The exponential growth of user traffic has been driving routers to run at higher capacity. In a traditional router, the centralized switching fabric is becoming the bottleneck for its limited number of ports and compl...The exponential growth of user traffic has been driving routers to run at higher capacity. In a traditional router, the centralized switching fabric is becoming the bottleneck for its limited number of ports and complicated scheduling algorithms. Direct networks, such as 3-D Torus topology, have been successfully applied to the design of scalable routers. They show good scalability and fault tolerance. Unfortunately, its scalability is limited in practice. In this paper, we introduce another type of direct network, called H-Torus. This network shows excellent topological properties. On its basis, the designs of line card and routing algorithms are introduced. Extensive simulations show that the routing algorithm is very important in such a system and results in low latency with high throughput.展开更多
WiNoC(Wireless Network-on-Chip)中的无线路由器面临着比传统片上路由器更加严峻的拥塞问题,平衡有线/无线链路负载是当前无线片上网络的研究热点之一.为此本文提出并设计了一种基于优先级的交叉开关仲裁方案PbSA(Priority based Switc...WiNoC(Wireless Network-on-Chip)中的无线路由器面临着比传统片上路由器更加严峻的拥塞问题,平衡有线/无线链路负载是当前无线片上网络的研究热点之一.为此本文提出并设计了一种基于优先级的交叉开关仲裁方案PbSA(Priority based Switch allocator),其将优先级更高的无线数据包优先路由至无线路由器;结合PbSA提出了拥塞感知的路由算法CARA(Congestion-Aware Routing Algorithm),该算法有效平衡有线/无线链路负载且避免死锁,提高了数据包在网络中的路由效率.此外,本文还提出了新颖的虚通道划分方法,它不仅减少了实现PbSA的硬件复杂度,而且缓解了无线路由器发生拥塞时对整个网络的影响.实验表明,虽然本文的方案引入了较小的面积和功耗开销,但是具有良好的流量自适应特性,从而在低注入率或高注入率的情况下较好的提升了网络性能.展开更多
求解开销最小组播树在数学上归结为Steiner树问题,但由于寻找最优的Steiner树问题是NP-Complete问题,因此在组播应用中,采用启发式算法获得次优的组播树是常见的方法。该文提出了一种新的的启发式组播路由算法(Shared Path First Heuris...求解开销最小组播树在数学上归结为Steiner树问题,但由于寻找最优的Steiner树问题是NP-Complete问题,因此在组播应用中,采用启发式算法获得次优的组播树是常见的方法。该文提出了一种新的的启发式组播路由算法(Shared Path First Heuristic,SPFH)该算法在选择目的节点加入组播树时,既考虑到目的节点到树上的距离,又考虑到先加入的节点对后续加入节点的影响。算法从距离当前组播树近的目的节点中挑选节点加入组播树,选择的规则是,把能够减小其它目的节点加入组播树开销的节点先加入树。仿真结果表明,SPFH算法能找到开销接近于最优解的组播树。展开更多
该文针对片上网络提出一种基于提前分配路径的低时延片上路由器结构(PAPR)。新路由器采用提前路由计算和提前分配路径来缩短路由器流水线深度。提前路由计算为虚信道提前分配提供了可靠保障,即使在虚信道路径提前分配失败的情况下,也不...该文针对片上网络提出一种基于提前分配路径的低时延片上路由器结构(PAPR)。新路由器采用提前路由计算和提前分配路径来缩短路由器流水线深度。提前路由计算为虚信道提前分配提供了可靠保障,即使在虚信道路径提前分配失败的情况下,也不影响分组在网络中的传输时延。该文提出基于缓存状态的仲裁算法BSTS(BufferStatus)综合考虑当前节点缓存信息和下游节点缓存信息,不但降低了分组等待时延,而且降低了缓存空闲的概率。仿真结果表明,新路由器能明显改善网络的时延和吞吐性能,相比采用滑动迭代轮询仲裁iSLIP(iterativeRound-Robin Matching with SLIP(Serial Line Interface Protocal))算法的经典虚信道路由器,网络平均端到端时延降低了24.5%,吞吐率提高了27.5%;与采用轮询迭代RRM(Round-Robin Matching)算法的经典虚信道路由器相比,平均端到端时延降低了39.2%,吞吐率提高了47.2%。路由器硬件开销和平均功耗分别增加仅为8.9%,5.9%。展开更多
基金the National Natural Science Foundation of China under Grant Nos,90604029 and 60773150the National Grand Fundamental Research 973 Program of China under Grant No.2003CB314801.
文摘The exponential growth of user traffic has been driving routers to run at higher capacity. In a traditional router, the centralized switching fabric is becoming the bottleneck for its limited number of ports and complicated scheduling algorithms. Direct networks, such as 3-D Torus topology, have been successfully applied to the design of scalable routers. They show good scalability and fault tolerance. Unfortunately, its scalability is limited in practice. In this paper, we introduce another type of direct network, called H-Torus. This network shows excellent topological properties. On its basis, the designs of line card and routing algorithms are introduced. Extensive simulations show that the routing algorithm is very important in such a system and results in low latency with high throughput.
基金~~Supported by the National Natural Science Foundation of China under Grant No.90104002 the National High Technology Development 863 Program of China under Grant No.863-306-ZD-07-01
文摘求解开销最小组播树在数学上归结为Steiner树问题,但由于寻找最优的Steiner树问题是NP-Complete问题,因此在组播应用中,采用启发式算法获得次优的组播树是常见的方法。该文提出了一种新的的启发式组播路由算法(Shared Path First Heuristic,SPFH)该算法在选择目的节点加入组播树时,既考虑到目的节点到树上的距离,又考虑到先加入的节点对后续加入节点的影响。算法从距离当前组播树近的目的节点中挑选节点加入组播树,选择的规则是,把能够减小其它目的节点加入组播树开销的节点先加入树。仿真结果表明,SPFH算法能找到开销接近于最优解的组播树。
文摘该文针对片上网络提出一种基于提前分配路径的低时延片上路由器结构(PAPR)。新路由器采用提前路由计算和提前分配路径来缩短路由器流水线深度。提前路由计算为虚信道提前分配提供了可靠保障,即使在虚信道路径提前分配失败的情况下,也不影响分组在网络中的传输时延。该文提出基于缓存状态的仲裁算法BSTS(BufferStatus)综合考虑当前节点缓存信息和下游节点缓存信息,不但降低了分组等待时延,而且降低了缓存空闲的概率。仿真结果表明,新路由器能明显改善网络的时延和吞吐性能,相比采用滑动迭代轮询仲裁iSLIP(iterativeRound-Robin Matching with SLIP(Serial Line Interface Protocal))算法的经典虚信道路由器,网络平均端到端时延降低了24.5%,吞吐率提高了27.5%;与采用轮询迭代RRM(Round-Robin Matching)算法的经典虚信道路由器相比,平均端到端时延降低了39.2%,吞吐率提高了47.2%。路由器硬件开销和平均功耗分别增加仅为8.9%,5.9%。
