Federated learning (FL) enables collaborative model training over distributed computing nodes without sharing their privacy-sensitive raw data. However, in FL, iterative exchanges of model updates between distributed nodes and the cloud server can result in significant communication cost, especially when the data distributions at distributed nodes are imbalanced with requiring more rounds of iterations. In this paper, with our in-depth empirical studies, we disclose that extensive cloud aggregations can be avoided without compromising the learning accuracy if frequent aggregations can be enabled at edge network. To this end, we shed light on the hierarchical federated learning (HFL) framework, where a subset of distributed nodes can play as edge aggregators to support edge aggregations. Under the HFL framework, we formulate a communication cost minimization (CCM) problem to minimize the total communication cost required for model learning with a target accuracy by making decisions on edge aggragator selection and node-edge associations. Inspired by our data-driven insights that the potential of HFL lies in the data distribution at edge aggregators, we propose ShapeFL, i.e., SHaping dAta distRibution at Edge, to transform and solve the CCM problem. In ShapeFL, we divide the original problem into two sub-problems to minimize the per-round communication cost and maximize the data distribution diversity of edge aggregator data, respectively, and devise two light-weight algorithms to solve them accordingly. Extensive experiments are carried out based on several opened datasets and real-world network topologies, and the results demonstrate the efficacy of ShapeFL in terms of both learning accuracy and communication efficiency.

The rapid proliferation of some real-time applications (e.g., video surveillance) has driven enormous interest in maximizing information freshness, quantified by the age of information (AoI). For some computation-intensive updates such as images or videos, the real-time update processing requires intensive resources, which edge servers can provide in mobile edge computing (MEC). In this paper, we investigate information updating scheduling with multiple users in MEC. Due to the centralized algorithms’ limitations in distributed systems where users are self-interested, we investigate an efficient distributed scheduling algorithm. We model the information updating scheduling as an uncooperative game and propose a distributed algorithm to compute the unique Nash equilibrium. Considering the unavailability of some global network information, we propose a learning algorithm where each user learns how to make decisions based on observable information in a distributed manner. Extensive evaluation results show the efficiency of the proposed algorithms.

Today’s intelligent services are built on well-trained deep neural network (DNN) models, which usually require large private datasets along with a high cost for model training. It consequently makes the model providers cherish the pre-trained DNN models and only distribute them to authorized users. However, malicious users can steal these valuable models for abuse, illegal copy and redistribution. Attackers can also extract private features from even authorized models to leak partial training datasets. They both violate privacy. Existing techniques from secure community attempt to avoid parameter leakage during model authorization but yet cannot solve privacy issues sufficiently. In this paper, we propose a privacy-preserving model authorization approach, AgAuth, to resist the aforementioned privacy threats. We devise a novel scheme called Information-Agnostic Conversion (IAC) for forwarding procedure to eliminate residual features in model parameters. Based on it, we then propose Inference-on-Ciphertext (CiFer) mechanism for DNN reasoning, which includes three stages in each forwarding. The Encrypt phase first converts the proprietary model parameters to demonstrate uniform distribution. The Forward stage per-forms forwarding function without decryption at authorized side. Specifically, this stage just computes over ciphertext. The Decrypt phase finally recovers the information-agnostic outputs to informative output tensor for real-world services. In addition, we implement a prototype and conduct extensive experiments to evaluate its performance. The qualitative and quantitative results demonstrate that our solution AgAuth is privacy-preserving to defend against model theft and feature leakage, without accuracy loss or notable performance decrease.

Time-efficient artificial intelligence (AI) service has recently witnessed increasing interest from academia and industry due to the urgent needs in massive smart applications such as self-driving cars, virtual reality, high-resolution video streaming, etc. Existing solutions to reduce AI latency, like edge computing and heterogeneous neural-network accelerators (NNAs), face high risk of privacy leakage. To achieve both low-latency and privacy-preserving purposes on edge servers (e.g., NNAs), this paper proposes ENIGMA that can exploit the trusted execution environment (TEE) and heterogeneous NNAs of edge servers for edge inference. The low-latency is supported by a new ahead-of-time analysis framework for analyzing the linearity of multilayer neural networks, which automatically slices forward-graph and assigns sub-graphs to TEE or NNA. To avoid privacy leakage issue, we then introduce a pre-forwarded cipher generation (PFCG) scheme for computing linear sub-forward-graphs on NNA. The input data is encrypted to ciphertext that can be computed directly by linear sub-graphs, and the output can be decrypted to obtain the correct output. To enable non-linear computation of sub-graphs on TEE, we use ring-cache and automatic vectorization optimization to address the memory limitation of TEE. Qualitative analysis and quantitative experiments on GPU, NPU and TPU demonstrate that ENIGMA is not only compatible with heterogeneous NNAs, but also can avoid leakages of private features with latency as low as 50-milliseconds.

With the increase of depth and width, Deep Neural Network has got the best results in the computer vision, but its massive calculation has brought a heavy burden to IOT devices. To speed up the inference of DNN models, Branchynet creatively puts forward the early exit, which means that samples exit from shallow layers to reduce the calculation amount of the model. But Branchynet has some unnecessary intermediate calculations in the inference process. We propose a dispense mode to solve this problem, which can optimize the accuracy and latency of BranchyNet at the same time. The dispense mode directly determines the exit position of the sample in the multi-branch network according to the difficulty of the sample without intermediate trial errors. Under the same accuracy requirements, the inference speed is improved by 30%-50% compared with the cascade mode of Branchynet. Moreover, while further reducing redundant calculation, it provides a method for dynamic adjustment of accuracy. Thus, our framework can easily adjust the accuracy of the model to meet higher throughputs.

Motivated by the popularity of edge intelligence, DNN services have been widely deployed at the edge, posing significant performance pressure on edge servers. How to improve the QoS of edge DNN services becomes a crucial and challenging problem. Previous works, however, did not fully consider the heterogeneous QoS requirements on urgent and non-urgent tasks, causing frequent QoS violations. Meanwhile, our empirical study shows that severe task interference exists in concurrent DNN tasks, further degrading the timeliness of urgent tasks and throughput of non-urgent tasks. To address these issues, we propose Kalmia, a heterogeneous QoS-aware framework for DNN inference task scheduling on edge servers. Specifically, Kalmia includes an offline profiling stage and an online scheduling policy. In offline profiling, we build a regression model to predict the execution time of tasks. During online scheduling, we classify the tasks into urgent and non-urgent tasks and distribute them into two CUDA contexts. By a tailored scheduling strategy, non-urgent tasks can fully utilize the computing resources for throughput improvement, while the timeliness of urgent tasks can be guaranteed via preemption. Experimental results demonstrate that Kalmia can achieve up to 2.8× improvement in throughput and significantly reduce the deadline violation rate compared with state-of-the-art methods.

Age of information (AoI) has been proposed to quantify data freshness. In some real-time applications such as surveillance systems, real-time analytics on source updates requires intensive computation resources and incurs high energy consumption. By providing computing resources at the network edge, mobile edge computing (MEC) can reduce update processing time and improve data freshness. In this paper, we investigate the age-optimal computation-intensive update scheduling for multiple sources in MEC-enabled IoT networks. Since the centralized algorithms may not apply due to the high computational complexity, we design an efficient decentralized scheduling mechanism for self-organized IoT networks. We provide a game-theoretic analysis and prove the existence of pure strategy Nash equilibrium. An efficient and decentralized algorithm based on the best-response dynamics is proposed to compute the equilibrium. We also provide the approximation ratio of the proposed algorithm. In particular, for the homogeneous source model, we show that the approximation ratio is at most 2.5. Extensive evaluation results show that the proposed decentralized algorithm is computationally efficient and closely approximates the centralized optimum in various settings.

Despite the significant development of mobile device SoCs, they are still inefficient in computing computation-intensive workloads, such as high-resolution image processing and AR/VR applications. Offloading offers a promising way to leverage cloud or edge servers for acceleration, but existing offloading is limited to specific tasks or specific hardware/software platforms, resulting in significant engineering overhead. To address this problem, we focus on the underlying layer of these applications (i.e., OpenCL) and propose Hyperion, a generic and distributed mobile offloading framework built on OpenCL. To achieve high-performance distributed execution for Hyperion, we first take a deep insight into the OpenCL data structures and design regularity-aware kernel analyzer to analyze the data dependency of work-groups and identify the essential data to offload. Then, context-aware execution time predictor is proposed to estimate the computing time of a given partitioned kernel workload that is highly impacted by many runtime factors. These techniques are integrated into pipeline-enabled and network-adaptive scheduler to make scheduling decisions, which coordinates the kernel partition and workload scheduling to form pipeline processing between data transmission and distributed execution with flexible adaptability to network dynamics. Extensive experimental results demonstrate that Hyperion achieves superior performance with an average 3.80× speedup compared with the best baseline and flexible adaptation to dynamic network conditions and available computing resources.

User-initiated mobile data trading, where mobile devices trade their mobile data quota via personal hotspots, is a promising approach to improve resource utilization. Most existing works only consider the data size, while ignoring the data rate and temporal requirements. To fill this void, we propose a novel data trading marketplace, where mobile users trade Internet access continuously for a time period with a specific data rate with neighboring mobile devices. Each request is characterized by an arrival time, departure time, the demanded data rate, and a value for getting services. To achieve the most system efficiency, we formulate an integer linear programming problem to maximize the total social welfare, which takes the data rate and temporal requirements into account. We next consider two request models: 1) a homogeneous request model and 2) a heterogeneous request model. In the homogeneous request model, all the requests demand the overall system lifetime, and we propose a computationally efficient auction that makes allocation decisions for all the requests simultaneously. In the heterogeneous request model, all the requests require different Internet access periods and dynamic arrive. Upon requests’ arrival, the system must make real-time allocations without the availability of future information. To jointly deal with requesters’ multidimensional private information (i.e., the arrival/departure time, demanded data rate, and the value), and the uncertainty about future arrival requests, we propose a multi-round online auction. Theoretical analysis shows that both the auctions satisfy the desired properties, including individual rationality, truthfulness, and computational efficiency. Simulation results show the efficiency of the proposed auctions.

Device-to-device (D2D) content sharing holds great promise to alleviate the growing strain on cellular networks, as it offloads popular content data onto direct peer-to-peer links. However, it is still unexplored how content sharing could benefit from utilizing multihop rather than the conventional single-hop D2D communications. As a step towards this end, this paper proposes a game theoretic approach to enable D2D content sharing with multihop communication capabilities. Given a subset of participants, a Nash bargaining game is modeled to provide the routing and pricing graphs, where a novel incentive mechanism is adopted to stimulate cooperation. By iteratively evaluating the solution of the Nash bargaining subgame, participants that include content sources and transmission relays are determined, which ensures that all participants make contributions to the content sharing process. An additional procedure termed pricing plan is introduced to make sure that the final pricing graph is practical and feasible in terms of D2D communication. Experimental results are presented to demonstrate that the proposed game theoretic approach could not only jointly deal with the participants selection, routing, and pricing problems in multihop D2D content sharing, but also effectively restrict utilities and transmission resources to only contributive participants.

Federated learning enables distributed model training over various computing nodes, e.g., mobile devices, where instead of sharing raw user data, computing nodes can solely commit model updates without compromising data privacy. The quality of federated learning relies on the model updates contributed by computing nodes training with their local data. However, with various factors (e.g., training data size, mislabeled data samples, skewed data distributions), the model update qualities of computing nodes can vary dramatically, while inclusively aggregating low-quality model updates can deteriorate the global model quality. To achieve efficient federated learning, in this paper, we propose a novel framework namedFAIR, i.e.,Federated leArning with qualIty awaReness. Particularly,FAIRintegrates three major components: 1) learning quality estimation: we adopt the model aggregation weight (learned in the third component) to reversely quantify the individual learning quality of nodes in a privacy-preserving manner, and leverage the historical learning records to infer the next-round learning quality; 2) quality-aware incentive mechanism: within the recruiting budget, we model a reverse auction problem to stimulate the participation of high-quality and low-cost computing nodes, and the method is proved to be truthful, individually rational, and computationally efficient; and 3) auto-weighted model aggregation: based on the gradient descent method, we devise an auto-weighted model aggregation algorithm to automatically learn the optimal aggregation weights to further enhance the global model quality. Based on real-world datasets and learning tasks, extensive experiments are conducted to demonstrate the efficacy ofFAIR.

The emergency of federated learning (FL) enables distributed data owners to collaboratively build a global model without sharing their raw data, which creates a new business chance for building data market. However, in practical FL scenarios, the hardware conditions and data resources of the participant clients can vary significantly, leading to different positive/negative effects on the FL performance, where the client selection problem becomes crucial. To this end, we proposeAUCTION, anAutomated and qUality-awareClient selecTIONframework for efficient FL, which can evaluate the learning quality of clients and select them automatically with quality-awareness for a given FL task within a limited budget. To designAUCTION, multiple factors such as data size, data quality, and learning budget that can affect the learning performance should be properly balanced. It is nontrivial since their impacts on the FL model are intricate and unquantifiable. Therefore,AUCTIONis designed to encode the client selection policy into a neural network and employ reinforcement learning to automatically learn client selection policies based on the observed client status and feedback rewards quantified by the federated learning performance. In particular, the policy network is built upon an encoder-decoder deep neural network with an attention mechanism, which can adapt to dynamic changes of the number of candidate clients and make sequential client selection actions to reduce the learning space significantly. Extensive experiments are carried out based on real-world datasets and well-known learning models to demonstrate the efficiency, robustness, and scalability ofAUCTION.

Deep learning plays an increasingly important role in human life. However, resource-constrained IoT devices are still inefficient in performing deep neural network (DNN) inference. Existing works have attempted to improve the performance by leveraging edge computing that partitions DNN and offloads a part of workloads to the edge server. However, most of them focus on scheduling workloads among different devices, ignoring the network costs. Thus, user experience easily suffers from inferior conditions such as network congestion. To address this issue, we jointly consider network conditions and computing capabilities of the IoT device and edge server, and then propose FastCoDNN, a co-execution framework that enables high-performance DNN inference. Specifically, under inferior conditions, we conduct redundant calculation instead of data synchronization to reduce network costs. Then, we orchestrate redundant calculations and data synchronizations among the local and edge, and flexibly adjust them according to new network conditions. We propose a new algorithm based on dynamic programming to achieve this adjustment. Experimental results show that FastCoDNN achieves fewer network costs and much performance improvement compared with existing methods.

Federated learning (FL) can enable distributed model training over mobile nodes without sharing privacy-sensitive raw data. However, to achieve efficient FL, one significant challenge is the prohibitive communication overhead to commit model updates since frequent cloud model aggregations are usually required to reach a target accuracy, especially when the data distributions at mobile nodes are imbalanced. With pilot experiments, it is verified that frequent cloud model aggregations can be avoided without performance degradation if model aggregations can be conducted at edge. To this end, we shed light on the hierarchical federated learning (HFL) framework, where a subset of distributed nodes are selected as edge aggregators to conduct edge aggregations. Particularly, under the HFL framework, we formulate a communication cost minimization (CCM) problem to minimize the communication cost raised by edge/cloud aggregations with making decisions on edge aggregator selection and distributed node association. Inspired by the insight that the potential of HFL lies in the data distribution at edge aggregators, we propose SHARE, i.e., SHaping dAta distRibution at Edge, to transform and solve the CCM problem. In SHARE, we divide the original problem into two sub-problems to minimize the per-round communication cost and mean Kullback-Leibler divergence of edge aggregator data, and devise two light-weight algorithms to solve them, respectively. Extensive experiments under various settings are carried out to corroborate the efficacy of SHARE.

Federated learning enables distributed learning in a privacy-protected manner, but two challenging reasons can affect learning performance significantly. First, mobile users are not willing to participate in learning due to computation and energy consumption. Second, with various factors (e.g., training data size/quality), the model update quality of mobile devices can vary dramatically, inclusively aggregating low-quality model updates can deteriorate the global model quality. In this paper, we propose a novel system named FAIR, i.e., Federated leArning with qualIty awaReness. FAIR integrates three major components: 1) learning quality estimation: we leverage historical learning records to estimate the user learning quality, where the record freshness is considered and the exponential forgetting function is utilized for weight assignment; 2) quality-aware incentive mechanism: within the recruiting budget, we model a reverse auction problem to encourage the participation of high-quality learning users, and the method is proved to be truthful, individually rational, and computationally efficient; and 3) model aggregation: we devise an aggregation algorithm that integrates the model quality into aggregation and filters out non-ideal model updates, to further optimize the global learning model. Based on real-world datasets and practical learning tasks, extensive experiments are carried out to demonstrate the efficacy of FAIR.

In cloud and edge networks, federated learning involves training statistical models over decentralized data, where servers aggregate models through intermediate updates trained from clients. By utilizing private and local data it improves quality of personalized services and reduces user's concern for privacy. However, federated learning still leaks multimedia features through trained intermediate updates and thereby is not privacy-preserving for multimedia. Existing techniques applied from secure community attempt to avoid multimedia features leakages for federated learning but yet cannot address issues of privacy. In this paper, we propose a privacy-preserving solution that avoids multimedia privacy leakages in federated learning. Firstly, we devise a novel encryption scheme called Non-Informative Transformation (NIT) for federated aggregation to eliminates residual multimedia features in intermediate updates. Based on the scheme, we then propose Just-Learn-over-Ciphertext (JLoC) mechanism for federated learning, which includes three stages in each model iteration. The Encrypt stage encrypts intermediate updates and makes it non-informative distribution at clients. The Aggregate stage performs model aggregation without decryption at servers. Specifically, this stage just computes over ciphertext, and its output of aggregation also keeps non-informative. The Decrypt stage converts non-informative outputs of aggregation to available parameters for the next iteration at clients. Moreover, we implement a prototype and conduct experiments to evaluate its privacy and performance on real devices. The experimental results demonstrate that our methods can defend against potential attacks for multimedia privacy leakages without accuracy loss in commercial off-the-shelf products.

Collaborative computation offloading in mobile edge computing where edge users offload tasks opportunistically to resourceful neighboring mobile devices (MDs), offers a promising solution to satisfy low-latency requirements. However, most existing works assume that those MDs volunteer to help edge users without an incentive mechanism. In this article, we propose an auction-based incentive mechanism, where users and MDs participate in the system dynamically. Our auction mechanism runs in the online fashion and optimizes the long-term system welfare without knowledge of future information, e.g., task start time, task length, resource demand, and valuation, etc. We prove that the proposed online mechanism achieves the desired properties, including individual rationality, truthfulness, and computational tractability. Moreover, the theoretical competitive ratio shows that our online mechanism achieves near-optimal long-term social welfare close to the offline optimum. Extensive experiments based on real-world traces demonstrate the efficiency of the proposed online mechanism.

Data rate trading, in which mobile devices trade their real-time data transmission rates to achieve cooperative mobile networks access, not only can meet the increasing data access demands of users but also can reduce the pressure on cellular networks. However, there is no directly available mechanism for data rate trading. In this paper, we propose a truthful online auction mechanism for data rate trading in mobile networks. In the designed auction, the data rate buyers submit their realtime data access requests, including the rate requirement, access time and payment. The auctioneer, which may be the network operator, assigns data rate requests to appropriate sellers who leverage their surplus cellular data plan or other networks to complete the data rate requests and benefit from them. To achieve this model, we first formulate the social welfare maximization problem in data rate trading as an integer linear programming and show its NP-hardness. Then, we resort to the Lagrangian relaxation technique to design an online approximation algorithm to assign data rate requests and compute the corresponding payments in polynomial time. Theoretical analysis and simulation experiments show that the proposed auction mechanism obtains a good competitive ratio and satisfies the desired properties, including individual rationality, truthfulness, and computational efficiency.

By exploiting the storage resources of end devices, local caching becomes a promising approach to reduce the latency and improve the throughput of content delivery. Considering battery constraints at end devices, this paper investigates energy-aware caching policy to minimize the power consumed in content delivery, taking into account the heterogeneity of interests and sharing willingness of different MSs. Specifically, MSs are divided into disjoint groups based on the preferences and sharing willingness, and each group customizes the caching policy accordingly. Both the non-coordinated and coordinated caching scenarios are considered. For the non-coordinated case, the problem is formulated as an optimization problem, which is proved to be concave and solved by Lagrange methods. For the coordinated caching case, a water-filling based iterative algorithm is proposed to get the optimal caching policies. Numerical results demonstrate that the designed caching policies can reduce the system energy consumption by around 10% - 20% comparing to the conventional ones without considering the heterogeneity of users.

Nowadays, multiple heterogeneous cellular networks coexist simultaneously, and mobile devices can freely select the appropriate network for data communication. Our measurement studies show that a barrier exists between heterogeneous cellular networks and applications. Triggering handovers between various cellular networks can break the barrier and present the promise of saving energy for cellular data communication. However, most existing network handover triggering methods do not adequately incorporate the characteristics of applications and may cause unnecessary handovers. In this paper, we propose an application-based handover triggering method called AHT. According to the applications that are used by users, AHT triggers handovers between high-performance and energy-efficient cellular networks, thus conserving energy. Based on the user experience (UX) requirements, AHT classifies applications into UX- sensitive and insensitive ones. AHT determines whether and when to switch to the high-performance network (e.g., LTE) in accordance with the predicted UX-sensitive application and estimated usage time, and triggers handovers to the energy- efficient network (e.g., UMTS) through an idle timer. We evaluate the performance of AHT with real application usage traces. Experimental results show that AHT saves up to 60.7% and 32.8% energy compared with the pure LTE network transmission and the screen-based handover triggering scheme Intelli3G, respectively.

In this paper, we study a small-cell caching system consisting of one privately-owned small base station (SBS) and multiple content providers (CPs), where CPs leverage the caching capabilities of SBSs to efficiently provide content delivery service to mobile subscribers. Specifically, an incentive cache mechanism is proposed, to stimulate the privately- owned SBS and CPs to participate in the caching system. A two-stage Stackelberg game is formulated for the interaction between the SBS and CPs. In the first stage, the private SBS first decides the price policy to maximize the profit. In the second stage, according to the charge price, each CP determines the amount of caching storage to maximize its utility. The impact of transmission congestion on CP utility is also taken into consideration, which also influences CPs' decisions. We prove the existence and uniqueness of the equilibrium, and design an optimal pricing algorithm to maximize the SBS's revenue. Simulation results are provided to evaluate the performance of the proposed mechanism, which demonstrates the efficiency and feasibility on the SBS resource sharing.

Device-to-device (D2D) content sharing holds great promise to alleviate the growing strain on cellular networks, as it offloads popular content data onto direct peer-to-peer links. However, it is still largely unexplored how content sharing could benefit from utilizing multi-hop rather than conventional single-hop D2D communications. As a step towards this end, this paper proposes a generalized two-level Stackelberg game theoretic framework to enable content sharing with multi-hop D2D communication capabilities. At the lower level, a Nash bargaining subgame is proposed to provide the routing and pricing graphs, where a novel incentive mechanism is adopted to stimulate cooperation. At the upper level, the set of participants is decided to ensure all participants contribute to the content sharing. An additional pricing plan is introduced to make sure that the final pricing is practical and feasible. Numerical results are presented to demonstrate that the proposed game theoretic framework could not only jointly deal with participants selection, routing and pricing in D2D content sharing, but also effectively restrict utilities and transmission resources to only contributive participants.

Mobile online social networks (OSNs) are emerging as the popular mainstream platform for information and content sharing among people. In order to provide Quality of Experience (QoE) support for mobile OSN services, in this paper we propose a socially-driven learning-based framework, namely Spice, for media content prefetching to reduce the access delay and enhance mobile user's satisfaction. Through a large-scale data-driven analysis over real-life mobile Twitter traces from over 17,000 users during a period of five months, we reveal that the social friendship has a great impact on user's media content click behavior. To capture this effect, we conduct social friendship clustering over the set of user's friends, and then develop a cluster-based Latent Bias Model for socially-driven learning-based prefetching prediction. We then propose a usage-adaptive prefetching scheduling scheme by taking into account that different users may possess heterogeneous patterns in the mobile OSN app usage. We comprehensively evaluate the performance of Spice framework using trace-driven emulations on smartphones. Evaluation results corroborate that the Spice can achieve superior performance, with an average 67.2% access delay reduction at the low cost of cellular data and energy consumption. Furthermore, by enabling users to offload their machine learning procedures to a cloud server, our design can achieve speed-up of a factor of 1000 over the local data training execution on smartphones.

The growing demand on the performance of transparent computing systems requires good cache schemes in order to overcome the prolonged network latency. However, evaluating cache schemes, especially measuring the performance of a transparent computing system with particular cache scheme remains challenging. This is because neither method is available to evaluate the effectiveness and efficiency of the cache schemes in transparent computing, nor the simulator has been developed to measure the system performance under particular cache schemes. In this paper, we propose TranSim, a full-featured, high-performance simulation framework for transparent computing. For the first time, TranSim introduces a methodology to evaluate the performance of multi-level cache hierarchies in transparent computing under different cache configurations and cache replacement policies. TranSim can also demonstrate the behavior and performance of the entire transparent computing system rather than only the cache miss/hit rate. Using TranSim, the system designer can quickly evaluate the effectiveness of cache schemes along with the system performance. We construct several experiments to evaluate the effectiveness and efficiency of TranSim. Results show that TranSim can accurately output the performance of both the cache hierarchy and the entire transparent computing system.

This study proposes a unidirectional‐valve (UDV) algorithm for energy‐efficient transmission, where the packets arrived at different times must be transmitted before a common deadline. The UDV algorithm divides the total transmission time into a series of segments. For a newly arrived packet, it is regarded as a new segment and its rate is compared with that of the former adjacent one, then the two segments are combined into one if the new segment has a lower or an equal rate. The comparison and combination are alternately performed until the rate on the whole actually becomes non‐decreasing. In the offline case that the sizes and arrival times of all packets are learned at the beginning, it has been demonstrated that the UDV scheduling achieves the optimal energy‐efficient transmission; while in the online case that only the mean size and mean interval of future packets are known, UDV achieves the energy efficiency approaching the optimality. In addition, the proposed UDV scheduling has a low computational complexity, and thus can be readily implemented in the system.

Elasticity has now become the elemental feature of cloud computing as it enables the ability to dynamically add or remove virtual machine instances when workload changes. However, effective virtualized resource management is still one of the most challenging tasks. When the workload of a service increases rapidly, existing approaches cannot respond to the growing performance requirement efficiently because of either inaccuracy of adaptation decisions or the slow process of adjustments, both of which may result in insufficient resource provisioning. As a consequence, the Quality of Service (QoS) of the hosted applications may degrade and the Service Level Objective (SLO) will be thus violated. In this paper, we introduce SPRNT, a novel resource management framework, to ensure high-level QoS in the cloud computing system. SPRNT utilizes an aggressive resource provisioning strategy which encourages SPRNT to substantially increase the resource allocation in each adaptation cycle when workload increases. This strategy first provisions resources which are possibly more than actual demands, and then reduces the over-provisioned resources if needed. By applying the aggressive strategy, SPRNT can satisfy the increasing performance requirement in the first place so that the QoS can be kept at a high level. The experimental results show that SPRNT achieves up to 7.7× speedup in adaptation time, compared with existing efforts. By enabling quick adaptation, SPRNT limits the SLO violation rate up to 1.3 percent even when dealing with rapidly increasing workload.

Accurately determining the geographic location of an Internet host is important for location-aware applications such as location-based advertising and network diagnostics. Despite their fast response time, widely used database-driven geolocation approaches provide only inaccurate locations. Delay measurement based approaches improve the estimation accuracy but still suffer from a limited precision (about 10 km) and a long response time (tens of seconds) to localize a single PC, which cannot meet the demand of precise and real-time geolocation for location-aware applications. In this paper, we propose a new geolocation approach, Checkin-Geo, which exploits geolocation resources fundamentally different from existing database-driven (using DNS, Whois, etc.) or network delay measurement based approaches. In particular, we leverage the location data that users are willing to share in location-sharing services and logs of user logins from PCs for real-time and accurate geolocation. Experimental results show that compared to existing geolocation techniques, Checkin-Geo achieves 1) a median estimation error of 799 meters (an order of magnitude smaller than existing approaches), and 2) a negligible response time, which are promising for accurate location-aware applications.

In cellular networks, inactivity timers are used to control the release of radio resources. However, during the timeout period of inactivity timers, known as the tail time, a large proportion of energy in user devices and a considerable amount of radio resources are wasted. In this paper, we propose TailTheft, a scheme that leverages the tail time for batching and prefetching to reduce energy consumption. For network requests from a number of applications that can be deferred or prefetched, TailTheft provides a customized application programming interface to distinguish requests and then schedules delay-tolerant and prefetchable requests in the tail time to save energy. TailTheft employs a virtual tail time mechanism to determine the amount of tail time that can be used and a dual queue scheduling algorithm to schedule transmissions. We implement TailTheft in the Network Simulator with a model for calculating energy consumption that is based on parameters measured from mobile phones. We evaluate TailTheft using real application traces, and the experimental results show that TailTheft can achieve significant savings on battery energy (up to 65%) and dedicated radio resources (up to 56%), compared to the default policy.

This paper presents the design, implementation, and evaluation of TransCom, a virtual disk (Vdisk) based cloud computing platform that supports heterogeneous services of operating systems (OSes) and their applications in enterprise environments. In TransCom, clients store all data and software, including OS and application software, on Vdisks that correspond to disk images located on centralized servers, while computing tasks are carried out by the clients. Users can choose to boot any client for using the desired OS, including Windows, and access software and data services from Vdisks as usual without consideration of any other tasks, such as installation, maintenance, and management. By centralizing storage yet distributing computing tasks, TransCom can greatly reduce the potential system maintenance and management costs. We have implemented a multi-platform TransCom prototype that supports both Windows and Linux services. The extensive evaluation based on both test-bed experiments and real-usage experiments has demonstrated that TransCom is a feasible, scalable, and efficient solution for successful real-world use.

Advancements in cloud computing enable the easy deployment of numerous services. However, the analysis of cloud service access platforms from a client perspective shows that maintaining and managing clients remain a challenge for end users. In this paper, we present the design, implementation, and evaluation of an asymmetric virtual machine monitor (AVMM), which is an asymmetric partitioning-based bare-metal approach that achieves near-native performance while supporting a new out-of-operating system mechanism for value-added services. To achieve these goals, AVMM divides underlying platforms into two asymmetric partitions: a user partition and a service partition. The user partition runs a commodity user OS, which is assigned to most of the underlying resources, maintaining end-user experience. The service partition runs a specialized OS, which consumes only the needed resources for its tasks and provides enhanced features to the user OS. AVMM considerably reduces virtualization overhead through two approaches: 1) Peripheral devices, such as graphics equipment, are assigned to be monopolized by a single user OS. 2) Efficient resource management mechanisms are leveraged to alleviate complicated resource sharing in existing virtualization technologies. We implement a prototype that supports Windows and Linux systems. Experimental results show that AVMM is a feasible and efficient approach to client virtualization.

New applications based on cloud computing, such as data synchronization for large chain departmental stores and bank transaction records, require very high-speed data transport. Although a number of high-bandwidth networks have been built, existing transport protocols or their variants over such networks cannot fully exploit the network bandwidth. Our experiments show that the fixed-size application level buffer employed in the receiver side is a major cause of this deficiency. A buffer that is either too small or too large impairs the transfer performance. Due to the varied natures of network conditions and of real-time packet processing (i.e., consuming) speed at the receiver, it is important to ensure that the buffer size is dynamically adjusted according to the perceived execution situation during runtime. In this paper, we propose Rada, a dynamic receiving buffer adaptation scheme for high-speed data transfer. Rada employs an exponential moving average aided scheme to quantify the data arrival rate and consumption rate in the buffer. Based on these two rates, we develop a linear aggressive increase conservative decrease scheme to adjust the buffer size dynamically. Moreover, a weighted mean function is employed to make the adjustment adaptive to the available memory in the receiver. Theoretical analysis is provided to demonstrate the rationale and parameter bounds of Rada. The performance of Rada is also theoretically compared with potential alternatives. We implement Rada in a Linux platform and extensively evaluate its performance in a variety of scenarios. Experimental results conform to the theoretical results, and show that Rada outperforms the static buffer scheme in terms of throughput, memory footprint, and fairness.

This paper considers the transmission of confidential data over a quasi-static fading wiretap channel where the main and eavesdropper channels are correlated. Assuming that before transmission the transmitter knows the channel state information (CSI) of the main channel, we derive the secrecy outage probability in a closed-form expression based on a new secrecy outage probability formula, which gives a more explicit measure on the level of security compared with the previous one. Remarkably, our results, which cover the corresponding results when the main channel and eavesdropper channels are independent as special cases, reveal that channel correlation has a significant impact on secrecy outage probability and such impact can be helpful or harmful depending on the relative channel conditions and transmission SNR threshold.

Cloud computing is an increasingly important solution for providing services deployed in dynamically scalable cloud networks. Services in the cloud computing networks may be virtualized with specific servers which host abstracted details. Some of the servers are active and available, while others are busy or heavy loaded, and the remaining are offline for various reasons. Users would expect the right and available servers to complete their application requirements. Therefore, in order to provide an effective control scheme with parameter guidance for cloud resource services, failure detection is essential to meet users' service expectations. It can resolve possible performance bottlenecks in providing the virtual service for the cloud computing networks. Most existing Failure Detector (FD) schemes do not automatically adjust their detection service parameters for the dynamic network conditions, thus they couldn't be used for actual application. This paper explores FD properties with relation to the actual and automatic fault-tolerant cloud computing networks, and find a general non-manual analysis method to self-tune the corresponding parameters to satisfy user requirements. Based on this general automatic method, we propose specific and dynamic Self-tuning Failure Detector, called SFD, as a major breakthrough in the existing schemes. We carry out actual and extensive experiments to compare the quality of service performance between the SFD and several other existing FDs. Our experimental results demonstrate that our scheme can automatically adjust SFD control parameters to obtain corresponding services and satisfy user requirements, while maintaining good performance. Such an SFD can be extensively applied to industrial and commercial usage, and it can also significantly benefit the cloud computing networks.

IP prefix hijacking is one of the top threats in the cloud computing Internets. Based on cryptography, many schemes for preventing prefix hijacks have been proposed. Securing binding between IP prefix and its owner underlies these schemes. We believe that a scheme for securing this binding should try to satisfy these seven critical requirements: no key escrow, no other secure channel, defending against Malicious Key Issuer (MKI) in the phase of prefix announcement, defending against MKI in the phase of key issuing, no certificate, in-band delegation attestation, and in-band public key witness. In this paper, we propose a new scheme, Origin Authentication based on Self-Certified public keys (SC-OA), using self-certified public keys to authenticate origin autonomous systems. To the best of our knowledge, it is the first work for securing prefix ownership using self-certified public keys to achieve an efficient and secure scheme that satisfies all seven requirements. The analyses show that SC-OA can defend against regular prefix, sub prefix, unassigned prefix, interception-based, and MKI hijacking, and improve performance in many aspects. It will be pushed ahead to practical deployment for preventing prefix hijacks.

Retrieving similar documents from a large-scale text corpus according to a given document is a fundamental technique for many applications. However, most of existing indexing techniques have difficulties to address this problem due to special properties of a document query, e.g. high dimensionality, sparse representation and semantic issue. Towards addressing this problem, we propose a two-level retrieval solution based on a document decomposition idea. A document is decomposed to a compact vector and a few document specific keywords by a dimension reduction approach. The compact vector embodies the major semantics of a document, and the document specific keywords complement the discriminative power lost in dimension reduction process. We adopt locality sensitive hashing (LSH) to index the compact vectors, which guarantees to quickly find a set of related documents according to the vector of a query document. Then we re-rank documents in this set by their document

It is regarded that the Internet-based computing has a lot of significant merits. However, the relatively long latency and relatively high cost of wide-area networking hamper the growth of its application. Currently, there are some computing nodes emerging in the Internet, such as the public data centers born in the initial Cloud computing environment, which can perform computations with large computational freedom and capability for the public use. This paper proposes and models the problem of optimal organization of the Internet-based computation which takes these emerging distributed computing nodes into account. Both the performance and cost are analytically formulized into the optimization problems. Because the solution space expands exponentially when directly addressing the problems, this paper develops two efficient approaches. Extensive evaluations on various networks and workloads show that the proposed model OIC outperforms the current Internet-based computing model in terms of both computational time and cost.

With the development of computing devices and networks, several efficient and high performance UDP-based protocols have been proposed and employed in recently emerging computing paradigms, e.g., pervasive or cloud computing, to transport large data. However, since the server in such protocols uses a fixed-size memory buffer to hold received packets before handling them, the buffer will be exhausted if these packets cannot be handled as fast as they arrive, impairing the performance dramatically even if there is plenty of free memory. To solve this problem, we propose a Rate and Resource Detection Based Buffer Adaptation Approach (RRDA). RRDA collects the difference between the server receiving and processing rate and the amount of free memory periodically. Based on these information, RRDA decides whether the receive buffer should be resized and if so, to what extent to adjust. RRDA can not only avoid the exhaustion of receive buffer when the server load is heavy, but also can free unnecessary memory when the load is low. Experimental results show that RRDA can reduce the occurrence of buffer exhaustion by a factor of 10 and improve the throughput remarkably, compared with the fixed-size buffer scheme.

The rapid development of computer network technologies and social informationalization has brought many new opportunities and challenges in information security. With improved information and service sharing enjoyed by more and more people, how to st

The recently proposed active queue management (AQM) is an effective method used in Internet routers for congestion control, and to achieve a tradeoff between link utilization and delay. The de facto standard, the random early detection (RED) AQM scheme, and most of its variants use average queue length as a congestion indicator to trigger packet dropping. In this paper, we propose a novel proportional and differential RED algorithm, called NPDRED, as an extension of RED. NPD-RED is based on a self-tuning proportional and differential controller, which not only considers the instantaneous queue length at the current time point, but also takes into consideration the ratio of the current differential error signal to the buffer size. Furthermore, we give theoretical analysis of the system stability and give guidelines for the selection of feedback gains for the TCP/RED system to stabilize the instantaneous queue length at a desirable level. Extensive simulations have been conducted with ns2. The simulation results have demonstrated that the proposed NPD-RED algorithm outperforms the existing AQM schemes in terms of average queue length,average throughput, and stability.

With the rapid improvements in hardware, software and networks, the computing paradigm has also shifted from mainframe computing to ubiquitous or pervasive computing, in which users can focus on their desired services rather than specific computing devices and technologies. However, the emerging of ubiquitous computing has brought many challenges, one of which is that it is hard to allow users to freely obtain desired services, such as heterogeneous OSes and applications via different light-weight devices. We have proposed a new paradigm, called Transparent Computing, to store and manage the commodity programs including OS codes centrally, while stream them to be run in non-state clients. This leads to a service-centric computing environment, in which users can select the desired services on demand, without concerning these services' administrations, such as their installation, maintenance, management, upgrade, and so on. In this paper, we introduce a novel concept: Meta OS to support such program streaming through a distributed 4VP+platform. Based on this platform, a pilot system has been implemented and it supports Windows and Linux environments. We verify the effectiveness of the platform through both real deployments and testbed experiments. The evaluation results suggest that 4VP+platform is a feasible and promising solution for future computing infrastructure in ubiquitous computing.

Service Integration plays an important role in Service-Oriented Architecture (SOA).With the thriving of SOA, many approaches have been proposed to address the issue. Several languages or specifications are introduced for the definition of integration plan, such as WSFL, BPML and BPEL4WS. And many recent researches employ semantic web techniques to enable the automatic service discovery and interoperation, OWL-S included. Most of the existing approaches are originally introduced for application developers or programmers to use, in which user requirement is usually expressed based on the services themselves. However, a service may be very complex, the functions of which are usually the composite and interconnection of several basic functions. It is sometimes difficult for normal users without developing skills to understand the services and then utilize them to define the integration solution. In the paper, we think that, the most familiar things for normal users are the functions they need and the working process they expect. Based on the intuitive idea, we propose a normal User-Centric Service Integration approach (UCSI). In the approach, a set of function elements are predefined for each specified service domain and three working process notations are introduced to model the relationships between function elements. Then user requirement and service function can both be represented as function elements interconnection based on the process notations. With the representation, the functionality relationship between user requirement and services can be resolved and the service integration solutions can be generated. Normal users with certain knowledge of specified service domains can try using the approach to achieve personalized service integration. And a prototype system is introduced in the end.