While the technology development towards microservices can significantly improve speed and agility of software service delivery, it also raises operational complexity associated with modern applications. This has led to emergence Service Mesh, a promising approach mitigate this situation by introducing dedicated infrastructure layer over without imposing modification on implementations. Aiming inspire more practical research work in exploited area, we paper present comprehensive review state...

Deep Convolutional Neural Networks (CNNs) are powerful models that have achieved excellent performance on difficult computer vision tasks. Although CNNs perform well whenever large labeled training samples available, they work badly video frame synthesis due to objects deforming and moving, scene lighting changes, cameras moving in sequence. In this paper, we present a novel general end-to-end architecture, called convolutional Transformer or ConvTransformer, for sequence learning synthesis....

Cloud brokerage mechanisms are fundamental to reduce the complexity of using multiple cloud infrastructures achieve optimal placement virtual machines and avoid potential vendor lock-in problems. However, current approaches restricted static scenarios, where changes in characteristics such as pricing schemes, machine types, service performance throughout life-cycle ignored. In this paper, we investigate dynamic scheduling use cases these parameters continuously changed, propose a linear...

In recent decades, virtualization as an abstraction from physical hardware has become a popular solution to resource isolation and server consolidation. With the surge in adoption of technologies, ensuring High Availability (HA) for applications hosted virtualized environments emerges important problem garnered substantial attention. this paper, we present brief comparison technologies HA perspective. The state-of-the-art solutions two mainstream types platforms (i.e., hypervisor-based...

This study investigates the influence of various factors on shear flocculation in apatite-dolomite mixed ore flotation system. The Box-Behnken response surface methodology is employed to evaluate interactions these behavior. Additionally, Focused Beam Reflectance Measurement (FBRM) and Particle Vision (PVM) techniques are applied investigate mechanisms by which affect flocculation. results show that sodium oleate (NaOL) concentration, pH, rate significantly size flocculated particles....

In this work, we present a novel approach that leverages Linux containers to achieve High Availability (HA) for cloud applications. A middleware is comprised of set HA agents defined compensate the limitations in achieving HA. our start modeling at application level, considering dependencies among components. We generate proper scheduling scheme and then deploy across cloud. For each container hosts critical component(s), continuously monitor its status checkpoint full state, react failure...

Virtual machine placement is the process of mapping virtual machines to available physical hosts within a data center or on remote in cloud federation. Normally, service owners cannot influence components beyond choosing provider and deployment zone at that provider. For some services, however, this lack hindrance adoption. example, services require specific geographical (due e.g. legislation), redundancy by avoiding co-location critical components. We present an approach for their...

Until now, most research on cloud service placement has focused static pricing scenarios, where providers offer fixed prices for their resources. However, with the recent trend of dynamic resources, price a compute resource can vary depending free capacity and load provider, new algorithms are needed. In this paper, we investigate in scenarios by evaluating set algorithms, tuned pricing. The range from simple heuristics to combinatorial optimization solutions. studied evaluated deploying...

Web Services communicate through XML-encoded messages and suffer from substantial overhead due to verbose encoding of transferred extensive (de)serialization at the end-points. We demonstrate that response caching is an effective approach reduce Internet latency server load. Our Tantivy middleware layer reduces volume data transmitted without semantic interpretation service requests or responses thus improves time. achieves this reduction combined use recent compression techniques decrease...

Efficient mapping of Virtual Machines~(VMs) onto physical servers is a key problem for cloud infrastructure providers as hardware utilization directly impacts profit. Today, this commonly only performed when new VMs are created, but VM workloads fluctuate and server availability varies, any initial bound to become suboptimal over time. We introduce set heuristic methods continuous optimization the VM-to-server based on combinations fundamental management actions, namely suspending resuming...

Most cities in the world promote bike-sharing services to encourage people decrease carbon exhausting and enhance their health. However, it is a big challenge for service supplying corporation re-balance bikes efficiently among different dockers without forecasting ability. For solving this problem, we contribute two new approaches based on standard Long short-term memory (LSTM), which can not only take advantages of multi features inputs multi-time steps outputs improve accuracy predicting...

Dynamic and zero-touch management is expected to be the key feature of next-generation 6G networks. Network Function Virtualization (NFV) one technologies for realizing such through software-based Despite great benefits offered by NFV, deploying network services (NSs) in NFV ecosystems remains a challenge, especially latency-sensitive NSs, as they demand stringent latency requirements fast service provisioning. Specifically, graphs should embedded into an infrastructure that these are...

The cloud computing landscape has recently developed into a spectrum of architectures, leading to broad range management tools for similar operations but specialized certain deployment scenarios. This both hinders the efficient reuse algorithmic innovations within and increases heterogeneity between different systems. Our overarching goal is overcome these problems by developing general enough support full popular architectures. In this contribution, we analyze commonalities in proposed...

We present a hybrid GPU-FPGA based computing platform to tackle the high-density problem of machine learning. In our platform, training part learning application is implemented on GPU and inferencing FPGA. It should also include model transplantation which can transplant from part. For evaluating this design methodology, we selected LeNet-5 as benchmark algorithm. During phase, TitanXp's speed was about 8.8x faster than CPU E-1620 in FPGA Arria-10's fastest, 44.4x 6341x TitanXp. Moreover, by...

Since multispectral images (MSIs) have much more sufficient spectral information than RGB (RGBs), reconstructing MS from is a severely underconstrained problem. We to generate colossally different between the two scopes. Almost all previous approaches are based on static and dependent neural networks, which fail explain how supplement massive lost information. This paper presents low-cost high-efficiency approach, "VAE-GAN", stochastic networks directly reconstruct high-quality MSIs RGBs....

In this paper, we present a REST-based system, \emph{Set It Right} (\emph{SIR}), where people can get feedback on and help with short texts correction. The rapid development of the SIR enabled by designing it as set services, also leveraging commercially offered illustrates strength SOA paradigm. Finally, evaluate cloud computing techniques infrastructures used to deploy system how technologies shorten time market lower initial costs.

We formulate a multi-matrices factorization model (MMF) for the missing sensor data estimation problem. The problem is adequately transformed into matrix completion one. With MMF, an n-by-t real matrix, R, adopted to represent collected by mobile sensors from n areas at time, T1, T2, ... , Tt, where entry, Rij, aggregate value of in ith area Tj . propose approximate R seeking family d-by-n probabilistic spatial feature matrices, U(1), U(2), U(t), and temporal V E Rdxt, Rj ≈ UT(j)Tj also...

Network Function Virtualization (NFV) is a network architecture that separates functions from dedicated hardware, implementing them as software modules known Virtual Functions (VNFs), which are executed in virtual machines or containers. NFV increases the deployment flexibility and agility within operator networks reduces operating capital expenditures significantly. In NFV, migration of VNFs can significantly reduce embedding cost. However, stringent latency requirements between make...

5G service providers consider Network Function Virtualization (NFV) as an enabler to foster new opportunities scale their business while reducing operational expenses. NFV builds on cloud native technologies, automation, and reusability. Central the success of is ability design templates once on-demand deploy those template services into specific contexts, e.g., network slices. Therefore, homing assigning designs with level requirements over distributed resources capabilities one main...

Abstract The flooding process is one of the main concerns damaged ship stability. This paper combines volume fluid (VOF) method incorporated in Navier-Stokes (NS) solver with dynamic mesh techniques to simulate a ship. VOF used capture interface, while are applied update as result transient motions. time-domain processes barge and rectangular cabin model carried out based on abovementioned method, computational results appear compatible experimental data. During process, motion flow at...