Kostas Pentikousis

Energy-Efficient Cloud Computing

Energy efficiency is increasingly important for future information and communication technologies (ICT), because the increased usage of ICT, together with increasing energy costs and the need to reduce green house gas emissions call for energy-efficient technologies that decrease the overall energy consumption of computation, storage and communications. Cloud computing has recently received considerable attention, as a promising approach for delivering ICT services by improving the utilization of data centre resources. In principle, cloud computing can be an inherently energy-efficient technology for ICT provided that its potential for significant energy savings that have so far focused on hardware aspects, can be fully explored with respect to system operation and networking aspects. Thus this paper, in the context of cloud computing, reviews the usage of methods and technologies currently used for energy-efficient operation of computer hardware and network infrastructure. After surveying some of the current best practice and relevant literature in this area, this paper identifies some of the remaining key research challenges that arise when such energy-saving techniques are extended for use in cloud computing environments.

Mobileflow: Toward software-defined mobile networks

Mobile carrier networks follow an architecture where network elements and their interfaces are defined in detail through standardization, but provide limited ways to develop new network features once deployed. In recent years we have witnessed rapid growth in over-the-top mobile applications and a 10-fold increase in subscriber traffic while ground-breaking network innovation took a back seat. We argue that carrier networks can benefit from advances in computer science and pertinent technology trends by incorporating a new way of thinking in their current toolbox. This article introduces a blueprint for implementing current as well as future network architectures based on a software-defined networking approach. Our architecture enables operators to capitalize on a flow-based forwarding model and fosters a rich environment for innovation inside the mobile network. In this article, we validate this concept in our wireless network research laboratory, demonstrate the programmability and flexibility of the architecture, and provide implementation and experimentation details.

Design considerations for a network of information

The existing Internet ecosystem is a result of decades of evolution. It has managed to scale well beyond the original aspirations. Evolution, though, highlighted a certain degree of inadequacies that is well documented. In this position paper we present the design considerations for a re-architected global networking architecture which delivers dissemination and non-dissemination objects only to consenting recipients, reducing unwanted traffic, linking information producers with consumers independently of the hosts involved, and connects the digital with the physical world. We consider issues ranging from the proposed object identifier/locator split to security and trust as we transition towards a Network of Information and relate our work with the emerging paradigm of publish/subscribe architectures. We introduce the fundamental components of a Network of Information, i.e., name resolution, routing, storage, and search, and close this paper with a discussion about future work.

In search of energy-efficient mobile networking

With the proliferation of mobile fixed-power devices, energy consumption emerged as a vibrant research and development subject area in networking. Mobile devices are designed with several hard constraints such as low cost and small geometries, as well as, low heat dissipation, and operation using fixed power sources. Manufacturers have been adding an ever increasing set of features to small mobile devices, which are no longer binary-use gadgets, but fully-fledged computers. With respect to power management, several mechanisms have been introduced; but, by and large, gains in power consumption at the hardware level have been essentially traded for extended functionality. All in all, the overall operational time has not increased. For example, early GSM cellular phones could only allow for less than an hour of talk time in a single battery charge. By the late 1990s, top models, introduced through better engineering and an evolutionary development approach, featured talk times increased by a factor of 3-5. This level of performance has remained the same over the last decade, although it is well below user expectations. This article reviews the evolution from simple cell phones toward the feature-rich mobile networked devices we have come to expect from manufacturers, and explains the factors that have led to stagnation in operational time. We then turn our attention to the multiaccess nature of modern mobile devices and the respective implications for power management. We find that the current host-centric mobile networking paradigm, based on end-to-end always on connectivity, leads to energy-inefficient operation. Finally, this article introduces information-centric networking and outlines open research issues in the design of energy-efficient future Internet architectures.

TCP in wired-cum-wireless environments

The Internet has evolved during the last decade, reaching a larger number of users and encompassing several new technologies. New classes of hosts such as mobile devices are gaining popularity, while the transmission media become more heterogeneous. Wireless networks exhibit different characteristics than wired ones. Mobile hosts have different needs and limitations than desktop computers. TCP has served well the wired Internet for almost 20 years, but is not ready for wired-cum-wireless environments. This article presents the challenges that must be met in order to provide reliable transport services to all hosts regardless of the type of network connectivity used. It surveys recently proposed solutions and evaluates them with respect to a wired-cum-wireless environment.

Trigger Management Mechanisms

Current protocol stacks can handle only a handful of events that may trigger actions for mobility management, such as signal strength indicators and cell load. We argue for new mechanisms that can deal with a greater variety of triggering events, which may originate from any component of the nodes protocol stack as well as mobility management entities within the network. We present the main concepts that govern our trigger management mechanisms (TMM), which aim at operating in a richer mobility management framework, and enable the deployment of new applications and services. After summarizing the architectural requirements with respect to trigger collection, processing, storage, and dissemination, we introduce a real implementation of TMM on commodity mobile devices. We briefly review our testbed environment and present experimental results using TMM to drive a lossless streaming video session handover between a laptop and a PDA. We position the current TMM design and implementation within the ambient networks architecture, centering in particular on the use of policies to steer TMM operation, and the role that TMM can play in an information service infrastructure. Finally, we outline current and future work items

Theory and Practice of RRC State Transitions in UMTS Networks

GPP has been enhancing UMTS networks by issu- ing new standard releases recurrently. However, the RRC state transition model has remained rather unchanged through sev- eral releases, namely, Rel. 99, Rel. 05 and Rel. 06. We review RRC state transitions and study them in practice using our nov- el 3G Transition Triggering Tool (3G3T), focusing on the net- work configuration parameters that prompt these transitions. We employ 3G3T in a landmark measurement study involving four UMTS networks in three countries and validate our find- ings using a state-of-the art proprietary measurement tool. Our results show that 3G3T is able to discover RRC configuration parameters without operator involvement or cooperation. We observe significant differences in UMTS network configura- tions, which directly impact end-user data service performance. We find that, in practice, the behavior of public UMTS net- works cannot be solely described based on the theoretical con- structs found in the literature, and that tools such as 3G3T are necessary in order to obtain a complete picture. The results pre- sented in this paper aim to assist simulationists in developing better models for UMTS networks. Moreover, we expect that operators can use 3G3T to configure their networks more effi- ciently. Finally, our methods and algorithms should be of great interest to application developers for mobile broadband net- works based on 3GPP standards.

Active goodput measurements from a public 3G/UMTS network

Although third generation cellular networks are being deployed in many countries, rigorous measurement studies from public networks have yet to be published. After extensive experimentation, we collect measurements from a commercial WCDMA 3G/UMTS network and observe that the goodput of the first of a series of back-to-back transfers is consistently below par. We compare these results with those obtained from an all-wired network and show that 3G signaling plays an important, albeit detrimental role in end-user goodput. Finally, we illustrate that near-nominal goodput is realizable for large payloads only.

Scenarios and Research Issues for a Network of Information

This paper describes ideas and items of work within the framework of the EU-funded 4WARD project. We present scenarios where the current host-centric approach to infor- mation storage and retrieval is ill-suited for and explain how a new networking paradigm emerges, by adopting the information-centric network architecture approach, which we call Network of Information (NetInf). NetInf capital- izes on a proposed identifier/locator split and allows users to create, distribute, and retrieve information using a com- mon infrastructure without tying data to particular hosts. NetInf introduces the concepts of information and data ob- jects. Data objects correspond to the particular bits and bytes of a digital object, such as text file, a specific encod- ing of a song or a video. Information objects can be used to identify other objects irrespective of their particular dig- ital representation. After discussing the benefits of such an indirection, we consider the impact of NetInf with respect to naming and governance in the Future Internet. Finally, we provide an outlook on the research scope of NetInf along with items for future work.

An experimental investigation of VoIP and video streaming over fixed WiMAX

Despite the significant interest in WiMAX technology and deployment, there are very few publicly reported measurements from testbeds and field trials. As such, most WiMAX studies employ simulation and modeling. This paper contributes to our understanding of what is realistically possible using off-the-shelf fixed WiMAX equipment today. We employ multiple competing traffic sources over a point-to-multipoint WiMAX topology and measure the capacity of the WiMAX equipment to handle a multitude of VoIP flows between subscriber stations while delivering a variable number of video streams. We measure throughput, packet loss, and one-way delay for both line-of-sight (LOS) and non-line-of-sight (NLOS) conditions. For the one-way delay measurements we synchronize the clocks of all testbed hosts with a software-only, open source implementation of the IEEE 1588 Precision Time Protocol. We compare these one-way delay measurements with those obtained when GPS-based synchronization is used.