We consider the problem of optimally mapping an edge computing service that is modeled as a tree with multiple processing sub-tasks and data flows onto the underlying physical network. As new computing and data analytics applications require more complicated data processing structures, and different types of data (e.g., images, videos, and numbers) sensed at geographically distributed locations must be collected and processed to obtain a complex and comprehensive result, highly intelligent algorithms are needed to solve this challenging problem. In this paper, we propose a learning-based hierarchical service tree placement strategy that aims to optimize the net utility, defined as achieved utility minus network congestion. The key idea is to decouple a service tree into appropriate sub-trees each containing a single computing sub-task as well as associated data flows and to recursively leverage Q-learning to place each sub-tree while maintaining the dependencies of sub-tasks in the service tree structure. It enables a scalable solution for large networks with unknown arrival statistics and complex service structures. Numerical results show that our solution can significantly outperform baseline heuristics in online service tree placement.

Accurately determining the network energy consumption of each software principal when multiple ones are active is the key to mobile energy optimization. Tail energy accounting, which attributes tail energy to individual software principals, remains an open problem. Besides, tail energy has also become a major energy drain, especially in mobile ad modules that generate frequent, intermittent network traffics by on-demand ad downloading. In this paper, we propose a systematic framework for mobile ad prefetching and energy optimization, based on a novel tail energy accounting policy using cooperative game theory. In particular, we maximize the sum of deadline- and energy-aware ad utility, by jointly determining apps' aggressiveness in ad prefetching. The proposed tail energy accounting not only characterizes the energy profile of each app's ad module, a crucial input in energy optimization, but also enables an efficient solution by decoupling decision making of individual apps. The proposed framework is implemented on Android with negligible performance/network overhead. Using real-world apps and usage traces, we demonstrate a significant reduction in mobile network energy consumption by up to 45 percent compared with existing approaches. To the best of our knowledge, it is the first fully implemented ad management system transparent to apps and ad ecosystem.

Mobile Edge Computing has quickly become a promising paradigm to meet the ever-increasing data-processing demands imposed by emerging applications, by shifting computations to network edge. In this paper, we address two problems unique in edge computing: How to determine the execution cost contributed by each computing task in an edge environment, and how to distributively assign tasks to heterogeneous edge nodes to minimize total execution cost? Cost accounting is a long-standing hard problem for multiprocessing systems, e.g., when edge nodes jointly process tasks from different users. We propose a new framework that models the problem as a multichoice cooperative game, and use Shapley Value for cost accounting. The result enables us to decouple the cost of concurrent task executions and to efficiently solve the task assignment problem through a distributive Hungarian algorithm. To evaluate the performance, we conduct hybrid experiments by collecting trace from a fully implemented edge testbed and generating cost profiles to drive extensive simulations. Numerical results show that our solution guided by multichoice Shapley value is able to consistently outperform two baseline strategies using oblivious cost accounting policies, for task assignment in heterogeneous edge networks. Our solution's advantages become more significant for edge networks with higher levels of heterogeneity (either computation- or network-wise).

Identifying vulnerabilities in software systems is crucial to minimizing the damages that result from malicious exploits and software failures. This often requires proper identification of vulnerable execution paths that contain program vulnerabilities or bugs. However, with rapid rise in software complexity, it has become notoriously difficult to identify such vulnerable paths through exhaustively searching the entire program execution space. In this paper, we propose StatSym, a novel, automated Statistics-Guided Symbolic Execution framework that integrates the swiftness of statistical inference and the rigorousness of symbolic execution techniques to achieve precision, agility and scalability in vulnerable program path discovery. Our solution first leverages statistical analysis of program runtime information to construct predicates that are indicative of potential vulnerability in programs. These statistically identified paths, along with the associated predicates, effectively drive a symbolic execution engine to verify the presence of vulnerable paths and reduce their time to solution. We evaluate StatSym on four real-world applications including polymorph, CTree, Grep and thttpd that come from diverse domains. Results show that StatSym is able to assist the symbolic executor, KLEE, in identifying the vulnerable paths for all of the four cases, whereas pure symbolic execution fails in three out of four applications due to memory space overrun.

Mobile edge computing aims at improving application response time and energy efficiency by deploying data processing at the edge of the network. Due to the proliferation of Internet of Things and interactive applications, the ever-increasing demand for low latency calls for novel approaches to further pushing the envelope of mobile edge computing beyond existing task offloading and distributed processing mechanisms. In this paper, we identify a new tradeoff between Quality-of-Result (QoR) and service response time in mobile edge computing. Our key idea is motivated by the observation that a growing set of edge applications involving media processing, machine learning, and data mining can tolerate some level of quality loss in the computed result. By relaxing the need for highest QoR, significant improvement in service response time can be achieved. Toward this end, we present a novel optimization framework, MobiQoR, which minimizes service response time and app energy consumption by jointly optimizing the QoR of all edge nodes and the offloading strategy. The proposed MobiQoR is prototyped using Parse, an open source mobile back-end tool, on Android smartphones. Using representative applications including face recognition and movie recommendation, our evaluation with real-world datasets shows that MobiQoR reduces response time and energy consumption by up to 77% (in face recognition) and 189.3% (in movie recommendation) over existing strategies under the same level of QoR relaxation.

In cellular networks, tail states are designed for a tradeoff between energy efficiency and latency. However, the energy consumed during tail states becomes a huge energy drainer itself. Traditional energy saving techniques by content prefetching cannot be directly applied to mobile ads, due to the deadline requirements of ads, randomness in user behaviors, different usage patterns of mobile apps and system services. In this paper, considering several significant runtime factors, we make a novel use of Markov Decision Process to model the energy minimization problem for ad prefetching (EMAP), and propose an algorithmic solution to the EMAP problem. Further, we implement the first mobile ad prefetching system that is fully compatible with contemporary ad libraries and mobile apps. By replaying real-world user traces on Android devices, we show our proposed solution consistently outperforms existing On-Demand policy on Android by up to 59% in saving ad-related energy, while a simple Fill-Up-Buffer policy can be even 2 times worse than the default On-Demand policy. Such findings provide critical insights regarding the promise of saving energy by ad prefetching in real-world mobile systems.

This paper presents a semantics-aware rule recommendation and enforcement (SARRE) system for taming information leakage on Android. SARRE leverages statistical analysis and a novel application of minimum path cover algorithm to identify system event paths from dynamic runtime monitoring. Then, an online recommendation system is developed to automatically assign a fine-grained security rule to each event path, capitalizing on both known security rules and application semantic information. The proposed SARRE system is prototyped on Android devices and evaluated using real-world malware samples and popular apps from Google Play spanning multiple categories. Our results show that SARRE achieves 93.8% precision and 96.4% recall in identifying the event paths, compared with tainting technique. Also, the average difference between rule recommendation and manual configuration is less than 5%, validating the effectiveness of the automatic rule recommendation. It is also demonstrated that by enforcing the recommended security rules through a camouflage engine, SARRE can effectively prevent information leakage and enable fine-grained protection over private data with very small performance overhead.