检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
机构地区:[1]浙江大学超大规模集成电路设计研究所,杭州310027
出 处:《电子与信息学报》2009年第5期1205-1209,共5页Journal of Electronics & Information Technology
基 金:国家863计划项目(2005AA1Z1271)资助课题
摘 要:该文提出了一种面向应用优化的片上总线调度策略。以系统通信事件信息为基础,使用最小任务松弛时间与最小总线空闲时间相结合的调度策略,在优先保证满足任务实时要求的基础上,最大限度利用总线空闲时间,提高调度效率。并提供了可配置的权重参数α用于总线时间消耗与片上缓冲区容量之间的设计折衷。在双核SoC平台上实现了本文调度方法,并以最新视频编解码标准H.264/AVC为目标应用,与FP(Fixed Priority),SBA(Slack Based Arbitration),RR(Round Robin)等调度方法进行了性能对比。实验结果表明,α=0.5时,较以上3种策略,分别平均可减少16.6%、13.2%与9.7%的总线时间,在实时性能方面,较最接近的SBA方法,未实时完成的任务数量减少了59.4%。额外缓冲区开销随α的变化关系表明,在最坏情况下(α=0)仅需435字节。An application-specific bus scheduling scheme was proposed in this paper. Two-fold optimization was considered in this scheme based on the communication events collected by system modeling and simulation. The first one, which had higher priority, was real time constraints of tasks while the other was making use of bus idle time to transfer data as much as possible. A configurable optimization parameter o was also proposed for the tradeoff between the total bus time consumed and the extra on-chip buffer requirements. This scheme was implemented in a dual-core SoC (System on Chip) for the H.264/AVC decoder and compared with RR (Round Robin), FP (Fixed Priority) and SBA (Slack Based Arbitration) schemes. The results showed that the proposed scheme had an average 16.6%, 13.2% and 9.7% less bus time when α was set to 0.5. The number of missed real time constraints tasks was 59.4% less than the SBA scheme, which was the closest to our scheme. The relationship between α and the extra on-chip buffer cost showed that under worst condition (α =0), it was only 435 bytes.
分 类 号:TN47[电子电信—微电子学与固体电子学] TP336[自动化与计算机技术—计算机系统结构]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.222