As mobile apps become more closely integrated into our everyday lives, mobile app interactions ought to be rapid and responsive. Unfortunately, even the basic primitive of launching a mobile app is sorrowfully sluggish: 20 seconds of delay is not uncommon even for very popular apps.We have designed and built FALCON to remedy slow app launch. FALCON uses contexts such as user location and temporal access patterns to predict app launches before they occur. FALCON then provides systems support for effective app-specific prelaunching, which can dramatically reduce perceived delay.FALCON uses novel features derived through extensive data analysis, and a novel cost-benefit learning algorithm that has strong predictive performance and low runtime overhead. Trace-based analysis shows that an average user saves around 6 seconds per app startup time with daily energy cost of no more than 2% battery life, and on average gets content that is only 3 minutes old at launch without needing to wait for content to update. FALCON is implemented as an OS modification to the Windows Phone OS.

The shortage of parking in crowded urban areas causes severe societal problems such as traffic congestion, environmental pollution, and many others. Recently, crowdsourced parking, where smartphone users are exploited to collect realtime parking availability information, has attracted significant attention. However, existing crowdsourced parking information systems suffer from low user participation rate and data quality due to the lack of carefully designed incentive schemes.In this paper, we address the incentive problem of trustworthy crowdsourced parking information systems by presenting an incentive platform named TruCentive, where high utility parking data can be obtained from unreliable crowds of mobile users. Our contribution is three-fold. First, we provide hierarchical incentives to stimulate the participation of mobile users for contributing parking information. Second, by introducing utility-related incentives, our platform encourages participants to contribute high utility data and thereby enhances the quality of collected data. Third, our active confirmation scheme validates the parking information utility by game-theoretically formulated incentive protocols. The active confirming not only validates the utility of contributed data but re-sells the high utility data as well. Our evaluation through user study on Amazon Mechanical Turk and simulation study demonstrate the feasibility and stability of TruCentive incentive platform.

Recent advances in sensor networks permit the use of a large number of relatively inexpensive distributed computational nodes with camera sensors linked in a network and possibly linked to one or more central servers. We argue that the full potential of such a distributed system can be realized if it is designed as a distributed search engine where images from different sensors can be captured, stored, searched and queried. However, unlike traditional image search engines that are focused on resource-rich situations, the resource limitations of camera sensor networks in terms of energy, bandwidth, computational power, and memory capacity present significant challenges. In this paper, we describe the design and implementation of a distributed search system over a camera sensor network where each node is a search engine that senses, stores and searches information. Our work involves innovation at many levels including local storage, local search, and distributed search, all of which are designed to be efficient under the resource constraints of sensor networks. We present an implementation of the search engine on a network of iMote2 sensor nodes equipped with low-power cameras and extended flash storage. We evaluate our system for a dataset comprising book images, and demonstrate more than two orders of magnitude reduction in the amount of data communicated and up to 5x reduction in overall energy consumption over alternate techniques.

No abstract

This paper presents the design and implementation of a dual-camera sensor network that can be used as a memory assistant tool for assisted living. Our system performs energy-efficient object detection and recognition of commonly misplaced objects. The novelty in our approach is the ability to tradeoff between recognition accuracy and computational efficiency by employing a combination of low complexity but less precise color histogram-based image recognition together with more complex image recognition using SIFT descriptors. In addition, our system can seamlessly integrate feedback from the user to improve the robustness of object recognition. Experimental results reveal that our system is computation-efficient and adaptive to slow changes of environmental conditions