Distributed Cooperative Laboratories: Networking, Instrumentation, and Measurements

This paper presents a novel approach on a joint team collaboration system. It combines two fields of applications that have been developed separately in the past: collaborative virtual environments and video-based teleconferencing. For this purpose technologies from both areas have been integrated to a common test platform. After a specific description of the initial technology branches, the paper mainly focuses on the conceptual work for merging them to a novel virtual team user environment.

Recently, centrally controlled information distribution systems are rapidly emerging to decentralized structures. This tendency can be observed in software distribution applications using BitTorrent or information sharing platforms based on distributed hash table structures like Chord or Kademlia. To ensure that the emerging platforms will function properly with a growing number of users and services the issue of

scalability

turned into one of the hottest research topics.

Traditionally, the term scalability often restricts to the functional scalability, which describes the scalability in terms of the system size. In this regard the basic structure is stationary, i.e. it does not fluctuate frequently. However, when the stochastic behavior of system components, the network structure, and user applications has to be taken into account, the

stochastic scalability

has to be investigated in the context of performance evaluation. In this paper we discuss the stochastic scalability of information sharing platforms. We give a classification of current information sharing platforms and define the terms functional and stochastic scalability in detail. A distributed phone book based on a Chord ring will be discussed as an example to motivate other areas of application and to show the potential of the evaluation of stochastic scalability.

The interest in creating multimedia support for remote learning is increasing explosively, thanks to the development of a variety of multimedia software packages for slide presentation, on the one hand, and the spread of the Internet in schools, Universities, research laboratories and the majority of student houses, on the other hand. However, most of the e-learning platforms used today are merely off-line. In other words, they consist of a multimedia site where teachers and students can exchange information and didactic material through file transfer and e-mail, or where the teacher can propose evaluation tests to the students. At the same time many videoconference hardware and software platforms have been deployed to support audio/video communications between two or more users. However, their main target is videoconference applications, and they are not suitable for teaching environments. The target of this paper is to provide an overview of the main aspects related to the development of multimedia e-learning tools which allow on-line remote learning, the main characteristic of which is close real-time interaction between teachers and students, and between students themselves. The authors conclude the paper by describing a relevant experience, the VIP-Teach tool, developed according to the principles illustrated in the paper.

The Hybrid Coordination Function (HCF) has been recently proposed by the 802.11e working group in order to provide real-time services in Wireless Local Area Networks (WLANs). HCF is made of a contention-based channel access, known as Enhanced Distributed Coordination Access (EDCA), and of a HCF Controlled Channel Access (HCCA), which requires a centralized controller, called Hybrid Coordinator (HC). This paper proposes two feedback-based bandwidth allocation algorithms exploiting HCCA for dynamically assigning the WLAN channel bandwidth to mobile stations hosting real-time traffic streams. Proposed algorithms, which have been referred to as Feedback Based Dynamic Scheduler (FBDS) and Proportional Integral (PI)-FBDS, have been designed using classic discrete-time feedback control theory. Simulation results, obtained using the

ns

-2 simulator, have shown that, unlike the simple scheduler proposed by the 802.11e working group, both FBDS and PI-FBDS provide a real-time service regardless of the network load. Moreover, when the PI-FBDS is used, the best trade-off between one-way packet delays and network utilization is achieved.

An immersive audio environment for videoconferencing or virtual reality applications is presented. The proposed system allows the spatialisation of 3D sound fields around listener positioned in front of a screen (PC or TV monitor) with headphones or only two loudspeakers in front of him. Furthermore, the presented application is real-time customizable on the basis of the particular listener anthropometry. The source sound can be a real audio signal captured at the remote side of a videoconference system by a microphone or a synthetic one. The loudspeaker signals are derived by the knowledge of the spatial coordinate of the speaker at the remote end of a videoconference session and a subsequent appropriate filtering of the audio streams with head related transfer functions (HRTFs). Virtual sources can be positioned in azimuth, in elevation and in distance, too. However, the reproduction area around the listener and the possible source space are restricted to a defined area. Nevertheless, the proposed algorithm implementation into a commercial videoconference system demonstrates an high quality of the audio rendering performance. A special Visual C++ development tool has been established to carry out system parameters optimization.

In the last decade, there has been an explosive growth in mobile computing technology and obviously an increase demand for platforms with multimedia application support. The desire to be connected “

any-time, any-where and any-way

” has led to an increasing array of heterogeneous systems, multimedia applications, devices, and service providers. In response to this issue, quality of service (QoS) is designed to hide low-level application variation and provide necessary service guarantees. In this contribution, motivation and operating guidelines for QoS assessment in real-time interactive multimedia communications are discussed and investigated. In particular, we consider the end-to-end service quality methodologies proposing a QoS evaluation technique by means of an unconventional use of fragile watermarking. Thus, by knowing the end-to-end QoS, it is possible to adopt optimal pricing strategies in terms of the QoS profiling for all the active users involved in the communication.

Recent work on hot-potato routing [

1

] has uncovered that large transit ASs can be sensitive to hot-potato disruptions. Designing a robust network is felt as overly important by transit providers as paths crossed by the traffic have both to be optimal and reliable. However, equipment failures and maintenance make this robustness non-trivial to achieve. To help understanding the robustness of large networks to internal failures, [

2

] proposed metrics aimed at capturing the sensitivity of ASs to internal failures. In this paper, we discuss the strengths and weaknesses of this approach to understand the robustness of the control plane of large networks, having carried this analysis on a large tier-1 ISP and smaller transit ASs. We argue that this sensitivity model is mainly useful for intradomain topology design, not for the design the whole routing plane of an AS. We claim that additional effort is required to understand the propagation of BGP routes inside large ASs. Complex iBGP structures, in particular route-reflection hierarchies [

3

], affect route diversity and optimality but it an unclear way.

CoMo (Continuous Monitoring) is a passive monitoring system. CoMo has been designed to be the basic building block of an open network monitoring infrastructure that would allow researchers and network operators to easily process and share network traffic statistics over multiple sites. This paper identifies the challenges that lie ahead in the deployment of such an open infrastructure. These main challenges are: (1) the system must allow any arbitrary traffic metric to be extracted from the packet streams, (2) it must provide privacy and security guarantees to the owner of the monitored link, the network users and the CoMo users, and (3) it must be robust to anomalous traffic patterns or traffic query loads. We describe the high-level architecture of CoMo and, in greater detail, the resource management, query processing and security aspects.

Passive measurements have recently received large attention from the scientific community as a mean, not only for traffic characterization, but also to infer critical protocol behaviors and network working conditions. In this paper we focus on passive measurements of TCP traffic, main component of nowadays traffic. In particular, we propose a heuristic technique for the classification of the anomalies that may occur during the lifetime of a connection. Since TCP is a closed-loop protocol that infers network conditions and reacts accordingly by means of losses, the possibility of carefully distinguishing the causes of anomalies in TCP traffic is very appealing and may be instrumental to the deep understanding of TCP behavior in real environments and the protocol engineering.

Traffic generation is essential in the processes of testing and developing new network elements, such as equipment, protocols and applications, regarding both the production and research area. Traditionally, two approaches have been followed for this purpose: the first is based on software applications that can be executed on inexpensive Personal Computers, while the second relies on dedicated hardware. Obviously, performance in the latter case (in terms of sustainable rates, precision in delays and jitters) outclasses that in the former, but also the costs usually grow of some order of magnitude. In this paper we describe a software IP traffic generator based on a hardware architecture specialized for packet processing (known as Network Processor), which we have developed and tested. Our approach is positioned between the two different philosophies listed above: it has a software (and then flexible) implementation running on a specific hardware only slightly more expensive than PCs.

Nowadays, networking equipment is realized by using decentralized architectures that often include special-purpose hardware elements. The latter considerably improve the performance on one hand, while on the other they limit the level of flexibility. Indeed, it is very difficult both to have access to details about internal operations and to perform any kind of interventions more complex than a configuration of parameters. Sometimes, the “experimental” nature of the Internet and its diffusion in many contexts suggest a different approach. This type of need is more evident inside the scientific community, which often encounters many difficulties in realizing experiments. Recent technological advances give a good chance to do something really effective in the field of open Internet equipment, also called Open Routers (ORs). The main target approached in this paper is to extend the evaluation of the OR forwarding performance proposed in [

1

], by analyzing the influence of the control plane functionalities.

Wireless sensor networks are often used for environmental monitoring applications. Sampling and reconstruction of a physical field is therefore one of the most important problems to solve. We focus on band-limited fields and investigate the relationship between the random topology of a sensor network and the quality of the reconstructed field. By reviewing irregular sampling theory, we derive some guidelines on how sensors should be deployed over a spatial area for efficient data acquisition and reconstruction. We analyze the problem using random matrix theory and show that even a very irregular spatial distribution of sensors may lead to a successful signal reconstruction, provided that the number of collected samples is large enough with respect to the field bandwidth.

In this paper we investigate the relationship between local next hop selection strategies and their efficiency in terms of both link related metrics, such as the mean packet delivery fraction, and network related metrics, such as the energy status of the node elected as relay. In standard geographical forwarding algorithms the relay selection is usually carried out by means of advancements toward the destination. However, channel attenuation phenomena often make pure geographical strategies ineffective as the quality of a transmission link is not necessarily deterministically related to the node coordinates. In order to achieve effective and cost efficient routing solutions, it is therefore crucial to couple advancements toward the destination with link quality aspects as well as network related metrics (e.g., node energies). This study is a preliminary step toward the design of local relay selection rules which jointly account for these aspects and whose aim is to cut the desired trade-off between delay and cost efficiency.

In this paper we describe a semi-probabilistic routing approach designed to enable content-based publish-subscribe on highly dynamic networks, e.g., mobile, peer-to-peer, or wireless sensor networks. We present the rationale and high level strategy of our approach, and then show its application in a link-based graph overlay as well as in a broadcast-based sensor network. Simulation results confirm that, in both scenarios, our semi-probabilistic approach strikes a balance between entirely deterministic and entirely probabilistic solutions, achieving high reliability with low overhead.

Sensor networks are composed of tiny, cheap, low power devices equipped with small batteries. Depending on the application scenarios, sensors collect some parameters and send them to the sink. This device processes and filters the information sending them to the remote network controller. In this paper we discuss the implementation and the results of the activity carried out and currently on work at the WiNe-Lab of the University of Catania. This activity consists in developing a sensor testbed, using MICAz motes, which can be remotely accessed via a web server by users. The latter can interact with the applications already running on the network and modify the operation mode of a single or all the sensor devices. Users can also run-time monitor the activity carried out by the sensors through a graphical interface and some plots of the sensed parameters. This activity aims either to allow sensor resources sharing among remote users or research groups that do not have the sensors available or to guarantee a high level of interaction between users and sensor devices accordingly this testbed introduces innovative features with respect to other existing and remotely accessible sensor network testbeds.

This paper presents a general approach to distributed detection with multiple sensors in network scenarios with noisy communication links between the sensors and the fusion center (or access point, AP). The sensors are independent and observe a common phenomenon. While in most of the literature the performance metrics usually considered are missed detection and false alarm probabilities, in this paper we follow a Bayesian approach for the evaluation of the probability of

decision error

at the AP. We first derive an optimized fusion rule at the AP in a scenario with ideal communication links. Then, we consider the presence of noisy links and model them as binary symmetric channels (BSCs). This assumption leads to a simple, yet meaningful, performance analysis. Under this assumption, we show, both analytically and through simulations, that if the noise intensity is above a critical level (i.e., the cross-over probability of the BSC is above a critical value), the lowest probability of decision error at the AP is obtained if the AP selectively discards the information transmitted by the sensors with noisy links.

We review several aspects of building real-time streaming data Grid applications. Building on general purpose messaging system software (NaradaBrokering) and generalized collaboration services (GlobalMMCS), we are developing a diverse set of interoperable capabilities. These include dynamic information systems for managing short-lived collaborative service collections (“gaggles”), stream filters to support the integration of Geographical Information Systems services with data analysis applications, streaming video to support collaborative geospatial maps with time-dependent data, and video stream playback and annotation services to enable scientific collaboration.

The GRIDCC [

1

] project will extend the use of Grid computing to include access to and control of distributed instrumentation. Access to the instruments will be via an interface to a Virtual Instrument Grid Service (VIGS). VIGS is a new concept and its design and implementation, together with middleware that can provide the appropriate quality of service, is a key part of the GRIDCC development plan. An overall architecture for GRIDCC has been defined and some of the application areas, which include distributed power systems, remote control of an accelerator and the remote monitoring of a large particle physics experiment, are briefly discussed.

The GRIDCC project will extend the use of Grid computing to include access to and control of distributed instrumentation. Access to the instruments will be via an interface to a Virtual Instrument Grid Service (VIGS). VIGS is a new concept and its design and implementation together with middleware that can provide the appropriate quality of service is a key part of the GRIDCC project. The paper describes the Multipurpose Collaborative Environment (MCE) which will provide the user interface to access Virtual Instrument Grid Services, as well as tools to support group-work. After an analysis of the status of the art in collaborative environments and technologies, taking into account both the Human Computer Interaction and the architectural and technological perspectives, we present the requirements of the MCE and more specific aspects including access and control of remote instruments.

The development and the deployment of wide-area Grid environments represent a new challenge for Next Generation Networks: not only is it necessary to implement network architectures supporting Quality of Service at IP level, but it is also essential to design and implement new network entities that monitor the network status, assure that network resources are allocated in an optimized way and interfaces Grid applications with network services, satisfying their SLSs. The paper presents the design, the development and the functional validation of a Grid Network Resource Broker (GNRB), a new architectural entity which handles the service requests coming from grid middleware (or from a single grid application) and, if necessary, allocates and reserves network resources.

In this paper we propose a p2p framework in order to organize the cooperation and the set-up phase of interconnected laboratories for didactical purposes through the Internet. Even if the application of standard p2p algorithms to distributed cooperative laboratories is a quite new methodology, we will base our approach on a solid foundation, such as the one provided by the Session Initiation Protocol (SIP). The proposed architecture solves some general problems such as the discovering of entities, and also provides, a general framework for devices accounting. In addition, our framework offers mechanisms in order to provide an abstract namespace for adding — removing facilities on the fly and invoking them.

This paper presents the rationale and the experience in exposing measurement instruments on the Grid. After explaining the design goals, we propose two implementation strategies targeted to access the instruments at different levels of abstraction. A performance evaluation of the realized services concludes the paper.

A telecommunication network is composed of a number of nodes. Each node can be either a terminal host or a web cache location. Objects are downloaded throughout the network among the nodes. An object can represent either a file or any other resource to be shared (e.g. machine time). For memory saving and for safety reasons, each object is composed of a number of portions and not all the portions are located within the same node, because no node should have the complete knowledge of each object. It means that a single node can have only a part of the file. It is strongly recommendable because if a single node should be accessed without authorization, the information retrieved would not be sufficient to detect the overall content.

Concerning networking viewpoint the following main issues will characterize the performance of object exchange: Position of the information, Strategy to reach the information, Algorithm to download information, Capacity Planning. The paper proposes a control architecture that considers the mentioned issues. It is composed of three layers: Local, Network and Planning Controller. The Local Controller acts locally to each node at object downloading time scale (seconds/minutes) and, after an “advanced flooding” signaling query to get informed about the object portions’ position, decides from which node (or nodes) each portion needs to be downloaded; the Network Controller may change the distribution of the object portions among the nodes; it acts with larger time scale (hours/day) and is centralized; the Planning Controller may change the dimension of each single portion and increase/decrease the physical link and node capacities. The order of magnitude of its intervention is weeks/months. The problem is modeled through a mathematical formulation and a specific cost function, which takes into account all the necessary details, is introduced for each controller as well as a minimization procedure.

A preliminary performance evaluation analyses the effect of the Local Controller.

This discussion paper present issues in the design of a networking measurement system. Key requisites are summarized at first. Then, architectures will be analyzed in order to let a designer aware of advantages and efforts of presented choices.

This paper deals with the traceability-assurance problems that are faced when a distributed measuring system is employed. These problems are highlighted for different distributed-system architecture, then a set of guide-lines are provided in order to correctly manage such systems from a metrological point of view. The proposed solution is based on a network-assisted calibration procedure, which requires suitable traveling standards that are sent to the nodes of a distributed system under calibration and are remotely controlled through the Internet. A prototype of a traveling standard is also described that has the capability to be synchronized with standards sent to the other nodes.

The paper presents a brief state of art of distributed laboratories for teaching of electric and electronic measurement and proposes an innovative platform for distance learning in such field. The platform consists of a learning management system (LMS) and a remotely accessible distributed laboratory to carry out experiments on real instrumentation without requiring specific software components on client side. In particular, the user authentication and management and the tracking of learning process are provided by the LMS. The experimental phase is carried out by means of Virtual Instruments executed on computers connected to the instrumentation included in a remote laboratory system distributed on geographical scale.

This paper addresses the concept of cooperative telemeasurements in which resources such as instruments and programmable devices are distributed in a network of different laboratories and can cooperate to set up augmented experiments. In particular, both the end user’s benefits and the technical implementation challenging issues in enabling cooperative distributed laboratories are discussed. An example of possible network architecture design guidelines able to enable cooperation and remote control of instrumentation is provided with particular regard to the access of a user to distributed resources. Finally, an implementation of a real cooperative telemeasurement platform at WiLAB, Bologna, Italy, is described.

The paper describes the Data Analysis Workstation developed by the IST project MOME (Monitoring and Measurement Cluster). The MOME project maintains the repository of information about publicly available measurement data, like e.g. packet traces, flow traces, routing data, and others. The collected information includes the detailed description of an assumed measurement scenario, as well as selected statistical parameters obtained by data analysis methods. The MOME repository aims at providing assistance to the research community in finding and retrieving raw measurement data, which is most appropriate from the point of view of different research objectives. In this paper we discuss the design and development of the MOME Data Analysis Workstation. It extends the MOME Database with capabilities for automatic statistical analysis of submitted measurement data. The included exemplary results of analysis performed on packet traces captured in a measurement site associated with MOME illustrate that the MOME Data base provides the researchers with useful information about available measurement data.

A networking laboratory contains expensive equipment and offers a small grade of scalability. Remote laboratories allow for much more efficient use of laboratory equipment, and give users the opportunity to batch-schedule their experiments and manage them from the comfort of their local clients. In this paper, LATERE, a new interactive laboratory which allows remote testing of complex satellite scenarios and wireless indoor networks on real equipment, is detailed. Users can select arbitrary network topologies; characteristics of each link and transport layers, used by the traffic generators, can be freely set. Moreover, LATERE offers a user-friendly web interface to define each experiment parameters and to asynchronously view all results, which are displayed and also emailed to the students; an experiments’ database is kept, allowing per-user profiles and history, plus logging and test reproducibility. LATERE is scalable, and built upon a lightweight middleware distributed and embedded on the interface with the equipment. LATERE is based on standard inexpensive hardware running open source software.

The availability of a wide range of instrumentation that can be remotely controlled, and the possibility to access the Internet with cheap and fast xDSL lines have encouraged the design of tele-measurement platforms by means of which researchers, instructors and students can exploit the facilities offered by complex and expensive laboratories. Although the market offers several solutions to remotely manage equipment, little attention has been paid to the hardware and software architectures devoted to distance learning experimental environments. The paper presents the architectural approach, and the related implementation, which was proposed and developed within two projects funded by the Italian Ministry of Education, University, and Research (MIUR). Both projects were aimed at building the experimental test-beds, which include very heterogeneous equipment, and at developing an open system architecture for their actual managing. Specifically, the attention is focused on the supervising central unit, which represents a crucial and critical element of the system. The software implementation of the central unit is described in detail, as concerns both the data structures adopted and the protocol machine. The results of some performance tests are shown to highlight the efficiency of the proposed solution.

Both educational and industrial institutions have a growing need to test experiments in highly complex, expensive structures. On the other hand, high-speed networking opens up the possibility of accessing remote sites and performing experiments with a high level of quality of service. The main goal of the LABNET (“ LABoratories on the NETwork” ) Project [1] was the realization of an Integrated Learning System that allows students and instructors, located in geographically dispersed areas, to access technological resources, such as sophisticated laboratory equipment, measurement devices and, in general, complex test systems, through a scalable networking infrastructure and a number of supporting multimedia technologies. The paper introduces a new perspective in the definition of a suitable interface to the instrumentation and proposes a web-service based instrument interface, as an evolution of the LABNET remote measurement framework.

In this paper, we describe our application framework (Multi-Site Navigator) for delivering context-aware services. The application framework is based on our middleware solution, named context Data Manager (XDM), using DHTs technologies as underlying substrate for integrating different contextual data generators or sensors.

In this paper we describe the Italian online learning project, called Teledoc2, financed by the Italian Ministry of Education, Universities and Research (MIUR) and realized by CNIT (National Inter-University Consortium for Telecommunications) [1]. The Project is running through 2003–2005 and aims to build a complete, multimedia, interactive and fully-featured online-learning service for ICT researchers and PhD students of Italian Research Centres, being branches of CNIT. The paper gives an overview of all the technical, didactical and organizational aspects implemented in order to achieve the successful activation of the Teledoc2 integrated services[2]. CNIT chose avant-garde and technologically updated solutions in all these aspects, in this way building an innovative and efficient didactic service and giving also a research tone to the Project. The main attention of the paper, in particular, is devoted to one of the key elements of the service: the technical aspects regarding the heterogeneous network architectures, designed, implemented and operated by CNIT to distribute multimedia information among its terminal sites, enabling the Teledoc2 learning service. A series of objective and subjective measurements are still in progress to evaluate the reliability of the different network solutions and the Quality of Service guaranteed by the different architectures for real-time transmission of audio-video and data contents. The pervasive use of avant-garde technological solutions, like multicast, satellite Ka band and NAT, further increases the efficiency of the widespread distribution of Teledoc2 learning service in all the Italian territory. Finally, the Project give special attention to the research and experimental activity, called “Network of Laboratories”, developed in particular at the WiLab (Wireless Communication Laboratories) of the University of Bologna.

This paper proposes the design of virtual teaching environments using the combination of VRML, Java and JavaScript languages for the construction of highly interactive and immersive 3D worlds, whose behaviour is real-time modified by users actions. A number of users involved in the animations are present in the classroom (their presence is captured by a camera or other kind of sensors) and for the rest they are remote internet students. In both cases they are represented within a VRML world as avatars into a virtual classroom designed to be available in an immersive way by everybody. They interact with each other through a Java program that is able to process and generate events determining the behaviour of scene elements. In particular, Java application establishes the communication among users and VRML environment and updates parameters stored in a MySQL database. The database, resident into the server, is necessary to share information among the clients and to guarantee real time changes in the virtual representation.

Recently, the need to provide location-based services has lead to an increasing interest in accurate location sensing techniques. For indoor environments, most of proposed solutions do not generally provide accurate location estimates due to propagation problems. In this paper, we present NeBULa, a novel and open framework to locate a mobile user within an indoor Wireless LAN environment. It uses signal-strength information extracted from the wireless LAN interface and doesn’t require additional and costly hardware. The performances of the proposed system have been evaluated in different operative conditions to show its effectiveness.

The Virtual Immersive Learning (VIL) test-bed intends to realize a virtual collaborative immersive environment, capable of integrating natural contexts and typical gestures, which may occur during traditional lectures, enhanced with advanced experimental sessions. The main aim of the paper is to describe the test-bed motivations, as well as the most significant strategies, both hardware and software, adopted for its implementation. The novelty of the authors’ approach essentially relies on its capability of remarking and emphasizing results that are the output of VICom project, and “putting the pieces together” in a well-integrated framework. These features, along with its high portability, good flexibility, and, above all, low cost, make the proposed test-bed appropriate for educational purposes, mainly concerning measurements on telecommunication systems and virtual restoration training courses at universities, as well as research centres.