Underlying network provides infrastructures for cloud computing in data centers. The server-centric architectures integrate network and compute, which place routing intelligence on servers. However, the existing multi-port server based architectures suffer from determined scale and large path length. In this paper, we propose FleCube, a flexibly-connected architecture on multi-port servers without using any switches. FleCube is recursively constructed on division of multiple ports in a server by means of complete graph. FleCube benefits data center networks by flexible scale and low diameter, as well as large bisection width and small bottleneck degree. Furthermore, we develop multi-path routing (MPR) to take advantage of parallel paths between any two servers. MPR adopts random forwarding to distribute traffic load and relieve network congestion. Analysis and comparisons with existing architectures show the advantages of FleCube. Evaluations under different degrees of network traffic demonstrate the merits of FleCube and the proposed routings.
Delay and stability are two key factors that affect the performance of multicast data transmission in a network. However, current algorithms of tree generation hardly meet the requirements of low delay and high stability simultaneously. Given a general network, the generation algorithm of a multicast tree with minimum delay and maximum stability is an NPhard problem, without a precise and efficient algorithm. To address these challenges, this paper studies the generation of low-delay and high-stability multicast trees under the model of spanning tree based on stability probability, degree-constrained, edge-weighted for multicast (T-SDE). A class of algorithms was proposed which creates the multicast tree greedy on the ratio of fan-out to delay (RFD) and probability of stability of terminal to obtain a high performance in multicast. The proposed algorithms greedily select terminals with a large RFD and a high probability of stability as forwarding nodes in the generation of the multicast tree, where the larger RFD and higher stability of upstream nodes are beneficial to achieve a low transmission delay and high stability in multicast. The proposed RFD can be compatible with the original model, which can take advantage of network connectivity during the generation of a multicast tree. This paper carries out simulation experiments on Matlab R2016b to measure the performance of the proposed algorithm. Experimental results show that the proposed algorithm can provide a smaller height, higher stability, and a lower transmission delay of the resulting multicast tree than other solutions. The spanning tree of the proposed algorithms can support low transmission delay and high stability in multicast transmission.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.