Giorgio Ventre

A channel and rate assignment algorithm and a layer-2.5 forwarding paradigm for multi-radio wireless mesh networks

The availability of cost-effective wireless network interface cards makes it practical to design network devices with multiple radios which can be exploited to simultaneously transmit/receive over different frequency channels. It has been shown that using multiple radios per node increases the throughput of multi-hop wireless mesh networks. However, multi-radios create several research challenges. A fundamental problem is the joint channel assignment and routing problem, i.e., how the channels can be assigned to radios and how a set of flow rates can be determined for every network link in order to achieve an anticipated objective. This joint problem is NP-complete. Thus, an approximate solution is developed by solving the channel assignment and the routing problems separately. The channel assignment problem turns out to be the problem to assign channels such that a given set of flow rates are schedulable and itself is shown to be also NP-complete. This paper shows that not only the channels but also the transmission rates of the links have to be properly selected to make a given set of flow rates schedulable. Thus, a greedy heuristic for the channel and rate assignment problem is developed. Algorithms to schedule the resulting set of flow rates have been proposed in the literature, which require synchronization among nodes and hence modified coordination functions. Unlike previous work, in this paper a forwarding paradigm is developed to achieve the resulting set of flow rates while using a standard MAC. A bi-dimensional Markov chain model of the proposed forwarding paradigm is presented to analyze its behavior. Thorough performance studies are conducted to: a) compare the proposed greedy heuristic to other channel assignment algorithms; b) analyze the behavior of the forwarding paradigm through numerical simulations based on the Markov chain model; c) simulate the operations of the forwarding paradigm and evaluate the achieved network throughput.

Internet traffic modeling by means of Hidden Markov Models

In this work, we propose a Hidden Markov Model for Internet traffic sources at packet level, jointly analyzing Inter Packet Time and Packet Size. We give an analytical basis and the mathematical details regarding the model, and we test the flexibility of the proposed modeling approach with real traffic traces related to common Internet services with strong differences in terms of both applications/users and protocol behavior: SMTP, HTTP, a network game, and an instant messaging platform. The presented experimental analysis shows that, even maintaining a simple structure, the model is able to achieve good results in terms of estimation of statistical parameters and synthetic series generation, taking into account marginal distributions, mutual, and temporal dependencies. Moreover we show how, by exploiting such temporal dependencies, the model is able to perform short-term prediction by observing traffic from real sources.

Interactive Multimedia on Next Generation Networks

Content delivery network architectures are initiatives designed to support the effective delivery of continuous and discrete media to end-users. Mobile devices are now capable of exploiting services such as content delivery, but with the protocols governing the content delivery networks designed for wired networked topologies; an assessment of the impact of mobile devices on the network has never been undertaken. Wireless devices access causes significant issues in the ability of the media transport system to function efficiently or as intended. This paper will highlight the current content delivery architectures available, their level of support for mobile devices and the issues therein.

Searching for invariants in network games traffic

Even if Internet traffic analysis and characterization is a fertile research area, a lot of work still must be done to study and understand the traffic characteristics of new emerging multimedia applications. Among them, an interesting category is that of multiplayer network games. This paper aims at demonstrating that, at packet level, spatial and temporal invariants exist in the traffic of such applications. For this purpose, we study Counter-Strike, a popular client/server network game, comparing results from two different networks. The effectiveness of the proposed approach is evaluated by studying statistics of both packet size and inter-packet time. Results provide a view on packet-level game traffic and they confirm that the main traffic dynamics present properties that can be generalized, independently of the observation point and time.

Definition of the DISC concurrent language

Although commercial distributed computers are now available, developing software for these systems is still a hard task. This prevents the users from getting full advantage out of the use of distributed computers. Among the techiniques proposed to make it easier to program the new architectures, concurrent languages have proved to be an effective approach both in system programming and in application development.The concurrent programming language DISC has been designed to promote software engineering techniques in the development of distributed programs. The language extends C with CSP-based concurrent mechanisms and it is especially suited for system programming of loosely coupled distributed architectures. A simple programming environment, running under UNIX System V operating system, allows the user to develop programs in a modular fashion. A first implementation of DISC has been developed on a network of UNIX machines. DISC programs are portable on a wide class of distributed computers, because of both the choice of UNIX as the host environment and the substantial independence of the language run time support from the hardware.In this article we report the definition of DISC and an example to show informally the practical use of the language.

Bringing service differentiation to the end system

A number of distributed applications require communication services with quality of service (QoS) guarantees. For this purpose, resource management is required both in the end system and inside the network. To date several approaches have been proposed, dealing either with end system issues or with those more strictly related to the network, but no unified view exists. Furthermore, while network QoS provisioning has reached a good level of standardization, no standard proposals exist for the end systems. We argue that a single architectural model is needed, taking into account a more exhaustive concept of resource management, where quality of service is defined at the level of user perception. We propose an architectural model that enhances the IETF Diffserv framework in order to provide service differentiation even inside network end-points. We also present a prototype implementation of such architecture along with experimental results validating the proposed solution.

Using CSP languages to program parallel workstation systems

Abstract During the last decade one of the most relevant events in the computer market has been the large diffusion of workstations. In both industrial and research environments a huge amount of computing is done on personal workstations. Despite the rapid growth in networking technologies, however, a network of workstations cannot be easily seen as a global computational resource, although it represents a large amount of computing power. Moreover, its inherent parallelism is not accessible without a heavy effort to modify existing software and/or to develop new code. It is our belief that the CSP model is suitable to develop distributed applications for a particular class of such systems that can be defined Parallel Workstation Systems. This thesis has been tested in the course of the DISC project. In DISC, the language implementation of the CSP model tries to minimize the programming effort toward the development of parallel applications, and a friendly programming environment, integrated in the native workbench of workstations, is provided.

Open Source Platforms for Internet Monitoring and Measurement

Understanding the ever-changing scenario of computer networks and how they operate in the real world implies measuring and analyzing their characteristics. This in turn requires a set of advanced tools and methodologies to be shared among researches, along with the data derived from such activities. In this paper we first present some of the main issues and challenges in the field of Internet Monitoring and Measurement, then we present several open source platforms we have developed in the last 10 years for monitoring heterogeneous and large scale networks. Finally, we describe some of the data sets we made publicly available to the research community.

A real time-based architecture for qos multimedia provisioning

Abstract According to the ITU-T, the quality of a multimedia service is defined by a set of user-related parameters: delay, delay variation, and information loss. To provide multimedia applications with end-to-end QoS guarantees, an efficient resource reservation and management strategy has to be adopted. This paper presents a schema for satisfying multimedia QoS parameters over a real-time operating system, which adopts the rate monotonic as a scheduling algorithm. Such a schema is implemented in a real-time based architecture for QoS multimedia provisioning. This architecture allows to define classes of services with different quality attributes concerning the multimedia data delivery.

A Service Differentiation Scheme for the End-System

A number of research studies show that the operating system has a substantial influence on communication delay in distributed environments. Thus, in order to provide applications with end-to-end QoS guarantees, network resource management alone is not sufficient. To date, several approaches have been proposed, addressing QoS issues from the end-system point of view. However, while network QoS provisioning has achieved a good level of standardization, no standard proposals exist for the end-systems. We claim that a single architectural model, taking into account both end-system and network resource management is needed. In this paper we propose a QoS architecture, which extends the concept of service differentiation inside the end-system. A system prototype has been developed and tested on a Diffserv network scenario. The prototype incorporates a priority-based communication mechanism inside the end-point operating system and a local marker so to implement an appropriate mapping between local classes of services and network QoS levels. Experimental results have also been provided in order to evaluate the impact of the proposed architecture on Diffserv routers.