

浏览全部资源
扫码关注微信
1. 南京航空航天大学电子信息工程学院,江苏,南京,210016
2. 台州学院计算机应用研究所, 浙江临海,317000
3. 南京航空航天大学电子信息工程学院,江苏,南京,210016
4. 台州学院计算机应用研究所 浙江临海,317000
Published:2016
移动端阅览
An Adaptive Hybrid Wireless NoC Topology Structure[J]. Acta Electronica Sinica, 2016, 44(6): 1420-1428.
An Adaptive Hybrid Wireless NoC Topology Structure[J]. Acta Electronica Sinica, 2016, 44(6): 1420-1428. DOI: 10.3969/j.issn.0372-2112.2016.06.024.
针对传统大规模片上网络(Network-on-Chip
NoC)远距离核间多跳通信所带来的高能耗与延时问题
提出了一种基于虚Torus的自适应的混合型无线NoC拓扑结构(VT-AWiNoC).该结构通过引入链路拥塞测度作为感知参数
基于此采用热点无线链路自动探测与带宽动态分配机制
并设计实现发送器动态分配的控制电路模块
以达到根据不同的通信流量模型
于片内自适应地调整拓扑结构及链路带宽的目的.通过建立混合型无线NoC的延时与功耗评估模型
对该结构的无线NoC进行性能评估.实验结果表明
该自适应拓扑与其它混合型无线NoC相比
在随机流量模型下
网络平均延时降低了16.52~23.27%;在20%的热点流量模型下
包平均能耗节省了39.19%;以真实应用FFT作为基准测试
平均延时降低了17.20%~21.68%
并节省了23.49%的包平均能耗.该结构以较小的面积开销获得了更优的性能.
Aimed at the high energy consumption and communication delay due to multi-hop wired links between distant cores of traditional and large-scale network-on-chip(NoC)
a virtual Torus-based adaptive wireless NoC(VT-AWiNoC) structure is proposed.The automatic detection and dynamic bandwidth allocation mechanism to hot wireless link based on the sensing parameter for link congestion measurement is adopted.Moreover
the dynamic allocation circuit for transmitter is designed to realize adaptive adjustment of the intra-chip topology and link bandwidth as different traffic patterns.We compare the performance of our proposed VT-AWiNoC to other hybrid wireless NoC topologies by setting up the energy consumption and delay estimate models.Experimental results show that the proposed topology obtains a delay improvement of 16.52% to 23.27% under random traffic pattern
an energy saving per packet of 39.19% under hotspot traffic pattern
a delay improvement of 17.20% to 21.68% and energy saving per packet of 23.49% on real application such benchmark as FFT(Fast Fourier Transform).The proposed topology achieves higher performance gains with small additional area cost.
0
Views
2
下载量
5
CSCD
Publicity Resources
Related Articles
Related Author
Related Institution
京公网安备11010802024621