The energy consumed by video streaming includes the energy consumed for data transmission and CPU processing, which are both affected by the CPU frequency. High CPU frequency can reduce the data transmission time but it consumes more CPU energy. Low CPU frequency reduces the CPU energy but increases the data transmission time and then increases the energy consumption. In this paper, we aim to reduce the total energy of mobile video streaming by adaptively adjusting the CPU frequency. Based on real measurement results, we model the effects of CPU frequency on TCP throughput and system power. Based on these models, we propose an Energy-aware CPU Frequency Scaling (EFS) algorithm which selects the CPU frequency that can achieve a balance between saving the data transmission energy and CPU energy. Since the downloading schedule of existing video streaming apps is not optimized in terms of energy, we also propose a method to determine when and how much data to download. Through trace-driven simulations and real measurement, we demonstrate that the EFS algorithm can reduce 30 percent of energy for the Youtube app, and the combination of our download method and EFS algorithm can save 50 percent of energy than the default Youtube app.

Wearable devices such as smartwatches do not have enough power and computation capability to process computationally intensive tasks. One viable solution is to offload these tasks to the connected smartphone. Existing Android smartphones allocate CPU resources to a task according to its performance requirement, which is determined by the context of the task. However, due to lack of context information, smartphones cannot properly allocate resources to tasks offloaded from wearable devices. Allocating too few resources to urgent tasks (related to user interaction) may cause high interaction latency on wearable devices, while allocating too many resources to unimportant tasks (unrelated to user interaction) may lead to energy waste on the smartphone. To solve this problem, we propose a context-aware task offloading (CATO) framework, in which offloaded tasks can be properly executed on the smartphone or further offloaded to the cloud based on their context, aiming to achieve a balance between good user experience on wearable devices and energy saving on the smartphone. To validate our design, we have implemented CATO on the Android platform and developed two applications on top of it. Experimental results show that CATO can significantly reduce latency for urgent tasks and save energy for other unimportant tasks.

The energy consumed by video streaming includes the energy consumed for data transmission and CPU processing, which are both affected by the CPU frequency. High CPU frequency can reduce the data transmission time but it consumes more CPU energy. Low CPU frequency reduces the CPU energy but increases the data transmission time and then increases the energy consumption. In this paper, we aim to reduce the total energy of mobile video streaming by adaptively adjusting the CPU frequency. Based on real measurement results, we model the effects of CPU frequency on TCP throughput and system power. Based on these models, we propose an Energy-aware CPU Frequency Scaling (EFS) algorithm which selects the CPU frequency that can achieve a balance between saving the data transmission energy and CPU energy. Since the downloading schedule of existing video streaming apps is not optimized in terms of energy, we also propose a method to determine when and how much data to download. Through trace-driven simulations and real measurement, we demonstrate that the EFS algorithm can reduce 30% of energy for the Youtube app, and the combination of our download method and EFS algorithm can save 50% of energy than the default Youtube app.

In cellular networks, due to many practical deployment issues, some areas have good wireless coverage while other areas may not. This results in significant throughput (service quality) difference between wireless carriers at some locations. We first analyze the factors that affect the service quality and then validate the existence of service quality difference between different carriers via extensive measurements. To deal with this problem, a mobile device (node) with low service quality can offload its data traffic to nearby nodes with better service quality through Device-to-Device interfaces, such as WiFi direct, to save energy and reduce delay. To achieve this goal, we propose a Quality-Aware Traffic Offloading (QATO) framework to offload network tasks to neighboring nodes with better service quality. QATO can identify neighbors with better service quality and motivate nodes to help each other using incentive schemes. To validate our design, we have implemented QATO on Android platform and have developed a web browser and a photo uploader on top of it. Experimental results show that QATO can significantly reduce energy and delay for both data downloading and uploading. Through trace-driven simulations, we also show that all users can benefit from data offloading in the long run.

Photo crowdsourcing with smartphone has attracted considerable attention recently due to the prevalence of smartphones and the rich information provided by photos. In scenarios such as disaster recovery or battlefield, where the cellular network is partly damaged or severely overloaded, Disruption Tolerant Networks (DTNs) become the best way to deliver the crowdsourced photos. Since the bandwidth and storage resources in DTN are very limited and not enough to deliver all the crowdsourced photos, it is important to prioritize more valuable photos to use the limited resources. In this paper, we design a resource-aware photo crowdsourcing framework in DTN, which uses photo metadata including the smartphone's location, orientation, and other built-in camera's parameters, to estimate the value of photos. We propose a photo selection algorithm to maximize the value of photos delivered to the command center considering bandwidth and storage constraints. Both prototype implementation and trace-driven simulations demonstrate the effectiveness of our design.

Photos obtained via crowdsourcing can be used in many critical applications. Due to the limitations of communication bandwidth, storage, and processing capability, it is a challenge to transfer the huge amount of crowdsourced photos. To address this problem, we propose a framework, calledSmartPhoto, to quantify the quality (utility) of crowdsourced photos based on the accessible geographical and geometrical information (calledmetadata) including the smartphone's orientation, position, and all related parameters of the built-in camera. From the metadata, we can infer where and how the photo is taken, and then only transmit the most useful photos. Four optimization problems regarding the tradeoffs between photo utility and resource constraints, namely Max-Utility, online Max-Utility, Min-Selection, and Min-Selection with$k$-coverage, are studied. Efficient algorithms are proposed and their performance bounds are theoretically proved. We have implemented SmartPhoto in a testbed using Android based smartphones, and proposed techniques to improve the accuracy of the collected metadata by reducing sensor reading errors and solving object occlusion issues. Results based on real implementations and extensive simulations demonstrate the effectiveness of the proposed algorithms.

This paper proposes SymDetector, a smartphone based application to unobtrusively detect the sound-related respiratory symptoms occurred in a user's daily life, including sneeze, cough, sniffle and throat clearing. SymDetector uses the built-in microphone on the smartphone to continuously monitor a user's acoustic data and uses multi-level processes to detect and classify the respiratory symptoms. Several practical issues are considered in developing SymDetector, such as users' privacy concerns about their acoustic data, resource constraints of the smartphone and different contexts of the smartphone. We have implemented SymDetector on Galaxy S3 and evaluated its performance in real experiments involving 16 users and 204 days. The experimental results show that SymDetector can detect these four types of respiratory symptoms with high accuracy under various conditions.

Computationally intensive applications may quickly drain mobile device batteries. One viable solution to address this problem utilizes computation offloading. The tradeoff is that computation offloading introduces additional communication, with a corresponding energy cost. Yet, previous research into computation offloading has failed to account for the special characteristics of cellular networks that impact mobile device energy consumption. In this paper, we aim to develop energy efficient computation offloading algorithms for cellular networks. We analyze the effects of the long tail problem on task offloading, formalize the computation offloading problem, and use Dijkstra's algorithm to find the optimal decision. Since this optimal solution relies on perfect knowledge of future tasks, we further propose an online algorithm for offloading. We have implemented this latter algorithm on Android-based smartphones. Both experimental results from this implementation and trace-driven simulation show that our algorithm can significantly reduce the energy of computation offloading in cellular networks.

Video streaming on smartphone consumes lots of energy. One common solution is to download and buffer future video data for playback so that the wireless interface can be turned off most of time and then save energy. However, this may waste energy and bandwidth if the user skips or quits before the end of the video. Using a small buffer can reduce the bandwidth wastage, but may consume more energy and introduce rebuffering delay. In this paper, we analyze the power consumption during video streaming considering user skip and early quit scenarios. We first propose an offline method to compute the minimum power consumption, and then introduce an online solution to save energy based on whether the user tends to watch video for a long time or tends to skip. We have implemented the online solution on Android based smartphones. Experimental results and trace-driven simulation results show that that our method can save energy while achieving a better tradeoff between delay and bandwidth compared to existing methods.

Smartphone based web browsing wastes a lot of power when downloading webpages due to the special characteristics of the wireless radio interface. In this paper, we identify these special characteristics, and address power consumption issues through two novel techniques. First, we reorganize the computation sequence of the web browser when loading a webpage, so that the web browser can first run the computations that will generate new data transmissions and retrieve these data from the web server. Then, the web browser can put the wireless radio interface into low power state, release the radio resource, and then run the remaining computations. Second, we introduce a practical data mining based method to predict the user reading time of webpages, based on which the smartphone can switch to low power state when the reading time is longer than a threshold. To demonstrate the effectiveness of our energy-aware approaches, we develop a testbed with Android phones on T-Mobile UMTS network. Experimental results show that our approach can reduce the power consumption of smartphone by more than 30 percent during web browsing. Moreover, our solution can further reduce the webpage loading time and increase the network capacity.

Cellular networks can provide pervasive data access for smartphones, but also consume lots of energy, because the cellular interface has to stay in high power state for a long time (called long tail problem) after a data transmission. In this paper, we propose to reduce the tail energy by aggregating the data traffic of multiple nodes using their P2P interfaces. This traffic aggregation problem is formalized as finding the best task schedule to minimize energy. We first propose an A* search algorithm, which can reduce the search space for finding the optimal schedule offline, and then introduce an online traffic aggregation algorithm. We have implemented the online traffic aggregation algorithm on Android smartphones, and have built a small testbed. Trace-driven simulations and Experimental results show that our traffic aggregation algorithm can significantly reduce the energy and delay.

Photos obtained via crowdsourcing can be used in many critical applications. Due to the limitations of communication bandwidth, storage and processing capability, it is a challenge to transfer the huge amount of crowdsourced photos. To address this problem, we propose a framework, called SmartPhoto, to quantify the quality (utility) of crowdsourced photos based on the accessible geographical and geometrical information (called metadata) including the smartphone's orientation, position and all related parameters of the built-in camera. From the metadata, we can infer where and how the photo is taken, and then only transmit the most useful photos. Three optimization problems regarding the tradeoffs between photo utility and resource constraints, namely the Max-Utility problem, the online Max-Utility problem and the Min-Selection problem, are studied. Efficient algorithms are proposed and their performance bounds are theoretically proved. We have implemented SmartPhoto in a testbed using Android based smartphones, and proposed techniques to improve the accuracy of the collected metadata by reducing sensor reading errors and solving object occlusion issues. Results based on real implementations and extensive simulations demonstrate the effectiveness of the proposed algorithms.

In cellular networks, due to practical deployment issues, some areas have good wireless coverage while others may not. This results in significant throughput (service quality) difference between wireless carriers at some locations. Through extensive measurements, we have validated the existence of such service quality difference. Then, through peer to peer interfaces such as WiFi direct, a mobile device (node) with low service quality can offload its data traffic to nodes with better service quality, to save energy and reduce delay. To achieve this goal, we propose a Quality-Aware Traffic Offloading (QATO) framework to offload network tasks to neighboring nodes with better service quality. QATO can identify neighbors with better service quality and provide incentive mechanisms to motivate nodes to help each other. To validate our design, we have implemented QATO on Android platform and have developed a web browser and a photo uploader on top of it. Experimental results show that QATO can significantly reduce energy and delay for both data downloading and uploading. Through trace-driven simulations, we also show that all users can benefit from data offloading in the long run.

Many practical problems in mobile social networks such as routing, community detection, and social behavior analysis, rely on accurate user contact detection. The frequently used method for detecting user contact is through Bluetooth on smartphones. However, Bluetooth scans consume lots of power. Although increasing the scan duty cycle can reduce the power consumption, it also reduces the accuracy of contact detection. In this paper, we address this problem based on the observation that user contact changes (i.e., starts and ends of user contacts) are mainly caused by user movement. Since most smartphones have accelerometers, we can use them to detect user movement with much less energy and then start Bluetooth scans to detect user contacts. By conducting experiments on smartphones, we discover three relationships between user movement and user contact changes. According to these relationships, we propose a Mobility-Assisted User Contact detection algorithm (MAUC), which triggers Bluetooth scans only when user movements have a high possibility to cause contact changes. Moreover, we propose energy-aware MAUC (E-MAUC) to further reduce energy consumption during Bluetooth discovery, while keeping the same detection accuracy as MAUC. Via trace driven simulations, we show that MAUC can reduce the number of Bluetooth scans by half while maintaining similar contact detection rates compared to existing algorithms, and E-MAUC can further reduce the energy consumption by 45% compared to MAUC.