We propose a new approach to scheduling virtual machines (VMs) on a provider CPU that is unique in that is based around the use of direct user input. In our system, a user’s VM is scheduled as a periodic real-time task. The user can instantaneously manipulate his VM’s schedule using a joystick. An on-screen display illustrates the current schedule’s cost and indicates when the user’s desired schedule is impossible due to the schedules of other VMs or resource constraints. We report on a user study of our prototype system that reveals that even a naive user is capable of using the interface to our system to find a schedule that balances cost and the comfort of his VM. Good schedules are user- and application-dependent to a large extent, illustrating the benefits of user involvement. 1

This paper presents the LOCUSTS framework. The aim of the LOCUSTS project is to enable easy use of cyber foraging techniques when developing for small, resource-constrained devices. Cyber foraging, construed as “living off the land”, enables resource poor devices to offload tasks to nearby computing machinery, thereby enabling the small devices to 1) save energy and time, 2) take on tasks that would normally not be possible on such small devices, and 3) co-operate to perform tasks. This paper is concerned with foraging for processing power, i.e. remote execution of tasks, and discusses how distribution and migration of tasks can be done in a highly mobile environment. The main contribution of LOCUSTS is the focus on highly mobile cyber foraging. Here highly mobile means two things: 1) that the mobile devices are physically moving through the environment, which calls for task migration, and 2) that this mobility moves the devices into unknown environments where they would still like to be able to perform cyber foraging, which calls for the use of mobile code. 1.

Mobile devices are increasingly being relied on for services that go beyond simple connectivity and require more complex processing. Fortunately, a mobile device encounters, possibly intermittently, many entities capable of lending it computational resources. At one extreme is the traditional cloud-computing context where a mobile device is connected to remote cloud resources maintained by a service provider with which it has an established relationship. In this paper we consider the other extreme, where a mobile device’s contacts are only with other mobile devices, where both the computation initiator and the remote computational resources are mobile, and where intermittent connectivity among these entities is the norm. We present the design and implementation of a system, Serendipity, that enables a mobile computation initiator to use remote computational resources available in other mobile systems in its environment to speedup computing and conserve energy. We propose a simple but powerful job structure that is suitable for such a system. Serendipity relies on the collaboration among mobile devices for task allocation and task progress monitoring functions. We develop algorithms that are designed to disseminate tasks among mobile devices by accounting for the specific properties of the available connectivity. We also undertake an extensive evaluation of our system, including experience with a prototype, that demonstrates Serendipity’s performance.

Adapting existing autonomous vehicle platforms to operate in new environments with additional capabilities through redesign and hardware augmentation can be expensive and risky; an alternative approach is to cooperatively utilize remote resources available from remote entities concurrently operating in the same environment. Polymorphic Control Systems research investigates highly dynamic control structures that can automatically reconfigure across vehicles to share remote sensors, actuators, and other resources available throughout the system of vehicles. In this paper, we investigate extending this framework to a smart space environment, allowing a ground vehicle without sufficient instrumentation or available onboard resources to navigate a complex indoor environment by coordinating with concurrently operating building control system agents. We develop a complex simulation of the environment, develop polymorphic control laws that restructure the control system over multiple entities during operation, and present a trajectory generation approach concurrently addresses topological optimization and dynamic control optimization of the collaborative control structure. Our results demonstrate viability of solving the two-part polymorphic control optimization problem utilizing pseudo-optimizing solvers that trade optimality for feasibility and real-time performance, and show that control polymorphism provides robust resource sharing between agents in a smart building infrastructure. I.

This paper presents the Locusts cyber foraging framework. Cyber foraging is the opportunistic use of computing resources available in the nearby environment, and using such resources thus fall into the category of distributed computing. Furthermore, for the resources to be used efficiently, parallel computing techniques must also be employed. Distributed and parallel computing are two concepts that are both notoriously known for being very hard for developers to grasp. Because of this one might think that techniques such as cyber foraging would have a hard time surviving outside of research environments. In this paper a framework is presented that has special focus on making cyber foraging accessible for all developers. 1

Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the NSF or Carnegie Mellon University. Keywords: Virtual machines, mobile computing, nomadic computing, pervasive computing, transient use, VirtualBox, VMM, performance, Diamond, OpenDiamond R ○ , discardbased This document describes three engineering contributions made to Diamond, a system for discard-based search, to improve its portability and maintainability, and add new functionality. First, core engineering work on Diamond’s RPC and content management subsystems improves the system’s maintainability. Secondly, a new mechanism supports “scoping ” a Diamond search through the use of external metadata sources. Scoping selects a subset of objects to perform content-based search on by executing a query on an external metadata source related to the data. After query execution, the scope is set for all subsequent searches performed by Diamond applications. The final contribution is Kimberley, a system that enables mobile application use by leveraging virtual machine technology. Kimberley separates application state

Mobile phones are becoming the convergent platform for personal sensing, computing, and communication. This paper attempts to exploit this convergence towards the problem of automatic image tagging. We envision TagSense, a mobile phone based collaborative system that senses the people, activity, and context in a picture, and merges them carefully to create tags on-the-fly. The main challenge pertains to discriminating phone users that are in the picture from those that are not. We deploy a prototype of TagSense on 8 Android phones, and demonstrate its effectiveness through 200 pictures, taken in various social settings. While research in face recognition continues to improve image tagging, TagSense is an attempt to embrace additional dimensions of sensing towards this end goal. Performance comparison with Apple iPhoto and Google Picasa shows that such an out-of-band approach is valuable, especially with increasing device density and greater sophistication in sensing/learning algorithms.

The availability of multiple sensors on mobile devices offers a significant new capability to enable rich user and context aware applications. Many of these applications run in the background to continuously sense user context. However, running these applications on mobile devices can impose a significant stress on the battery life, and the use of supplementary low-power processors has been proposed on mobile devices for continuous background activities. In this paper, we experimentally and analytically investigate the design considerations that arise in the efficient use of the low power processor and provide a thorough understanding of the problem space. We answer fundamental questions such as which segments of the application are most efficient to be hosted on the low power processor, and how to select an appropriate low power processor. We discuss our measurements, analysis, and results using multiple low power processors and existing phone platforms.

A technical limitation of today’s mobile phones is that they are designed for a narrow set of primary use cases. This is one of the reasons why the majority of mobile phones are disposed of or relegated to the drawer even though they are fully functional. In this paper we argue that future mobile devices should be designed to support use cases beyond their primary purpose. We motivate and explore the implications of two design principles that can help us move towards this vision – loose association and device composition. We discuss an assortment of mobile applications that are enabled by these principles and present our early experiences in applying these principles in the context of a scheduled sensing system deployed on used phones. 1.

Mobile devices are increasingly being relied on for tasks that go beyond simple connectivity and demand more complex processing. The primary approach in wide use today uses cloud computing resources to off-load the “heavy lifting ” to specially designated servers when they are well connected. In reality, a mobile device often encounters, albeit intermittently, many entities capable of lending computational resources. In this work-in-progress paper we first give an overview of this environment, which we call a Cirrus Cloud due to its intermittent connectivity feature, and explain how it provides a spectrum of computational contexts for remote computation in a mobile environment. An ultimately successful system will need to have the flexibility to handle intermittent connectivity and use a mix of options on that spectrum. We investigate two scenarios at the extremes of the spectrum: 1) a scenario where a mobile device experiences intermittent connectivity to a central cloud computing resource, and 2) a scenario where a mobile device off-loads computation to other mobile devices it might meet intermittently. We present preliminary designs, implementations, and evaluations of systems that enable a mobile application to use remote computational resources to speedup computing and conserve energy in these scenarios. The preliminary results show the effectiveness of our systems and demonstrate the potential of computing