In this paper, we highlight a potential privacy threat in the current smartphone platforms, which allows any third party to collect a snapshot of installed applications without the user's consent. This can be exploited by third parties to infer various user attributes similar to what is done through tracking. We show that using only installed apps, user's gender, a demographic attribute that is frequently used in targeted advertising, can be instantly predicted with an accuracy around 70%, by training a classifier using established supervised learning techniques.

We consider an architecture in which the same WiFi infrastructure can be dynamically shared among multiple operators. Our system, ViFi, virtualizes WLAN resources, allowing for controlled sharing of both the uplink and downlink bandwidth. ViFi operates with stock 802.11 clients, and can be implemented entirely as a software add-on for commodity 802.11 APs. ViFi puts users (customers) of different operators in separate groups, each creating a virtual WLAN. ViFi guarantees proportional fair share of channel access time at group level, and isolates traffic between groups. The key technical contribution of ViFi is a useful form of virtualization without requiring changes to the underlying WiFi protocol.

Delay Tolerant Networks (DTNs), especially the Bundle Protocol (BP), transmit data in self-contained bundles each of which carrying all necessary information to process it and route it to its destination. While this allows for long delays, link disruptions and higher loss rates and makes the BP well-suited for networks such as Wireless Sensor Networks (WSNs), it imposes a significant overhead in terms of the header sizes, as e. g. node addresses are denoted as full URIs, called Endpoint Identifiers (EIDs). While the Compressed Bundle Header Encoding introduces a special naming scheme to eradicate these URIs which reduces the header size to some extent, in this paper we present a novel, full-fledged header compression for the BP that can be applied stateless or stateful, i. e. without or with storing part of the bundle headers on forwarding nodes. The gains achievable with this approach are extensively evaluated with the simulation of bundles carefully generated from real-world network traffic on the one hand, and of realistically moving public transport vehicles with a traffic pattern often found in such Delay Tolerant Wireless Sensor Networks (DTWSNs) on the other hand.

Mobile opportunistic networking utilizes device-to-device communication to provide messaging and content sharing mechanisms between mobile users without the need for supporting infrastructure networks. However, enabling opportunistic networking in practice requires a sufficient number of users to download, install, and run the respective routing and application software to provide sufficient node density, and thus connectivity for the network to actually function. In this paper, we explore reaching out to nodes that have not (yet) installed any dedicated software to: (1) allow them to access public content in an opportunistic network to possibly seed their interest and (2) instrument them to assist as (limited) message carriers to improve connectivity. We report on our system design and implementation and offer performance insights gained from simulations and initial experiments.

We introduce Harmonia, a new RF-based localization scheme that provides the simplicity, cost, and power advantages of traditional narrowband radios with the decimeter-scale accuracy of ultra wideband localization techniques. Harmonia is an asymmetric tag and anchor system, requiring minimal modifications to existing low-power wireless devices to support highfidelity localization with comparatively modest infrastructure costs. A prototype Harmonia design offers location estimates with an average-case error of 53.4 cm in complex, heavymultipath, indoor environments and captures location estimates at 56 Hz while requiring only 1.7 mA additional power draw for each tag and complying with all US UWB regulations. We believe this architecture's combination of accuracy, update rate, power draw, and system complexity will lead to a new point in the design space.

We present BeHop, a wireless testbed for dense WiFi networks often seen in residential and enterprise settings. BeHop aims to provide insights on the operation of dense deployments, and evaluate how different WiFi management strategies affect user experience and network behavior. It has sufficient flexibility to let us try different management techniques and setups (e.g. residential or enterprise, client or infrastructure-driven operation). It is deployed at a university dorm, where it acts as the main network for a diverse set of users and devices, exposing practical insights and implications on the operation of the network. In this paper we discuss the design and implementation of BeHop, and share our early experience over a five-month period.

We present a solution for enabling standard compliant channel access for a fully softwarebased Software Defined Radio (SDR) architecture. With the availability of a GNURadio implementation of an Orthogonal Frequency Division Multiplexing (OFDM) transceiver, there is substantial demand for standard compliant channel access. It has been shown that implementation of CSMA on a host PC is infeasible due to system-inherent delays. The common approach is to fully implement the protocol stack on the FPGA, which makes further updates or modifications to the protocols a complex and time consuming task. We take another approach and investigate the feasibility of a fully software-based solution and show that standard compliant broadcast transmissions are possible with marginal modifications of the FPGA. We envision the use of our system for example in the vehicular networking domain, where broadcast is the main communication paradigm. We show that our SDR solution exactly complies with the IEEE 802.11 Distributed Coordination Function (DCF) as well as Enhanced Distributed Channel Access (EDCA) timings. We were even able to identify shortcomings of commercial systems and prototypes.

We present HiLight, a new form of unobtrusive screen-camera communication for off-theshelf smart devices. HiLight hides information underlying any images shown on an LED or OLED screen, and camera-equipped smart devices can fetch the information by turning their cameras to the screen. HiLight achieves this by leveraging the transparency (alpha) channel, a well-known concept in computer graphics, to encode bits into pixel translucency changes without modifying pixel color (RGB) values. We demonstrated HiLight's feasibility using smartphones. By offering an unobtrusive, flexible, and lightweight communication channel between screens and cameras, HiLight allows new HCI and context-aware applications to emerge.

One of the central challenges in backscatter is how to enable concurrent transmissions. Most backscatter protocols operate in a sequential TDMA-like manner due to the fact that most nodes cannot overhear each other's transmissions, which is detrimental for throughout and energy consumption. Recent efforts to separate concurrent signals by inverting a system of linear equations is also problematic due to varying channel coefficients caused by system and environmental dynamics. In this paper, we introduce BST, a novel physical layer for backscatter communication that enables concurrent transmission by leveraging vintra-bit multiplexing of OOK signals from multiple tags. The key idea underlying BST is that the reader can sample at considerably higher rates than the tags, hence it can extract time-domain signal edges that result from interleaved transmissions of several tags. Our preliminary experiment results show that BST can achieve 5x the throughput of Buzz and 10x the throughput of TDMA-based solutions, such as EPC Gen 2.

Sophisticated mobile computing, sensing and recording devices are commonplace. Smart phones have achieved significant penetration and novel devices like Google Glass are imminent. These devices can serve most functions of a conventional notebook computer, but also have a range of additional capabilities, including image/audio/video recording, GPS location, compass, accelerometer, nearrange radio (NFC and Bluetooth), and soon health and fitness monitors. Moreover, these devices are carried by their users virtually around the clock, blurring the distinction between the online and offline world and enabling transformative new applications and services. For instance, mobile apps can provide location and activity-sensitive services and information, in the case of Google Glass overlaid right onto a user’s field of view. They can record what the user does, sees and hears for future reference; and they can keep track of a user’s encounters with nearby users’ devices to enable communication related to a shared experience or event. However, these applications and services also introduce a range of new threats to users’ privacy. While a user carries it, a mobile device can capture a complete record of the user’s location, online and offline activities, and social encounters, potentially including an audio-visual record. While such a record is very useful to a user for their own reference and to enable new applications, it is also highly sensitive and inherently private. Unlike information users post on Facebook or Twitter, most users would likely not want to share such a comprehensive record with anyone. In this paper, we catalog privacy threats introduced by these devices and applications. Our survey of threats underlines how privacy threats from mobile devices are fundamentally different and inherently more dangerous than in prior systems. For each specific risk vector, we describe technical challenges that, if solved, can mitigate its effects. We note that technical innovations merely provide a starting point: an end-to-end privacy-preserving infrastructure will require changes in how basic services are deployed, how laws