INFOCOM 2020 论文列表

In Mobile Edge Computing (MEC), many tasks require specific service support for execution and in addition, have a dependent order of execution among the tasks. However, previous works often ignore the impact of having limited services cached at the edge nodes on (dependent) task offloading, thus may lead to an infeasible offloading decision or a longer completion time. To bridge the gap, this paper studies how to efficiently offload dependent tasks to edge nodes with limited (and predetermined) service caching. We formally define the problem of offloading dependent tasks with service caching (ODT-SC), and prove that there exists no algorithm with constant approximation for this hard problem. Then, we design an efficient convex programming based algorithm (CP) to solve this problem. Moreover, we study a special case with a homogeneous MEC and propose a favorite successor based algorithm (FS) to solve this special case with a competitive ratio of O(1). Extensive simulation results using Google data traces show that our proposed algorithms can significantly reduce applications' completion time by about 27-51% compared with other alternatives.

Point-of-Interest (POI) demand modeling in urban regions is critical for building smart cities with various applications, e.g., business location selection and urban planning. However, existing work does not fully utilize human mobility data and ignores the interactive-aware information. In this work, we design a refined POI demand modeling framework, named RePiDeM, to identify region POI demands based on multi-source data, including cellular data, POI data, satellite image, geographic data, etc. Specifically, we introduce a Cellular Data (CD) based visit inference algorithm to estimate the POI visit probability based on human mobility and POI data. Further, to address the data sparsity issue, we design a multi-source attention neural collaborative filtering (MANCF) model to output region POI demands considering various aspect attention. We conduct extensive experiments on real-world data collected in the Chinese city Shenyang, which show that RePiDeM is effective for modeling region POI demands.

The Internet can be made more efficient and robust with hop-by-hop multipath routing: Each router on the path can split packets between multiple nexthops in order to 1) avoid failed links and 2) reduce traffic on congested links. Before deciding how to split traffic, one first needs to decide which nexthops to allow at each step. In this paper, we investigate the requirements and trade-offs for making this choice.Most related work chooses the viable nexthops by applying the "Downward Criterion", i.e., only adding nexthops that lead closer to the destination; or more generally by creating a Directed Acyclic Graph (DAG) for each destination. We show that a DAG's nexthop options are necessarily limited, and that, by using certain links in both directions (per destination), we can add further nexthops while still avoiding loops. Our solution LFID (LoopFree Inport-Dependent) routing, though having a slightly higher time complexity, leads to both a higher number of and shorter potential paths than related work. LFID thus protects against a higher percentage of single and multiple failures (or congestions) and comes close to the performance of arbitrary source routing.

As the Wi-Fi technology goes through its sixth generation (Wi-Fi 6), there is a growing consensus on the need to support security and coordination functions at the Physical (PHY) layer, beyond traditional functions such as frame detection and rate adaptation. In contrast to the costly approach of extending the PHY-layer header to support new functions (e.g., Target Wake Time field in 802.11ax), we propose to turn a specific part of the frame preamble into a data field while maintaining its primary functions. Specifically, in this paper, we develop a scheme called extensible preamble modulation (ePMod) for the MIMO-based 802.11ac protocol. For each frame, eP-Mod can embed up to 20 bits into the 802.11ac preamble under 1 × 2 or 2 × 1 MIMO transmission modes to support the operations of a given PHY-layer function. The proposed scheme enables several promising PHY-layer services, such as PHY-layer encryption and channel/device authentication, PHYlayer signaling, etc. At the same time, it allows legacy (eP-Modunaware) devices to continue to process the received preamble as normal by guaranteeing that our proposed preamble waveforms satisfy the structural properties of a standardized preamble. Through numerical analysis, extensive simulations, and hardware experiments, we validate the practicality and reliability of ePMod.

The widespread deployment of machine learning applications in ubiquitous environments has sparked interests in exploiting the vast amount of data stored on mobile devices. To preserve data privacy, Federated Learning has been proposed to learn a shared model by performing distributed training locally on participating devices and aggregating the local models into a global one. However, due to the limited network connectivity of mobile devices, it is not practical for federated learning to perform model updates and aggregation on all participating devices in parallel. Besides, data samples across all devices are usually not independent and identically distributed (IID), posing additional challenges to the convergence and speed of federated learning. In this paper, we propose Favor, an experience-driven control framework that intelligently chooses the client devices to participate in each round of federated learning to counterbalance the bias introduced by non-IID data and to speed up convergence. Through both empirical and mathematical analysis, we observe an implicit connection between the distribution of training data on a device and the model weights trained based on those data, which enables us to profile the data distribution on that device based on its uploaded model weights. We then propose a mechanism based on deep Q-learning that learns to select a subset of devices in each communication round to maximize a reward that encourages the increase of validation accuracy and penalizes the use of more communication rounds. With extensive experiments performed in PyTorch, we show that the number of communication rounds required in federated learning can be reduced by up to 49% on the MNIST dataset, 23% on FashionMNIST, and 42% on CIFAR-10, as compared to the Federated Averaging algorithm.

Magnetic resonant coupling wireless power transfer (MRC-WPT) enables convenient device-charging. When MIMO MRC-WPT system incorporated with multiple relay components, both relay On-Off state (i.e., power routing) and TX current (i.e., current scheduling) could be adjusted for improving charging efficiency and distance. Previous approaches need the collaboration and feedback from the energy receiver (RX), achieved using side-channels, e.g., Bluetooth, which is time/energy-consuming. In this work we propose, design, and implement a multi-relay MIMO MRC-WPT system, and design an almost optimum joint optimization of power routing and current scheduling method, without relying on any feedback from RX. We carefully decompose the joint optimization problem into two subproblems without affecting the overall optimality of the combined solution. For current scheduling subproblem, we propose an almost-optimum RX-feedback independent solution. For power routing subproblem, we first design a greedy algorithm with ½ approximation ratio, and then design a DQN based method to further improve its effectiveness. We prototype our system and evaluate it with extensive experiments. Our results demonstrate the effectiveness of the proposed algorithms. The achieved power transfer efficiency (PTE) on average is 3.2X, 1.43X, 1.34X, and 7.3X over the other four strategies: without relay, with non-adjustable relays, greed based, and shortest-path based ones.

The next frontier in communications is teleoperation – manipulation and control of remote environments. Compared to conventional networked applications, teleoperation poses widely different requirements, ultra-low latency (ULL) being the primary one. Teleoperation, along with a host of other applications requiring ULL communication, is termed as Tactile Internet (TI). A significant redesign of conventional networking techniques is necessary to realize TI applications. Further, these advancements can be evaluated only when meaningful performance metrics are available. However, existing TI performance metrics fall severely short of comprehensively characterizing TI performance. In this paper, we take the first step towards bridging this gap. To this end, we propose a method that captures the fine-grained performance of TI in terms of delay and precision. We take Dynamic Time Warping (DTW) as the basis of our work and identify whether it is sufficient in characterizing TI systems. We refine DTW by developing a framework called Effective Time- and Value-Offset (ETVO) that extracts fine-grained time and value offsets between input and output signals of TI. Using ETVO, we present two quantitative metrics for TI – Effective Delay-Derivative (EDD) and Effective Root Mean Square Error. Through rigorous experiments conducted on a realistic TI setup, we demonstrate the potential of the proposed metrics to precisely characterize TI interactions.

In this paper, we consider the problem of scheduling real-time traffic in wireless networks under a conflict-graph interference model and single-hop traffic. The objective is to guarantee that at least a certain fraction of packets of each link are delivered within their deadlines, which is referred to as delivery ratio. This problem has been studied before under restrictive frame-based traffic models, or greedy maximal scheduling schemes like LDF (Largest-Deficit First) that can lead to poor delivery ratio for general traffic patterns. In this paper, we pursue a different approach through randomization over the choice of maximal links that can transmit at each time. We design randomized policies in collocated networks, multipartite networks, and general networks, that can achieve delivery ratios much higher than what is achievable by LDF. Further, our results apply to traffic (arrival and deadline) processes that evolve as positive recurrent Markov chains. Hence, this work is an improvement with respect to both efficiency and traffic assumptions compared to the past work. We further present extensive simulation results over various traffic patterns and interference graphs to illustrate the gains of our randomized policies over LDF variants.

Federated learning (FL), as a privacy-preserving machine learning framework, draws growing attention in both industry and academia. It obtains a jointly accurate model by distributing training tasks into data owners and aggregating their model updates. However, FL faces new security problems, as it losses direct control to training processes. One fundamental demand is to ensure whether participants execute training tasks as intended.In this paper, we propose TrustFL, a practical scheme that leverages Trusted Execution Environments (TEEs) to build assurance of participants' training executions with high confidence. Specifically, we use TEE to randomly check a small fraction of all training processes for tunable levels of assurance, while all computations are executed on the co-located faster yet insecure processor (e.g., GPU) for efficiency. To prevent various cheating behaviors like only processing TEE-requested computations or uploading old results, we devise a commitment-based method with specific data selection. We prototype TrustFL using GPU and SGX and evaluate its performance. The results show that TrustFL achieves one/two orders of magnitude speedups compared with naive training with SGX, when assuring correct training with a confidence level of 99%.

Recently, data-driven prediction strategies have shown the potential of shepherding the optimization strategies for end viewer's Quality-of-Experience in practical streaming applications. The current prediction-based designs have largely focused on optimizing the last-mile, i.e., viewer-side, which 1) need the real-time feedback from viewers to improve the prediction accuracy; and 2) need quick responses to guarantee the effectiveness of optimization strategies in the future. Thanks to the emerged crowdsourced livecast services, e.g., Twitch.tv, we for the first time exploit the opportunity to realize the long-term prediction and optimization with the assistance derived from the first-mile, i.e., source broadcasters.In this paper, we propose a novel framework CastFlag, which analyzes the broadcasters' operations and interactions, predicts the key events (i.e., highlights), and optimizes the transcoding stage in the corresponding live streams, even before the encoding stage. Taking the most popular eSports gamecast as an example, we illustrate the effectiveness of this framework in the game highlight prediction and transcoding workload allocation. The trace-driven evaluation shows the superiority of CastFlag as it: (1) improves the prediction accuracy over other learning-based approaches by up to 30%; (2) achieves an average of 10% saving of the transcoding latency at less cost.

Identification and tracking of multiple objects are essential in many applications. As a key enabler of automatic ID technology, RFID has got widespread adoption with item-level tagging in everyday life. However, restricted to the computation capability of passive RFID systems, locating or tracking tags has always been a challenging task. Meanwhile, as a fundamental problem in the field of computer vision, object tracking in images has progressed to a remarkable state especially with the rapid development of deep learning in the past few years. To enable lightweight tracking of a specific target, researchers try to complement computer vision to existing RFID architecture and achieves fine granularity. However, such solution requires calibration of the cameras extrinsic parameters at each new setup, which is not convenient for usage. In this work, we propose Tagview, a pervasive identifying and tracking system that can work in various settings without repetitive calibration efforts. It addresses the challenge by skillfully deploying the RFID antenna and video camera at the identical position and devising a multi-target recognition schema with only the image-level trajectory information. We have implemented Tagview with commercial RFID and camera devices and evaluated it extensively. Experimental results show that our method can archive high accuracy and robustness.

The dynamic nature of software-defined networking requires frequent updates to the flow table in the data plane of switches. Therefore, the ternary content-addressable memory (TCAM) used in switches to match packet header fields against forwarding rules needs to support high rates of updates. Existing off-the-shelf switches update rules in batches for efficiency, but may suffer from forwarding inconsistencies during the batch update. In this paper, we design and evaluate a TCAM update optimization framework that can guarantee consistent forwarding during the entire update process while making use of a layered TCAM structure. Our approach is based on a modified-entry-first write-back strategy that significantly reduces the overhead from movements of TCAM entries. In addition, our approach detects reordering cases, which are handled using efficient solutions. Based on our evaluation results, we can reduce the cost of TCAM updates by 30%-88% compared to state-of-the-art techniques.

Serverless computing has emerged as a new compelling paradigm for the deployment of applications and services. It represents an evolution of cloud computing with a simplified programming model, that aims to abstract away most operational concerns. Running serverless functions requires users to configure multiple parameters, such as memory, CPU, cloud provider, etc. While relatively simpler, configuring such parameters correctly while minimizing cost and meeting delay constraints is not trivial. In this paper, we present COSE, a framework that uses Bayesian Optimization to find the optimal configuration for serverless functions. COSE uses statistical learning techniques to intelligently collect samples and predict the cost and execution time of a serverless function across unseen configuration values. Our framework uses the predicted cost and execution time, to select the "best" configuration parameters for running a single or a chain of functions, while satisfying customer objectives. In addition, COSE has the ability to adapt to changes in the execution time of a serverless function. We evaluate COSE not only on a commercial cloud provider, where we successfully found optimal/near-optimal configurations in as few as five samples, but also over a wide range of simulated distributed cloud environments that confirm the efficacy of our approach.

5G technologies promise to revolutionize mobile networks and push them to the limits of resource utilization. Besides better capacity, we also need better resource management via virtualization. End-to-end network slicing not only involves the core but also the Radio Access Network (RAN) which makes this a challenging problem. This is because multiple alternative radio access technologies exist (e. g. ,LTE, WLAN, and WiMAX), and there is no unifying abstraction to compare and compose from diverse technologies. In addition, existing work assumes that all RAN infrastructure exists under a single administrative domain. Software-Defined Radio Access Network (SD-RAN) offers programmability that facilitates a unified abstraction for resource sharing and composition across multiple providers harnessing different technology stacks. In this paper we propose a new architecture for heterogeneous RAN slicing across multiple providers. A central component in our architecture is a service orchestrator that interacts with multiple network providers and service providers to negotiate resource allocations that are jointly optimal. We propose a double auction mechanism that captures the interaction among selfish parties and guarantees convergence to optimal social welfare in finite time. We then demonstrate the feasibility of our proposed system by using open source SD-RAN systems such as EmPOWER (WiFi) and FlexRAN (LTE).

At the foundation of every network security architecture lies the premise that formulated network flow policies are reliably deployed and enforced by the network infrastructure. However, software-defined networks (SDNs) add a particular challenge to satisfying this premise, as for SDNs the flow pol-icy implementation spans multiple applications and abstraction layers across the SDN stack. In this paper, we focus on the question of how to automatically identify cases in which the SDN stack fails to prevent policy inconsistencies from arising among these components. This question is rather essential, as when such inconsistencies arise the implications to the security and reliability of the network are devastating. We present AudiSDN, an automated fuzz-testing framework designed to formulate test cases in which policy inconsistencies can arise in OpenFlow networks, the most prevalent SDN protocol used today. We also present results from applying AudiSDN to two widely used SDN controllers, Floodlight and ONOS. In fact, our test results have led to the filing of 3 separate CVE reports. We believe that the approach presented in this paper is applicable to the breadth of OpenFlow platforms used today, and that its broader usage will help to address a serious but yet understudied pragmatic concern.

Understanding in-network information diffusion is a fundamental problem in many application domains and one of the primary challenges is to predict the size of the information cascade. Most of the existing models rely either on hypothesized point process (e.g., Poisson and Hawkes process), or simply predict the information propagation via deep neural networks. However, they fail to simultaneously capture the underlying structure of a cascade graph and the propagation of uncertainty in the diffusion, which may result in unsatisfactory prediction performance. To address these, in this work we propose a novel probabilistic cascade prediction framework: Variational Cascade (VaCas) graph learning networks. VaCas allows a non-linear information diffusion inference and models the information diffusion process by learning the latent representation of both the structural and temporal information. It is a pattern-agnostic model leveraging variational inference to learn the node-level and cascade-level latent factors in an unsupervised manner. In addition, VaCas is capable of capturing both the cascade representation uncertainty and node infection uncertainty, while enabling hierarchical pattern learning of information diffusion. Extensive experiments conducted on real-world datasets demonstrate that VaCas significantly improves the prediction accuracy, compared to state-of-the-art approaches, while also enabling interpretability.

Benefitting from the recent breakthrough of wireless power transfer technology, Wireless Rechargeable Sensor Networks (WRSNs) have become an important research topic. Most prior arts focus on system performance enhancement in an ideal environment that ignores impacts of obstacles. This contradicts with practical applications in which obstacles can be found almost anywhere and have dramatic impacts on energy transmission. In this paper, we concentrate on the problem of charging a practical WRSN in the presence of obstacles to maximize the charging utility under specific energy constraints. First, we propose a new theoretical charging model with obstacles based on Fresnel diffraction model, and conduct experiments to verify its effectiveness. Then, we propose a spatial discretization scheme to obtain a finite feasible charging position set for MC, which largely reduces computation overhead. Afterwards, we reformalize charging utility maximization with energy constraints as a submodular function maximization problem and propose a cost-efficient algorithm with approximation ratio (e-1)/2e (1 - ε) to solve it. Lastly, we demonstrate that our scheme outperforms other algorithms by at least 14.8% in terms of charging utility through test-bed experiments and extensive simulations.

We detail a practical compression scheme for lossless compression of time-series data, based on the emerging concept of generalized deduplication. As data is no longer stored for just archival purposes, but needs to be continuously accessed in many applications, the scheme is designed for low-cost random access to its compressed data, avoiding decompression. With this method, an arbitrary bit of the original data can be read by accessing only a few hundred bits in the worst case, several orders of magnitude fewer than state-of-the-art compression schemes. Subsequent retrieval of bits requires visiting at most a few tens of bits. A comprehensive evaluation of the compressor on eight real-life data sets from various domains is provided. The cost of this random access capability is a loss in compression ratio compared with the state-of-the-art compression schemes BZIP2 and 7z, which can be as low as 5% depending on the data set. Compared to GZIP, the proposed scheme has a better compression ratio for most of the data sets. Our method has massive potential for applications requiring frequent random accesses, as the only existing approach with comparable random access cost is to store the data without compression.

Once considered a luxury, Home Security Cameras (HSCs) are now commonplace and constitute a growing part of the wider online video ecosystem. This paper argues that their expanding coverage and close integration with daily life may result in not only unique behavioral patterns, but also key privacy concerns. This motivates us to perform a detailed measurement study of a major HSC provider, covering 15.4M streams and 211K users. Our study takes two perspectives: (i) we explore the per-user behaviour, identifying core clusters of users; and (ii) we build on this analysis to extract and predict privacy-compromising insight. Key observations include a highly asymmetrical traffic distribution, distinct usage patterns, wasted resources and fixed viewing locations. Furthermore, we identify three privacy risks and explore them in detail. We find that paid users are more likely to be exposed to attacks due to their heavier usage patterns. We conclude by proposing simple mitigations that can alleviate these risks.