Object manipulation capabilities are essential skills that set apart embodied agents engaging with the world, especially in the realm of robotics. The ability to predict outcomes of interactions with objects is paramount in this setting. Although model-based control methods have started to be employed to tackle manipulation tasks, they have faced challenges in accurately manipulating objects. As we analyze the causes of this limitation, we identify the cause of underperformance in the way current world models represent crucial positional information, especially about the target’s goal specification for object positioning tasks. We introduce a general approach that empowers world model-based agents to effectively solve object positioning tasks. We propose two declinations of this approach for generative world models: position-conditioned (PCP) and latent-conditioned (LCP) policy learning. In particular, LCP employs object-centric latent representations that explicitly capture object positional information for goal specification. This naturally leads to the emergence of multimodal capabilities, enabling the specification of goals through spatial coordinates or a visual goal. Our methods are rigorously evaluated across several manipulation environments, showing favorable performance compared to current model-based control approaches.

Learning generalist embodied agents, able to solve multitudes of tasks in different domains is a long-standing problem. Reinforcement learning (RL) is hard to scale up as it requires a complex reward design for each task. In contrast, language can specify tasks in a more natural way. Current foundation vision-language models (VLMs) generally require fine-tuning or other adaptations to be adopted in embodied contexts, due to the significant domain gap. However, the lack of multimodal data in such domains represents an obstacle to developing foundation models for embodied applications. In this work, we overcome these problems by presenting multimodal-foundation world models, able to connect and align the representation of foundation VLMs with the latent space of generative world models for RL, without any language annotations. The resulting agent learning framework, GenRL, allows one to specify tasks through vision and/or language prompts, ground them in the embodied domain’s dynamics, and learn the corresponding behaviors in imagination. As assessed through large-scale multi-task benchmarking in locomotion and manipulation domains, GenRL enables multi-task generalization from language and visual prompts. Furthermore, by introducing a data-free policy learning strategy, our approach lays the groundwork for foundational policy learning using generative world models. Website, code and data: https://mazpie.github.io/genrl/

Representing a scene and its constituent objects from raw sensory data is a core ability for enabling robots to interact with their environment. In this letter, we propose a novel approach for scene understanding, leveraging an object-centric generative model that enables an agent to infer object category and pose in an allocentric reference frame using active inference, a neuro-inspired framework for action and perception. For evaluating the behavior of an active vision agent, we also propose a new benchmark where, given a target viewpoint of a particular object, the agent needs to find the best matching viewpoint given a workspace with randomly positioned objects in 3D. We demonstrate that our active inference agent is able to balance epistemic foraging and goal-driven behavior, and quantitatively outperforms both supervised and reinforcement learning baselines by more than a factor of two in terms of success rate.

Controlling artificial agents from visual sensory data is an arduous task. Reinforcement learning (RL) algorithms can succeed but require large amounts of interactions between the agent and the environment. To alleviate the issue, unsupervised RL proposes to employ self-supervised interaction and learning, for adapting faster to future tasks. Yet, as shown in the Unsupervised RL Benchmark (URLB; Laskin et al. 2021), whether current unsupervised strategies can improve generalization capabilities is still unclear, especially in visual control settings. In this work, we study the URLB and propose a new method to solve it, using unsupervised model-based RL, for pre-training the agent, and a task-aware fine-tuning strategy combined with a new proposed hybrid planner, Dyna-MPC, to adapt the agent for downstream tasks. On URLB, our method obtains 93.59% overall normalized performance, surpassing previous baselines by a staggering margin. The approach is empirically evaluated through a large-scale empirical study, which we use to validate our design choices and analyze our models. We also show robust performance on the Real-Word RL benchmark, hinting at resiliency to environment perturbations during adaptation. Project website: https://masteringurlb.github.io/

The human intrinsic desire to pursue knowledge, also known as curiosity, is considered essential in the process of skill acquisition. With the aid of artificial curiosity, we could equip current techniques for control, such as Reinforcement Learning, with more natural exploration capabilities. A promising approach in this respect has consisted of using Bayesian surprise on model parameters, i.e. a metric for the difference between prior and posterior beliefs, to favour exploration. In this contribution, we propose to apply Bayesian surprise in a latent space representing the agent’s current understanding of the dynamics of the system, drastically reducing the computational costs. We extensively evaluate our method by measuring the agent's performance in terms of environment exploration, for continuous tasks, and looking at the game scores achieved, for video games. Our model is computationally cheap and compares positively with current state-of-the-art methods on several problems. We also investigate the effects caused by stochasticity in the environment, which is often a failure case for curiosity-driven agents. In this regime, the results suggest that our approach is resilient to stochastic transitions.

Biologically inspired algorithms for simultaneous localization and mapping (SLAM) such as RatSLAM have been shown to yield effective and robust robot navigation in both indoor and outdoor environments. One drawback however is the sensitivity to perceptual aliasing due to the template matching of low-dimensional sensory templates. In this paper, we propose an unsupervised representation learning method that yields low-dimensional latent state descriptors that can be used for RatSLAM. Our method is sensor agnostic and can be applied to any sensor modality, as we illustrate for camera images, radar range-doppler maps and lidar scans. We also show how combining multiple sensors can increase the robustness, by reducing the number of false matches. We evaluate on a dataset captured with a mobile robot navigating in a warehouse-like environment, moving through different aisles with similar appearance, making it hard for the SLAM algorithms to disambiguate locations.

When designing variational autoencoders (VAEs) or other types of latent space models, the dimensionality of the latent space is typically defined upfront. In this process, it is possible that the number of dimensions is under- or overprovisioned for the application at hand. In case the dimensionality is not predefined, this parameter is usually determined using time- and resource-consuming cross-validation. For these reasons we have developed a technique to shrink the latent space dimensionality of VAEs automatically and on-the-fty during training using Generalized ELBO with Constrained Optimization (GECO) and the$L_{0}$-Augment-REINFORcE-Merge ($L_{0}$-ARM) gradient estimator. The GECO optimizer ensures that we are not violating a predefined upper bound on the reconstruction error. This paper presents the algorithmic details of our method along with experimental results on five different datasets. We find that our training procedure is stable and that the latent space can be pruned effectively without violating the GECO constraints.

Active inference is a unifying theory for perception and action resting upon the idea that the brain maintains an internal model of the world by minimizing free energy. From a behavioral perspective, active inference agents can be seen as self-evidencing beings that act to fulfill their optimistic predictions, namely preferred outcomes or goals. In contrast, reinforcement learning requires human-designed rewards to accomplish any desired outcome. Although active inference could provide a more natural self-supervised objective for control, its applicability has been limited because of the shortcomings in scaling the approach to complex environments. In this work, we propose a contrastive objective for active inference that strongly reduces the computational burden in learning the agent's generative model and planning future actions. Our method performs notably better than likelihood-based active inference in image-based tasks, while also being computationally cheaper and easier to train. We compare to reinforcement learning agents that have access to human-designed reward functions, showing that our approach closely matches their performance. Finally, we also show that contrastive methods perform significantly better in the case of distractors in the environment and that our method is able to generalize goals to variations in the background.

Abstract Modern deep learning models achieve state-of-the-art results for many tasks in computer vision, such as image classification and segmentation. However, its adoption into high-risk applications, e.g. automated medical diagnosis systems, happens at a slow pace. One of the main reasons for this is that regular neural networks do not capture uncertainty. To assess uncertainty in classification, several techniques have been proposed casting neural network approaches in a Bayesian setting. Amongst these techniques, Monte Carlo dropout is by far the most popular. This particular technique estimates the moments of the output distribution through sampling with different dropout masks. The output uncertainty of a neural network is then approximated as the sample variance. In this paper, we highlight the limitations of such a variance-based uncertainty metric and propose an novel approach. Our approach is based on the overlap between output distributions of different classes. We show that our technique leads to a better approximation of the inter-class output confusion. We illustrate the advantages of our method using benchmark datasets. In addition, we apply our metric to skin lesion classification—a real-world use case—and show that this yields promising results.

Active inference is a process theory of the brain that states that all living organisms infer actions in order to minimize their (expected) free energy. However, current experiments are limited to predefined, often discrete, state spaces. In this paper we use recent advances in deep learning to learn the state space and approximate the necessary probability distributions to engage in active inference.

As warehouses, storage facilities and factories become more expanded and equipped with smart devices, there is a substantial need for rapid, intelligent and autonomous detection of unusual and potentially hazardous situations, also called anomalies. In particular for Autonomous Guided Vehicles (AGVs) that drive around these premises independently, unforeseen obstructions along their path-e.g. a cardboard box in the middle of a corridor or bumps in the floor-and sudden or unexpected actions executed by personnel-e.g. someone walking in a restricted area-make it hard for AGVs to navigate safely. We therefore propose a novel approach to detect such anomalies in an unsupervised manner by measuring Bayesian surprise: whenever an event is observed that does not align with the agent's prior knowledge of the world, this event is deemed surprising and could indicate an anomaly. This paper lays out the details on how to learn both the prior and posterior models of an AGV that drives around a warehouse and observes the environment through an RGBD camera. In the experiments we show that our Bayesian surprise approach outperforms a baseline that is traditionally used to detect anomalies in sequences of images.

Upon the advent of Industry 4.0, collaborative robotics and intelligent automation gain more and more traction for enterprises to improve their production processes. In order to adapt to this trend, new programming, learning and collaborative techniques are investigated. Program-bydemonstration is one of the techniques that aim to reduce the burden of manually programming collaborative robots. However, this is often limited to teaching to grasp at a certain position, rather than grasping a certain object. In this paper, we propose a method that learns to grasp an arbitrary object from visual input. While other learning-based approaches for robotic grasping require collecting a large dataset, manually or automatically labeled in a real or simulated world, our approach requires a single demonstration. We present results on grasping various objects with the Franka Panda collaborative robot after capturing a single image from a wrist mounted RGB camera. From this image we learn a robot controller with a convolutional neural network to adapt to changes in the object's position and rotation with less than 5 minutes of training time on a NVIDIA Titan X GPU, achieving over 90% grasp success rate.

Internet applications rely on strong encryption techniques to protect the content of all communications between client and server. These encryption algorithms ensure that third parties are unable to obtain the plain text data but also make it hard for the network administrator to enforce restrictions on the types of traffic that are allowed. In this paper we show that we can train accurate machine learning models which can predict the type of traffic going through an IPsec or TOR tunnel based on features extracted from the encrypted streams. We use small, fast to execute machine learning models that work on small windows of data. This makes it possible to use our approach in real-time, for example as part of a Quality of Service (QoS) system.

Driven by the increasing popularity of the microservice architecture, we see an increase in services with unknown demand pattern located in the edge network. Predeployed instances of such services would be idle most of the time, which is economically infeasible. Also, the finite storage capacity limits the amount of deployed instances we can offer. Instead, we present an on-demand deployment scheme using the Docker platform. In Docker, service images consist of layers, each layer adding specific functionality. This allows different services to reuse layers, avoiding cluttering the storages with redundant replicas. We propose a layer placement method which allows users to connect to a server, retrieve all necessary layers -possibly from multiple locationsand deploy an instance of the requested service within the desired response time. We search for the best layer placement which maximizes the satisfied demand given the storage and delay constraints. We developed an iterative optimization heuristic which is less exhaustive by dividing the global problem in smaller subproblems. Our simulation results show that our heuristic is able to solve the problem with less system resources. Last, we present interesting use-cases to use this approach in real-life scenarios.

Deep reinforcement learning is becoming increasingly popular for robot control algorithms, with the aim for a robot to self-learn useful feature representations from unstructured sensory input leading to the optimal actuation policy. In addition to sensors mounted on the robot, sensors might also be deployed in the environment, although these might need to be accessed via an unreliable wireless connection. In this paper, we demonstrate deep neural network architectures that are able to fuse information generated by multiple sensors and are robust to sensor failures at runtime. We evaluate our method on a search and pick task for a robot both in simulation and the real world.

Using deep neural networks on resource constrained devices is a trending topic in neural network research. Various techniques for compressing neural networks have been proposed that allow evaluating a large neural network on a device with limited memory and processing power. These approaches usually generate a single compressed student network based on a larger teacher network. In some cases a more dynamic trade-off may be desired. In this paper we trained a sequence of increasingly large networks where each network is constrained to contain the unmodified features of all smaller networks. The weight matrix of the largest network has submatrices that correspond to the weight matrices of each of the smaller networks. This technique allows us to keep the parameters of several networks in memory while having the same memory footprint as the single largest network. A trade-off between accuracy and speed can be made at runtime. The proposed approach is validated on two image classification tasks running on a real-world Internet-of-Things (IoT) device.

In recent years, there has been a rising interest in cloud computing, which is often used to offer Software-as-a-Service (SaaS) over the Internet. SaaS can be offered to clients at a lower cost as it is usually multi-tenant: many end users make use of a single application instance, even when they are from different organizations, reducing resource consumption. This however makes it increasingly hard to offer customization and tailoring options for clients, resulting in limited SaaS customizability. Some applications require high customizability, making it impossible to cost-efficiently offer them using cloud computing. In this thesis we design, develop and evaluate multiple approaches for the development and management of highly customizable multi-tenant SaaS, focusing on an approach that decomposes applications into components and on the modeling of the relations between application components, and present a methodology for managing these applications in cloud datacenters and (traditional and Network Function Virtualization-aware) network environments. Extensive evaluations using simulations based on existing commercial applications and generated scenarios show that the proposed approach can be used to increase multi-tenancy and save costs in the management of highly customizable SaaS.

We see a trend to design services as a suite of small service components instead of the typical monolithic nature of classic web services, which led to an increasing amount of long-tail services on the Internet. Deploying instances everywhere to achieve a fast response time results in high costs, especially when these services are used infrequently and remain idle most of the time. One way to avoid needless over-provisioning is to deploy instances on-demand but this requires every component to be available upon request arrival. We propose a placement algorithm to maximize the amount of clients we can serve on-demand using the Docker layered filesystem. Docker facilitates automated deployment of services in lightweight software containers, allowing almost instantaneous deployment. Our algorithm finds the optimal storage location for layers so we can retrieve all service layers, deploy a service instance and provide a first response to a request within the desired time. We solve this problem using integer linear programming (ILP) and present techniques to improve the scalability of ILP while minimizing the performance loss. Results show that our approximation performs better with large scale problems than the classic ILP case.

The amount of services located in the network has drastically increased over the last decade which is why more and more datacenters are located at the network edge, closer to the users. In the current Internet it is up to the client to select a destination using a resolution service (Domain Name System, Content Delivery Networks). In the last few years, research on Information-Centric Networking (ICN) suggests to put this selection responsibility at the network components, routers find the closest copy of a content object using the content name as input. We extend the principle of ICN to services, service routers forward requests to service instances located in datacenters spread across the network edge. To solve this problem, we first present a service selection algorithm based on both server and network metrics. Next, we describe a method to reduce the state required in service routers while minimizing the performance loss caused by this data reduction. Simulation results based on real-life networks show that we are able to find a near-optimal load distribution with only minimal state required in the service routers.

Even though the problem of event detection from social media has been well studied in recent years, few authors have looked at deriving structured representations for their detected events. We envision the use of social media for extracting large-scale structured event databases, which could in turn be used for answering complex (historical) queries. As a key stepping-stone towards this goal, we introduce a method for discovering the semantic type of extracted events, focusing in particular on how this type is influenced by the spatio-temporal grounding of the event, the profile of its attendees, and the semantic type of the venue and other entities which are associated with the event. We estimate the aforementioned characteristics from metadata associated with Flickr photos of the event and then use an ensemble learner to identify its most likely semantic type. Experimental results based on an event dataset from Upcoming.org and Last.fm show a marked improvement over bag-of-words based methods.

Deep neural networks are the state of the art technique for a wide variety of classification problems. Although deeper networks are able to make more accurate classifications, the value brought by an additional hidden layer diminishes rapidly. Even shallow networks are able to achieve relatively good results on various classification problems. Only for a small subset of the samples do the deeper layers make a significant difference. We describe an architecture in which only the samples that can not be classified with a sufficient confidence by a shallow network have to be processed by the deeper layers. Instead of training a network with one output layer at the end of the network, we train several output layers, one for each hidden layer. When an output layer is sufficiently confident in this result, we stop propagating at this layer and the deeper layers need not be evaluated. The choice of a threshold confidence value allows us to trade-off accuracy and speed.

MOTIVATION: The accurate discovery and annotation of regulatory elements remains a challenging problem. The growing number of sequenced genomes creates new opportunities for comparative approaches to motif discovery. Putative binding sites are then considered to be functional if they are conserved in orthologous promoter sequences of multiple related species. Existing methods for comparative motif discovery usually rely on pregenerated multiple sequence alignments, which are difficult to obtain for more diverged species such as plants. As a consequence, misaligned regulatory elements often remain undetected. RESULTS: We present a novel algorithm that supports both alignment-free and alignment-based motif discovery in the promoter sequences of related species. Putative motifs are exhaustively enumerated as words over the IUPAC alphabet and screened for conservation using the branch length score. Additionally, a confidence score is established in a genome-wide fashion. In order to take advantage of a cloud computing infrastructure, the MapReduce programming model is adopted. The method is applied to four monocotyledon plant species and it is shown that high-scoring motifs are significantly enriched for open chromatin regions in Oryza sativa and for transcription factor binding sites inferred through protein-binding microarrays in O.sativa and Zea mays. Furthermore, the method is shown to recover experimentally profiled ga2ox1-like KN1 binding sites in Z.mays. AVAILABILITY AND IMPLEMENTATION: BLSSpeller was written in Java. Source code and manual are available at http://bioinformatics.intec.ugent.be/blsspeller CONTACT: [email protected] or [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

We present a framework for large-scale crowdsourcing of first-person viewpoint videos recorded on mobile devices. Collecting videos at a massive scale poses a number of major issues in terms of network planning. To improve the scalability with regards to the number of users, videos and geographical area and better cope with restrictions on storage, bandwidth and processing power, the framework is distributed and based on the two-layer cloudlet architecture. To mitigate the limited bandwidth in the access network, a set of decision algorithms is constructed and evaluated that are able to filter out irrelevant videos based on their metadata and given selection criteria. To illustrate the crowdsourcing framework, we present Street View Live, an application for presenting videos based on location, similar to the popular Google Street View but with up-to-date videos covering the location instead of possibly outdated images. In order to have an up-to-date view of every location, the video collection is continuously extended and updated by crowdsourcing videos from mobile devices.

Security is not taken into account by default in the Representational State Transfer (REST) architecture, but its layered architecture provides many opportunities for implementing it. In this paper, a security mechanism for Web Service communication through mobile clients devices is proposed, that conforms to the REST architecture as much as possible. This approach has been inspired by some known security mechanisms, but implemented in such a way that it focusses on statelessness and aims to be lightweight. Results indicate that the custom security mechanism outperforms the Transport Layered Security (TLS) based system. Because of the genericness of REST, the proposed security mechanism can be adopted by a wide variety of other RESTful Web Services.

In cloud environments, resources can be requested on-demand when they are needed. A cloud management system is responsible for determining which physical machines are responsible for processing the requests. The problem of determining which servers are used for which services is referred to as the Cloud Application Placement Problem (CAPP), and multiple criteria such as cost and number of migrations must be taken into account. When applications are constructed as a collection of communicating services, such as in Service-Oriented Architectures, it becomes important to take the underlying network properties into account when these placement decisions are made. In this paper, we propose an Integer Linear Programming (ILP) formulation for the CAPP, which optimizes multiple criteria such as cost, latency and number of migrations between subsequent invocations by using multiple optimization criteria. We also present hierarchical algorithms based on particle swarm optimization and genetic algorithms to solve the CAPP. These algorithms are be executed within a management hierarchy, which reduces the amount of information needed for the algorithms to function, increasing scalability of the management system. Finally, we evaluate the hierarchical algorithms by comparing them to an optimal algorithm based on the ILP formulation.

The task of assigning geographic coordinates to textual resources plays an increasingly central role in geographic information retrieval. The ability to select those terms from a given collection that are most indicative of geographic location is of key importance in successfully addressing this task. However, this process of selecting spatially relevant terms is at present not well understood, and the majority of current systems are based on standard term selection techniques, such as $(\chi^2)$ or information gain, and thus fail to exploit the spatial nature of the domain. In this paper, we propose two classes of term selection techniques based on standard geostatistical methods. First, to implement the idea of spatial smoothing of term occurrences, we investigate the use of kernel density estimation (KDE) to model each term as a two-dimensional probability distribution over the surface of the Earth. The second class of term selection methods we consider is based on Ripley's K statistic, which measures the deviation of a point set from spatial homogeneity. We provide experimental results which compare these classes of methods against existing baseline techniques on the tasks of assigning coordinates to Flickr photos and to Wikipedia articles, revealing marked improvements in cases where only a relatively small number of terms can be selected.

Social media sources such as Flickr and Twitter continuously generate large amounts of textual information (tags on Flickr and short messages on Twitter). This textual information is increasingly linked to geographical coordinates, which makes it possible to learn how people refer to places by identifying correlations between the occurrence of terms and the locations of the corresponding social media objects. Recent work has focused on how this potentially rich source of geographic information can be used to estimate geographic coordinates for previously unseen Flickr photos or Twitter messages. In this article, we extend this work by analysing to what extent probabilistic language models trained on Flickr and Twitter can be used to assign coordinates to Wikipedia articles. Our results show that exploiting these language models substantially outperforms both (i) classical gazetteer-based methods (in particular, using Yahoo! Placemaker and Geonames) and (ii) language modelling approaches trained on Wikipedia alone. This supports the hypothesis that social media are important sources of geographic information, which are valuable beyond the scope of individual applications.

We address the dimensioning of infrastructure, comprising both network and server resources, for large-scale decentralized distributed systems such as grids or clouds. We design the resulting grid/cloud to be resilient against network link or server failures. To this end, we exploit relocation: Under failure conditions, a grid job or cloud virtual machine may be served at an alternate destination (i.e., different from the one under failure-free conditions). We thus consider grid/cloud requests to have a known origin, but assume a degree of freedom as to where they end up being served, which is the case for grid applications of the bag-of-tasks (BoT) type or hosted virtual machines in the cloud case. We present a generic methodology based on integer linear programming (ILP) that: chooses a given number of sites in a given network topology where to install server infrastructure; and determines the amount of both network and server capacity to cater for both the failure-free scenario and failures of links or nodes. For the latter, we consider either failure-independent (FID) or failure-dependent (FD) recovery. Case studies on European-scale networks show that relocation allows considerable reduction of the total amount of network and server resources, especially in sparse topologies and for higher numbers of server sites. Adopting a failure-dependent backup routing strategy does lead to lower resource dimensions, but only when we adopt relocation (especially for a high number of server sites): Without exploiting relocation, potential savings of FD versus FID are not meaningful.

Databases of places have become increasingly popular to identify places of a given type that are close to a user-specified location. As it is important for these systems to use an up-to-date database with a broad coverage, there is a need for techniques that are capable of expanding place databases in an automated way. In this paper the authors discuss how geographically annotated information obtained from social media can be used to discover new places. In particular, the authors first determine potential places of interest by clustering the locations where Flickr photos have been taken. The tags from the Flickr photos and the terms of the Twitter messages posted in the vicinity of the obtained candidate places of interest are then used to rank them based on the likelihood that they belong to a given type. For several place types, their methodology finds places that are not yet contained in the databases used by Foursquare, Google, LinkedGeoData and Geonames. Furthermore, the authors’ experimental results show that the proposed method can successfully identify errors in existing place databases such as Foursquare.

In this paper we address the problem of dimensioning infrastructure, comprising both network and server resources, for large-scale decentralized distributed systems such as grids or clouds. We will provide an overview of our work in this area, and in particular focus on how to design the resulting grid/cloud to be resilient against network link and/or server site failures. To this end, we will exploit relocation: under failure conditions, a request may be sent to an alternate destination than the one under failure-free conditions. We will provide a comprehensive overview of related work in this area, and focus in some detail on our own most recent work. The latter comprises a case study where traffic has a known origin, but we assume a degree of freedom as to where its end up being processed, which is typically the case for e.g., grid applications of the bag-of-tasks (BoT) type or for providing cloud services. In particular, we will provide in this paper a new integer linear programming (ILP) formulation to solve the resilient grid/cloud dimensioning problem using failure-dependent backup routes. Our algorithm will simultaneously decide on server and network capacity. We find that in the anycast routing problem we address, the benefit of using failure-dependent (FD) rerouting is limited compared to failure-independent (FID) backup routing. We confirm our earlier findings in terms of network capacity savings achieved by relocation compared to not exploiting relocation (order of 6-10% in the current case studies).

Having numerous sensors on-board, smartphones have rapidly become a very attractive platform for augmented reality applications. Although the computational resources of mobile devices grow, they still cannot match commonly available desktop hardware, which results in downscaled versions of well known computer vision techniques that sacrifice accuracy for speed. We propose a component-based approach towards mobile augmented reality applications, where components can be configured and distributed at runtime, resulting in a performance increase by offloading CPU intensive tasks to a server in the network. By sharing distributed components between multiple users, collaborative AR applications can easily be developed. In this poster, we present a component-based implementation of the Parallel Tracking And Mapping (PTAM) algorithm, enabling to distribute components to achieve a mobile, distributed version of the original PTAM algorithm, as well as a collaborative scenario.

While the use of cloud computing is on the rise, many obstacles to its adoption remain. One of the weaknesses of current cloud offerings is the difficulty of developing highly customizable applications while retaining the increased scalability and lower cost offered by the multi-tenant nature of cloud applications. In this paper we describe a Software Product Line Engineering (SPLE) approach to the modelling and deployment of customizable Software as a Service (SaaS) applications. Afterwards we define a formal feature placement problem to manage these applications, and compare several heuristic approaches to solve the problem. The scalability and performance of the algorithms is investigated in detail. Our experiments show that the heuristics scale and perform well for systems with a reasonable load.

In recent years, there has been a growing interest in cloud technologies. Using current cloud solutions, it is however difficult to create customizable multi-tenant applications, especially if the application must support varying Quality of Service (QoS) guarantees. Software Product Line Engineering (SPLE) and feature modeling techniques are commonly used to address these issues in non-cloud applications, but these techniques cannot be ported directly to a cloud context, as the common approaches are geared towards customization of on-premise deployed applications, and do not support multi-tenancy. In this paper, we propose an architecture for the development and management of customizable Software as a Service (SaaS) applications, built using SPLE techniques. In our approach, each application is a composition of services, where individual services correspond to specific application functionalities, referred to as features. A feature-based methodology is described to abstract and convert the application information required at different stages of the application life-cycle: development, customization and deployment. We specifically focus on how development feature models can be adapted ensuring a one-to-one correspondence between features and services exists, ensuring the composition of services yields an application containing the corresponding features. These runtime features can then be managed using feature placement techniques. The proposed approach enables developers to define significantly less features, while limiting the amount of automatically generated features in the application runtime stage. Conversion times between models are shown to be in the order of milliseconds, while execution times of management algorithms are shown to improve by 5 to 17% depending on the application case.

Place recommender systems are increasingly being used to find places of a given type that are close to a user-specified location. As it is important for these systems to use an up-to-date database with a wide coverage, there is a need for techniques that are capable of expanding place databases in an automated way. On the other hand, social media are a rich source of geographically distributed information. In this paper, we therefore propose an approach to discover new instances of a given place type by exploiting correlations between terms and locations in geotagged social media. For a variety of place types, our approach is able to find places which are not yet included in popular place databases such as Foursquare or Google Places.

The evolution toward grid and cloud computing as observed for over a decennium illustrates the crucial role played by (optical) networks in supporting today's applications. In this paper, we start from an overview of the challenging applications in both academic (further referred to as scientific), enterprise (business) and nonprofessional user (consumer) domains. They pose novel challenges, calling for efficient interworking of IT resources, for both processing and storage, as well as the network that interconnects them and provides access to their users. We outline those novel applications' requirements, including sheer performance attributes (which will determine the quality as perceived by end users of the cloud applications), as well as the ability to adapt to changing demands (usually referred to as elasticity) and possible failures (i.e., resilience). In outlining the foundational concepts that provide the building blocks for grid/cloud solutions that meet the stringent application requirements we highlight, a prominent role is played by optical networking. The pieces of the solution studied in this respect span the optical transport layer as well as mechanisms located in higher layers (e.g., anycast routing, virtualization) and their interworking (e.g., through appropriate control plane extensions and middleware). Based on this study, we conclude by identifying challenges and research opportunities that can enable future-proof optical cloud systems (e.g., pushing the virtualization paradigms to optical networks).

Grids can efficiently deal with challenging computational and data processing tasks which cutting edge science is generating today. So-called e-Science grids cope with these complex task by deploying geographically distributed server infrastructure, interconnected by high speed networks. The latter benefit from optical technology, offering low latencies and high bandwidths, thus giving rise to so-called optical grids or lambda grids. In this paper, we address the dimensioning problem of such grids: how to decide how much server infrastructure to deploy, at which locations in a given topology, the amount of network capacity to provide and which routes to follow along them. Compared to earlier work, we propose an integrated solution solving these questions in an integrated way, i.e., we jointly optimize network and server capacity, and incorporate resiliency against both network and server failures. Assuming we are given the amount of resource reservation requests arriving at each network node (where a resource reservation implies to reserve both processing capacity at a server site, and a network connection towards it), we solve the problem of first choosing a predetermined number of server locations to use, and subsequently determine the routes to follow while minimizing resource requirements. In a case study on a meshed European network comprising 28 nodes and 41 links, we show that compared to classical (i.e. without relocation) shared path protection against link failures only, we can offer resilience against both single link and network failures by adding about 55% extra server capacity, and 26% extra wavelengths.

This paper is aiming at facilitating the energy-efficient operation of an integrated optical network and IT infrastructure. In this context we propose an energy-efficient routing algorithm for provisioning of IT services that originate from specific source sites and which need to be executed by suitable IT resources (e.g. data centers). The routing approach followed is anycast, since the requirement for the IT services is the delivery of results, while the exact location of the execution of the job can be freely chosen. In this scenario, energy efficiency is achieved by identifying the least energy consuming IT and network resources required to support the services, enabling the switching off of any unused network and IT resources. Our results show significant energy savings that can reach up to 55% compared to energy-unaware schemes, depending on the granularity with which a data center is able to switch on/off servers.

As the requirements and scale of cloud environments increase, scalable management of the cloud is needed. Centralized solutions lack scalability and fully distributed management systems only have a limited overview of the system. One of the often-studied problems in cloud environments is the application placement problem, used to decide where application instances are instantiated and how many resources to allocate to the instances. In this paper a general approach is introduced for using centralized cloud resource management algorithms in a hierarchical context, increasing the scalability of the management system while maintaining a high placement quality. The management system itself is executed on the cloud, further increasing scalability and robustness. The proposed method uses aggregation techniques to generate input values for a centralized application placement algorithm which is run in all management nodes. Decoupling ensures management nodes can function independently. Subsequently, we compare the performance of hierarchical application placement method with that of a fully centralized algorithm. The results show that a solution, within 5% of the optimum placement when using the centralized algorithm, can be achieved hierarchically in less than 25% of the time needed for execution of the centralized algorithm.

We present a two-step approach to estimate where a given photo or video was taken, using only the tags that a user has assigned to it. In the first step, a language modeling approach is adopted to find the area which most likely contains the geographic location of the resource. In the subsequent second step, a precise location is determined within the area that was found to be most plausible. The main idea of this step is to compare the multimedia object under consideration with resources from the training set, for which the exact coordinates are known, and which were taken in that area. Our final estimation is then determined as a function of the coordinates of the most similar among these resources. Experimental results show this two-step approach to improve substantially over either language models or similarity search alone.

MOTIVATION: Many comparative genomics studies rely on the correct identification of homologous genomic regions using accurate alignment tools. In such case, the alphabet of the input sequences consists of complete genes, rather than nucleotides or amino acids. As optimal multiple sequence alignment is computationally impractical, a progressive alignment strategy is often employed. However, such an approach is susceptible to the propagation of alignment errors in early pairwise alignment steps, especially when dealing with strongly diverged genomic regions. In this article, we present a novel accurate and efficient greedy, graph-based algorithm for the alignment of multiple homologous genomic segments, represented as ordered gene lists. RESULTS: Based on provable properties of the graph structure, several heuristics are developed to resolve local alignment conflicts that occur due to gene duplication and/or rearrangement events on the different genomic segments. The performance of the algorithm is assessed by comparing the alignment results of homologous genomic segments in Arabidopsis thaliana to those obtained by using both a progressive alignment method and an earlier graph-based implementation. Especially for datasets that contain strongly diverged segments, the proposed method achieves a substantially higher alignment accuracy, and proves to be sufficiently fast for large datasets including a few dozens of eukaryotic genomes. AVAILABILITY: http://bioinformatics.psb.ugent.be/software. The algorithm is implemented as a part of the i-ADHoRe 3.0 package.

The use of semantic technology has recently witnessed a huge increase. One of the areas in which this technology is being used increasingly more often is that of context-aware agents. However, the use of ontologies in general and reasoning in particular can rapidly become resource intensive. Certainly if the data set, called the A-Box, used by these agents grows considerably over time. Moreover, in order to create context-aware applications, taking into account a wide range of different data sets and context parameters, agents have to be provided to expose that data. The collaboration between the agents in the system is necessary to correlate the information and augment the intelligence and added value of the context-aware agents. Therefore, there is a need to have a distributed approach by means of a service-platform, where the different agents in a context-aware environment can collaborate. The main focus of this article is on the research on the design of a service-platform for semantic ontology-based context-aware collaboration. The platform architecture to allow the collaboration and scheduling, together with the associated algorithms, will be presented. The engineering and implementation details will be highlighted. By means of detailed UML sequence diagrams, we will present the workflow and collaboration between the different modules in the platform. Additionally, supporting developments, such as the meta-ontology and our ontology generator, OTAGen, will be presented. Furthermore, we will detail how the platform can operate in an autonomous way, taking into account the changing context of the agents in the platform.

Nowadays, users consume a lot of functionality in their home coming from a service provider located in the Internet. While the home network is typically shielded off as much as possible from the `outside world', the supplied services could be greatly extended if it was possible to use local information. In this article, an extended service is presented that integrates the user's multimedia content, scattered over multiple devices in the home network, into the Electronic Program Guide (EPG) of the Digital TV. We propose to virtualize the set-top box, by migrating all functionality except user interfacing to the service provider infrastructure. The media in the home network is discovered through standard Universal Plug and Play (UPnP), of which the QoS functionality is exploited to ensure high quality playback over the home network, that basically is out of the control of the service provider. The performance of the subsystems are analysed.

This paper presents the results of a user study carried out to evaluate an application prototype in which an audio-visual chat conversation between two users is augmented by pictures related to the topics of that conversation. The prototype analyses the conversation and deducts the topic of conversation by means of a keyword tree, augmented by an ontology. Then it retrieves pictures from Flickr based on this topic, after which the pictures are shown to the users. This mechanism is called conversation-based image retrieval. 15 participants were recruited for this user study; the duration of one session was approximately 30 minutes. Eye tracking and questionnaires were used to evaluate participants' experiences. We found that participants value the use of pictures to augment an audio-visual chat application. Furthermore, participants claimed they would use it in a social context: talking to family, friends and acquaintances. One significant improvement over the prototype would be to use their own pictures (personal user-generated content) instead of just random pictures.

As the adoption of the Service Oriented Architecture paradigm has dramatically increased over the past few years, proper coordination of loosely coupled services becomes an important issue when building state-of-the-art applications. This coordination is typically organized through orchestration (requiring a central coordinating entity) or through choreographies. While the latter approach allows for a fully distributed coordination, the need also arises for a distributed conformance check, ensuring that each participant of the choreography behaves according to the general choreography. In this paper, a formalism is presented to ensure this conformance at design time, with possible extensions to deploy time and to runtime conformance checking. This formalism is referred to as the piX-model and it will be shown that the approach taken is inherently less complex, both in time and space, than the conventional π-calculus-based approach, whilst offering the same conformance guarantees. This gain in performance allows for a small design turnaround time, and also opens the avenue to runtime conformance checking by resource constrained devices.

Deflection routing is a much-studied contention resolution technique in the context of Optical Burst/Packet Switching networks, as it promises to improve burst blocking performance and may reduce or even eliminate buffer requirements. An OBS-based Grid is frequently based on anycast routing, which h

Service choreographies describe the interactions that take place in a distributed service collaboration without central entity orchestrating these interactions. It is obvious that each partner will execute parts of the choreography to fulfill the global collaborative effort. This paper focuses on translating the global choreography to local projections at design time. These projections need to be implemented by each participating partner. The process is decomposed in two steps: instantiation and extraction. In the instantiation step the abstraction levels are automatically determined, ranging from the choreography level to its smallest building blocks, the channel instances. In the extraction step, we present a way to map these channel instances to WS-BPEL. It is shown that this results in small WS-BPEL processes with a very straightforward correlation set, allowing for even resource-limited devices to participate in the choreography.

Massively online virtual environments (MOVEs) have been gaining popularity for several years. Today, these complex networked applications are serving thousands of clients simultaneously. However, these MOVEs are typically hosted on specialized server clusters and rely on internal knowledge of the services to optimize the load balancing. This makes running MOVEs an expensive undertaking as it cannot be outsourced to third party hosting providers. This paper details two integer linear programming approaches to optimize the MOVE deployment through load balancing and minimizing the delay experienced by the end-users. Optimization includes assigning MOVE components to resources and replication of components to increase the scalability. One approach assuming full application knowledge of a dedicated MOVE and one with no internal knowledge and geared toward a generic MOVE hosting platform. For both cases an optimizing heuristic is evaluated and the obtained results are compared.

A generic and self-managing service hosting infrastructure, provides a means to offer a large variety of services to users across the Internet. Such an infrastructure provides mechanisms to automatically allocate resources to services, discover the location of these services, and route client requests to a suitable service instance. In this paper we propose a dynamic and latency-aware algorithm for assigning resources to services. Additionally, the proposed service hosting architecture and its protocols to support the service placement algorithm, are described in detail. Extensive simulations were performed to compare the solution of our latency-aware algorithm to the latency-unaware variant, in terms of system efficiency and scalability.

In thin client computing, applications are executed on a network server instead of on the user terminal. Since the amount of processing at the terminal is reduced, thin clients are potentially energy efficient devices. However, a network connection between client and server is required for the transmission of user input and display updates. The energy needed for this intense network communication might undo or even exceed the power savings achieved by the reduction in client-side processing. In this paper, we present experimental results on power efficiency of the wireless platform on the thin client in case of thin client traffic. The discussion is focused on VNC-RFB, a widespread thin client protocol, over an IEEE 802.11 link in three typical user scenarios. The results indicate that a cross-layer approach between application and wireless link layer could potentially lead to important power savings.

A grid is a distributed computational and storage environment often composed of heterogeneous autonomously managed subsystems. As a result, varying resource availability becomes commonplace, often resulting in loss and delay of executing jobs. To ensure good grid performance, fault tolerance should be taken into account. Commonly utilized techniques for providing fault tolerance in distributed systems are periodic job checkpointing and replication. While very robust, both techniques can delay job execution if inappropriate checkpointing intervals and replica numbers are chosen. This paper introduces several heuristics that dynamically adapt the above mentioned parameters based on information on grid status to provide high job throughput in the presence of failure while reducing the system overhead. Furthermore, a novel fault-tolerant algorithm combining checkpointing and replication is presented. The proposed methods are evaluated in a newly developed grid simulation environment dynamic scheduling in distributed environments (DSiDE), which allows for easy modeling of dynamic system and job behavior. Simulations are run employing workload and system parameters derived from logs that were collected from several large-scale parallel production systems. Experiments have shown that adaptive approaches can considerably improve system performance, while the preference for one of the solutions depends on particular system characteristics, such as load, job submission patterns, and failure frequency.

To cope with ever-increasing traffic demands in transport networks, all-optical switching is currently perceived as a potential solution to remove bottlenecks caused by optoelectronic conversions. An effective realization of this concept must support a wide range of traffic patterns, while remaining feasible to construct and deploy both in an economical and practical sense. In this paper, we propose the use of multi-granular optical cross-connects (MG-OXC), which support switching on both the wavelength and sub-wavelength level. To this end, the MG-OXCs are equipped with cheap, highly scalable slow switching fabrics, as well as a small number of expensive fast switching ports. The main goal of this work is to motivate the use of multi-granular switching, as this can reduce total network installation costs. To this end, we introduce an Integer Linear Programming model, and our evaluation demonstrates that multi-granular optical switching can be a cost-effective solution on the network level, in comparison to slow only or fast only approaches. Furthermore, we can achieve reduced costs of individual OXC nodes, which allows us to minimize scalability problems corresponding to emerging fast switching fabrics.

On the one hand, agent-based software platforms are commonly used these days, while on the other hand Semantic Web technologies are also maturing. It is obvious that the combination of these two technologies can bring added value through the creation of Semantic Agent-based frameworks. However, it is also known that these Semantic Web technologies, and the reasoning on ontologies in particular, can rapidly become resource intensive. In order to get a clear view on this problem, we have developed OTAGen, a highly tunable tool to generate customized ontologies and corresponding queries. The generated ontologies can then be used to evaluate at design-time the performance of the Semantic Agent-based platform as a function of the number of ontologies, users and queries.

In this paper, an MILP-based methodology is presented that allows to optimize the deployment of a set of software components over a set of computing resources, with respect to fault tolerance and response times. The MILP model takes into account the reliability and performance parameters of hardware nodes and links, and optimizes a (configurable) trade-off between reliability and performance by replicating software components where necessary and finding an optimal deployment for them. The complete system can be modeled using UML component diagrams and activity diagrams, and an algorithm is presented to transform the UML model to the MILP model. The resulting deployment can then be fed back into the UML model. The applicability of the approach is demonstrated through a case study.

More and more, multimedia services are being accessed via fixed and mobile networks. These services are typically much more sensitive to packet loss, delay and/or congestion than traditional services. In particular, multimedia data is often time critical and, as a result, network issues are not well tolerated and significantly deteriorate the userpsilas quality of experience (QoE). We therefore propose a QoE optimization platform that is able to mitigate problems that might occur at any location in the delivery path from service provider to customer. More specifically, the distributed architecture supports overlay routing to circumvent erratic parts of the network core. In addition, it comprises proxy components that realize last mile optimization through automatic bandwidth management and the application of processing on multimedia flows. In this paper we introduce a transcoding service for this proxy component which enables the transformation of H.264/AVC video flows to an arbitrary bit rate. Through representative experimental results, we illustrate how this addition enhances the QoE optimization capabilities of the proposed platform by allowing the proxy component to compute more flexible and effective bandwidth distributions.

Grid computing aims to realize a high-performance computing environment, while increasing the usage efficiency of installed resources. This puts considerable constraints on the network technology, and ultimately has led to the development of Grids over optical networks. In this paper, we investigate the fundamental question of how to optimize the performance of such Grid networks. We start with an analysis of different architectural approaches (and their respective technological choices) to integrate Grid computing with optical networks. This results in models and algorithms to design optical Grid networks, and we show the importance to combine both dimensioning (offline) and scheduling (online) in the design phase of such systems. Finally, the concept of anycast routing is introduced and motivated. Both exact and heuristic algorithms are proposed, and their performance in terms of blocking probability and latency is presented.

As the complexity of business processes grows, the shift towards workflow-based programming becomes more attractive. The typical long-running characteristic of workflows imposes new challenges such as dynamic adaptation of running workflow instances. Windows Workflow Foundation (in short WF) was released by Microsoft as their solution for workflow-driven application development. Although WF contains features that allow dynamic workflow adaptation, the framework lacks an instrumentation framework to make such adaptations more manageable. Therefore, we built an instrumentation framework that provides more flexibility for applying workflow adaptation batches to workflow instances, both at creation time and during an instance's lifecycle. In this paper we present this workflow instrumentation framework and performance implications caused by dynamic workflow adaptation are detailed.

In order to deliver intelligent context-aware services (e.g. notifications of nearby points of interest), location based services are getting a lot of interest. However, deploying these services efficiently is currently hampered by the lack of enabling platforms, especially platforms taking advantage of easy service composition and flexible management. In this paper we detail the design of pCASE, a platform for enabling context aware services. This platform allows easy deployment in both home and business premises and flexible management of complex services taking into account context information such as location and presence. A use case is presented: an intelligent call redirection service which dynamically (re)routes communication sessions depending on the location, occupation and social networks of users.

Notwithstanding IP anycast's introduction in Internet standards dates back to 1993 and its more recent adoption in IPv6 standards, its use in production environments is limited to date. This is mainly because native IP anycast lacks routing scalability and does not support session-based communications, thereby limiting its applicability to single request-response services such as DNS. For this reason, we propose a transparent anycast overlay architecture that retains the strengths of native anycast and neutralizes above-mentioned limitations. The resulting proxy infrastructure unleashes the power of anycast by opening up new opportunities for transparent distributed service provisioning. Taking into account user demands, available resources, network overhead and anycast infrastructure costs, we provide near- optimal heuristics for the placement of proxy nodes and dimensioning the infrastructure in large networks. We show that even modest overlay infrastructures, consisting of a small number of proxy routers, provide an effective stateful anycast solution where the detour via the proxy routers is negligible in terms of extra network load. Furthermore, simulation results illustrate that server state aggregation in the proxy nodes lessens control plane overhead, which contributes significantly to service robustness.

As the popularity of resource-constrained devices such as hand-held computers increases, a new network service off loading complex processing tasks towards computational resources located in the access- or core network, sounds very promising. In a consumer-oriented environment, characterized by a large diversity in connected devices, a transparent network-based request processing strategy offers a clear flexibility advantage, as the installation and configuration of extra software components on all client devices can be avoided. In this work, this is achieved by linking computational resources to an any cast group, which allows intermediate router nodes to decide upon the target server. It is shown in the paper that the anycast routing problem can be reduced to unicast routing. Consequently, unicast multiple constraints routing algorithms can be applied to compute an optimal path based on several server selection criteria, including server load, path delay, path cost, etc. For this purpose, we envision the SAMCRA algorithm. A new evaluation ordering strategy for previously computed sub-paths is introduced, which guarantees optimality for the complete SAMCRA path between source and destination. Simulation results show that an effective distribution of the job scheduling requests over the available resources can be achieved by applying the described algorithm.

While the Java programming language and the J2EE platform are increasingly popular for implementing business logic on backend platforms, new emerging Java technologies such as JAIN SLEE and SIP Servlet are focusing on the development of low latency Java applications. As J2EE mainly focuses on enterprise applications with complex long lasting transactions, this technology is considered unsuitable for applications with low latency and high throughput characteristics. This paper compares these telecom oriented Java technologies to J2EE both in terms of functionality and through a detailed performance evaluation. JVM performance tuning has been studied as well and is explained in the paper. We performed a SIP proxy benchmark with strict low latency requirements of which the results are presented. Furthermore, design guidelines for J2EE applications are discussed to optimize for low latency behavior together with an interpretation of the obtained performance results.

More and more, content providers offer multimedia services such as Internet TV, multimedia conferencing and online gaming to their customers. These services are characterized by their high sensitivity to network delay and a multicast nature. An overlay network allows for supporting QoS by making reservations in the underlying networks and for multicasting the multimedia streams towards their targets at the overlay layer, without requiring multicast support from the underlying networks. This paper outlines the architecture of a dynamic QoS enabled multicast overlay network and also introduces a set of algorithms to determine an overlay distribution tree that connects a multimedia server to a number of clients. The algorithms construct a tree with a bounded end-to-end delay and minimize the bandwidth that is used. These algorithms are evaluated in terms of bandwidth cost, overlay cost and end-to-end delay. We show that one of our heuristics finds overlay multicast trees that approximate the optimal result in terms of cost and that have a small diameter and a low average delay.

A recent important evolution in broadband access network design is the deployment of IP aware access network elements, which allow to introduce access network services beyond basic triple-play. The focus of this paper is on the management of time-shifted television (tsTV), an IPTV service which allows for watching the broadcast content at real-time or with a (small) time shift. An architecture for a large-scale tsTV service deployment is presented, using co-operating transparent diskless proxy caches in broadband access networks, with an implementation based on the IETF's real-time streaming protocol (RTSP). Caching algorithms have been designed to take into account content popularity and distance metrics. The algorithms make use of the sliding window concept and calculate the optimal trade-off between bandwidth usage efficiency and storage cost. A prototype implementation of a transparent tsTV proxy is presented and evaluated through performance measurements.

A current important trend is the introduction of new services in the access and aggregation network, close to the end user. A major opportunity for such service enabled access networks is providing users with fast and reliable storage, allowing them to transparently access and share their digital media library anytime, anywhere, while guaranteeing data retention. An important issue in deploying such a service is where to put the storage servers, in order to minimize deployment cost without sacrificing performance. This paper presents and evaluates two algorithms for solving the storage server placement problem, minimizing the deployment cost while guaranteeing a low delay for accessing the digital media libraries from any access node. The base algorithm, referred to as SSPA (storage server placement algorithm), assumes the access and aggregation network has unlimited bandwidth. SSPA provides a lower bound on the required number of servers, respecting a maximum delay constraint. The extended algorithm, SSPA, solves the storage server placement problem, respecting both a maximum delay constraint and bandwidth constraints of the network links. It will be shown that the algorithms produce close to optimal results relatively fast

The increasing use of wireless networks and the popularity of multimedia applications, leads to the need of Quality of Service support in a mobile IP-based environment. In [1], we investigated the reasons for path changes in the access network and we presented a framework to support these path changes regardless their cause and to avoid router inconsistencies. In this paper, the performance of this framework is thoroughly evaluated. To this end, we added the functionality of the framework to our Q-MEHROM handoff scheme [2]. However, the ideas and results presented in this paper are not restricted to this micromobility protocol. The performance with and without the defined functionalities is compared and tested under different network topologies, in terms of received service, packet loss and end-to-end delay.

Content distribution networks (CDN) have been increasingly used to deliver bandwidth-intensive multimedia content to a large amount of users. In a CDN, the content is replicated from the origin server to so-called surrogate servers in order to improve the quality of service experienced by the end-users and decrease the network load. However, despite the promising concept, current centralized and distributed CDN architectures lack placement and retrieval algorithms that are both scalable and provide a close to optimal placement. In this article, we propose a novel replica placement algorithm called COCOA (cooperative cost optimization algorithm), suited for a self-organizing hybrid CDN architecture. Our results show that COCOA achieves a performance comparable to the less scalable centralized algorithms, while maintaining the benefits of distributed approaches. Contrary to the more common off-line content replication and management strategies, on-line replication in a self-optimizing CDN puts an additional strain on the network. We explore techniques to control this traffic and study its implications on the performance of the CDN. Because in a CDN content is replicated to geographically distributed surrogate servers, one of the main benefits is its ability to recover from network failures and increase the availability of content during flash crowds. As illustrated in this article, we succeed in making the CDN more robust by effectively reducing the convergence time of the network after the occurrence of such disruptive events

As Massively Multiplayer Online Games enjoy a huge popularity and are played by tens of thousands of players simultaneously, an efficient software architecture is needed to cope with the dynamically changing loads at the server side. In this paper we discuss a novel way to support this kind of application by dividing the virtual world into several parts, called microcells. Every server is assigned a number of microcells and by dynamically redeploying these microcells when the load in a region of the world suddenly increases, the platform is able to adapt to changing load distributions. The software architecture for this system is described and we also provide some evaluation results that indicate the performance of our platform.

Despite the growing popularity and advantages of thin-client systems, they still have some important shortcomings. Current thin-client systems are ideally suited to be used with classic office-applications but as soon as multimedia and 3D gaming applications are used they require a large amount of bandwidth and processing power. Furthermore, most of these applications heavily rely on the Graphical Processing Unit (GPU). Due to the architectural design of thin-client systems, they cannot profit from the GPU resulting in slow performance and bad image quality. In this paper, we propose a thin-client system which addresses these problems: we introduce a realtime desktopstreamer using a videocodec to stream the graphical output of applications after GPU-processing to a thin-client device, capable of decoding a videostream. We compare this approach to a number of popular classic thin-client systems in terms of bandwidth, delay and image quality. The outcome is an architecture for a hybrid protocol, which can dynamically switch between a classic thin-client protocol and realtime desktopstreaming.

Due to natural disasters or intentional attacks, large parts of our telecommunication network can be destroyed. The setup of a wireless shadow network can restore the damaged network connectivity. This paper investigates the need for a network layer solution to integrate such a wireless shadow network with the wired network. It presents the combination of the micromobility protocols MEHROM and mobile IP regional registration as an efficient solution to support terminal mobility. The proposed solution is evaluated in terms of control overhead, required storage and calculation capacity, and functional complexity

In this paper, we propose the use of a distributed service management architecture for state-of-the-art service-enabled grids. The architecture is capable of performing automated resource and network bandwidth partitioning based on registered grid resource properties and monitored grid service demand. A main characteristic is that it enables the use of different service priority schemes and allows for policy-based differentiation between local and foreign service offerings. Resource and network bandwidth partitioning algorithms are introduced and their performance is evaluated on a sample grid topology using NSGrid, an ns-2 based grid simulator. Our results show that the use of this Service Management Architecture improves resource efficiency, simplifies schedule making decisions, reduces the overall complexity of managing the grid system, and at the same time improves grid service QoS support (with regard to job response times) by automatically making grid resource and network service reservations prior to scheduling.

The increasing use of wireless networks and the popularity of multimedia applications, leads to the need of QoS (quality of service) support in a mobile IP-based environment. This paper considers the reservation of resources in the micromobility scenario. The reasons for path changes in the access network are investigated. Based upon this, we present the functionality that should be present in the different access network elements to support path changes regardless their cause and to avoid router inconsistencies. For the simulations we added this functionality to our previously developed Q-MEHROM handoff scheme (L. Peters et al., 2005). However, the ideas and results presented in this paper are not restricted to this micromobility protocol

Currently, a lot of research has been devoted to (i) content distribution, (ii) traffic engineering, (iii) network monitoring and (iv) service enabling platforms. However, the integration of these four individual concepts in a single platform has not yet been studied in enough detail. In this paper we present an architecture for such a robust content delivery service. The combination of both distributed replication of videos and multi-source traffic engineering tackles specific problems such as congested network parts, overloaded servers and the occurrence of flash crowds. Contrary to most existing systems, the content placement and retrieval algorithms in the presented CDN obtain precise network state information from an integrated monitoring system, allowing even a higher efficiency. To validate the performance of the CDN, an exact placement ILP formulation and various RPA heuristics are implemented and simulated.

With the deployment of broadband Internet access, a new set of applications that is becoming more and more popular are highly interactive services such as online gaming. When the delay experienced in an online game passes a certain threshold, the quality of service (QoS) degrades enormously, as a result the clients are no longer satisfied and might decide to leave the game or to cancel the subscription. Thus, delay is critical to the success of this type of service. Network delay is a result of the way the data is distributed between the players. In this paper we propose the use of a set of game servers. These servers not only play the role of a traditional game server but also form an overlay network which can be used as an application layer routing infrastructure, allowing us to route along low delay paths and thereby increasing the QoS of the game. We study algorithms that determine the ideal location of the servers in a network and look at the performance of the resulting overlay networks in terms of end-to-end delay and the relationship between the number of servers in the overlay network and the offered QoS. We also compare our distributed architecture with both a single server and a peer-to-peer architecture.

Distributed systems often use a form of communication middleware to cope with different forms of heterogeneity, including geographical spreading of the components, different programming languages and platform architectures, etc. The middleware, of course, impact the architecture and the performance of the system. This paper presents a model transformation framework to automatically include the architectural impact and the overhead incurred by using a middleware layer between several system components. Using this framework, architects can model the system in a middleware-independent fashion. Accurate, middleware-aware models can then be obtained automatically using a middleware model repository. The actual transformation algorithm is presented in more detail. The resulting models can be used to obtain performance models of the system. From those performance models, early indications of the system performance can be extracted.

An ad hoc network is an autonomous system of heterogeneous, mobile nodes that communicate with each other over wireless links. Routing protocols for these networks are inherently based on broadcasting control information and are therefore very bandwidth consuming. In order to limit the amount of routing information that has to be stored and maintained by the individual nodes, the technique of clustering is used. The network is partitioned into nonoverlapping sub networks, referred to as clusters, and one cluster node, the clusterhead, will take a leading role in the dissemination of control information. In this paper we approach the problem of finding an optimal partition that explicitly takes into account the heterogeneity of the network, as an integer linear programming (ILP) problem. In a second phase we have developed a new heuristic that approximates our ILP solution that is used in our clustering algorithm. It is shown that this heuristic tends to be more stable than existing clustering techniques that are solely based on ID number and/or connectivity and that do not take into account the heterogeneity of the network.

The combination of mobile IP with frequent handovers results in high handoff latencies and control overhead in the core network. Micromobility protocols like cellular IP. Hawaii and MIPv4 regional registration were developed to avoid these problems, supporting local mobility within one IP domain. Hereby, tree access topologies were considered to evaluate the protocol performance A. Campbell et al., (2000) (2002) although meshed topologies are desired for reasons of robustness against link failures and load balancing L. Peters et al., (2003). The mentioned micromobility protocols do not use the extra links or result in a bad performance in terms of end-to-end delay. This paper proposes a new micromobility protocol for random access topologies. During handoff a router in the access network detects if it has the function of cross-over node. Hereby, control traffic is concentrated near the involved access routers, while optimal paths in the access network are achieved. The basic concept of the protocol is presented and the handoff performance is compared to other micromobility protocols.

Active networking is one of the suggested technologies to introduce additional intelligence and programmability in the network and its services. In this paper, the use of active networking to support advanced multicast services providing media transcoding inside the network is investigated. In the multicast service different versions of the streamed data are made available and customers can select a specific version according to their wishes or their capabilities. Based on the active networking facilities of the underlying framework the different versions of the streamed data can be created inside the network, through transformations or transcodings of the original data. Both design and performance issues of the detailed service are discussed. A new multicast tree set-up protocol, taking into account the required transcodings, is introduced. A number of different strategies are discussed optimizing the location of the transcodings as well as the use of bandwidth in the network, while considering the availability of sufficient processing power in the nodes. The performance analysis is done for a voice stream multicast service, addressing the efficiency of the tree set-up strategies, the optimization of network resource utilization, the use of processing power for transcodings, and the resulting quality of streamed voice signals after multiple transcodings.

In this paper, we address the design and implementation of a generic and scalable platform for efficient management of computational resources. The developed platform is called the Intelligent Agent Platform. Its architecture is based on middleware technology in order to ensure easy distribution of the software components between the participating workstations and to exploit advanced software techniques. The computational tasks are referred to as agents, defined as software components that are capable of executing particular algorithms on input data. The platform offers advanced features such as transparent task management, load balancing, run time compilation of agent code and task migration and is therefore denoted by the adjective "Intelligent". The architecture of the platform will be outlined from a computational point of view and each component will be described in detail. Furthermore, some important design issues of the platform are covered and a performance evaluation is presented.

The growing widespread use of advanced multimedia and interactive real-time applications is setting forth new challenges such as end-to-end quality-of-service (QoS) and broadband Internet access. The high bandwidth needs are pushing fiber closer and closer to the home, and as such WDM (wavelength division multiplexing) seems ideally suited to be used in the broadband access feeder network which interconnects the Internet core networks and the last mile networks. In the HARMONICS project (Hybrid Access Reconfigurable Multi-wavelength Optical Networks for IP-based Communication Services), a novel DWDM (dense wavelength division multiplexing) based optical access feeder network is investigated. This feeder network transports IP, guarantees QoS and can feed various last mile networks, stimulating the convergence of access networks. VDSL and Hiperlan/2 are studied within the project as last mile access networks. To support end-to-end QoS, a distributed CORBA-based generic network management framework is being developed as part of the project. This paper elaborates on the framework which is aligned with TINA, although adapted to be more consistent and applicable. End-to-end QoS is based on Differentiated Services (DiffServ) at layer 3, various QoS supporting technologies at layer 2 and QoS mappings between both layers.