Rajesh Palit

Energy-as-a-Service (EaaS): On the Efficacy of Multimedia Cloud Computing to Save Smartphone Energy

In spite of the dramatic growth in the number of smartphones in recent years, the challenge of limited energy capacity of these devices has not been solved satisfactorily. However, in the era of cloud computing, the limitation on energy capacity can be eased off in an efficient way by offloading heavy tasks to the cloud. It is important for smartphone and cloud computing developers to have insights into the energy cost of smartphone applications before implementing the offloading techniques. In this paper, we evaluate the energy cost of multimedia applications on smartphones that are connected to Multimedia Cloud Computing (MCC). We have conducted an extensive set of experiments to measure the energy costs to investigate whether or not smartphones save energy by using MCC services. In other words, we investigate the feasibility of MCC to provide the Energy-as-a-Service (EaaS). Specifically, we compared the energy costs for uploading and downloading a video file to and from MCC with the energy costs of encoding the same video file on a smartphone. The aforementioned comparison was performed by using HTTP and FTP Internet protocols with 3G and WiFi network interfaces. All the experiments were conducted on an Android based HTC Nexus One smartphone. Our results show that MCC provides the smartphones with many multimedia functionalities and saves smartphone energy from 30% to 70%.

Selection and execution of user level test cases for energy cost evaluation of smartphones

Smartphones are emerging as a preferred communication device of users. In this paper, we provide a methodology to select user level test cases for performing energy cost evaluation of smartphone applications. We define the concept of a user level test case for smartphones and show that, due to configuration settings, there exist millions of such test cases. Next, we discuss a test selection technique to reduce the number of test cases. We apply the technique to five different smartphones and evaluate their energy costs for running common network related applications. We have developed a test bench to execute those test cases for real applications on smartphones and measure their actual energy costs. This work provides a framework for researchers and developers to conduct experiments for measuring the energy cost of applications on smartphones.

Impact of packet aggregation on energy consumption in smartphones

With the tremendous growth in mobile applications, communication accounts for a significant portion of a smartphones total energy consumption. We studied the traffic pattern of smartphones and observed that a good portion of the packets are of small size and the generated traffic is bursty in nature. Motivated by these observations, we propose a Low Energy Data-packet Aggregation Scheme (LEDAS) in this paper. It accumulates a number of upper layer packets into a burst at medium access control (MAC) level, based on accumulation time, size, and number of packets. With this scheme, larger bursts lead to longer inactivity periods during which the communication module can be kept in doze mode. In addition, fewer MAC frames lead to less overheads and contentions in the wireless medium. However, the data packets incur delays due to the accumulation process. We have given a detail flowchart description of the technique. By means of analysis, we have derived expressions for the average values of burst size, burst inter-arrival times, and number of packets in a burst. We also evaluated the efficacy of the technique by simulations and showed the energy-delay trade-offs. Finally, we explained a test-bench to evaluate the energy saving potential of LEDAS on a smartphone.

Modeling the energy cost of applications on portable wireless devices

Extending the battery life of portable wireless devices has been in the focus of researchers for close to a decade. Several energy management techniques have been investigated at different levels of system design -- starting from silicon at the bottom to application design at the top, with communication protocols and operating system in between. In this paper, we present a model to estimate the energy cost of an application running on a portable wireless device. To develop the cost model, we partition a wireless device into two components, namely, computation and communication. Each component is modeled by a state-transition diagram. Two attributes are associated with each state: an average power cost and a state residence time. The cost of each state of the state-transition diagrams is validated by actual measurements. For a constant voltage supply, the average power cost of a state is denoted by the average current drawn by the component. The state residence times are estimated from the behavior of applications. The cost model has been validated by performing actual measurement of energy cost. We find that the estimated cost and the actual energy cost are within 5-10% of each other. This study will help us in improving the design of energy efficient software for portable devices. Moreover, the energy consumption breakdown into components will be an essential guide for future research in energy management of hardware and software systems.

A methodology for selecting experiments to measure energy costs in smartphones

Smartphones equipped with latest applications and features are the reality of the modern telecommunication world. Researchers have studied the energy consumption in smartphones while running some network related applications (NRAs) but they have neither fully covered the wide pool of NRAs nor provided a methodology to measure the energy consumption in smartphones. In this paper, we identified the most popular NRAs and configurable parameters which can impact the energy consumption while running these NRAs. We further propose a methodology to measure the energy consumption in smartphones while conducting a feasible set of experiments. We present a measurement bench for measuring the energy consumption in smartphones. We conducted selected experiments on latest smartphones (BB 9700, Nokia E71, HTC Nexus One and HTC HD2) to support our methodology. Our methodology evaluates the impact of configurable parameters and NRAs on energy consumption and provide a base to compare the energy consumption across smartphones.

Estimating the energy cost of communication on portable wireless devices

Software applications running on portable wireless devices communicate with the rest of the network over a wireless link. In these portable devices, the communication cost is a large fraction of the total energy consumption. The amount of energy consumed by the communication component of a portable device mostly depends on different parameters such as packet size and packet rate (or, bit rate). In this paper, we present the results of our investigation of the impacts of these communication parameters on energy consumption. First we build a simple analytic model to estimate the energy consumption due to receiving and transmitting data packets, and then we validate our model by conducting experiments. Results show that the analytical model is effective and gives accurate results. By varying data packet lengths, a communication device consumes different levels of energy to achieve the same data rate. When the packet size is very small compared to the maximum transmission unit (MTU), the device consumes more energy. However, large packets do not necessarily save energy. They rather add some other types of overheads, such as segmentation, recombination, and packet drop. Thus, for a given set of network parameters, an application can choose a suitable data packet length to minimize energy consumption. We also present the impact of data rate and packet delays on energy consumption. These results help us in understanding the energy consumption behavior of a communication device. They also facilitate us in optimizing the energy cost while designing a wireless application.

Enhancing the capability and energy efficiency of smartphones using WPAN

Smartphones are constrained in battery-energy and resources such as processing power, storage, and Internet bandwidth. The capabilities of the smartphones can be enhanced if they can access the hardware, software resources and data on a laptop in an energy efficient way. This enables the owner of a smartphone and a laptop to use his/her devices with flexibility. With this view, we propose Universal Computing and Communication Interface (UCCI) to facilitate such sharing of resources between two wireless portable devices using wireless personal area network (WPAN). We developed prototypes of UCCI using BlackBerry 9700 and HTC Nexus One smartphones, and performed experiments on an energy measurement testbed. We show the efficacy of the proposed model through experiments, and evaluated the energy costs of an application which transfers file through 3G, WiFi and Bluetooth links. We also measured the costs of file compression and decompression on a HTC Nexus One smartphone, and discuss how they affect the file transfer costs.

Energy-Aware Co-Operative (ECO) Relay-Based Packet Transmission in Wireless Networks

In infrastructure wireless networks, nodes at the edge of the coverage area need to spend more energy to transmit their packets than those close to the Access Point (AP). Less energy is required if intermediate nodes can be used to forward data. To enable this, intermediate nodes need an incentive and there must be a mechanism for selecting these nodes. In this paper we propose Energy-aware Co-Operative (ECO) relaying for selecting relays to forward packets. The technique is based on the idea of Relative Energy Usage (REU), which reflects the proportion of energy that a node saves by forwarding its packets through relays. A node which saves more energy by using relays is more likely to be chosen as a relay. Conversely, nodes are only permitted to use relays proportionate to the amount of energy they themselves have spent as relays. We compare our scheme with direct transmission, minimum energy path (MnEP), and maximum residual energy path (MxRE). We show that ECO can transmit 50% more data than direct transmission, while using less energy on average. Although MnEP and MxRE can also transmit more data than direct transmission, they do so at severe energy cost to a small number of nodes, doubling the average energy usage, making them ill-suited to commercial networks.