Julien Gedeon

Noisemap: multi-tier incentive mechanisms for participative urban sensing

Noise pollution is a problem increasingly acknowledged by authorities and governments around the globe. At last years PhoneSense we presented Noisemap, a participating sensing application to accurately measure noise pollution. Noisemap incorporated frequency calibration to overcome the limited microphone hardware. The challenge remaining is how to motivate smartphone users to sacrifice their time and battery on measuring noise. A user study was conducted with 49 users divided into three groups. As expected the average measurements taken per user increased from 402 to 3,357 as the number of incentive schemes increased. Over the course of 7 weeks the users captured more than 85, 000 measurements, measuring for more than six hours on average.

Decision Support for Computational Offloading by Probing Unknown Services

Mobile Cloud Computing (MCC) leverages resourceful data centers that are distant (aka the cloud) or closely located (aka edge servers) for computational offloading to overcome resource limitations of modern mobile systems like smartphones or IoT devices. Many research works investigate context-aware offloading decision algorithms aiming to find the best offloading system at runtime. However, all approaches require prior knowledge of the offloading systems or a running service profiler on the backend system. In this paper, we present a novel approach that overcomes this issue by first probing available unknown services such as nearby cloudlets or the distant cloud, and networks in an energy-efficient way at runtime to make better offloading decisions. For that, we investigate a probing strategy to assess these unknown services by offloading micro tasks and accurately predicting the performance for larger offloading tasks using regression models. Our evaluation on three algorithms with different time complexities shows that we achieve high prediction accuracies up to 85.5%, already after probing of two micro tasks running in the range of few milliseconds. To the best of our knowledge, this is the first supplement approach for offloading decision support that can handle unknown third-party services requiring no prior knowledge about these offloading systems and making no assumptions for real-world deployments.

Router-Based Brokering for Surrogate Discovery in Edge Computing

In-network processing pushes computational capabilities closer to the edge of the network, enabling new kinds of location-aware, real-time applications, while preserving bandwidth in the core network. This is done by offloading computations to more powerful or energy-efficient surrogates that are opportunistically available at the network edge. In mobile and heterogeneous usage contexts, the question arises how a client can discover the most appropriate surrogate in the network for offloading a task. In this paper, we propose a brokering mechanism that matches a client with the best available surrogate, based on specified requirements and capabilities. The broker is implemented on standard home routers, and thus, leverages the ubiquity of such devices in urban environments. To motivate the feasibility of this approach, we conduct a coverage analysis based on collected access point locations in a major city. Furthermore, the brokering functionality introduces only a minimal resource overhead on the routers and can significantly reduce the latency compared to distant, cloud-based solutions.

Upgrading Wireless Home Routers as Emergency Cloudlet and Secure DTN Communication Bridge

Reliable communications are crucial for the success of emergency response and management. However, todays technologies used by rescuers and civilians mainly rely on either centralized or specialized emergency approaches, which reveal individual issues especially in infrastructure- less emergency situations (e.g., blackout). In this paper, we present a customary home router upgraded as self- sustaining emergency device which can ad hoc network with nearby devices (e.g., other upgraded routers, smartphones) using wireless communication technologies. On top of the ad-hoc networking, an upgraded router provides (1) personal computing capacities for low-latency offloading from mobile devices (aka cloudlet) using isolated lightweight containers; and (2) store-and-forward delay-tolerant data exchanges to serve as secure communication bridge for cooperation between involved or affected people (e.g., rescuers, civilians). We believe that upgrading ubiquitous routers is a very promising concept for a scalable ad-hoc networking and energy-efficient computing infrastructure in urban emergency situations.

vStore: A Context-Aware Framework for Mobile Micro-Storage at the Edge

The way mobile users store and share their data today is completely decoupled from their current usage context and actual intentions. Furthermore, the paradigm of cloud computing, where all data is placed in distant cloud data centers is seldom questioned. As a result, we are faced with congested networks and high latencies when retrieving data stored at distant locations. The emergence of edge computing provides an opportunity to overcome this issue. In this paper, we present vStore, a framework that provides the capabilities for context-aware micro-storage. The framework is targeted at mobile users and leverages small-scale, decentralized storage nodes at the extreme edge of the network. The decision where to store data is made based on rules that can either be pushed globally to the framework or created individually by users. We motivate our approach with different use cases, one of which is the sharing of data at events where cellular networks tend to be congested. To demonstrate the feasibility of our approach, we implement a demo application on the Android platform, leveraging storage nodes placed at different locations in a major city. By conducting a field trial, we demonstrate the key functionalities of vStore and report on first usage insights.

A Multi-Cloudlet Infrastructure for Future Smart Cities: An Empirical Study

The emerging paradigm of edge computing has proposed cloudlets to offload data and computations from mobile, resource-constrained devices. However, little attention has been paid to the question on where to deploy cloudlets in the context of smart city environments. In this vision paper, we propose to deploy cloudlets on a city-wide scale by leveraging three kinds of existing infrastructures: cellular base stations, routers and street lamps. We motivate the use of this infrastructure with real location data of nearly 50,000 access points from a major city. We provide an analysis on the potential coverage for the different cloudlet types. Besides spatial coverage, we also consider user traces from two mobile applications. Our results show that upgrading only a relatively small number of access points can lead to a city-scale cloudlet coverage. This is especially true for the coverage analysis of the mobility traces, where mobile users are within the communication range of a cloudlet-enabled access point most of the time.

Understanding spatial and temporal coverage in participatory sensor networks

Sensor coverage is a well established problem in sensor networks. Most of the work is focused on optimizing coverage in stationary networks or by controlling the movement of mobile nodes (e.g. robots) in order to maximize their coverage. In participatory sensor networks, we are faced with noncontrollable mobility. Humans move freely and there is no central nor distributed algorithm that optimizes coverage. There is no work in literature yet that explores coverage in the context of non-controllable mobility. To this end, we report results of a study applying an adapted greedy coverage algorithm onto three different data sets. Given these datasets, we report results studying the effect of different mobility characteristics on the spatial and temporal coverage. Our results show that high coverage can be achieved by a relatively small subset of nodes. Also, given a real-world participatory sensing system, turn-around times are relevant for continuous temporal coverage.