Publications from UMOBILE activities
Mobile edge-cloud (MEC) aims to support low latency mobile services by bringing remote cloud services nearer to mobile users. However, in order to deal with dynamic workloads, MEC is deployed in a large number of fixed-location microclouds, leading to resource wastage during stable/low workload periods. Limiting the number of micro-clouds improves resource utilization and saves operational costs, but faces service performance degradations due to insufficient physical capacity during peak time from nearby micro-clouds. To efficiently support services with low latency requirement under varying workload conditions, we adopt the emerging Network Function Virtualization (NFV)-enabled MEC, which offers new flexibility in hosting MEC services in any virtualized network node, e.g., access points, routers, etc. This flexibility overcomes the limitations imposed by fixed-location solutions, providing new freedom in terms of MEC service-hosting locations. In this paper, we address the questions on where and when to allocate resources as well as how many resources to be allocated among NFVenabled MECs, such that both the low latency requirements of mobile services and MEC cost efficiency are achieved. We propose a dynamic resource allocation framework that consists of a fast heuristic-based incremental allocation mechanism that dynamically performs resource allocation and a reoptimization algorithm that periodically adjusts allocation to maintain a nearoptimal MEC operational cost over time. We show through extensive simulations that our flexible framework always manages to allocate sufficient resources in time to guarantee continuous satisfaction of applications’ low latency requirements. At the same time, our proposal saves up to 33% of cost in comparison to existing fixed-location MEC solutions.
Edge computing has emerged as a new paradigm to bring cloud applications closer to users for increased performance. ISPs have the opportunity to deploy private edge-clouds in their infrastructure to generate additional revenue by providing ultra-low latency applications to local users. We envision a rapid increase in the number of such applications for “edge” networks in the near future with virtual/augmented reality (VR/AR), networked gaming, wearable cognitive assistance, autonomous driving and IoT analytics having already been proposed for edgeclouds instead of the central clouds to improve performance. This raises new challenges as the complexity of the resource allocation problem for multiple services with latency deadlines (i.e., which service to place at which node of the edge-cloud in order to satisfy the latency constraints) becomes significant. In this paper, we propose a set of practical, uncoordinated strategies for service placement in edge-clouds. Through extensive simulations using both synthetic and real-world trace data, we demonstrate that uncoordinated strategies can perform comparatively well with the optimal placement solution, which satisfies the maximum amount of user requests.
Emerging mobile multimedia applications, such as augmented reality, have stringent latency requirements and high computational cost. To address this, mobile edge-cloud (MEC) has been proposed as an approach to bring resources closer to users. Recently, in contrast to conventional fixed cloud locations, the advent of network function virtualization (NFV) has, with some added cost due to the necessary decentralization, enhanced MEC with new flexibility in placing MEC services to any nodes capable of virtualizing their resources. In this work, we address the question on how to optimally place resources among NFVenabled nodes to support mobile multimedia applications with low latency requirement and when to adapt the current resource placements to address workload changes. We first show that the placement optimization problem is NP-hard and propose an online dynamic resource allocation scheme that consists of an adaptive greedy heuristic algorithm and a detection mechanism to identify the time when the system will no longer be able to satisfy the applications’ delay requirement. Our scheme takes into account the effect of current existing techniques (i.e., autoscaling and load balancing). We design and implement a realistic NFV-enabled MEC simulated framework and show through extensive simulations that our proposal always manages to allocate sufficient resources on time to guarantee continuous satisfaction of the application latency requirements under changing workload while incurring up to 40% less cost in comparison to existing overprovisioning approaches.
In this demo we present a NDN-based approach to deploy dockerised services closer to end-users when the network is impaired. We further increase resiliency, employing DTN to tunnel traffic between intermittently connected NDN nodes
This technical report describes the architecture of Oi! , a novel short messaging Android application for the Named-Data Networking (NDN) framework. Oi! can be used on top of NDN Android, and also on top of NDN-OPP [1,4,5], an extension of the Named-Data Networking (NDN) framework for opportunistic networking scenarios. Oi! code is open and is available as source code on https://github.com/COPELABS-SITI/ and its APK can be downloaded on Google play 1 . The available Oi! APK works on top of NDN-OPP, allowing Oi! to be used even in the presence of intermittent wireless connectivity.
Opportunistic networking is rising due to the capability of end-to-end devices to exchange data directly, based on physical proximity as well as based on encounters of the persons carrying such devices around. The development of opportunistic solutions is often based on short-range wireless technology such as Bluetooth, and most of the times is validated via simulators or emulators. Today, devices are enabled not only with Bluetooth, but also with Wi-Fi Direct. Opportunistic networking solutions can therefore take into consideration these aspects (proximity among users and their encounter), to provide support to intermittent internet access as well as to assist traffic locality. This abstract describes the NDN framework for Opportunistic Networks (NDN-Opp), which is being developed aiming to support opportunistic forwarding based on users' interests and their dynamic social behavior.
Sharing content has become part of our lives; Twitter for instance, is one of the most popular application in this area with millions of users in the entire world. At the same time, in the recent years, Named-Data Networking has become a promising network infrastructure, with continuous growth and collaborating teams that are working on it. In this paper we describe Now@, aiming to increase the impact of NDN near the end user with an Android application that allows them to exchange data based on their interests. To achieve this goal, we have developed Now@ based on synchronization of data. Now@ can operate on top of NFD Android allowing data exchange via wireless Internet and on top of NDN-Opp allowing data to be exchanged even in the presence of intermittent connectivity.
This document describes the demo of our NDN-Opp framework which brings Named-Data Networking to Opportunistic Networks. Our implementation attempts to leverage all communication opportunities, supports intermittently connected device-to-device communication links and push models. We are also experimenting with acknowledgement mechanisms and connection-less transfer of packets.
This technical report describes the architecture of NDN-OPP, an extension of the Named-Data Networking (NDN) framework for opportunistic networking scenarios. NDN-OPP was developed based on the NDN Android implementation. NDN-OPP code is open and is available as source code on https://github.com/COPELABS-SITI/ and will be released as APK on Google play. The research leading to these results has received funding from the European Union (EU) Horizon 2020 research and innovation programmer under grant agreement No 645124 (Action full title: Universal, mobile-centric and opportunistic communications architecture, Action Acronym: UMOBILE). This paper reflects only the authors’ views and the Community is not liable for any use that may be made of the information contained therein.
Recent research has considered various approaches for discovering content in the cache-enabled nodes of an Autonomous System (AS) to reduce the costly inter-AS traffic. Such approaches include i) searching content opportunistically (on-path) along the default intra-AS path towards the content origin for limited gain, and ii) actively coordinate nodes when caching content for significantly higher gains, but also higher overhead. In this paper, we try to combine the merits of both worlds by using traditional opportunistic caching mechanisms enhanced with a lightweight content discovery approach. Particularly, a content retrieved through an inter-AS link is cached only once along the intra-AS delivery path to maximize network storage utilization, and ephemeral forwarding state to locate temporarily stored content is established opportunistically at each node along that path during the processing of Data packets. The ephemeral forwarding state either points to the arriving or the destination face of the Data packet depending on whether the content has already been cached along the path or not. The challenge in such an approach is to appropriately use and maintain the ephemeral forwarding state to minimize inter-AS content retrieval, while keeping retrieval latency and overhead at acceptable levels. We propose several forwarding strategies to use and manage ephemeral state and evaluate our mechanism using an ISP topology for various system parameters. Our results indicate that our opportunistic content discovery mechanism can achieve near-optimal performance and significantly reduce inter-AS traffic.
Information-Centric Networking (ICN) has been proposed as a promising solution for the Internet of Things (IoT), due to its focus on naming data, rather than endpoints, which can greatly simplify applications. The hierarchical naming of the Named-Data Networking (NDN) architecture can be used to name groups of data values, for example, all temperature sensors in a building. However, the use of a single naming hierarchy for all kinds of different applications is inflexible. Moreover, IoT data are typically retrieved from multiple sources at the same time, allowing applications to aggregate similar information items, something not natively supported by NDN. To this end, in this paper we propose (a) locating IoT data using (unordered) keywords combined with NDN names and (b) processing multiple such items at the edge of the network with arbitrary functions. We describe and evaluate three different strategies for retrieving data and placing the calculations in the edge IoT network, thus combining connectivity, storage and computing.
The vast majority of mobile data transfers today follow the traditional client-server model. Although in the fixed network P2P approaches have been exploited and shown to be very efficient, in the mobile domain there has been limited attempt to leverage on P2P (D2D) for large-scale content distribution (i.e., not DTN-like, point-to-point message transfers). In this paper, we explore the potential of a user-operated, smartphonecentric content distribution model for smartphone applications. In particular, we assume source nodes that are updated directly from the content provider (e.g., BBC, CNN), whenever updates are available; destination nodes are then directly updated by source nodes in a D2D manner. We leverage on sophisticated information-aware and application-centric connectivity techniques to distribute content between mobile devices in densely-populated urban environments. Our target is to investigate the feasibility of an opportunistic content distribution network in an attempt to achieve widespread distribution of heavy content (e.g., video files) to the majority of the destination nodes. We propose ubiCDN as a ubiquitous, user-operated and distributed CDN for mobile applications.
In the context of social well-being and context awareness several eHealth applications have been focused on tracking activities, such as sleep or specific fitness habits, with the purpose of promoting physical well-being with increasing success. Sensing technology can, however, be applied to improve social well-being, in addition to physical well-being. This paper addresses NSense, a tool that has been developed to capture and to infer social interaction patterns aiming to assist in the promotion of social well-being. Experiments carried out under realistic settings validate the NSense performance in terms of its capability to infer social interaction context based on our proposed computational utility functions. Traces obtained during the experiments are available via the CRAWDAD international trace repository.
The Future Internet Architecture for Developing region (FI4D) is about bringing together academic participants working on innovative solutions for bridging the gap in Internet coverage for everyone (especially in remote regions) using cutting-edge technologies such as ICN, SDN, NFV within modern motivational frameworks such as 5G networks. FI4D is an initiative originally covering the need for synergy between RIFE (http://rife-project.eu) and UMOBILE (http://www.umobile-project.eu); these two projects have the common objective of fostering Internet connectivity in developing regions. In RIFE we are developing a solution for bringing Internet connection to rural and remote areas at a lowcost, whereas in UMOBILE we are developing a solution to bring the Internet in disaster scenarios; both of them using an Information Centric Networking approach.
Opportunistic networks can increase network capacity, support collaborative downloading of content and oﬄoad traﬃc from a cellular to a cellular-assisted, device-to-device network.
They can also support communication and content exchange when the cellular infrastructure is under severe stress and when the network is down or inaccessible. Fountain coding has been considered as especially suitable for lossy networks, providing reliable multicast transport without requiring feedback from receivers.
It is also ideal for multi-path and multi-source communication that ﬁts exceptionally well with opportunistic networks.
In this paper, we propose a content-centric approach for disseminating content in opportunistic networks eﬃciently and reliably.
Our approach is based on Information-Centric Networking (ICN) and employs fountain coding. When tied together, ICN and fountain coding provide a comprehensive solution that overcomes signiﬁcant limitations of existing approaches. Extensive network simulations indicate that our approach is viable.
Recent research in Information-Centric Networks has considered various approaches for discovering content in the cache-enabled nodes of the network. Such approaches include scoped flooding and deploying a control plane protocol to disseminate the cache contents in the network, to name a few. In this work, we consider an opportunistic approach that uses trails left behind by data packets from the content origin to the sources in order to discover off-path cached content. We evaluate our approach using an ISP topology for various system parameters. We propose two new forwarding strategies built on top of our approach. Our results indicate that the opportunistic discovery mechanism can significantly increase cache hit rate compared to NDN’s default forwarding strategy, while limiting the overhead at acceptable levels.
Mobile devices are increasingly presented with multiple connectivity options, including WiFi hotspots, micro-/macrocells or even other devices in device-to-device communications. In this environment, connectivity management has been focused on contention, congestion and wireless medium conditions. In this paper, we assess the role of informationcentrism in mobile device connectivity management. Motivated by the increasing availability of content and services in in-network caches and micro-data centers, we design an access network selection scheme that takes into account information availability within each connectivity option. Using extensive simulations, we show that information-awareness results in a significant increase of experienced cache hit ratios, i.e., up to 115% in certain scenarios.
The current momentum gained by the Internet of Things (IoT) is encouraging the scientific community and industry to evolve the current silo-based IoT platform models towards a globally unified IoT platform—a platform capable of efficiently supporting 50-100 billions networked objects. To meet the challenge, several proposals have emerged, such as building a unified host-centric IoT platform overlaid on top of today’s host-centric Internet. Alternatively, the networking community suggests the use of Information Centric Networks (ICNs), such as NDN (Named Data Networking), to face the challenge. NDN uses name–based routing to deliver data packets in a simple stateful forwarding plane. It offers developers great flexibility in naming, security, caching and inherent multicast support. It avoids dependencies on separate protocols and various middleware used in IoT networks. In this paper, we use NDN to provide an initial evaluation of a home smart lighting system and demonstrate some of the advantages of using the ICN paradigm to meet the IoT challenge. We developed a prototype and benchmarked it against the IP cloud based approach.
The advent and wide adoption of smartphones in the second half of ’00s has completely changed our everyday mobile computing experience. Tens of applications are being introduced every day in the application markets. Given the technology progress and the fact that mobile devices are becoming strong computing devices, mobile applications are expected to follow suit and become computation-heavy, bandwidth-hungry and latency-sensitive. In this paper, we introduce a new mobile computing paradigm to alleviate some of the network stress that mobile applications are already putting into the network, e.g., in case of crowded areas and events, where the mobile network effectively collapses. According to this paradigm, users can share the applications that they have on their mobile devices with nearby users that want access to processed information, which their own applications cannot provide.
This paper presents Cloudrone- a preliminary idea of deploying a lightweight micro cloud infrastructure in the sky using indigenously built low cost drones, single board computers and lightweight Operating System virtualization technologies. Our paper lays out the preliminary ideas on such a system that can be instantaneously deployed on demand. We describe an initial design of Cloudrone and provide a preliminary evaluation of the proposed system mainly focussed on the scalability issues of supporting multiple services and users.