一种新的矢量量化索引值分配算法  

A Novel Index Assignment Algorithm on Noisy Channel

作　　者：王粤[1] 余松煜[1] 钱团结[1] 

机构地区：[1]上海交通大学图像通信与信息处理研究所,上海200030

出　　处：《电子学报》2004年第10期1734-1737,共4页Acta Electronica Sinica

摘　　要：借鉴于生物免疫系统强大的防御能力 ,本文提出了一种免疫克隆 (MCIAA)算法 .该算法能更好地在群体的收敛性和个体的多样性之间保持动态平衡 ,有效地克服了局部收敛和早期收敛问题 .将该算法应用于噪声信道矢量量化索引值分配寻优中 ,在存在信道噪声时 ,可较好地提高矢量量化器的性能 .模拟试验表明该算法比其他算法有更好的增益和收敛速度 .Enlightened by the defence ability of immune system,a novel immune colon index assignment algorithm (MCIAA) is presented.The new algorithm has a good effort to adjust the dynamic balance between colony's convergence and individual's diversity,and effectively overcomes the problem of local optimal resolution.In this paper,we apply this algorithm to the vector quantization(VQ) index assignment on noisy channel.Simulation result conforms that the algorithm yields a significant reduction in average distortion and converges in less short time compared to other algorithms.

关 键 词：矢量量化 索引值分配 噪声信道 免疫系统 克隆 信源压缩编码技术 

分 类 号：TN911.21[电子电信—通信与信息系统]