Showing papers by "Fabrizio Granelli published in 2010"

Standardization and research in cognitive and dynamic spectrum access networks: IEEE SCC41 efforts and other activities

Abstract: Spectrum crowding, spectrum management, quality of service, and user support are the topics of vigorous research in the cognitive and dynamic spectrum access network communities. As research matures, standardization provides a bridge between research results, implementation, and widespread deployment of such networks. This article reports recent developments within the IEEE Standardization Coordinating Committee 41, "Dynamic Spectrum Access Networks." It outlines possible future standardization topics for IEEE SCC41, in the framework of other related standardization activities, and discusses open research issues that present future challenges for the standardization community.

Guest editors' introduction: Special issue on modeling and simulation of cross-layer interactions in communication networks

Abstract: Cross-layering represents an important emerging design approach to overcome limitations and lack of flexibility of the layering principle, which is the basis of today’s communication networks. Indeed, in specific scenarios, such as Qualityof-Service provisioning and wireless networks, deeper integration and interaction among different layers is envisioned. Despite the widespread adoption of the general cross-layer design principle, not much attention has been placed on solid, quantitative, systematic study of the system behavior and interactions between different protocols at different layers. In fact, even though several cross-layering architectures are available in the literature, very few works address the problem of providing frameworks for modeling cross-layering interactions and support the proper definition of cross-layering solutions. Quantitative cross-layer formulations and models are still needed to provide effective guidelines on the impact of the design parameters in terms of overall system performance. Simulation-based methods represent indispensable tools to provide further insight in this direction given the analytical intractability of the complex cross-layer system. This special issue of ACM Transactions on Modeling and Computer Simulation (TOMACS) focuses on relevant novel contributions in the development of models for cross-layering interactions and computational solutions. We received several worthy articles that we considered for inclusion in this issue. In addition to the ones appearing in the special issue, certain articles required more extensive revision than could be completed in time for publication of the special issue. Therefore some of these articles will appear in future regular issues of TOMACS. The first article is titled “Steepest-Ascent Constrained Simultaneous Perturbation for Multi-Objective Optimization” and is authored by Daniel W. McClary, Violet R. Syrotiuk, and Murat Kulahci. The authors consider the problem of optimizing multiple responses in which the gradient is known for only one response. They propose a hybrid approach for this problem, called simultaneous perturbation stochastic approximation steepest ascent (SPSA-SA), which is abbreviated as SP(SA)2. SP(SA)2 is an SPSA technique that leverages information about the known gradient to constrain the perturbations used to approximate the others. The authors apply SP(SA)2 to the cross-layer optimization of throughput, packet loss, and end-to-end delay in a mobile adhoc network (MANET), a self-organizing wireless network. The results show that SP(SA)2 achieves higher throughput, lower packet loss, and lower endto-end delay than the steepest ascent, SPSA, and the Nelder-Mead stochastic

Architectures and Cross-Layer Design for Cognitive Networks

Abstract: Network evolution towards self-aware autonomous adaptive networking attempts to overcome the inefficiency of configuring and managing networks, which leads to performancedegradation. In order to optimize network operations, the introduction of self-awareness, self-management, and self-healing into the network was proposed. This created a new paradigm in networking, known as cognitive networking. This chapter describes state-of-the-art, as well as future directions in cognitive networking. Fundamental techniques for enabling cognitive properties, such as, adaptation, learning, and goal optimization processes are detailed in this text. A comparison of available research proposals leads to the design of a promising cognitive network architecture capable of incorporating cognitive network techniques. Finally, a discussion on the required properties of the cross-layer design for cognitive networks and deployment issues are specified.

WING/WORLD: an open experimental toolkit for the design and deployment of IEEE 802.11-based wireless mesh networks testbeds

Abstract: Wireless Mesh Networks represent an interesting instance of light-infrastructure wireless networks. Due to their flexibility and resiliency to network failures, wireless mesh networks are particularly suitable for incremental and rapid deployments of wireless access networks in both metropolitan and rural areas. This paper illustrates the design and development of an open toolkit aimed at supporting the design of different solutions for wireless mesh networking by enabling real evaluation, validation, and demonstration. The resulting testbed is based on off-the-shelf hardware components and open-source software and is focused on IEEE 802.11 commodity devices. The software toolkit is based on an "open" philosophy and aims at providing the scientific community with a tool for effective and reproducible performance analysis of WMNs. The paper describes the architecture of the toolkit, and its core functionalities, as well as its potential evolutions.

CogProt: A Framework for Cognitive Configuration and Optimization of Communication Protocols

Abstract: Advancements in network technologies dramatically increased management complexity. Cognitive networking was introduced to deal with this problem, by providing algorithms for autonomous network management and protocol reconfiguration. In this paper, we propose a framework for cognitive configuration and optimization of communication protocols called CogProt. CogProt is a distributed framework that allows dynamic reconfiguration of operational protocol stack parameters for optimizing protocol performance under changing network conditions. As a proof of concept, the framework is illustrated for the cognitive configuration of TCP congestion window evolution. In this setup, the TCP window increase factor is adjusted in runtime based on the TCP goodput experienced in the immediate past. Simulation results demonstrate that the proposed cognitive framework is able to improve average TCP performance under changing network conditions.

TCP-Aware Forward Error Correction for Wireless Networks

Abstract: This paper studies TCP performance improvement in wireless and heterogeneous networks using Forward Error Correction (FEC) technique driven by TCP semantics. In the proposed scheme, called TCP-aware FEC, the amount of redundancy added to a packet at the sender node corresponds to the level of error protection and is computed as a function of TCP congestion window. The TCP-aware FEC becomes stronger for low congestion window values (experienced after a packet loss is detected), while the amount of added redundancy is reduced for the large window values approaching the capacity of the end-to-end link.

Identifying Relevant Cross-Layer Interactions in Cognitive Processes

Abstract: Cognitive networks were recently proposed to cope with the complexity and the dynamics of network managements, exploiting reasoning to adapt the behavior of protocols. Among the reasoning formalisms that can be employed, Fuzzy Cognitive Maps seem to be very promising, as they potentially allow the cognitive process to consider cross-layer interactions in the characterization of the performance of a network node. However, when considering a high number of cross-layer interactions, reasoning schemes can be too time consuming and may not provide a suitable solution as environmental conditions change. In order to decrease the demand of reasoning time it is of utmost importance to discover which cross-layer relationships carry relevant information to the cognitive process. This paper discusses how to make such differentiation. Moreover, it proposes a metric to evaluate the influence a cross-layer interaction has on the cognitive process.

A Cognitive Approach for Cross-Layer Performance Management

Abstract: The evolution of network technologies brought increasing management complexity of networking infrastructure and protocols. Cognitive networking was introduced to deal with such complexity. This work presents a cognitive algorithm for cross-layer performance management which is the core of a decentralized framework for self-configuration of communication protocols. We illustrate the proposed solution for the joint reconfiguration of protocol parameters at different layers. The cognitive joint adaptation of TCP congestion window and MAC layer data rate is carried out as a proof of concept. Simulation results show performance improvement given by the proposed approach under changing network conditions.

QoS-driven radio resource allocation for OFDMA networks based on a game theoretical approach

Abstract: In this paper, we propose a cooperative strategy for OFDMA radio resource allocation based on game theory The main novelty with respect to state-of-the-art is that the utility function considers the application-oriented Mean Opinion Score (MOS) rather than the gross data rate attributed to each user flow In such a scenario, data flows compete in cooperative way in order to maximize the perceived Quality-of-Service (QoS) Experimental results show that the MOS achievable by the proposed resource allocation strategy is better than that one provided both by uncoordinated strategies based on water-filling and by cooperative strategies based on pure data rate maximization

Characterizing User Behavior and Network Load on a Large-Scale Wireless Mesh Network

Abstract: Wireless mesh networks represent a promising paradigm to provide a scalable infrastructure for Internet access in metropolitan areas. In this paper, a large-scale wireless mesh testbed deployed in three cities in the Trentino region is described and experimentation results obtained from the public use of the testbed are reported and analyzed. The large-scale of the deployment and high number of users ensure to have proper traces which can capture the trends in user traffic based on the applications used and realistic mobility patterns.

Service-aware retransmission control in cellular networks

Abstract: This paper proposes a service-aware cross-layer approach between application/transport layers on the mobile terminal and link layer on the wireless base station to enable dynamic control on the level of per-packet error protection for multimedia data streams. Specifically, in the context of cellular networks, the proposed scheme enables the mobile terminal to specify to the base station the desired level of Hybrid ARQ (HARQ) protection by using an in-band control feedback channel. Such protection is dynamically adapted on a per-packet basis and depends on the perceptual importance of different packets as well as on the reception history of the flow. Experimental results demonstrate the potential benefits deriving from the proposed strategy either for audio and video real-time streams as well as for TCP-based data transfers.

Lightweight mobile and wireless systems: technologies, architectures, and services

Abstract: 1Department of Information and Communication Systems Engineering (ICSE), University of the Aegean, 81100 Mytilene, Greece 2Department of Information Engineering and Computer Science (DISI), University of Trento, 38123 Trento, Italy 3Department of Informatics, Alexander Technological Educational Institute of Thessaloniki, Thessaloniki, 574 00 Macedonia, Greece 4Centre Tecnologic de Telecomunicacions de Catalunya (CTTC), 08860 Barcelona, Spain 5North Carolina State University (NCSU), Raleigh, NC 27695, USA

Emulating Multi–hop Wireless Networks over a Planetary Scale Testbed

Abstract: The most commonly used technique to evaluate novel solutions is to leverage on simulation studies which are largely based on a simplified model of the system behavior. Such an approach provides an approximate evaluation of the system’s performances, which can be potentially far away from the behavior on a real deployment. Another solution is to exploit real world prototypes and testbeds. Such testbeds are generally based on a limited number of nodes and, due to their experimental purposes, they only partially present the challenges of a real operational environment with hundred or thousand users. Furthermore, the limited number of nodes limits the possibility to study scalability issues.

TCP Kikolo for performance improvement in highspeed networks in presence of packet losses

Abstract: TCP shows poor performance in high speed networks with large latencies as well as in wireless networks where link error rates are high. The main two factors degrading TCP performance are related to its conservative policies for sending rate increase and treating all packet losses as congestion-related. In order to address these limitations a number of solutions have been proposed. Most of them either adapt to the large bandwidth-delay product when the loss rate is low (below 10−7) or face the problem of high error rate (above 10−3) not accounting for high-speed connections. In this paper we propose a novel approach Kikolo TCP. It is based on Markov Modulated Poisson Process (MMPP) to control the congestion window evolution. Kikolo TCP offers improved performance in environments with high error rates. At the same time it ensures bandwidth scalability in low loss environments, TCP friendliness and fairness.