Showing papers by "Daqing Zhang published in 2017"

RT-Fall: A Real-Time and Contactless Fall Detection System with Commodity WiFi Devices

Abstract: This paper presents the design and implementation of RT-Fall, a real-time, contactless, low-cost yet accurate indoor fall detection system using the commodity WiFi devices. RT-Fall exploits the phase and amplitude of the fine-grained Channel State Information (CSI) accessible in commodity WiFi devices, and for the first time fulfills the goal of segmenting and detecting the falls automatically in real-time, which allows users to perform daily activities naturally and continuously without wearing any devices on the body. This work makes two key technical contributions. First, we find that the CSI phase difference over two antennas is a more sensitive base signal than amplitude for activity recognition, which can enable very reliable segmentation of fall and fall-like activities. Second, we discover the sharp power profile decline pattern of the fall in the time-frequency domain and further exploit the insight for new feature extraction and accurate fall segmentation/detection. Experimental results in four indoor scenarios demonstrate that RT-fall consistently outperforms the state-of-the-art approach WiFall with 14 percent higher sensitivity and 10 percent higher specificity on average.

IndoTrack: Device-Free Indoor Human Tracking with Commodity Wi-Fi

Abstract: Indoor human tracking is fundamental to many real-world applications such as security surveillance, behavioral analysis, and elderly care. Previous solutions usually require dedicated device being carried by the human target, which is inconvenient or even infeasible in scenarios such as elderly care and break-ins. However, compared with device-based tracking, device-free tracking is particularly challenging because the much weaker reflection signals are employed for tracking. The problem becomes even more difficult with commodity Wi-Fi devices, which have limited number of antennas, small bandwidth size, and severe hardware noise. In this work, we propose IndoTrack, a device-free indoor human tracking system that utilizes only commodity Wi-Fi devices. IndoTrack is composed of two innovative methods: (1) Doppler-MUSIC is able to extract accurate Doppler velocity information from noisy Wi-Fi Channel State Information (CSI) samples; and (2) Doppler-AoA is able to determine the absolute trajectory of the target by jointly estimating target velocity and location via probabilistic co-modeling of spatial-temporal Doppler and AoA information. Extensive experiments demonstrate that IndoTrack can achieve a 35cm median error in human trajectory estimation, outperforming the state-of-the-art systems and provide accurate location and velocity information for indoor human mobility and behavioral analysis.

Device-Free WiFi Human Sensing: From Pattern-Based to Model-Based Approaches

Abstract: Recently, device-free WiFi CSI-based human behavior recognition has attracted a great amount of interest as it promises to provide a ubiquitous sensing solution by using the pervasive WiFi infrastructure. While most existing solutions are pattern-based, applying machine learning techniques, there is a recent trend of developing accurate models to reveal the underlining radio propagation properties and exploit models for fine-grained human behavior recognition. In this article, we first classify the existing work into two categories: pattern-based and model-based recognition solutions. Then we review and examine the two approaches together with their enabled applications. Finally, we show the favorable properties of model-based approaches by comparing them using human respiration detection as a case study, and argue that our proposed Fresnel zone model could be a generic one with great potential for device-free human sensing using fine-grained WiFi CSI.

Toward Centimeter-Scale Human Activity Sensing with Wi-Fi Signals

Abstract: By allowing centimeter-scale human activity sensing with Wi-Fi signals, the Fresnel zone model could revolutionize wireless sensing and Internet of Things applications. The web extra at www.youtube.com/watch?v=R_vR6O8706g demonstrates how the Fresnel zone model can be leveraged for device-free sensing of human activities such as respiration and walking.

crowddeliver : Planning City-Wide Package Delivery Paths Leveraging the Crowd of Taxis

Abstract: Despite the great demand on and attempts at package express shipping services, online retailers have not yet had a practical solution to make such services profitable. In this paper, we propose an economical approach to express package delivery, i.e., exploiting relays of taxis with passengers to help transport package collectively, without degrading the quality of passenger services. Specifically, we propose a two-phase framework called crowddeliver for the package delivery path planning. In the first phase, we mine the historical taxi trajectory data offline to identify the shortest package delivery paths with estimated travel time given any Origin–Destination pairs. Using the paths and travel time as the reference, in the second phase we develop an online adaptive taxi scheduling algorithm to find the near-optimal delivery paths iteratively upon real-time requests and direct the package routing accordingly. Finally, we evaluate the two-phase framework using the real-world data sets, which consist of a point of interest, a road network, and the large-scale trajectory data, respectively, that are generated by 7614 taxis in a month in the city of Hangzhou, China. Results show that over 85% of packages can be delivered within 8 hours, with around 4.2 relays of taxis on average.

Location Privacy-Preserving Task Allocation for Mobile Crowdsensing with Differential Geo-Obfuscation

Abstract: In traditional mobile crowdsensing applications, organizers need participants' precise locations for optimal task allocation, e.g., minimizing selected workers' travel distance to task locations. However, the exposure of their locations raises privacy concerns. Especially for those who are not eventually selected for any task, their location privacy is sacrificed in vain. Hence, in this paper, we propose a location privacy-preserving task allocation framework with geo-obfuscation to protect users' locations during task assignments. Specifically, we make participants obfuscate their reported locations under the guarantee of differential privacy, which can provide privacy protection regardless of adversaries' prior knowledge and without the involvement of any third-part entity. In order to achieve optimal task allocation with such differential geo-obfuscation, we formulate a mixed-integer non-linear programming problem to minimize the expected travel distance of the selected workers under the constraint of differential privacy. Evaluation results on both simulation and real-world user mobility traces show the effectiveness of our proposed framework. Particularly, our framework outperforms Laplace obfuscation, a state-of-the-art differential geo-obfuscation mechanism, by achieving 45% less average travel distance on the real-world data.

PSAllocator: Multi-Task Allocation for Participatory Sensing with Sensing Capability Constraints

Abstract: This paper proposes a novel multi-task allocation framework, named PSAllocator, for participatory sensing (PS). Different from previous single-task oriented approaches, which select an optimal set of users for each single task independently, PSAllocator attempts to coordinate the allocation of multiple tasks to maximize the overall system utility on a multi-task PS platform. Furthermore, PSAllocator takes the maximum number of sensing tasks allowed for each participant and the sensor availability of each mobile device into consideration. PSAllocator utilizes a two-phase offline multi-task allocation approach to achieve the near-optimal goal. First, it predicts the participants' connections to cell towers and locations based on historical data from the telecom operator; Then, it converts the multi-task allocation problem into the representation of a bipartite graph, and employs an iterative greedy process to optimize the task allocation. Extensive evaluations based on real-world mobility traces show that PSAllocator outperforms the baseline methods under various settings.

TaskMe: toward a dynamic and quality-enhanced incentive mechanism for mobile crowd sensing

Abstract: Incentive is crucial to the success of mobile crowd sensing (MCS) systems. Over the different manners of incentives, providing monetary rewards has been proved quite useful. However, existing monetary-based incentive studies (e.g., the reverse auction based methods) mainly encourage user participation, whereas sensing quality is often neglected. First, the budget setting is static and may not meet the sensing contexts or user anticipation. Second, they do not measure the quality of data contributed. Third, the design of most incentive schemes is quantity- or cost-focused and not quality-oriented. To address these issues, we propose a novel MCS incentive mechanism called TaskMe. An LBSN (location-based social network)-powered model is leveraged for dynamic budgeting and proper worker selection, and a combination of multi-facet quality measurements and a multi-payment-enhanced reverse auction scheme are used to improve sensing quality. Experiments on several user studies and the crawled dataset validate TaskMe's effectiveness.

Worker-Contributed Data Utility Measurement for Visual Crowdsensing Systems

Abstract: Visual crowdsensing is successfully applied in numerous application areas, yet little work has been done on measuring and improving the quality of worker contributed visual data. Rather than evaluating the visual quality based on traditional metrics such as resolution, we focus on data diversity, which is crucial for a broad stream of visual crowdsensing tasks. Two representative diversity-oriented task types are studied, namely static object imagery and evolving event photography. The former aims to collect multi-facet/aspect yet low redundant data about a stationary object, while the latter wants to detect and collect details of key scenes throughout an event. We link these quality needs with data utility and propose a unified visual crowdsensing framework called UtiPay. Data utility is characterized by the macro and micro diversity needs: at the macro level, the pyramid-tree approach is proposed for multi-attribute-based data grouping; at the micro level, we use several strategies for intra-group data selection and worker contribution measurement. To study the impact of our proposed utility measurement approaches, we propose two utility-enhanced payment schemes as incentive mechanisms: Uti and Uti-Bid . Experiments over several user studies with a total of 43 subjects validate the performance of UtiPay for measuring and enhancing the data quality of visual crowdsensing tasks.

SPACE-TA: Cost-Effective Task Allocation Exploiting Intradata and Interdata Correlations in Sparse Crowdsensing

Abstract: Data quality and budget are two primary concerns in urban-scale mobile crowdsensing. Traditional research on mobile crowdsensing mainly takes sensing coverage ratio as the data quality metric rather than the overall sensed data error in the target-sensing area. In this article, we propose to leverage spatiotemporal correlations among the sensed data in the target-sensing area to significantly reduce the number of sensing task assignments. In particular, we exploit both intradata correlations within the same type of sensed data and interdata correlations among different types of sensed data in the sensing task. We propose a novel crowdsensing task allocation framework called SPACE-TA (SPArse Cost-Effective Task Allocation), combining compressive sensing, statistical analysis, active learning, and transfer learning, to dynamically select a small set of subareas for sensing in each timeslot (cycle), while inferring the data of unsensed subareas under a probabilistic data quality guarantee. Evaluations on real-life temperature, humidity, air quality, and traffic monitoring datasets verify the effectiveness of SPACE-TA. In the temperature-monitoring task leveraging intradata correlations, SPACE-TA requires data from only 15.5% of the subareas while keeping the inference error below 0.25°C in 95% of the cycles, reducing the number of sensed subareas by 18.0% to 26.5% compared to baselines. When multiple tasks run simultaneously, for example, for temperature and humidity monitoring, SPACE-TA can further reduce ∼10% of the sensed subareas by exploiting interdata correlations.

Fine-Grained Urban Event Detection and Characterization Based on Tensor Cofactorization

Abstract: Understanding the irregular crowd movement and social activities caused by urban events such as city festivals and concerts can benefit event management and city planning. Although various urban data can be exploited to detect such irregularities, the crowd mobility data (e.g., bike trip records) are usually in a mixed state with several basic patterns (e.g., eating, working, and recreation), making it difficult to separate concurrent events happening in the same region. The social activity data (e.g., social network check-ins) are usually oversparse, hindering the fine-grained characterization of urban events. In this paper, we propose a tensor cofactorization-based data fusion framework for fine-grained urban event detection and characterization leveraging crowd mobility data and social activity data. First, we adopt a nonnegative tensor cofactorization approach to decompose the crowd mobility tensor into several basic patterns, with the help of the auxiliary social activity tensor. We then use a multivariate-outlier-detection-based method to identify irregularities from the decomposed basic patterns and aggregate them to detect and characterize the associated urban events. We evaluate the performance of our framework using real-world bike trip data and check-in data from New York City and Washington, DC, respectively. Results show that by fusing the two types of urban data, our method achieves fine-grained urban event detection and characterization in both cities and consistently outperforms the baselines.

Real-time and generic queue time estimation based on mobile crowdsensing

Abstract: People often have to queue for a busy service in many places around a city, and knowing the queue time can be helpful for making better activity plans to avoid long queues. Traditional solutions to the queue time monitoring are based on pre-deployed infrastructures, such as cameras and infrared sensors, which are costly and fail to deliver the queue time information to scattered citizens. This paper presents CrowdQTE, a mobile crowdsensing system, which utilizes the sensor-enhanced mobile devices and crowd human intelligence to monitor and provide real-time queue time information for various queuing scenarios. When people are waiting in a line, we utilize the accelerometer sensor data and ambient contexts to automatically detect the queueing behavior and calculate the queue time. When people are not waiting in a line, it estimates the queue time based on the information reported manually by participants. We evaluate the performance of the system with a two-week and 12-person deployment using commercially-available smartphones. The results demonstrate that CrowdQTE is effective in estimating queuing status.

AR-Alarm: An Adaptive and Robust Intrusion Detection System Leveraging CSI from Commodity Wi-Fi

Abstract: Device-free human intrusion detection holds great potential and multiple challenges for applications ranging from asset protection to elder care. In this paper, leveraging the fine-grained Channel State Information (CSI) in commodity WiFi devices, we design and implement an adaptive and robust human intrusion detection system, called AR-Alarm. By utilizing a robust feature and self-adaptive learning mechanism, AR-Alarm achieves real-time intrusion detection in different environments without calibration efforts. To further increase the system robustness, we propose a few novel methods to distinguish real human intrusion from object motion in daily life such as object dropping, curtain swinging and pets moving. As demonstrated in the experiments, AR-Alarm achieves a high detection rate and low false alarm rate.

ecoSense: Minimize Participants’ Total 3G Data Cost in Mobile Crowdsensing Using Opportunistic Relays

Abstract: In mobile crowdsensing (MCS), one of the participants’ main concerns is the cost for 3G data usage, which affects their willingness to participate in a crowdsensing task. In this paper, we present the design and implementation of an MCS data uploading mechanism—ecoSense—to help reduce additional 3G data cost incurred by the whole crowd of sensing participants. By considering the two most common real-life 3G price plans—unlimited data plan (UnDP) and pay as you go (PAYG), ecoSense partitions all the users into two groups corresponding to these two price plans at the beginning of each month, with the objective of minimizing the total refunding budget for all participants. The partitioning is based on predicting users’ mobility patterns and sensed data size. The ecoSense mechanism is designed inspired by the observation that during the data uploading cycles, UnDP users could opportunistically relay PAYG users’ data to the crowdsensing server without extra 3G cost, provided the two types of users are able to “meet” on a common local cost-free network (e.g., Bluetooth or WiFi direct). We conduct our experiments using both the Massachusetts Institute of Technology reality mining and the Small World In Motion (SWIM) simulation data sets. Evaluation results show that ecoSense could reduce total 3G data cost by up to $ {\sim }50$ %, when compared to the direct-assignment method that assigns each participant to UnDP or PAYG directly according to the size of her sensed data.

ScenicPlanner: planning scenic travel routes leveraging heterogeneous user-generated digital footprints

Abstract: To facilitate the travel preparation process to a city, a lot of work has been done to recommend a POI or a sequence of POIs automatically to satisfy users' needs. However, most of the existing work ignores the issue of planning the detailed travel routes between POIs, leaving the task to online map services or commercial GPS navigators. Such a service or navigator in terms of suggesting the shortest travel distance or time, which cannot meet the diverse requirements of users. For instance, in the case of traveling by driving for leisure purpose, the scenic view along the travel routes would be of great importance to users, and a good planning service should put the sceneries of the route in higher priority rather than the distance or time taken. To this end, in this paper, we propose a novel framework called ScenicPlanner for route recommendation, leveraging a combination of geotagged image and check-in digital footprints from locationbased social networks (LBSNs). First, we enrich the road network and assign a proper scenic view score to each road segment to model the scenic road network, by extracting relevant information from geo-tagged images and check-ins. Then, we apply heuristic algorithms to iteratively add road segment and determine the travelling order of added road segments with the objective of maximizing the total scenic view score while satisfying the user-specified constraints (i.e., origin, destination and the total travel distance). Finally, to validate the efficiency and effectiveness of the proposed framework, we conduct extensive experiments on three real-world data sets from the Bay Area in the city of San Francisco, which contain a road network crawled from OpenStreetMap, more than 31 000 geo-tagged images generated by 1 571 Flickr users in one year, and 110 214 check-ins left by 15 680 Foursquare users in six months.

MFDL: A Multicarrier Fresnel Penetration Model based Device-Free Localization System leveraging Commodity Wi-Fi Cards

Abstract: Device-free localization plays an important role in many ubiquitous applications. Among the different technologies proposed, Wi-Fi based technology using commercial devices has attracted much attention due to its low cost, ease of deployment, and high potential for accurate localization. Existing solutions use either fingerprints that require labor-intensive radio-map survey and updates, or models constructed from empirical studies with dense deployment of Wi-Fi transceivers. In this work, we explore the Fresnel Zone Theory in physics and propose a generic Fresnel Penetration Model (FPM), which reveals the linear relationship between specific Fresnel zones and multicarrier Fresnel phase difference, along with the Fresnel phase offset caused by static multipath environments. We validate FPM in both outdoor and complex indoor environments. Furthermore, we design a multicarrier FPM based device-free localization system (MFDL), which overcomes a number of practical challenges, particularly the Fresnel phase difference estimation and phase offset calibration in multipath-rich indoor environments. Extensive experimental results show that compared with the state-of-the-art work (LiFS), our MFDL system achieves better localization accuracy with much fewer number of Wi-Fi transceivers. Specifically, using only three transceivers, the median localization error of MFDL is as low as 45$cm$ in an outdoor environment of 36$m^2$, and 55$cm$ in indoor settings of 25$m^2$. Increasing the number of transceivers to four allows us to achieve 75$cm$ median localization error in a 72$m^2$ indoor area, compared with the 1.1$m$ median localization error achieved by LiFS using 11 transceivers in a 70$m^2$ area.

Near-Optimal Incentive Allocation for Piggyback Crowdsensing

Abstract: Piggyback crowdsensing (PCS) is a novel energy- efficient mobile crowdsensing paradigm that reduces the energy consumption of crowdsensing tasks by leveraging smartphone app opportunities (SAOs). This article, based on several fundamental assumptions of incentive payment for PCS task participation and spatial-temporal coverage assessment for collected sensor data, first proposes two alternating data collection goals. Goal 1 is maximizing overall spatial-temporal coverage under a predefined incentive budget constraint; goal 2 is minimizing total incentive payment while ensuring predefined spatial-temporal coverage for collected sensor data, all on top of the PCS task model. With all of the above assumptions, settings, and models, we introduce CrowdMind -- a generic incentive allocation framework for the two optimal data collection goals, on top of the PCS model. We evaluated CrowdMind extensively using a large-scale real-world SAO dataset for the two incentive allocation problems. The results demonstrate that compared to baseline algorithms, CrowdMind achieves better spatial-temporal coverage under the same incentive budget constraint, while costing less in total incentive payments and ensuring the same spatial-temporal coverage, under various coverage/incentive settings. Further, a short theoretical analysis is presented to analyze the performance of Crowd- Mind in terms of the optimization with total incentive cost and overall spatial-temporal coverage objectives/constraints.

Poster: FooDNet: Optimized On Demand Take-out Food Delivery using Spatial Crowdsourcing

Abstract: This paper builds a Food Delivery Network (FooDNet) that investigates the usage of urban taxis to support on demand take-out food delivery by leveraging spatial crowdsourcing. Unlike existing service sharing systems (e.g., ridesharing), the delivery of food in FooDNet is more time-sensitive and the optimization problem is more complex regarding high-efficiency, huge-number of delivery needs. In particular, we study the food delivery problem in association with the Opportunistic Online Takeout Ordering & Delivery service (O-OTOD). Specifically, the food is delivered incidentally by taxis when carrying passengers in the O-OTOD problem, and the optimization goal is to minimize the number of selected taxis to maintain a relative high incentive to the participated drivers. The two-stage method is proposed to solve the problem, consisting of the construction algorithm and the Large Neighborhood Search (LNS) algorithm. Preliminary experiments based on real-world taxi trajectory datasets verify that our proposed algorithms are effective and efficient.

Toward Context-Aware Mobile Social Networks

Abstract: CA-MSNs are more intelligent and user-friendly than conventional online or mobile social networks. We first classify CA-MSNs into four categories, and divide their life cycle into four phases: discovery, connection, interaction, and organization. We then introduce personal and community context, and discuss the corresponding taxonomy. Subsequently, we elaborate how such context can be leveraged to enhance each life cycle phase. We also present our practices on designing various CA-MSN applications. Finally, future research directions are identified to shed light on the next generation MSNs from the context awareness perspective.

Gesture-enabled remote control for healthcare

Abstract: In recent years, wearable sensor-based gesture recognition is proliferating in the field of healthcare. It could be used to enable remote control of medical devices, contactless navigation of X-ray display and Magnetic Resonance Imaging (MRI), and largely enhance patients' daily living capabilities. However, even though a few commercial or prototype devices are available for wearable gesture recognition, none of them provides a combination of (1) fully open API for various healthcare application development, (2) appropriate form factor for comfortable daily wear, and (3) affordable cost for large scale adoption. In addition, the existing gesture recognition algorithms are mainly designed for discrete gestures. Accurate recognition of continuous gestures is still a significant challenge, which prevents the wide usage of existing wearable gesture recognition technology. In this paper, we present Gemote, a smart wristband-based hardware/software platform for gesture recognition and remote control. Due to its affordability, small size, and comfortable profile, Gemote is an attractive option for mass consumption. Gemote provides full open API access for third party research and application development. In addition, it employs a novel continuous gesture segmentation and recognition algorithm, which accurately and automatically separates hand movements into segments, and merges adjacent segments if needed, so that each gesture only exists in one segment. Experiments with human subjects show that the recognition accuracy is 99.4% when users perform gestures discretely, and 94.6% when users perform gestures continuously.

Mining Spatial-temporal Correlation of Sensory Data for Estimating Traffic Volumes on Highways

Abstract: Sensory data are often of low quality, for example, data are incomplete, ambiguous, or indirect, which has become the bottleneck of many data-driven applications. Two kinds of data which are handled in the paper for estimating traffic volumes on highways are no exception. In particular, the traffic volume data obtained from the loop detectors are accurate but sparse, and the mobile signaling data for estimating relative traffic volumes are wide in coverage and low in cost, but they are indirect and inaccurate. Keeping the characteristics of data in mind, the paper proposes a data fusion approach named Polaris which extends compressive sensing to estimate traffic volumes on highways. The Polaris analyzes the sparsity of the traffic volumes reported by detectors, mines the spatial-temporal correlations between the two kinds of data, and then gives the computational steps in the light of compressive sensing. Experiments are conducted on the large-scale real signaling data and the loop detector data. The experimental results show that the Polaris has the lowest estimation errors in comparison with several other methods. The corresponding Polaris system has been built and deployed in Fujian Province, China. It can obtain real-time traffic volumes on the highways with full coverage at a very low cost.1

A model based decimeter-scale device-free localization system using COTS wi-fi devices

Abstract: Commodity Wi-Fi based device-free localization has attracted a great attention in recent years. Previous related work is either fingerprint-based or model-based. In this demo, we will demonstrate a generic Fresnel Penetration Model (FPM) based real-time device-free localization system called MFDL. Using only three to four commodity Wi-Fi devices, it can localize a metal plate reflector with 6cm median error in the open space and localize a moving person with 45cm median error in an outdoor space of 36m2 and 50-75cm median error in indoor environments with a size ranging from 25m2 to 72m2, outperforming the state-of-the-art device-free localization approaches in similar settings.

Ridesourcing Car Detection by Transfer Learning

Abstract: Ridesourcing platforms like Uber and Didi are getting more and more popular around the world. However, unauthorized ridesourcing activities taking advantages of the sharing economy can greatly impair the healthy development of this emerging industry. As the first step to regulate on-demand ride services and eliminate black market, we design a method to detect ridesourcing cars from a pool of cars based on their trajectories. Since licensed ridesourcing car traces are not openly available and may be completely missing in some cities due to legal issues, we turn to transferring knowledge from public transport open data, i.e, taxis and buses, to ridesourcing detection among ordinary vehicles. We propose a two-stage transfer learning framework. In Stage 1, we take taxi and bus data as input to learn a random forest (RF) classifier using trajectory features shared by taxis/buses and ridesourcing/other cars. Then, we use the RF to label all the candidate cars. In Stage 2, leveraging the subset of high confident labels from the previous stage as input, we further learn a convolutional neural network (CNN) classifier for ridesourcing detection, and iteratively refine RF and CNN, as well as the feature set, via a co-training process. Finally, we use the resulting ensemble of RF and CNN to identify the ridesourcing cars in the candidate pool. Experiments on real car, taxi and bus traces show that our transfer learning framework, with no need of a pre-labeled ridesourcing dataset, can achieve similar accuracy as the supervised learning methods.

Selective Traffic Offloading on the Fly: A Machine Learning Approach

Abstract: It has been well recognized that network transmission constitutes a large portion of smartphone energy consumption, mainly because of the tail energy caused by cellular network interface. Traffic offloading has been proposed to reduce energy by letting a smartphone offload network traffic to its neighbors in vicinity via low-power direct connections (e.g., WiFi Direct or Bluetooth). Our experiments conducted in a realistic environment reveal that energy efficiency cannot be improved or even deteriorates without a carefully designed offloading strategy. In this paper, we propose a selective traffic offloading scheme implemented as a smartphone middleware in a software-defined fashion, which consists of a packet classifier and a traffic scheduler. Using a light-weight machine learning approach exploiting unique smartphone context information, the packet classifieridentifies packets generated on the fly as offloadable or notwith substantially improved efficiency and feasibility on resource limited smartphones compared to traditional approaches. Both testbed and simulation based experiments are conducted and the results show that our proposal always attains the superior performance on a number of comparison metrics.