AI tools like ChatGPT and Be-My-AI are increasingly being used by blind individuals. Although prior work has explored their use in some Do-It-Yourself (DIY) tasks by blind individuals, little is known about how they use these tools and the available product-manual resources to assemble, operate, and troubleshoot physical/tangible products – tasks requiring spatial reasoning, structural understanding, and precise execution. We address this knowledge gap via an interview study and a usability study with blind participants, investigating how they leverage AI tools and product manuals for DIY tasks with physical products. Findings show that manuals are essential resources, but product-manual instructions are often inadequate for blind users. AI tools presently do not adequately address this insufficiency, in fact, we observed that they often exacerbate this issue with incomplete, incoherent, or misleading guidance. Lastly, we suggest improvements to AI tools for generating tailored instructions for blind users’ DIY tasks involving tangible products.

Text editing on smartphones presents substantial difficulties for blind users, particularly in mobile situations where using the smartphone touch screen is challenging.While voice input allows for hands-free text creation, editing the text typically requires physical interaction with the touchscreen, negating the benefits of the hands-free input mechanism.This paper introduces GestureVoice, a novel multimodal approach that enables screen-free text editing for blind users.By leveraging smartwatch-based hand gestures for navigation and voice commands for correction, GestureVoice allows users to edit text without any contact with their smartphones.GestureVoice replaces cumbersome screen-based interaction for choosing the navigation granularity with an intuitive mid-air hand gesture.It also introduces an adaptive crown cursor (rotating the physical dial of the watch) to smoothly navigate to the edit location.A preliminary study highlighted the significant time spent by blind users correcting text errors using traditional methods.In contrast, our evaluation with 8 blind users demonstrates that Ges-tureVoice achieves a 53.80% reduction in text editing time, offering a more efficient, intuitive, and screen-free solution for blind users.

Museum visits often lack personalized and interactive experiences, limiting visitor engagement with art and historical artifacts. To address this, we present ArtSpeak, a standalone augmented reality (AR) application that transforms traditional art viewing into an interactive storytelling experience. When users point their mobile cameras at an artwork, the system responds to their questions with lifelike, talking-head video narratives generated from historical portraits. However, generating such talking-head videos at runtime is computationally expensive, often requiring over a minute per response. To address this challenge, ArtSpeak introduces two major contributions. First, it employs a collection of frequently asked questions (FAQ) to generate a set of lifelike video responses for various art portraits. Second, it introduces a novel retrieval-based approach that uses GPT-based embeddings and cosine similarity to select the most relevant response. As a result, the system dynamically presents the video reply that best aligns with the user's inquiry, reducing computational overhead and ensuring a real-time, low-latency experience. More precisely, ArtSpeak achieves over 30 x lower latency and reduces energy consumption by approximately 81 % compared to the real-time video generation method. User studies further validate the system's effectiveness, with 85 % of participants rating the retrieved responses as relevant to their queries and 90 % reporting smooth video playback. These results highlight the efficiency and user satisfaction enabled by our retrieval-based approach.

Hand gestures provide an alternate interaction modality for blind users and can be supported using commodity smartwatches without requiring specialized sensors. The enabling technology is an accurate gesture recognition algorithm, but almost all algorithms are designed for sighted users. Our study shows that blind user gestures are considerably diferent from sighted users, rendering current recognition algorithms unsuitable. Blind user gestures have high inter-user variance, making learning gesture patterns difcult without large-scale training data. Instead, we design a gesture recognition algorithm that works on a 3D representation of the gesture trajectory, capturing motion in free space. Our insight is to extract a micro-movement in the gesture that is user-invariant and use this micro-movement for gesture classifcation. To this end, we develop an ensemble classifer that combines image classifcation with geometric properties of the gesture. Our evaluation demonstrates a 92% classifcation accuracy, surpassing the next best state-of-the-art which has an accuracy of 82%.

Ethereum Name Service (ENS) domains allow users to map human-readable names (such as gold.eth) to their cryptocurrency addresses, simplifying cryptocurrency transactions. Like traditional DNS domains, ENS domains must be periodically renewed. Failure to renew leads to expiration, making them available for others to register (a phenomenon known as dropcatching). This presents a security risk where attackers can register expired domains to leverage the residual trust associated with them and, in the context of ENS, receive transactions intended for their previous owners. In this paper, we conduct the first large-scale study on dropcatching in ENS domains. We curate and analyze a dataset comprising 3.1M ENS domains and 9.7M Ethereum transactions, finding that 241K of these domains were re-registered by new owners after expiration. Our findings indicate a preference for domains linked to high-income wallets in re-registrations. We identify 2,633 transactions that were misdirected to new owners, averaging the equivalent of thousands of US dollars. Lastly, we highlight the lack of countermeasures by digital wallet providers, and suggest straightforward approaches that they can use to minimize financial losses due to ENS dropcatching.

The proliferation of video analytics in applications like autonomous driving, traffic surveillance, and teleoperated vehicles requires on-premise (on edge) execution of deep learning models to meet latency requirements and curb bandwidth usage by limiting frequent offloading of inference tasks. However, constrained by the compute and power availability on the edge, a cheaper model is typically deployed. These shallower models have two major associated problems: 1) using the same model for all cameras/vehicles gives inconsistent accuracy, and 2) trained models are prone to data drift.

In this paper, we present AccessWear, a system that improves the accessibility of smartphone touchscreen interactions for blind users using smartwatch gestures. Our system design is human-centered, namely, it incorporates the design goals that were learned from a formative user study with 9 blind participants. The formative study showed that blind users liked the idea of using smartwatch gestures as an alternative: 4 participants liked that when using smart-watch gestures, they did not have to bring their expensive phones out in public and 6 participants liked that smart-watch gestures can be performed with one-hand, as the other hand is usually occupied in holding a cane or a guide dog. Even though there are several advantages to smartwatch gestures, our study also shows that gestures performed by blind users have different patterns compared to sighted users, making gesture recognition more challenging. To this end, AccessWear makes two contributions. The first is a gesture recognition system that works specifically for blind users that is lightweight and does not require per-person training. The second is a near-zero-effort gesture replacement system that does not require any changes to the original application. AccessWear uses input virtualization techniques so that a given gesture can replace the touchscreen input seamlessly. We implement AccessWear on an Android smartphone and Android watch. We perform a quantitative and qualitative study with 8 blind participants. Our study shows that AccessWear can recognize gestures with a 92% accuracy and the end-to-end latency when using an alternate gesture was 53 msec on average. The qualitative study shows that when participants perform a task, consisting of a series of gestures, the system is robust, does not have perceived delays, and does not add physical or mental load on the users.

InterPlanetary File System (IPFS) is a peer-to-peer protocol for decentralized content storage and retrieval. The IPFS platform has the potential to help users evade censorship and avoid a central point of failure. IPFS is seeing increasing adoption for distributing various kinds of files, including video. However, the performance of video streaming on IPFS has not been well-studied. We conduct a measurement study with over 28,000 videos hosted on the IPFS network and find that video streaming experiences high stall rates due to relatively high Round Trip Times (RTT). Further, videos are encoded using a single static quality, because of which streaming cannot adapt to different network conditions.

We present a model for predicting visual attention during the free viewing of graphic design documents. While existing works on this topic have aimed at predicting static saliency of graphic designs, our work is the first attempt to predict both spatial attention and dynamic temporal order in which the document regions are fixated by gaze using a deep learning based model. We propose a two-stage model for predicting dynamic attention on such documents, with webpages being our primary choice of document design for demonstration. In the first stage, we predict the saliency maps for each of the document components (e.g. logos, banners, texts, etc. for webpages) conditioned on the type of document layout. These component saliency maps are then jointly used to predict the overall document saliency. In the second stage, we use these layout-specific component saliency maps as the state representation for an inverse reinforcement learning model of fixation scanpath prediction during document viewing. To test our model, we collected a new dataset consisting of eye movements from 41 people freely viewing 450 webpages (the largest dataset of its kind). Experimental results show that our model outperforms existing models in both saliency and scanpath prediction for webpages, and also generalizes very well to other graphic design documents such as comics, posters, mobile UIs, etc. and natural images.

BBR is a new congestion control algorithm that has seen widespread Internet adoption in recent years with an estimated 40% of Internet traffic volume as BBR traffic. While many studies examine the performance and fairness of BBR on desktops and servers, there is still a question of how BBR would behave on mobile devices. This is especially important because mobiles represent a large segment of Internet devices. In this work, we study the potential performance bottlenecks of BBR if it were to be deployed on Android devices. We compare the performance of BBR and the default congestion control algorithm Cubic for different devices and device configurations. We find that BBR performs poorly compared to Cubic, especially under low-end device configurations. Further investigation reveals that this poor performance is because of packet pacing which is enabled in BBR by default. Pacing increases the computational overhead, which can affect performance for low-end devices. To address this problem, we propose a first cut solution that modifies BBR's pacing behavior to improve performance while still retaining the benefits of packet pacing.

Modern mobile OSes support to display Web pages in the native apps, which we call embedded Web pages. In this paper, we conduct, to the best of our knowledge, the first measurement study on browsing embedded Web pages on Android. Our study on 22,521 popular Android apps shows that 57.9% and 73.8% of apps embed Web pages on two popular app markets: Google Play and Wandoujia, respectively. To analyze the embedded Web browsing performance at scale, we design and implement EWProfiler, a tool that can automatically search for embedded Web pages inside apps, trigger page loads, and retrieve performance metrics. Based on 445 embedded Web pages obtained by EWProfiler in 99 popular apps from the two app markets, we investigate the characteristics and performance of embedded Web pages, and find that embedded Web pages significantly impede the app user experience. To optimize the performance of embedded Web browsing, we investigate the effectiveness of three techniques, i.e., separating the browser kernel to a different process, loading pages from local storage, and pre-rendering. We believe that our findings could draw attentions to Web developers, browser vendors, app developers, and mobile OS vendors together towards better performance of embedded Web browsing.

Qingqing Cao, Yash Kumar Lal, Harsh Trivedi, Aruna Balasubramanian, Niranjan Balasubramanian. Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2021.

Users are surrounded by sensors that are available through various devices beyond their smartphones. However, these sensors are not fully utilized by current end-user applications. A key reason sensor use is so limited is that application developers must exactly identify how the sensor data can be used by smartphone apps. To mitigate this problem, we present SenseWear, a sensor-sharing platform that extends the functionality of a smartphone to use remote sensors with limited additional developer effort. Sensor sharing has several uses, including augmenting the hardware in smartphones, creating new gestural interactions with smartphone applications, and improving application's Quality of Experience via higher-quality sensors from other devices, such as wearables. We developed and present six use cases that use remote sensors in various smartphone applications. Each extension requires adding fewer than 20 lines of code on average. Furthermore, using remote sensors did not introduce a perceptible increase in latency, and creates more convenient interaction options for smartphone apps.

BBR is a new congestion control algorithm and is seeing increased adoption especially for video traffic. BBR solves the bufferbloat problem in legacy loss-based congestion control algorithms where application performance drops considerably when router buffers are deep. BBR regulates traffic such that router queues don’t build up to avoid the bufferbloat problem while still maintaining high throughput. However, our analysis shows that video applications experience significantly poor performance when using BBR under deep buffers. In fact, we find that video traffic sees inflated latencies because of long queues at the router, ultimately degrading video performance. To understand this dichotomy, we study the interaction between BBR and DASH video. Our investigation reveals that BBR under deep buffers and high network burstiness severely overestimates available bandwidth and does not converge to steady state, both of which results in BBR sending substantially more data into the network, causing a queue buildup. This elevated packet sending rate under BBR is ultimately caused by the router’s ability to absorb bursts in traffic, which destabilizes BBR’s bandwidth estimation and overrides BBR’s expected logic for exiting the startup phase. We design a new bandwidth estimation algorithm and apply it to BBR (and a still-unreleased, newer version of BBR called BBR2). Our modified BBR and BBR2 both see significantly improved video QoE even under deep buffers.

Transformer-based QA models use input-wide self-attention -i.e.across both the question and the input passage -at all layers, causing them to be slow and memory-intensive.It turns out that we can get by without inputwide self-attention at all layers, especially in the lower layers.We introduce DeFormer, a decomposed transformer, which substitutes the full self-attention with question-wide and passage-wide self-attentions in the lower layers.This allows for question-independent processing of the input text representations, which in turn enables pre-computing passage representations reducing runtime compute drastically.Furthermore, because DeFormer is largely similar to the original model, we can initialize DeFormer with the pre-training weights of a standard transformer, and directly fine-tune on the target QA dataset.We show DeFormer versions of BERT and XLNet can be used to speed up QA by over 4.3x and with simple distillation-based losses they incur only a 1% drop in accuracy.We open source the code at https://github.com/ StonyBrookNLP/deformer.

360° videos provide an immersive experience to users, but require considerably more bandwidth to stream compared to regular videos. State-of-the-art 360° video streaming systems use viewport prediction to reduce bandwidth requirement, that involves predicting which part of the video the user will view and only fetching that content. However, viewport prediction is error prone resulting in poor user Quality of Experience (QoE). We design PARSEC, a 360° video streaming system that reduces bandwidth requirement while improving video quality. PARSEC trades off bandwidth for additional client-side computation to achieve its goals. PARSEC uses an approach based on super-resolution, where the video is significantly compressed at the server and the client runs a deep learning model to enhance the video to a much higher quality. PARSEC addresses a set of challenges associated with using super-resolution for 360° video streaming: large deep learning models, slow inference rate, and variance in the quality of the enhanced videos. To this end, PAR-SEC trains small micro-models over shorter video segments, and then combines traditional video encoding with super-resolution techniques to overcome the challenges. We evaluate PARSEC on a real WiFi network, over a broadband network trace released by FCC, and over a 4G/LTE network trace. PARSEC significantly outperforms the state-of-art 360° video streaming systems while reducing the bandwidth requirement.

Page Load Time (PLT) is critical in measuring web page load performance. However, the existing PLT metrics are designed to measure the Web page load performance on desktops/laptops and do not consider user interactions on mobile browsers. As a result, they are ill-suited to measure mobile page load performance from the perspective of the user. In this work, we present the Mobile User-Centered Page Load Time Estimator (muPLTest), a model that estimates the PLT of users on Web pages for mobile browsers. We show that traditional methods to measure user PLT for desktops are unsuited to mobiles because they only consider the initial viewport, which is the part of the screen that is in the user’s view when they first begin to load the page. However, mobile users view multiple viewports during the page load process since they start to scroll even before the page is loaded. We thus construct the muPLTest to account for page load activities across viewports. We train our model with crowdsourced scrolling behavior from live users. We show that muPLTest predicts ground truth user-centered PLT, or the muPLT, obtained from live users with an error of 10-15% across 50 Web pages. Comparatively, traditional PLT metrics perform within 44-90% of the muPLT. Finally, we show how developers can use the muPLTest to scalably estimate changes in user experience when applying different Web optimizations.

Wearable devices, such as smart watches and fitness trackers are growing in popularity, creating a need for application developers to adapt or extend a UI, typically from a smartphone, onto these devices. Wearables generally have a smaller form factor than a phone; thus, porting an app to the watch necessarily involves reworking the UI. An open problem is identifying best practices for adapting UIs to wearable devices.

Web page performance is crucial in today's Internet ecosystem, and Web developers use various developer tools to analyze their page load performance. However, existing tools cannot be used to identify the critical bottlenecks during the page load process. In this work, we design an online tool called WProfX that allows Web developers to visually identify bottlenecks in their page structure. The key to WProfX is that unlike existing Web performance tools, WProfX not only visualizes the page load activity timings, but also extracts the dependencies between the activities. Using the dependency structure, WProfX identifies the critical bottleneck activities. This lets a developer quickly identify why their page is loading slow and conduct what-if analyses to study the effect of different optimizations. WProfX uses low-level tracing information exposed by most major browsers to extract the relationship between page load activities. The result is that WProfX works with most major browsers and newer browser versions. WProfX visualizes the page load process as a dependency graph of semantically meaningful Web activities and identifies the critical bottlenecks. We evaluate WProfX with 14 Web developers who perform three what-if analysis tasks involving identifying the page load bottleneck and evaluating the effect of a page optimization. All the participants were able to complete the tasks with WProfX, compared to less than 60% when using the popular developer tools available today. WProfX is currently being used by Web developers in a large telecom and at a Silicon Valley startup.

This paper presents a user study with 10 blind participants to understand their perception of power consumption in smartphones. We found that a widely used power saving mechanism for smartphones--pressing the power button to put the smartphone to sleep--has a serious usability issue for blind screen reader users. Among other findings, our study also unearthed several usage patterns and misconceptions of blind users that contribute to excessive battery drainage. Informed by the first user study, this paper proposes DarkReader, a screen reader developed in Android that bridges users' perception of power consumption to reality. DarkReader darkens the screen by truly turning it off, but allows users to interact with their smartphones. A second user study with 10 blind participants shows that participants perceived no difference in completion times in performing routine tasks using DarkReader and default screen reader. Yet DarkReader saves 24% to 52% power depending on tasks and screen brightness.

Observable reading behavior, the act of moving the eyes over lines of text, is highly stereotyped among the users of a language, and this has led to the development of reading detectors-methods that input windows of sequential fixations and output predictions of the fixation behavior during those windows being reading or skimming. The present study introduces a new method for reading detection using Region Ranking SVM (RRSVM). An SVM-based classifier learns the local oculomotor features that are important for real-time reading detection while it is optimizing for the global reading/skimming classification, making it unnecessary to hand-label local fixation windows for model training. This RRSVM reading detector was trained and evaluated using eye movement data collected in a laboratory context, where participants viewed modified web news articles and had to either read them carefully for comprehension or skim them quickly for the selection of keywords (separate groups). Ground truth labels were known at the global level (the instructed reading or skimming task), and obtained at the local level in a separate rating task. The RRSVM reading detector accurately predicted 82.5% of the global (article-level) reading/skimming behavior, with accuracy in predicting local window labels ranging from 72-95%, depending on how tuned the RRSVM was for local and global weights. With this RRSVM reading detector, a method now exists for near real-time reading detection without the need for hand-labeling of local fixation windows. With real-time reading detection capability comes the potential for applications ranging from education and training to intelligent interfaces that learn what a user is likely to know based on previous detection of their reading behavior.

This short paper presents a detailed empirical study of BBR's performance under different real-world and emulated testbeds across a range of network operating conditions. Our empirical results help to identify network conditions under which BBR outperforms, in terms of goodput, contemporary TCP congestion control algorithms. We find that BBR is well suited for networks with shallow buffers, despite its high retransmissions, whereas existing loss-based algorithms are better suited for deep buffers.

Given the growing significance of network performance, it is crucial to examine how to make the most of available network options and protocols. We propose ECON, a model that predicts performance of applications under different protocols and network conditions to scalably make better network choices. ECON is built on an analytical framework to predict TCP performance, and uses the TCP model as a building block for predicting application performance. ECON infers a relationship between loss and congestion using empirical data that drives an online model to predict TCP performance. ECON then builds on the TCP model to predict latency and HTTP performance. Across four wired and one wireless network, our model outperforms seven alternative TCP models. We demonstrate how ECON (i) can be used by a Web server application to choose between HTTP/1.1 and HTTP/2 for a given Web page and network condition, and (ii) can be used by a video application to choose the optimal bitrate that maximizes video quality without rebuffering.

Content delivery networks serve a major fraction of the Internet traffic, and their geographically deployed infrastructure makes them a good vantage point to observe traffic access patterns. We perform a large-scale investigation to characterize Web traffic patterns observed from a major CDN infrastructure. Specifically, we discover that responses with application/json content-type form a growing majority of all HTTP requests. As a result, we seek to understand what types of devices and applications are requesting JSON objects and explore opportunities to optimize CDN delivery of JSON traffic. Our study shows that mobile applications account for at least 52% of JSON traffic on the CDN and embedded devices account for another 12% of all JSON traffic. We also find that more than 55% of JSON traffic on the CDN is uncacheable, showing that a large portion of JSON traffic on the CDN is dynamic. By further looking at patterns of periodicity in requests, we find that 6.3% of JSON traffic is periodically requested and reflects the use of (partially) autonomous software systems, IoT devices, and other kinds of machine-to-machine communication. Finally, we explore dependencies in JSON traffic through the lens of ngram models and find that these models can capture patterns between subsequent requests. We can potentially leverage this to prefetch requests, improving the cache hit ratio.

Today there is no effective support for device-wide question answering on mobile devices. State-of-the-art QA models are deep learning behemoths designed for the cloud which run extremely slow and require more memory than available on phones. We present DeQA, a suite of latency- and memory- optimizations that adapts existing QA systems to run completely locally on mobile phones. Specifically, we design two latency optimizations that (1) stops processing documents if further processing cannot improve answer quality, and (2) identifies computation that does not depend on the question and moves it offline. These optimizations do not depend on the QA model internals and can be applied to several existing QA models. DeQA also implements a set of memory optimizations by (i) loading partial indexes in memory, (ii) working with smaller units of data, and (iii) replacing in-memory lookups with a key-value database. We use DeQA to port three state-of-the-art QA systems to the mobile device and evaluate over three datasets. The first is a large scale SQuAD dataset defined over Wikipedia collection. We also create two on-device QA datasets, one over a publicly available email data collection and the other using a cross-app data collection we obtain from two users. Our evaluations show that DeQA can run QA models with only a few hundred MBs of memory and provides at least 13x speedup on average on the mobile phone across all three datasets.% with less than a 1% drop in accuracy.

Smartphones are now a central technology in the daily lives of billions, but it relies on its battery to perform. Battery optimization is thereby a crucial design constraint in any mobile OS and device. However, even with new low-power methods, the ever-growing touchscreen remains the most power-hungry component. We propose an Ultra-Low-Power Mode (ULPM) for mobile devices that allows for touch interaction without visual feedback and exhibits significant power savings of up to 60% while allowing to complete interactive tasks. We demonstrate the effectiveness of the screenless ULPM in text-entry tasks, camera usage, and listening to videos, showing only a small decrease in usability for typical users.

Wearable devices, such as smart watches, offer exciting new opportunities for users to interact with their applications. The current state of the art for wearable devices is for a developer to write a custom {\em companion app}, which is a variant of the smartphone app, tailored to the wearable form factor. A developer puts a non-trivial amount of effort to write these companion apps and the programming model does not scale to an increasing diversity of form factors. In this demo, we show a working prototype of our system UIWear that allows a developer to easily extend a smartphone application to other wearable interfaces. Our system, UIWear, extracts the application GUI as a UI tree, which preserves the semantics of the GUI. The developer (or the user) only writes a {\em metaprogram} to encode the GUI design for the wearable device; no effort is needed beyond the design phase. UIWear executes the metaprogram by performing all the underlying tasks to virtualize the application GUI, adapt it, and recreate it on the wearable. A metaprogram can create the same functionality as existing companion apps with an order-of-magnitude less programming effort.

Wearable devices such as smartwatches offer exciting new opportunities for users to interact with their applications. However, the current wearable programming model requires the developer to write a custom companion app for each wearable form factor; the companion app extends the smartphone display onto the wearable, relays user interactions from the wearable to the phone, and updates the wearable display as needed. The development effort required to write a companion app is significant and will not scale to an increasing diversity of form factors. This paper argues for a different programming model for wearable devices. The developer writes an application for the smartphone, but only specifies a UI design for the wearable. Our UIWear system abstracts a logical model of the smartphone GUI, re-tailors the GUI for the wearable device based on the specified UI design, and compiles it into a companion app that we call the UICompanion app. We implemented UIWear on Android smartphones, AndroidWear smartwatches, and Sony SmartEyeGlasses. We evaluate 20 developer-written companion apps from the AndroidWear category on Google Play against the UIWear-created UICompanion apps. The lines-of-code required for the developer to specify the UI design in UIWear is an order-of-magnitude smaller compared to the companion app lines-of-code. Further, in most cases, the UICompanion app performed comparably or better than the corresponding companion app both in terms of qualitative metrics, including latency and energy, and quantitative metrics, including look-and-feel.

No abstract available.

Mobile page load times are an order of magnitude slower compared to non-mobile pages. It is not clear what causes the poor performance: the slower network, the slower computational speeds, or other reasons. Further, most Web optimizations are designed for non-mobile browsers and do not translate well to the mobile browser. Towards understanding mobile Web page load times, in this paper we: (1) perform an in-depth pairwise comparison of loading a page on a mobile versus a non-mobile browser, and (2) characterize the bottlenecks in the mobile browser {\em vis-a-vis} non-mobile browsers. To this end, we build a testbed that allows us to directly compare the low-level page load activities and bottlenecks when loading a page on a mobile versus a non-mobile browser. We find that computation is the main bottleneck when loading a page on mobile browsers. This is in contrast to non-mobile browsers where networking is the main bottleneck. We also find that the composition of the critical path during page load is different when loading pages on the mobile versus the non-mobile browser. A key takeaway of our work is that we need to fundamentally rethink optimizations for mobile browsers.

Always-on continuous sensing apps drain the battery quickly because they prevent the main processor from sleeping. Instead, sensor hub hardware, available in many smartphones today, can run continuous sensing at lower power while keeping the main processor idle. However, developers have to divide functionality between the main processor and the sensor hub. We implement MobileHub, a system that automatically rewrites applications to leverage the sensor hub without additional programming effort. MobileHub uses a combination of dynamic taint tracking and machine learning to learn when it is safe to leverage the sensor hub without affecting application semantics. We implement MobileHub in Android and prototype a sensor hub on a 8-bit AVR micro-controller. We experiment with 20 applications from Google Play. Our evaluation shows that MobileHub significantly reduces power consumption for continuous sensing apps.

Due to advances in low power sensors, energy harvesting, and disruption tolerant networking, we can now build mobile systems that operate perpetually, sensing and streaming data directly to scientists. However, factors such as energy harvesting variability and unpredictable network connectivity make building robust and perpetual systems difficult. In this paper, we present a system, Tula, that balances sensing with data delivery, to allow perpetual and robust operation across highly dynamic and mobile networks. This balance is especially important in unpredictable environments; sensing more data than can be delivered by the network is not useful, while gathering less underutilizes the system's potential. Tula is decentralized, fair and automatically adapts across different mobility patterns. We evaluate Tula using mobility and energy traces from TurtleNet-a mobile sensor network we deployed to study Gopher tortoises-and publicly available traces from the UMass DieselNet testbed. Our evaluations show that Tula senses and delivers data at up to 85 percent of an optimal, oracular system that perfectly replicates data and has foreknowledge of future energy harvesting. We also demonstrate that Tula can be implemented on a small microcontroller with modest code, memory, and processing requirements.

We present the design and evaluation of FindAll, a local search engine that lets users search and retrieve web pages, even in the absence of connectivity. Our user study with 23 users show that mobile users often search for web pages that they have previously visited, known as re-finding. This re-finding behavior makes the case for a local solution. FindAll goes beyond caching and using keyword search, and instead, implements a full blown search engine. The key challenge in FindAll is in designing a search engine, which is both memory- and energy-intensive, on the constrained phone environment. To this end, FindAll balances the cost of running the search engine with the expected benefits of serving a web page locally. FindAll estimates the benefits of local search, by learning the re-finding behavior of users. We implement FindAll on Android by adapting a publicly available search engine. Our evaluations, based on the traces collected from our user study, shows that FindAll reduces search latency by two-folds for users who re-find often, and reduces 3G data usage by up to 100 MB a month.

Our work is motivated by a simple question: can we design a simple routing protocol that ensures robust performance across networks with diverse connectivity characteristics such as meshes, MANETs, and DTNs? We identify packet replication as a key structural difference between protocols designed for opposite ends of the connectivity spectrum---DTNs and meshes. We develop a model to quantify under what conditions and by how much replication improves packet delays, and use these insights to drive the design of R3, a routing protocol that self-adapts replication to the extent of uncertainty in network path delays. We implement and deploy R3 on a mesh testbed and a DTN testbed. To the best of our knowledge, R3 is the first routing protocol to be deployed and evaluated on both a DTN testbed and a mesh testbed. We evaluate its performance through deployment, trace-driven simulations, and emulation experiments. Our results show that R3 achieves significantly better delay and goodput over existing protocols in a variety of network connectivity and load conditions.

We investigate if WiFi access can be used to augment 3G capacity in mobile environments. We rst conduct a detailed study of 3G and WiFi access from moving vehicles, in three different cities. We find that the average 3G and WiFi availability across the cities is 87% and 11%, respectively. WiFi throughput is lower than 3G through-put, and WiFi loss rates are higher. We then design a system, called Wiffler, to augments mobile 3G capacity. It uses two key ideas leveraging delay tolerance and fast switching -- to overcome the poor availability and performance of WiFi. For delay tolerant applications, Wiffler uses a simple model of the environment to predict WiFi connectivity. It uses these predictions to delays transfers to offload more data on WiFi, but only if delaying reduces 3G usage and the transfers can be completed within the application's tolerance threshold. For applications that are extremely sensitive to delay or loss (e.g., VoIP), Wiffler quickly switches to 3G if WiFi is unable to successfully transmit the packet within a small time window. We implement and deploy Wiffler in our vehicular testbed. Our experiments show that Wiffler significantly reduces 3G usage. For a realistic workload, the reduction is 45% for a delay tolerance of 60 seconds.

This paper presents MAUI, a system that enables fine-grained energy-aware offload of mobile code to the infrastructure. Previous approaches to these problems either relied heavily on programmer support to partition an application, or they were coarse-grained requiring full process (or full VM) migration. MAUI uses the benefits of a managed code environment to offer the best of both worlds: it supports fine-grained code offload to maximize energy savings with minimal burden on the programmer. MAUI decides at run-time which methods should be remotely executed, driven by an optimization engine that achieves the best energy savings possible under the mobile device's current connectivity constrains. In our evaluation, we show that MAUI enables: 1) a resource-intensive face recognition application that consumes an order of magnitude less energy, 2) a latency-sensitive arcade game application that doubles its refresh rate, and 3) a voice-based language translation application that bypasses the limitations of the smartphone environment by executing unsupported components remotely.

In this paper, we present a measurement study of the energy consumption characteristics of three widespread mobile networking technologies: 3G, GSM, and WiFi. We find that 3G and GSM incur a high tail energy overhead because of lingering in high power states after completing a transfer. Based on these measurements, we develop a model for the energy consumed by network activity for each technology.Using this model, we develop TailEnder, a protocol that reduces energy consumption of common mobile applications. For applications that can tolerate a small delay such as e-mail, TailEnder schedules transfers so as to minimize the cumulative energy consumed meeting user-specified deadlines. We show that the TailEnder scheduling algorithm is within a factor 2x of the optimal and show that any online algorithm can at best be within a factor 1.62x of the optimal. For applications like web search that can benefit from prefetching, TailEnder aggressively prefetches several times more data and improves user-specified response times while consuming less energy. We evaluate the benefits of TailEnder for three different case study applications - email, news feeds, and web search - based on real user logs and show significant reduction in energy consumption in each case. Experiments conducted on the mobile phone show that TailEnder can download 60% more news feed updates and download search results for more than 50% of web queries, compared to using the default policy.

Mobile Internet users have several options today including high bandwidth cellular data services such as 3G, that may be the choice for many. However, the ubiquity and low cost of WiFi suggests an attractive alternative, namely, opportunistic use of open WiFi access points (APs) or planned municipal mesh networks. Unfortunately, for vehicular users, the intermittent nature of WiFi connectivity makes it challenging to support popular interactive applications such as Web search and browsing.

We ask if the ubiquity of WiFi can be leveraged to provide cheap connectivity from moving vehicles for common applications such as Web browsing and VoIP. Driven by this question, we conduct a study of connection quality available to vehicular WiFi clients based on measurements from testbeds in two different cities. We find that current WiFi handoff methods, in which clients communicate with one basestation at a time, lead to frequent disruptions in connectivity. We also find that clients can overcome many disruptions by communicating with multiple basestations simultaneously. These findings lead us to develop ViFi, a protocol that opportunistically exploits basestation diversity to minimize disruptions and support interactive applications for mobile clients. ViFi uses a decentralized and lightweight probabilistic algorithm for coordination between participating basestations. Our evaluation using a two-month long deployment and trace-driven simulations shows that its link-layer performance comes close to an ideal diversity-based protocol. Using two applications, VoIP and short TCP transfers, we show that the link layer performance improvement translates to better application performance. In our deployment, ViFi doubles the number of successful short TCP transfers and doubles the length of disruption-free VoIP sessions compared to an existing WiFi-style handoff protocol.

Many DTN routing protocols use a variety of mechanisms, including discovering the meeting probabilities among nodes, packet replication, and network coding. The primary focus of these mechanisms is to increase the likelihood of finding a path with limited information, so these approaches have only an incidental effect on such routing metrics as maximum or average delivery latency. In this paper, we present RAPID, an intentional DTN routing protocol that can optimize a specific routing metric such as worst-case delivery latency or the fraction of packets that are delivered within a deadline. The key insight is to treat DTN routing as a resource allocation problem that translates the routing metric into per-packet utilities which determine how packets should be replicated in the system.

Wireless ad-hoc networks are vulnerable to various kinds of security threats and attacks due to relative ease of access to wireless medium and lack of a centralized infrastructure. In this paper, we seek to detect and mitigate the denial of service (DoS) attacks that prevent authorized users from gaining access to the networks. These attacks affect the service availability and connectivity of the wireless networks and hence reduce the network performance. To this end, we propose a novel cross-layer based intrusion detection system (CIDS) to identify the malicious node(s). Exploiting the information available across different layers of the protocol stack by triggering multiple levels of detection, enhances the accuracy of detection. We validate our design through simulations and also demonstrate lower occurrence of false positives

Designing a key management system is both important and challenging for wireless ad hoc networks. We have developed a secure, scalable, decentralized and robust key management solution using a hybrid (symmetric/asymmetric) key based methodology that is well suited for ad hoc networks. The nodes are grouped into clusters, and keys are distributed such that intra-cluster communication is secured using a symmetric cryptosystem and inter-cluster communication is secured using an asymmetric cryptosystem. We present a detailed analysis of the solution and simulation results. We observe that the hybrid solution provides a significant improvement in the performance of the key management solution in a highly hostile environment, and scales well to large networks.