INFOCOM 2018 论文列表

Given the increasing scarcity of IPv4 addresses, network operators are resorting to measures to expand their address pool or prolong the life of existing addresses. One such approach is Carrier-Grade NAT (CGN), where many end-users in a network share a single public IPv4 address. There is limited data about the prevalence of CGN, despite the implications on performance, security, and ultimately, the adoption of IPv6. In this work, we present passive measurement-based techniques for detecting CGN deployments across the entire Internet, without the requirement of access to machines behind a CGN. Specifically, we identify patterns in how client IP addresses are observed at M-Lab servers and at the UCSD network telescope to infer whether those clients are behind a CGN. We apply our methods on data collected from 2014 to 2016. We find that CGN deployment is increasing rapidly. Overall, we infer that 4.1K autonomous systems are deploying CGN, 6 times the number inferred by the most recent studies.

With the rapid growth of congestion-sensitive and data-intensive applications, traditional settlement-free peering agreements with best-effort delivery often do not meet the QoS requirements of content providers (CPs). Meanwhile, Internet access providers (IAPs) feel that revenues from end-users are not sufficient to recoup the upgrade costs of network infrastructures. Consequently, some IAPs have begun to offer CPs a new type of peering agreement, called paid peering, under which they provide CPs with better data delivery quality for a fee. In this paper, we model a network platform where an IAP makes decisions on the peering types offered to CPs and the prices charged to CPs and end-users. We study the optimal peering schemes for the IAP, i.e., to offer CPs both the paid and settlement-free peering to choose from or only one of them, as the objective is profit or welfare maximization. Our results show that 1) the IAP should always offer the paid and settlement-free peering under the profit-optimal and welfare-optimal schemes, respectively, 2) whether to simultaneously offer the other peering type is largely driven by the type of data traffic, e.g., text or video, and 3) regulators might want to encourage the IAP to allocate more network capacity to the settlement-free peering for increasing user welfare.

Battery-Free Wireless Sensor Network (BF-WSN) is a newly proposed network architecture to address the limitation of traditional Wireless Sensor Networks (WSNs). The special features of BF-WSNs make the coverage problem quite different and even more challenging from and than that in traditional WSNs. This paper defines a new coverage problem in BF-WSNs which aims at maximizing coverage quality rather than prolonging network lifetime. The newly defined coverage problem is proved to be at least NP-Hard. Two sufficient conditions, under which the optimal solution of the problem can be derived in polynomial time, are given in this paper. Furthermore, two approximate algorithms are proposed to derive nearly optimal coverage when the sufficient conditions are unsatisfied. The time complexity and approximate ratio of the two algorithms are analyzed. Extensive simulations are carried out to examine the performance of the proposed algorithms. The simulation results show that these algorithms are efficient and effective.

As the U.S. mobile data market matures, Internet service providers (ISPs) generally charge their users with some variation on a quota-based data plan with overage charges. Common variants include unlimited, prepaid, and usage-based data plans. However, despite a recent flurry of research on optimizing mobile data pricing, few works have considered how these data plans affect users' consumption behavior. In particular, while users with such plans have a strong incentive to plan their usage over the month, they also face uncertainty in their future data usage needs that would make such planning difficult. In this work, we develop a dynamic programming model of users' consumption decisions over the month that takes this uncertainty into account. We use this model to quantify which types of users would benefit from different types of data plans, using these conditions to extrapolate the optimal types of data plans that ISPs should offer. Our theoretical findings are complemented by numerical simulations on a dataset of user usage from a large U.S. ISP. The results help mobile users to choose data plans that maximize their utilities and ISPs to gain profit by understanding their user behavior while choosing what data plans to offer.

Many emerging sensor network applications operate in challenging environments wherein sensor nodes do not always have connected paths to the base station. Data generated from such intermittently connected sensor networks therefore must be stored inside the network for some unpredictable period of time before uploading opportunities become available. Consequently, sensory data could overflow limited storage capacity available in the entire network, making discarding valuable data inevitable. To overcome such overall storage overflow in intermittently connected sensor networks, we propose and study a new algorithmic problem called data aggregation for overall storage overflow (DA02). Utilizing spatial data correlation that commonly exists among sensory data, DAO2employs data aggregation techniques to reduce the overflow data size while minimizing the total energy consumption. To solve DAO2, we uncover a new graph theoretic problem called multiple traveling salesman walks (MTSW), and show that with proper graph transformation, the DAO2is equivalent to the MTSW. We prove that MTSW is NP-hard and design a (2-1.)-approximation q algorithm, where q is the number of nodes to visit (i.e., the number of sensor nodes that aggregate their overflow data). The approximation algorithm is based on a novel routing structure called minimum q-edge forest that accurately captures information needed for energy-efficient data aggregation. We further put forward a heuristic algorithm and empirically show that it constantly outperforms the approximation algorithm by 15% -30% in energy consumption. Finally, we propose a distributed data aggregation algorithm that can achieve the same approximation ratio as the centralized algorithm under some condition, while incurring comparable energy consumption.

Smart home IoT devices have been more prevalent than ever before but the relevant security considerations fail to keep up with due to device and technology heterogeneity and resource constraints, making IoT systems susceptible to various attacks. In this paper, we propose a novel graph-based mechanism to identify the vulnerabilities in communication of IoT devices for smart home systems. Our approach takes one or more packet capture files as inputs to construct a traffic graph by passing the captured messages, identify the correlated subgraphs by examining the attribute-value pairs associated with each message, and then quantify their vulnerabilities based on the sensitivity levels of different keywords. To test the effectiveness of our approach, we setup a smart home system that can control a smart bulb LB100 via either the smartphone APP for LB100 or the Google Home speaker. We collected and analyzed 58,714 messages and exploited 6 vulnerable correlated sub graphs, based on which we implemented 6 attack cases that can be easily reproduced by attackers with little knowledge of IoT. This study is novel as our approach takes only the collected traffic files as inputs without requiring the knowledge of the device firmware while being able to identify new vulnerabilities. With this approach, we won the third prize out of 20 teams in a hacking competition.

The black-box nature of clouds introduces a lack of trusts in clouds. Since provenance can provide a complete history of an entity, trustworthy provenance management for data, application, or workflow can make the cloud more accountable. Current research on cloud provenance mainly focuses on collecting provenance records and trusting the cloud providers in managing the provenance records. However, a dishonest cloud provider can alter the provenance records, as the records are stored within the control of the cloud provider. To solve this problem, we first propose CloProv - a provenance model to capture the complete provenance of any type of entities in the cloud. We analyze the threats on the CloProv model considering collusion among malicious users and dishonest cloud providers. Based on the threat model, we propose a secure data provenance scheme - SECProv for cloud-based, multi-user, shared data storage systems. We integrate SECProv with the object storage module of an open source cloud framework - OpenStack Swift and analyze the efficiency of the proposed scheme.

Recently, Edge Computing has emerged as a new computing paradigm dedicated for mobile applications for performance enhancement and energy efficiency purposes. Specifically, it benefits today's interactive applications on power-constrained devices by offloading compute-intensive tasks to the edge nodes which is in close proximity. Meanwhile, Field Programmable Gate Array (FPGA) is well known for its excellence in accelerating compute-intensive tasks such as deep learning algorithms in a high performance and energy efficiency manner due to its hardware-customizable nature. In this paper, we make the first attempt to leverage and combine the advantages of these two, and proposed a new network-assisted computing model, namely FPGA-based edge computing. As a case study, we choose three computer vision (CV)-based interactive mobile applications, and implement their backend computation parts on FPGA. By deploying such application-customized accelerator modules for computation offloading at the network edge, we experimentally demonstrate that this approach can effectively reduce response time for the applications and energy consumption for the entire system in comparison with traditional CPU-based edge/cloud offloading approach.

The current paradigm of exclusive spectrum assignment and allocation is creating artificial spectrum scarcity that has a dramatic impact on network performance and user experience. Thus, governments, industry and academia have endeavored to create novel spectrum management mechanisms that allow multi-tiered access. A key component of such an approach is deep understanding of spectrum utilization in time, frequency and space. To address this challenge, we propose AirVIEW, a one-pass, unsupervised spectrum characterization approach for rapid transmitter detection with high tolerance to noise. AirVIEW autonomously learns its parameters and employs wavelet decomposition in order to amplify and reliably detect transmissions at a given time instant. We show that AirVIEW can robustly identify transmitters even when their power is only 5dBm above the noise floor. Furthermore, we demonstrate AirVIEW's ability to inform next-generation Dynamic Spectrum Access by characterizing essential transmitter properties in wideband spectrum measurements from 50MHz to 4.4GHz.

Emerging 5G systems will need to efficiently support both broadband traffic (eMBB) and ultra-low-latency (URLLC) traffic. In these systems, time is divided into slots which are further sub-divided into minislots. From a scheduling perspective, eMBB resource allocations occur at slot boundaries, whereas to reduce latency URLLC traffic is pre-emptively overlapped at the minislot timescale, resulting in selective superposition/puncturing of eMBB allocations. This approach enables minimal URLLC latency at a potential rate loss to eMBB traffic. We study joint eMBB and URLLC schedulers for such systems, with the dual objectives of maximizing utility for eMBB traffic while satisfying instantaneous URLLC demands. For a linear rate loss model (loss to eMBB is linear in the amount of superposition/puncturing), we derive an optimal joint scheduler. Somewhat counter-intuitively, our results show that our dual objectives can be met by an iterative gradient scheduler for eMBB traffic that anticipates the expected loss from URLLC traffic, along with an URLLC demand scheduler that is oblivious to eMBB channel states, utility functions and allocations decisions of the eMBB scheduler. Next we consider a more general class of (convex) loss models and study optimal online joint eMBB/URLLC schedulers within the broad class of channel state dependent but time-homogeneous policies. We validate the characteristics and benefits of our schedulers via simulation.

In this paper we study the queuing system that describes the operations of data services in cellular networks, e.g., UMTS, LTE/LTE-A, and most likely the forthcoming 5G standard. The main characteristic of all these systems is that after service access, resources remain allocated to the end user for some time before release, so that if the same user requests access to service again, before a system timeout, the same resources are still available. For the resulting queuing model, we express the blocking probability in closed form, and we also provide recursive expressions in the number of connections that can be handled by the base station. Closed form expressions are also derived for other useful performance metrics, i.e., throughput and network service time. Analytical results are validated against results of a detailed simulation model, and compared to traditional queueing models results, such as the Erlang B formula iteratively applied to the resources that are not blocked by potentially returning users. Our analysis complements the performance evaluation of the other key mechanism used to access data services in cellular networks, namely the random access, which precedes the resource allocation and utilization phase studied in this paper.

Millimeter-Wave (mmWave) bands have become the de-facto candidate for 5G vehicle-to-everything (V2X) since future vehicular systems demand Gbps links to acquire the necessary sensory information for (semi)-autonomous driving. Nevertheless, the directionality of mmWave communications and its susceptibility to blockage raise severe questions on the feasibility of mmWave vehicular communications. The dynamic nature of 5G vehicular scenarios, and the complexity of directional mmWave communication calls for higher context-awareness and adaptability. To this aim, we propose the first online learning algorithm addressing the problem of beam selection with environment-awareness in mmWave vehicular systems. In particular, we model this problem as a contextual multi-armed bandit problem. Next, we propose a lightweight context-aware online learning algorithm, namely FML, with proven performance bound and guaranteed convergence. FML exploits coarse user location information and aggregates received data to learn from and adapt to its environment. We also perform an extensive evaluation using realistic traffic patterns derived from Google Maps. Our evaluation shows that FML enables mmWave base stations to achieve near-optimal performance on average within 33 minutes of deployment by learning from the available context. Moreover, FML remains within ~ 5% of the optimal performance by swift adaptation to system changes such as blockage and traffic.

As infrastructure-as-a-service clouds become more popular, cloud providers face the complicated problem of maximizing their resource utilization by handling the dynamics of user demand. Auction-based pricing, such as Amazon EC2 spot pricing, provides an option for users to use idle resources at highly reduced yet dynamic prices; under such a pricing scheme, users place bids for cloud resources, and the provider chooses a threshold “spot” price above which bids are admitted. In this paper, we propose a nonlinear dynamical system model for the time-evolution of the spot price as a function of latent states that characterize user demand in the spot and on-demand markets. This model enables us to adaptively predict future spot prices given past spot price observations, allowing us to derive user bidding strategies for heterogeneous cloud resources that minimize the cost to complete a job with negligible probability of interruption. Along the way, the model also yields novel, empirically verifiable insights into cloud provider behavior. We experimentally validate our model and bidding strategy on two months of Amazon EC2 spot price data and find that our proposed bidding strategy is up to 4 times closer to the optimal strategy in hindsight compared to a baseline regression approach while incurring the same negligible probability of interruption.

The Blockchain paradigm provides a popular mechanism for establishing trust and consensus in distributed environments. While Blockchain technology is currently primarily deployed in crypto-currency systems like Bitcoin, the concept is also expected to emerge as a key component of the Internet-of-Things (IoT), enabling novel applications in digital health, smart energy, asset tracking and smart transportation. As Blockchain networks evolve to industrial deployments with large numbers of geographically distributed nodes, the block transfer and processing delays arise as a critical issue which may create greater potential for forks and vulnerability to adversarial attacks. Motivated by these issues, we develop stochastic network models to capture the Blockchain evolution and dynamics and analyze the impact of the block dissemination delay and hashing power of the member nodes on Blockchain performance in terms of the overall block generation rate and required computational power for launching a successful attack. The results provide useful insight in crucial design issues, e.g., how to adjust the `difficulty-of-work' in the presence of delay so as to achieve a target block generation rate or appropriate level of immunity from adversarial attacks. We employ a combination of analytical calculations and simulation experiments to investigate both stationary and transient performance features, and demonstrate close agreement with measurements on a wide-area network testbed running the Ethereum protocol.

Image retrieval is crucial for social media sites such as Instagram to identify similar images and make recommendations for users who share similar interests. To get rid of the storage burden and computation for image retrieval, outsourcing to a remote cloud is now a trend. Yet, privacy concerns mandate the use of encryption before outsourcing the images. We need a secure way for retrieving images from a not-fully-trusted server. This paper proposes InstantCryptoGram, a secure image retrieval service. We first design a new data structure called sub-simhash, which fits for the inverted index used by many searchable symmetric encryption schemes. It leads to our modular solution that supports efficient similarity queries and updates over encrypted images. Our experiments on Amazon AWS EC2 over representative datasets show that our scheme is efficient and accurate in finding similar images while preserving privacy.

Many important applications in the extreme environment require wireless communications to connect smart devices. Metamaterial-enhanced magnetic induction (M2I) has been proposed as a promising solution thanks to its long communication range in the lossy medium. M 2I communication relies on magnetic coupling, which makes it intrinsically full-duplex without self-interference. Moreover, the engineered active metamaterial provides reconfigurability in communication range and interference. In this paper, the new networking paradigm based on the reconfigurable and full-duplex M 2I communication technique is investigated. In particular, the theoretical analysis and electromagnetic simulation are first provided to prove the feasibility. Then, a medium access control protocol is proposed to avoid collisions. Finally, the capacity and delay of the full-duplex M2I network are derived to show the advantage of the new networking paradigm. The analysis in this paper indicates that in a full-duplex M 2I network, the distance between the source and destination can be arbitrarily long and the end-to-end delay can be as short as a single hop delay. As a result, each node in such network can reach any other node by one hop, which can greatly enhance the network robustness and efficiency. It is important for timely transmission of emergent information or real-time control signals.

Telecommunication networks are converging to a massively distributed cloud infrastructure interconnected with software defined networks. In the envisioned architecture, services will be deployed flexibly and quickly as network slices. Our paper addresses a major bottleneck in this context, namely the challenge of computing the best resource provisioning for network slices in a robust and efficient manner. With tractability in mind, we propose a novel optimization framework which allows fine-grained resource allocation for slices both in terms of network bandwidth and cloud processing. The slices can be further provisioned and auto-scaled optimally based on a large class of utility functions in real-time. Furthermore, by tuning a slice-specific parameter, system designers can trade off traffic-fairness with computing-fairness to provide a mixed fairness strategy. We also propose an iterative algorithm based on the alternating direction method of multipliers (ADMM) that provably converges to the optimal resource allocation and we demonstrate the method's fast convergence in a wide range of quasi-stationary and dynamic settings.

To adapt to the increasing storage demands and varying storage redundancy requirements, practical distributed storage systems need to support storage scaling by relocating currently stored data to different storage nodes. However, the scaling process inevitably transfers substantial data traffic over the network. Thus, minimizing the bandwidth cost of the scaling process is critical in distributed settings. In this paper, we show that optimal storage scaling is achievable in erasure-coded distributed storage based on network coding, by allowing storage nodes to send encoded data during scaling. We formally prove the information-theoretically minimum scaling bandwidth. Based on our theoretical findings, we also build a distributed storage system prototype NCScale, which realizes network-coding-based scaling while preserving the necessary properties for practical deployment. Experiments on Amazon EC2 show that the scaling time can be reduced by up to 50% over the state-of-the-art.

The emerging deployment of IoT devices increasingly requires more energy-and-cost-efficient wireless links between devices. Nowadays backscatter networks, one of the most feasible technology to meet the requirement in IoT spaces, have evolved for better usability and wider coverage. The most likely consequence of the evolution is Wi-Fi backscatter networks that harmonize with widely deployed commercial devices. However recent Wi-Fi backscatter techniques are stuck in front of several hurdles. The backscatter techniques along with 802.11b devices impair the spectral efficiency of wireless channels and break backward compatibility. Meanwhile, the backscatter techniques along with the other types of 802.11 devices support only per-packet backscatter resulting poor performance. To tackle all these problems, we propose a flicker detector that achieves per-symbol in-band backscatter by exploiting residual channel of Wi-Fi packets. Our approach shows robust performance without any modification on the hardware and any side effect on wireless channels. Extensive experiments on a software-defined radio testbed demonstrate that our approach overcomes the hurdles of existing Wi-Fi backscatter networks.