Parallel Heuristic Methods to Accelerate Best Equivocation Code Generation

In this paper, we propose parallel heuristic methods to accelerate the generation of $(n,m)$ best equivocation code (BEC), where notation="LaTeX">$n$ and notation="LaTeX">$m$ are message lengths, respectively. The proposed dynamic programming (DP) method greedy extend a previous heuristics by reducing time complexity process. DP produces same codes as but incurs an overhead for data reuse. contrast, avoids generates slightly different due its approach. We parallelize exploiting coarse-grained fine-grained parallelisms, which achieve further acceleration on multicore CPU graphics processing unit (GPU) systems, Experimental results demonstrate that reduce sequential quarter, indicated theoretical analysis. addition, implementation achieves linear speedup system, GPU realizes coalesced memory accesses, resulting in notation="LaTeX">$17\times $ over eight-core implementation. found produced differ from methods; however, generated had higher rates than those naive random method. believe can effectively BEC large values, especially with larger values relative .