decoding throughput (40-50 Mb/s), the hardware cost is still affordable for 3GPP's interleaver, 5 iterations and 80-100 MHz system clock.