论文部分内容阅读
提出了一个全局优化算法(GOA)对RS(255,239)解码器中的并行钱氏搜索电路进行面积优化.通过查找钱氏搜索电路中GF (Galois field)常数乘法器的公共模2加运算并进行预运算,GOA能够有效地减少电路中异或门的数量,从而减少电路面积.与原有局部优化方法不同,GOA是一个全局优化算法.当每次迭代中同时有多个最大匹配对时,GOA通过选取与其他匹配对关系最小的一对作为最优匹配而不是随机地选择一对,使得当前结果对最终的优化结果影响最小.进一步将基于组的GOA用于GF乘法器组的优化,结果显示相对于直接实现方法,可使并行钱氏搜索电路的面积减少51%,而对GF乘法器的单独优化也能使电路面积减少26%.该优化方法可广泛地用于含有大量模2加运算的并行结构中.
A global optimization algorithm (GOA) is proposed to optimize the area of parallel Qianshi search circuit in RS (255,239) decoder. The common mode 2 addition operation of GF (Galois field) constant multiplier in Qian search circuit is performed and carried out Pre-computation, GOA can effectively reduce the number of XOR gates in the circuit, thereby reducing the circuit area.Against the original local optimization method, GOA is a global optimization algorithm.When there are multiple maximum matching pairs in each iteration, GOA selects the pair that has the least relationship with other matching pairs as the best match instead of selecting a pair randomly so that the current result has the least influence on the final optimization result.Group-based GOA is further used for the optimization of the GF multiplier group, The results show that compared to the direct implementation method, the parallel Qianshi search circuit area can be reduced by 51%, while the GF multiplier alone optimization can also make the circuit area reduced by 26% .The optimization method can be widely used in a large number of modules 2 Plus the parallel structure of the operation.