This paper presents BCube, a new network architecture specifically designed for shipping-container based, modular data centers. At the core of BCube is its server-centric structure, where servers with multiple ports connect to layers COTS (commodity off-the-shelf) mini-switches. Servers act as not only end hosts, but also relay nodes each other. supports various bandwidth-intensive applications by speeding-up one-to-one, one-to-several, and one-to-all traffic patterns, providing high...

This paper presents BCube, a new network architecture specifically designed for shipping-container based, modular data centers. At the core of BCube is its server-centric structure, where servers with multiple ports connect to layers COTS (commodity off-the-shelf) mini-switches. Servers act as not only end hosts, but also relay nodes each other. supports various bandwidth-intensive applications by speeding-up one-to-one, one-to-several, and one-to-all traffic patterns, providing high...

Shipping-container-based data centers have been introduced as building blocks for constructing mega-data centers. However, it is a challenge on how to interconnect those containers together with reasonable cost and cabling complexity, due the fact that center can hundreds or even thousands of aggregate bandwidth among easily reach tera-bit per second. As new inner-container server-centric network architecture, BCube [9] interconnects servers inside container provides high support typical...

The explosive expansion of data center sizes aggravates the power consumption and carbon footprint, which has restricted sustainable growth cloud services seriously troubled operators. In recent years, plenty advanced network architectures have been proposed. They usually employ richly-connected topologies multi-path routing to provide high capacity. Unfortunately, they also undergo inefficient energy usage during traffic valley time. To address problem, many energy-aware flow scheduling...

The emergence of new capabilities such as virtualization and elastic (private or public) cloud computing infrastructures has made it possible to deploy multiple applications, on demand, the same infrastructure. A major challenge achieve this possibility, however, is that modern applications are typically distributed, structured systems include not only computational storage entities, but also policy entities (e.g., load balancers, firewalls, intrusion prevention boxes). Deploying a...

Multicast benefits group communications in saving network traffic and improving application throughput, both of which are important for data center applications. However, the technical trend future design poses new challenges efficient scalable routing. First, densely connected networks make traditional receiver-driven routing protocols inefficient tree formation. Second, it is quite difficult low-end switches largely used centers to hold entries massive groups.

The network resource competition of today' data enters is extremely intense between long-lived elephant flows and latency-sensitive mice flows. Achieving both goals high throughput low latency respectively for the two types requires compromise, which recent research has not successfully solved mainly due to transfer on shared links without any differentiation. However, current usually adopt clos-based topology, e.g. Fat-tree/VL2, so there exist multiple shortest paths pair source...

Enterprises deploy their middlebox services in cloud seeking for easy management, flexible scalability and economic savings. However, existing elastic virtual network function(VNF) placement strategy often leads to an unpredictable placing location due the ever-changing workload, which may waste much precious bandwidth resource bring a lot of VM operation overhead(e.g. launch, termination migration). A key problem providers is how conduct effective service provide provision according various...

Network performance aware optimizations have long been an effective approach to optimizing distributed applications on traditional network environments. However, the assumptions of topology or direct use several measurements pair-wise for are no longer valid IaaS clouds. Virtualization hides from users, and may not represent long-term performance. To enable existing clouds, we propose decouple constant component dynamic while minimizing difference by a mathematical method called RPCA (Robust...

Data centers are becoming a big consumer of the world's electricity because explosive growth their scale, which has become serious concern. In recent years, many advanced data center network architectures have been proposed, employ richly-connected topologies and multi-path routing to achieve high bisection bandwidth. Unfortunately, they undergo inefficient energy usage. To address this issue, network-wide conservation techniques proposed in community. paper, we explore new dimension saving...

With the emergence of new enabling technologies including software-defined networking and network function virtualization, virtualization has been reemphasized in recent years, such as slicing 5G networks. As an important aspect topology is studied this work-from traditional embedding toward full virtualization. By tenant virtual can specify whatever without knowing underlying physical get same experience building a by his own. Based on switches with late-binding key extractor, not only...

In this paper, robust transceiver designs for point-to-point MIMO-OFDM systems are investigated. A general performance metric of weighted mean-square-error (MSE) minimization is adopted as the optimization objective function. Exploiting signal structure systems, a simple and readily comprehensive derivation procedure optimal solution given. It revealed that design an interesting variant famous water-filling solution, which named cluster water-filling. This significantly different from over...

A popular solution to internet performance problems is the widespread caching of data. Many algorithms have been proposed in literature, most attempting optimize for one criteria or another, and recent efforts explored automation self-tuning response observed workloads. We extend these consider goal optimizing selectable criteria. With our algorithm, we shown matching exceeding best known greedy dual-size either object byte hit ratios across different web GD-GhOST consistently outperforms...

Dataflow architecture has native advantages in achieving high instruction parallelism and power efficiency for today's emerging applications such as performance computing deep neural network. For the dataflow computing, execution of instructions is driven by data, so data transfer network on chip (NoC) a key factor affecting performance. In NoC, latest router uses multicast routing scheme output buffer structure to improve efficiency. However, effective utilization rate router's low due...

While large layer-2 networks are widely accepted as the network fabric for modern data centers and virtualization is required to support multi-tenant cloud computing, existing solutions not specifically designed networks. In this paper, we design <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">SVDC</i> , a highly-scalable low-overhead architecture center By leveraging emerging software defined networking (SDN) framework, SVDC decouples global...

With the rapid development of smartphones and mobile Internet technology, we witness an overwhelming growth social networks (MSN), which is a type network, forming virtual communities among people with similar interests or commonalities. In MSNs, users play crucial role for their development, deployment success. Understanding MSN user behavior therefore attracts different entities — ISPs, service providers, researchers. However, it hard to gather comprehensive real data set, little known...

RFID finds wide-ranging applications in many fields. When customers decide to implement a application,they are possibly confused with some technical problems while allocating tags and readers satisfy their business requirements. Due the environment complexity, general deployment strategy is inapplicable that means must create own specific one. A simulator for design,view, simulation evaluation helpful walking through problems. In order improve efficiency application, research has been done...

Nowadays, the influence of machine learning algorithms on our lives is booming gradually, and gradient descent (GD) method a typical big data analysis that can be applied to various algorithms. Traditional GD has its advantages in dealing with optimal parameter finding issue for partial loss functions, but process may trapped into local minimum sometimes if function certain convexity; moreover, traditional cannot meet requirement high accuracy. Fractional order GD, which core fixed memory...

NAT gateway is an important network system in today's IPv4 when translating a private address to public address. However, traditional based on Linux Netfilter cannot achieve high throughput meet modern requirements such as data centers. To this challenge, we improve the performance of by three ways. First, leverage DPDK enable polling and zero-copy delivery, so reduce cost interrupt packet copies. Second, multiple CPU cores process parallel use lock-free hash table minimize contention...

Traditional security detection methods are difficult to meet the real-time, accurate, energy-saving, and scalability requirements of wireless sensor networks. Therefore, design a multi-layer scheme for NAWL-ILSTM apply it The concept mobile Agent is introduced, estimate node spacing using signal attenuation characteristics. network divided into structures, static state nodes transfer during transmission task analyzed, combined with LSTM NAWL optimization algorithm online updating mechanism....

Let $\emph{spex}_{\mathcal{OP}}(n,F)$ and $\emph{spex}_{\mathcal{P}}(n,F)$ be the maximum spectral radius over all $n$-vertex $F$-free outerplanar graphs planar graphs, respectively. Define $tC_l$ as $t$ vertex-disjoint $l$-cycles, $B_{tl}$ graph obtained by sharing a common vertex among edge-disjoint $l$-cycles %$B_{tl}$ connecting cycles in at single vertex, $(t+1)K_{2}$ disjoint union of $t+1$ copies $K_2$. In 1990s, Cvetkovi\'c Rowlinson conjectured $K_1 \vee P_{n-1}$ maximizes on $n$...

Abstract This paper proposes a random access network admission control algorithm under delay and jitter constraints to efficiently guarantee the quality of service visible light communication system. The multi-packet reception technology is considered enhance service, interrupted Bernoulli process used model arrival packets. queuing system constructed from perspective single transmission link, this derives probability successful for terminals. performance are evaluated according state...