Mobile Computing, Applications, and Services

We propose a novel paradigm of consuming rich web content in a mobile setting (which are often eyes-free), through a predominantly 3D audio interface. Web content, formatted in audio, is streamed via the mobile device’s network connection, and placed virtually in a 3D audio space. The user moves in the virtual space, using a variety of human computer interaction (HCI) means, such as voice input, touching, rotating, and shaking the device, as well as hand and head gesturing. We provide applications benefitting from this paradigm of rich mobile 3D audio web consumption. We provide system designs, and a system architecture for mobile 3D audio. Finally, we implement our ideas in a system prototype using the Apple iOS mobile platform.

This paper proposes a challenge-response authentication system for web applications called Snap2Pass that is easy to use, provides strong security guarantees, and requires no browser extensions. The system uses QR codes which are small two-dimensional pictures that encode digital data. When logging in to a site, the web server sends the PC browser a QR code that encodes a cryptographic challenge; the user takes a picture of the QR code with his cell phone camera which results in a cryptographic response sent to the server; the web server then logs the PC browser in. Our user study shows that authentication using Snap2Pass is easy to learn and considerably faster than existing one-time password and challenge-response systems. By implementing our solution as an OpenID provider, we have made this scheme available to over 30,000 websites that use OpenID today. This paper also proposes Snap2Pay, an extension of Snap2Pass, to improve the usability and security of online payments. Snap2Pay allows a consumer to use one-time credit cards as well as the Verified by Visa or Mastercard SecureCode services securely and easily with just a snap of a QR code.

Offloading computation from smartphones to remote cloud resources has recently been rediscovered as a technique to enhance the performance of smartphone applications, while reducing the energy usage.

In this paper we present the first practical implementation of this idea for Android: the Cuckoo framework, which simplifies the development of smartphone applications that benefit from computation offloading and provides a dynamic runtime system, that can, at runtime, decide whether a part of an application will be executed locally or remotely. We evaluate the framework using two real life applications.

Mobile Java development using CLDC and MIDP can be very restricting, not only because of the more restricted libraries and older Java language, but also because some very basic development tools are not available in many situations. One of the biggest problems when debugging a midlet – a CLDC/MIDP application – is that when running a mobile Java application in a real device, stack traces are not available. Also other tools, like profiling tools, only work in certain emulators. In this paper, a set of improved tools for mobile Java development is introduced. Instrumentation, a well-known technique is used to work around the restrictions of the Java sandbox. Consequently no special support is required from the platform.

In distributed virtual environments such as multiplayer games, where many users interact in real time while communicating through a network, the users may have an inconsistent view of the game world because of the communication delays across the network. Consistency maintenance algorithms must be used to have a uniform view of the game world. The majority of these algorithms use rollback mechanisms to correct the inconsistencies that occur because of the disorder of the arrival of update messages. These rollbacks are very costly, especially when playing a game, using high-latency wireless networks, on mobile terminals which have limited memory and processing speed. In this paper, we present a dynamic and adaptive approach for reducing the number of rollbacks in distributed virtual environments on wireless mobile devices. This approach takes into account the underlying network latency and the semantics of the game virtual world to dynamically decide whether a rollback is needed in case inconsistencies have occurred or can be possibly avoided. We evaluate our approach on a simplified version of a Football game on hand-held devices and show that this dynamic rollbacks’ reduction approach improves the responsiveness of the game and maintains consistency of the game state while limiting the use of processing power and memory space.

This paper proposes an algorithm to identify groups of users connected to a mobile ad hoc network that remain stable over time. Several similar algorithms have been proposed to manage mobility either by predicting disconnections or by identifying groups of peers stable over time. They all rely on information such as GPS, signal strength or routes and result in message overhead. The algorithm proposed here uses information from the routing layer to detect groups that are stable over time. This algorithm is fully distributed and creates no message overhead as the result of using cross-layer information.

We present the analysis and design of a Small World VoIP system (SW-VoIP) which is geared towards customers that are communicating with their Small World of social contacts. We use the term Small World to refer to the Peer-to-Peer (P2P) network of a client and his contacts both incoming and outbound. We reconstruct the small world of a user by collecting calling patterns over a configurable period of time. We enable user mobility by using a stepwise social identity to an IP address binding propagation model. We propose an efficient algorithm to locate users by electing popular users and leveraging the users closeness. We also introduce a self-stabilized load balancing mechanism to optimize the system performance under heavy network traffic. We evaluate our SW-VoIP system performance by simulating the user’s lookup process using real-world telephone logs. Our experimental results show that our SW-VoIP system offers a better performance in optimizing the required routing path and reducing the average lookup delay when compared to traditional, non small-world P2P VoIP systems.

The purpose of this paper is to predict people’s future locations or when they will be at given locations. These predictions support proactive, context-aware and social applications. Markov models have been shown to be effective predictors of someone’s next location [1]. This paper incorporates temporal information in order to predict future locations or the times when someone will be at a given location. Previous models use sequences of location symbols and apply Markov-based algorithms to predict the next location symbol. In our model, we embed temporal information within the sequence of location symbols. To predict a future location, we use the temporal information as the previous state (or context) in the Markov model to predict the location that is most likely at that given time. To predict when someone will be at a location, we use the location as the context and predict the time(s) the person will be at that location. The model produces up to 91% accuracy for predicting locations, and less than 10% accuracy for predicting times. We show that prediction of location and prediction of time are two very different problems, because the number of predictions produced by the Markov model differ greatly between the two variables. A heuristic algorithm is proposed which incorporates additional context to improve predictions of future times to 43%.

GPS positioning does not provide pervasive coverage and the accuracy depends on the local environment. When deploying and managing position-based applications it is important to know when to depend on GPS and when to deploy supplementary means of positioning, such as local or inertial positioning. This paper proposes PosQ, a system for unsupervised fingerprinting and visualization of GPS positioning quality. PosQ provides quality maps to position-based applications and visual overlays to users and managers to reveal the positioning quality in a local environment. The system reveals the quality both as it changes over time, in 2D and 3D, and for each type of GPS receiver. Our evaluation provides evidence that the collected quality maps are accurate, that they remain informative over time, that they capture the differences among GPS receivers, and that they can be efficiently collected by participating devices.

Symbolic location of a user, like a store name in a mall, is essential for context-based mobile advertising. Existing fingerprint-based localization using only a single phone is susceptible to noise, and has a major limitation in that the phone has to be held in the hand at all times. In this paper, we present SensOrchestra, a collaborative sensing framework for symbolic location recognition that groups nearby phones to recognize ambient sounds and images of a location collaboratively. We investigated audio and image features, and designed a classifier fusion model to integrate estimates from different phones. We also evaluated the energy consumption, bandwidth, and response time of the system. Experimental results show that SensOrchestra achieved 87.7% recognition accuracy, which reduces the error rate of single-phone approach by 2X, and eliminates the limitations on how users carry their phones. We believe general location or activity recognition systems can all benefit from this collaborative framework.

Continuous stress monitoring may help users better understand their stress patterns and provide physicians with more reliable data for interventions. Previously, studies on mental stress detection were limited to a laboratory environment where participants generally rested in a sedentary position. However, it is impractical to exclude the effects of physical activity while developing a pervasive stress monitoring application for everyday use. The physiological responses caused by mental stress can be masked by variations due to physical activity.

We present an activity-aware mental stress detection scheme. Electrocardiogram (ECG), galvanic skin response (GSR), and accelerometer data were gathered from 20 participants across three activities: sitting, standing, and walking. For each activity, we gathered baseline physiological measurements and measurements while users were subjected to mental stressors. The activity information derived from the accelerometer enabled us to achieve 92.4% accuracy of mental stress classification for 10-fold cross validation and 80.9% accuracy for between-subjects classification.

In this paper we present our initial work on a mobile phone application for assisting stroke rehabilitation. We believe that using a mobile phone to administer and track stroke rehabilitation is novel. We call our system Dr. Droid and focus on the automated scoring of motions performed by patients being administered the Wolf Motor Function Test (WMFT) by placing a smart phone in a holster at the patients wrist. We have developed a complete software application that administers the test by giving audio and visual instructions. We collect a motion trace by sampling the 3-axis accelerometer available on the phone. We double-integrate the acceleration data and apply a novel reorientation algorithm to correct for mis-alignment of the accelerometer. Using dynamic time warping and hidden Markov models we assign an objective, quantitative score to the patient’s exercises. We validate our method by performing experiments designed to simulate the motions of a stroke patient.

In order to understand the value of social information in the context of mobile commerce, we created the Open Transaction Network (OTN), a collaborative, social transaction system. OTN uses voluntarily contributed transactions to index personal and social experiences in the physical world and to form dynamic communities around purchases. We use mobile phones and Open Spaces as portals to facilitate sharing of transactions. Through real world deployment we investigate the design elements and analyze users’ tolerance to sharing such experiences. The sociability threshold, introduced as a measure of user’s willingness to share for different categories of products, is found to correlate with price. The system was deployed to over 20 users over a 5 month period to allow the participants to share their in-store purchases. The analysis of empirical data shows that second degree connections are valuable for obtaining recommendations.

Lifelog system involves capturing personal experiences in the form of digital multimedia during an entire lifespan. Recent advancements in mobile sensor technologies have helped to develop these systems using commercial smart phones. These systems have the potential to act as a secondary memory and also aid people who struggle with episodic memory impairment (EMI). Despite their huge potential, there are major challenges that need to be addressed to make them useful. One of them is how to index the inherently large lifelog data so that the person can efficiently retrieve the log segments that interest him / her most. In this paper, we present an ongoing research of using mobile phones to record and index lifelogs using activity language. By converting sensory data such as accelerometer and GPS readings into activity language, we are able to apply statistical natural language processing techniques to index, recognize, segment, cluster, retrieve, and infer high-level semantic meanings of the collected lifelogs. Based on this indexing approach, our lifelog system supports easy retrieval of log segments representing past similar activities and automatic lifelog segmentation for efficient browsing and activity summarization.

Continuous stress monitoring may help users better under- stand their stress patterns and provide physicians with more reliable data for interventions. Previously, studies on mental stress detection were lim- ited to a laboratory environment where participants generally rested in a sedentary position. However, it is impractical to exclude the effects of physical activity while developing a pervasive stress monitoring appli- cation for everyday use. The physiological responses caused by mental stress can be masked by variations due to physical activity.

We present an activity-aware mental stress detection scheme. Electrocar- diogram (ECG), galvanic skin response (GSR), and accelerometer data were gathered from 20 participants across three activities: sitting, stand- ing, and walking. For each activity, we gathered baseline physiological measurements and measurements while users were subjected to mental stressors. The activity information derived from the accelerometer en- abled us to achieve 92.4% accuracy of mental stress classification for 10-fold cross validation and 80.9% accuracy for between-subjects classi- fication.

With this demonstration we present Decisionlib, a Distributed Decision Support System that runs on mobile devices. Decisionlib is a flexible, durable, and robust voting system which places emphasis on communication and error handling in order to provide reliable and easy to program group-based decisions on distributed mobile devices. Written on top of the Ibis [4] distributed communication system, Decisionlib represents the state of the art for group decision making on mobile devices.

Thanks to the decreasing cost, increasing mobility, and wider use of sensors, a great number of possible applications have recently emerged, including applications that may impact the high-level goals in people’s lives. Applications can be found in many areas ranging from medical devices to consumer devices. Information about activity and context can be inferred from sensors and be used to provide automated recommendations.

This poster describes the MagicWand framework for developing web applications that can be controlled remotely by smart phone running on android OS.

This paper addresses media sharing via an approach that offers ‘fungible’ storage, where storage services implement virtual stores that are dynamically mapped to suitable ‘nearby’ or otherwise available physical devices. In particular, the novel VStore++ system provides seamless and flexible data storage, access, and sharing services, by exploiting virtualization technology to aggregate and make use of both ‘nearby’ and private storage (e.g., in a mobile user’s home), and public storage resources offered on remote cloud platforms.

Cloud has become a promising service model for mobile devices. Using cloud services, mobile devices can outsource its computationally intensive operations to the cloud, such as searching, data mining, and multimedia processing. In this service computing model, how to build an economic service provisioning scheme is critical for mobile cloud service providers. Particularly when the mobile cloud resource is restricted. In this paper, we present an economic mobile cloud computing model using Semi-Markov Decision Process for mobile cloud resource allocation. Our model takes the considerations the cloud computing capacity, the overall cloud system gain, and expenses of mobile users using cloud services. Based on the best of our knowledge, our presented model is the first to address the economic service provisioning for mobile cloud services. In the performance evaluation, we showed that the presented economic mobile cloud computing model can produce the optimal system gain with a given cloud service inter-domain transfer probability.

In this paper we study how smartphones can benefit from the resources available in clouds. Unfortunately, integrating smartphones with cloud resources is challenging and comes with dangers for the user in terms of loss of control of applications and data as portions move into the cloud. In this paper we outline our work on the Interdroid project, where we are building a framework for smart applications for smartphones which includes components for integration with the cloud.

The advancements in computing have resulted in a boom of cheap, ubiquitous, connected mobile devices as well as seemingly unlimited, utility style, pay as you go computing resources, commonly referred to as Cloud computing. Taking advantage of this computing landscape, however, has been hampered by the many heterogeneities that exist in the mobile space as well as the Cloud space.

This research attempts to introduce a disciplined methodology to develop Cloud-mobile hybrid applications by using a Domain Specific Language (DSL) centric approach to generate applications. A Cloud-mobile hybrid is an application that is split between a Cloud based back-end and a mobile device based front-end. We present mobicloud, our prototype system we built based on a DSL that is capable of developing these hybrid applications. This not only reduces the learning curve but also shields the developers from the native complexities of the target platforms. We also present our vision on propelling this research forward by enriching the DSLs with semantics. The high-level vision is outline in the ambitious Cirrocumulus project, the driving principle being write once - run on any device.

Despite the rapid advances in mobile technology, many constraints still prohibit smartphones to run resource-demanding applications in pervasive environments. Emerging cloud computing opens an access to unlimited resources for mobile devices. However, the combination of both technologies to deliver sound mobile cloud applications and services raises new challenges and requirements. Based on a scenario-based requirement analysis and a comprehensive study on existing work for augmenting mobile devices, we propose a XMPP-based mobile cloud computing architecture employing module partitioning and adaptive offloading to nearby computing infrastructure. Research has also been done in the underlying offloading mechanism based on context-aware cost model. Further problems related to this approach are discussed as well, including selection of most optimal offloading plan, application partitioning and issues with XMPP on mobile systems.

Mobile phones have become a primary communication device nowadays. In order to maintain proper functionality, various existing security solutions are being integrated into mobile devices. Some of the more sophisticated solutions, such as host-based intrusion detection systems (HIDS) are based on continuously monitoring many parameters in the device such as CPU and memory consumption. Since the continuous monitoring of many parameters consumes considerable computational resources it is necessary to reduce consumption in order to efficiently use HIDS. One way to achieve this is to collect less parameters by means of cost-sensitive feature selection techniques. In this study, we evaluate ProCASH, a new cost-sensitive feature selection algorithm which considers resources consumption, misclassification costs and feature grouping. ProCASH was evaluated on an Android-based mobile device. The data mining task was to distinguish between benign and malicious applications. The evaluation demonstrated the effectiveness of ProCASH compared to other cost sensitive algorithms.

This paper concerns the implementation of our online signature verification system on a mobile device. Verification involves confirming or denying a person‘s claimed identity. Our system is based on a Hidden Markov Model and outputs two complementary scores: the first one is related to the likelihood given by the HMM of the claimed identity; the second one is related to the segmentation given by such an HMM on the input signature. A claimed identity is confirmed when the arithmetic mean of the two scores obtained on such an input signature is higher than a threshold. Also, a personal normalization of the local parameters of the signature is carried out to make the system robust to changes of platforms. A patent was submitted with special emphasis on the latter claim. This system is implemented on a mobile platform PDA Qtek 2020 ARM 400 MHz. An acquisition interface is developed allowing an enrollment step of a person by acquisition of 5 of his/her signatures, and a verification step of a given signature of a registered person. Enrolment speed depends on the complexity of the signature, while verification is performed in real time. Performance assessment of our system, carried out on two databases acquired on a PDA, shows a degradation of system performance on mobile platform compared to a fixed platform. In order to improve the performance in the case of mobility, we propose a strategy for enhancing the quality of the reference signatures at the enrolment phase.

Among the most significant smartphone operating systems that have arisen recently is Google’s Android framework. Google’s Android is a software framework for mobile communication devices. The Android framework includes an operating system, middleware and a set of key applications. Designed as open, programmable, networked devices, Android is vulnerable to various types of threats. This paper provides a security assessment of the Android framework and the security mechanisms incorporated into it. In addition, a review of recent academic and commercial solutions in the area of smartphone security in general and Android in particular is presented.

The availability of free Software Development Kits for recent mobile device platforms challenges many developers in realizing applications for the growing Smartphone market. In many cases such applications may interoperate in their working environment using mechanisms similar to the inter-process communication (IPC) and made available by the mobile operating system. Unfortunately, mobile devices lack in flexible solutions for making these communications secure. In this paper we propose a framework to secure the message exchange with the services installed on Google Android mobile devices. VASs realized by different providers are discovered, used and composed by an Application Frame designed for realizing complex goals. We implemented a prototype of our proposed framework on a real device and we performed extensive testing to measure the overhead introduced by the cryptographic operations required to protect the inter process communication.