Embeddings)

As the interest on Network Virtualization continues to grow, so does the awareness of the many technical obstacles to transpose before the envisioned virtualized network environment may become a reality. A significant obstacle lies on the efficient assignment of virtual resources into physical ones. Performing the so-called mapping of a virtual network into a substrate network is a computationally intensive task, due to the dual optimization required for nodes and links placement. The purpose of this paper is to tackle this problem taking into consideration real-life scenarios of network operators, where the limitations imposed by the heterogeneity of the virtual and substrate networks must be accounted for. To that end, this paper proposes a heuristic algorithm for virtual resources mapping in the physical infrastructure that supports the heterogeneity of networks, in both links and nodes. The mapping heuristic was evaluated both through simulation and in a real experimental virtualization platform. Through the simulation results, it is shown that the mapping approach is able to embed a high percentage of the Virtual Network (VNet) requests respecting all links and node constraints. With respect to the experimental results, the proposed algorithm was shown to be fast, requiring a mapping time in the order of low tens of milliseconds, and linearly scalable with the increase in the number of existing VNets.

In motion analysis and understanding it is important to be able to fit a suitable model or structure to the temporal series of observed data, in order to describe motion patterns in a compact way, and to discriminate between them. In an unsupervised context, i.e., no prior model of the moving object(s) is available, such a structure has to be learned from the data in a bottom-up fashion. In recent times, volumetric approaches in which the motion is captured from a number of cameras and a voxel-set representation of the body is built from the camera views, have gained ground due to attractive features such as inherent view-invariance and robustness to occlusions. Automatic, unsupervised segmentation of moving bodies along entire sequences, in a temporally-coherent and robust way, has the potential to provide a means of constructing a bottom-up model of the moving body, and track motion cues that may be later exploited for motion classification. Spectral methods such as locally linear embedding (LLE) can be useful in this context, as they preserve ''protrusions", i.e., high-curvature regions of the 3D volume, of articulated shapes, while improving their separation in a lower dimensional space, making them in this way easier to cluster. In this paper we therefore propose a spectral approach to unsupervised and temporally-coherent body-protrusion segmentation along time sequences. Volumetric shapes are clustered in an embedding space, clusters are propagated in time to ensure coherence, and merged or split to accommodate changes in the body's topology. Experiments on both synthetic and real sequences of dense voxel-set data are shown. This supports the ability of the proposed method to cluster body-parts consistently over time in a totally unsupervised fashion, its robustness to sampling density and shape quality, and its potential for bottom-up model construction.

Abstract: Cloud computing presupposes on-demand network access to pool of computing resources. However, network access through the WAN is usually not compliant with any kind of service guarantees, including reliability, security and performance. In this work, two types of network services able to fulfill cloud requirements are presented. In addition, an extension of the Virtual Private Cloud concept is proposed by integrating these network services. Managing cloud and network resources in an integrated way is a need and an obvious ...

Abstract: Cloud computing presupposes on-demand network access to pool of computing resources. However, network access through the WAN is usually not compliant with any kind of service guarantees, including reliability, security and performance. In this work, two types of network services able to fulfill cloud requirements are presented. In addition, an extension of the Virtual Private Cloud concept is proposed by integrating these network services.