INFOCOM 2021 论文列表

Traffic policing is widely used by ISPs to limit their customers' traffic rates. It has long been believed that a well-tuned traffic policer offers a satisfactory performance for TCP. However, we find this belief breaks with the emergence of new congestion control (CC) algorithms like BBR: flows using these new CC algorithms can easily occupy the majority of the bandwidth, starving traditional TCP flows. We confirm this problem with experiments and reveal its root cause as follows. Without buffer in traffic policers, congestion only causes packet losses, while new CC algorithms are loss-resilient, i.e. they adjust the sending rate based on other network feedback like delay. Thus, when being policed they will not reduce the sending rate until an unacceptable loss ratio for TCP is reached, resulting in low throughput for TCP. Simply adding buffer to the traffic policer improves fairness but incurs high latency. To this end, we propose FairPolicer, which can achieve fair bandwidth allocation without sacrificing latency. FairPolicer regards token as a basic unit of bandwidth and fairly allocates tokens to active flows in a round-robin manner. Testbed experiments show that FairPolicer can significantly improve the fairness and achieve much lower latency than other kinds of rate-limiters.

Java reflection technique is pervasively used in the Android system. To reduce the risk of reflection abuse, Android restricts the use of reflection at the Android Runtime (ART) to hide potentially dangerous methods/fields. We perform the first comprehensive study of the reflection restrictions and have discovered three novel approaches to bypass the reflection restrictions. Novel reflection-based attacks are also presented, including the password stealing attack. To mitigate the threats, we analyze these restriction bypassing approaches and find three techniques crucial to these approaches, i.e., double reflection, memory manipulation, and inline hook. We propose a defense mechanism that consists of classloader double checker, ART variable protector, and ART method protector, to prohibit the reflection restriction bypassing. Finally, we design and implement an automatic reflection detection framework and have discovered 5,531 reflection powered apps out of 100,000 downloaded apps, which are installed on our defense enabled Android system of a Google Pixel 2 to evaluate the effectiveness and efficiency of our defense mechanism. Extensive empirical experiment results demonstrate that our defense enabled system can accurately obstruct the malicious reflection attempts.

We learn optimal user association policies for traffic from different locations to Access Points(APs), in the presence of unknown dynamic traffic demand. We aim at minimizing a broad family of α-fair cost functions that express various objectives in load assignment in the wireless downlink, such as total load or total delay minimization. Finding an optimal user association policy in dynamic environments is challenging because traffic demand fluctuations over time are non-stationary and difficult to characterize statistically, which obstructs the computation of cost-efficient associations. Assuming arbitrary traffic patterns over time, we formulate the problem of online learning of optimal user association policies using the Online Convex Optimization (OCO) framework. We introduce a periodic benchmark for OCO problems that generalizes state-of-the-art benchmarks. We exploit inherent properties of the online user association problem and propose PerOnE, a simple online learning scheme that dynamically adapts the association policy to arbitrary traffic demand variations. We compare PerOnE against our periodic benchmark and prove that it enjoys the no-regret property, with additional sublinear dependence of the network size. To the best of our knowledge, this is the first work that introduces a periodic benchmark for OCO problems and a no-regret algorithm for the online user association problem. Our theoretical findings are validated through results on a real-trace dataset.

The wide deployment of machine learning (ML) models and service APIs exposes the sensitive training data to untrusted and unknown parties, such as end-users and corporations. It is important to preserve data privacy in the released ML models. An essential issue with today's privacy-preserving ML platforms is a lack of concern on the tradeoff between data privacy and model utility: a private datablock can only be accessed a finite number of times as each access is privacy-leaking. However, it has never been interrogated whether such privacy leaked in the training brings good utility. We propose a differentially-private access control mechanism on the ML platform to assign datablocks to queries. Each datablock arrives at the platform with a privacy budget, which would be consumed at each query access. We aim to make the most use of the data under the privacy budget constraints. In practice, both datablocks and queries arrive continuously so that each access decision has to be made without knowledge about the future. Hence we propose online algorithms with a worst-case performance guarantee. Experiments on a variety of settings show our privacy budgeting scheme yields high utility on ML platforms.

Boolean Network Tomography (BNT) allows to localize network failures by means of end-to-end monitoring paths. Nevertheless, it falls short of providing efficient failure identification in real scenarios, due to the large combinatorial size of the solution space, especially when multiple failures occur concurrently. We aim at maximizing the identification capabilities of a bounded number of monitoring probes. To tackle this problem we propose a progressive approach to failure localization based on stochastic optimization, whose solution is the optimal sequence of monitoring paths to probe. We address the complexity of the problem by proposing a greedy strategy in two variants: one considers exact calculation of posterior probabilities of node failures given the observation, whereas the other approximates these values through a novel failure centrality metric. We discuss the approximation of the proposed approaches. Then, by means of numerical experiments conducted on real network topologies, we demonstrate the practical applicability of our approach. The performance evaluation evidences the superiority of our algorithms with respect to state of the art solutions based on classic Boolean Network Tomography as well as approaches based on sequential group testing.

The ability to detect outliers is crucial in data mining, with widespread usage in many fields, including fraud detection, malicious behavior monitoring, health diagnosis, etc. With the tremendous volume of data becoming more distributed than ever, global outlier detection for a group of distributed datasets is particularly desirable. In this work, we propose PIF (Privacy-preserving Isolation Forest), which can detect outliers for multiple distributed data providers with high efficiency and accuracy while giving certain security guarantees. To achieve the goal, PIF makes an innovative improvement to the traditional iForest algorithm, enabling it in distributed environments. With a series of carefully-designed algorithms, each participating party collaborates to build an ensemble of isolation trees efficiently without disclosing sensitive information of data. Besides, to deal with complicated real-world scenarios where different kinds of partitioned data are involved, we propose a comprehensive schema that can work for both horizontally and vertically partitioned data models. We have implemented our method and evaluated it with extensive experiments. It is demonstrated that PIF can achieve comparable AUC to existing iForest on average and maintains a linear time complexity without privacy violation.

We propose CoRE, a 360° video streaming approach that reduces bandwidth requirements compared to transferring the entire 360° video. CoRE uses non-linear sampling in both the spatial and temporal domains to achieve robustness to view direction prediction error and to transient wireless network bandwidth fluctuation. Each CoRE frame samples the environment in all directions, with full resolution over the predicted field of view and gradually decreasing resolution at the periphery, so that missing pixels are avoided, irrespective of the view prediction error magnitude. A CoRE video chunk has a main part at full frame rate, and an extension part at a gradually decreasing frame rate, which avoids stalls while waiting for a delayed transfer. We evaluate a prototype implementation of CoRE through trace-based experiments and a user study, and find that, compared to tiling with low-resolution padding, CoRE reduces data transfer amounts, stalls, and H.264 decoding overhead, increases frame rates, and eliminates missing pixels.

The multi-armed bandit (MAB) model has been deeply studied to solve many online learning problems, such as rate allocation in communication networks, Ad recommendation in social networks, etc. In an MAB model, given N arms whose rewards are unknown in advance, the player selects exactly one arm in each round, and his goal is to maximize the cumulative rewards over a fixed horizon. In this paper, we study the budget-constrained auction-based combinatorial multi-armed bandit mechanism with strategic arms, where the player can select K (<; N) arms in a round and pulling each arm has a unique cost. In addition, each arm might strategically report its cost in the auction. To this end, we combine the upper confidence bound (UCB) with auction to define the UCB-based rewards and then devise an auction-based UCB algorithm (called AUCB). In each round, AUCB selects the top K arms according to the ratios of UCB-based rewards to bids and further determines the critical payment for each arm. For AUCB, we derive an upper bound on regret and prove the truthfulness, individual rationality, and computational efficiency. Extensive simulations show that the rewards achieved by AUCB are at least 12.49% higher than those of state-of-the-art algorithms.

As a key enabler of future 5G network, Service Function Chain (SFC) forwards the traffic flow along a chain of Virtual Network Functions (VNFs) to provide network services flexibility. One of the most important problems in SFC is to deploy the VNFs and schedule arriving requests among computing nodes to achieve low latency and high reliability. Existing works consider a static network and assume that all SFC requests are known in advance, which is impractical. In this paper, we focus on the dynamic 5G network environment where the SFC requests arrive randomly and the computing nodes can redeploy all types of VNF with a time cost. We formulate the problem of SFC scheduling in NFV-enabled 5G network as a mixed integer non-linear programing. The objective is to maximize the number of requests satisfying the latency and reliability constraints. To solve the problem, we propose an efficient algorithm to decide the redundancy of the VNFs while minimizing delay. Then we present a state-of-art Reinforcement Learning (RL) to learn SFC scheduling policy to increase the success rate of SFC requests. The effectiveness of our method is evaluated through extensive simulations. The result shows that our proposed RL solution can increase the success rate by 18.7% over the benchmarks.

Automatic Modulation Classification (AMC) is significant for the practical support of a plethora of emerging spectrum applications, such as Dynamic Spectrum Access (DSA) in 5G and beyond, resource allocation, jammer identification, intruder detection, and in general, automated interference analysis. Although a well-known problem, most of the existing AMC work has been done under the assumption that the classifier has prior knowledge about the signal and channel parameters. This paper shows that unknown signal and channel parameters significantly degrade the performance of two of the most popular research streams in modulation classification: expert feature-based and data-driven. By understanding why and where those methods fail, in such unknown scenarios, we propose two possible directions to make AMC more robust to signal shape transformations introduced by unknown signal and channel parameters. We show that Spatial Transformer Networks (STN) and Transfer Learning (TL) embedded into a light ResNeXt-based classifier can improve average classification accuracy up to 10-30% for specific unseen scenarios with only 5% labeled data for a large dataset of 20 complex higher-order modulations.

Long Range (LoRa) is a low-power wireless communication technology for long-range connectivity, extensively used in the Internet of Things. Several works in the literature have analytically characterized the performance of LoRa networks, with particular focus on scalability and reliability. However, most of the related models are limited, as they cannot account for factors that occur in practice, or make strong assumptions on how devices are deployed in the network. This article proposes an analytical model that describes the delivery ratio in a LoRa network with device-level granularity. Specifically, it considers the impact of several key factors that affect real deployments, including multiple gateways and channel variation. Therefore, the proposed model can effectively evaluate the delivery ratio in realistic network topologies, without any restrictions on device deployment or configuration. It also accurately characterizes the delivery ratio of each device in a network, as demonstrated by extensive simulations in a wide variety of conditions, including diverse networks in terms of node deployment and link-level parameter settings. The proposed model provides a level of detail that is not available in the state of the art, and it matches the simulation results within an error of a few percentage points.

The number of devices that the edge of the Internet accommodates and the volume of the data these devices generate are expected to grow dramatically in the years to come. As a result, managing and processing such massive data amounts at the edge becomes a vital issue. This paper proposes "Store Edge Networked Data" (SEND), a novel framework for in-network storage management realized through data repositories deployed at the network edge. SEND considers different criteria (e.g., data popularity, data proximity from processing functions at the edge) to intelligently place different categories of raw and processed data at the edge based on system-wide identifiers of the data context, called labels. We implement a data repository prototype on top of the Google file system, which we evaluate based on real-world datasets of images and Internet of Things device measurements. To scale up our experiments, we perform a network simulation study based on synthetic and real-world datasets evaluating the performance and trade-offs of the SEND design as a whole. Our results demonstrate that SEND achieves data insertion times of 0.06ms-0.9ms, data lookup times of 0.5ms-5.3ms, and on-time completion of up to 92% of user requests for the retrieval of raw and processed data.

This paper studies the problem of source detection, which is to infer the source node out of an aftermath of a cascade, i.e., the observed infected graph GN of the network at some time. Prior arts have adopted various statistical quantities such as degree, distance or infection size to reflect the structural centrality of the source. In this paper, we propose a new metric which we call the infected tree entropy (ITE), to utilize richer underlying structural features for source detection. Our idea of ITE is inspired by the conception of structural entropy [1], which demonstrated that the minimization of average bits to encode the network structures with different partitions is the principle for detecting the natural or true structures in real-world networks. Accordingly, our proposed ITE based estimator for the source tries to minimize the coding of network partitions brought by the infected tree rooted at all the potential sources, thus minimizing the structural deviation between the cascades from the potential sources and the actual infection process included in GN. On polynomially growing geometric trees, with increasing tree heterogeneity, the ITE estimator remarkably yields more reliable detection under only moderate infection sizes. In contrast, for regular expanding trees, we still observe guaranteed detection probability of ITE estimator even with an infinite infection size, thanks to the degree regularity property. We also algorithmically realize the ITE based detection that enjoys linear time complexity via a message-passing scheme, and further extend it to general graphs. Experiments on various network topologies confirm the superiority of ITE to the baselines. For example, ITE returns an accuracy of 75% ranking the source among top 5%, far exceeding 45% of the classic algorithms on scale-free networks.

While most clouds use on-server rate limiters for bandwidth allocation, we propose to implement them on switches. On-switch rate limiters can simplify network management and promote the performance of control-plane rate limiting applications. We leverage the recent progress of programmable switches to implement on-switch rate limiters, named SwRL. In the design of SwRL, we make design choices according to the programmable hardware characteristics, we deeply optimize the memory usage of the algorithm so as to fit a cloud-scale (one million) rate limiters in a single switch, and we complement the missing computation primitives of the hardware using a pre-computed approximate table. We further developed three control-plane applications and integrate them with SwRL, showing the control-plane interoperability of SwRL. We prototype and evaluate SwRL in both testbed and production environments, demonstrating its good properties of precision rate control, scalability, interoperability, and manageability (execution environmental isolation).

Smart home IoT systems utilize trigger-action platforms, e.g., IFTTT, to manage devices from various vendors. These platforms allow users to define rules for automatically triggering operations on devices. However, they may be abused by triggering malicious rule execution with forged IoT devices or events violating the execution integrity and the intentions of the users. To address this issue, we propose a ledger based IoT platform called Ruledger, which ensures the correct execution of rules by verifying the authenticity of the corresponding information. Ruledger utilizes smart contracts to enforce verifying the information associated with rule executions, e.g., the user and configuration information from users, device events, and triggers in the trigger-action platforms. In particular, we develop three algorithms to enable ledger-wallet based applications for Ruledger and guarantee that the records used for verification are stateful and correct. Thus, the execution integrity of rules is ensured even if devices and platforms in the smart home systems are compromised. We prototype Ruledger in a real IoT platform, i.e., IFTTT, and evaluate the performance with various settings. The experimental results demonstrate Ruledger incurs an average of 12.53% delay, which is acceptable for smart home systems.

While energy consumption is the primary concern for the design of real-time embedded systems, fault-tolerance and quality of service (QoS) are becoming increasingly important in the development of today's pervasive computing systems. In this work, we study the problem of energy-aware standby-sparing for weakly hard real-time embedded systems. The standby-sparing systems adopt a primary processor and a spare processor to provide fault tolerance for both permanent and transient faults. In order to reduce energy consumption for such kind of systems, we proposed two novel scheduling schemes: one for (1,1)-hard tasks and one for general (m,k)-hard tasks which require that at least m out of any k consecutive jobs of a task meet their deadlines. Through extensive evaluations, our results demonstrate that the proposed techniques significantly outperform the previous research in reducing energy consumption for both (1,1)-hard task sets and general (m,k)-hard task sets while assuring fault tolerance through standby-sparing.

Network functions (NFs) are critical components in the network data plane. Their efficiency is important to the whole network's end-to-end performance. We identify three types of runtime redundant logic in individual NF and NF chains when they are deployed with concrete configured rules. We use program analysis techniques to optimize away the redundancy where we also overcome the NF specific challenges - we combine symbolic execution and dead code elimination to eliminate unused logic, we customize the common sub-expression elimination to eliminate duplicated logic, and we add network semantics to the dead code elimination to eliminate overwritten logic. We implement a prototype named NFReducer using LLVM. Our evaluation on both legacy and platform NFs shows that after eliminating the redundant logic, the packet processing rate of the NFs can be significantly improved and the operational overhead is small.

The physical layer secret key generation exploiting wireless channel reciprocity has attracted considerable attention in the past two decades. On-going research have demonstrated its viability in various radio frequency (RF) systems. Most of existing work rely on quantization technique to convert channel measurements into digital binaries that are suitable for secret key generation. However, non-simultaneous packet exchanges in time division duplex systems and noise effects in practice usually create random channel measurements between two users, leading to inconsistent quantization results and mismatched secret bits. While significant efforts were spent in recent research to mitigate such non-reciprocity, no efficient method has been found yet. Unlike existing quantization-based approaches, we take a different viewpoint and perform the secret key agreement by solving a bipartite graph matching problem. Specifically, an efficient dual-permutation secret key generation method, DP-SKG, is developed to match the randomly permuted channel measurements between a pair of users by minimizing their discrepancy holistically. DP-SKG allows two users to generate the same secret key based on the permutation order of channel measurements despite the non-reciprocity over wireless channels. Extensive experimental results show that DP-SKG could achieve error-free key agreement on received signal strength (RSS) with a low cost under various scenarios.

The quality of experience (QoE) is an important issue for users when accessing the web. Although many metrics have been designed to estimate the QoE in the desktop environment, few studies have confirmed whether the QoE metrics are valid in the mobile environment. In this paper, we ask questions regarding the validity of using desktop-based QoE metrics for the mobile web and find answers. We first classify the existing QoE metrics into several groups according to three criteria and then identify the differences between the mobile and desktop environments. Based on the analysis, we ask three research questions and develop a system, called WebMythBusters, for collecting and analyzing mobile web experiences. Through an extensive analysis of the collected user data, we find that (1) the metrics focusing on fast completion or fast initiation of the page loading process cannot estimate the actual QoE, (2) the conventional scheme of calculating visual progress is not appropriate, and (3) focusing only on the above-the-fold area is not sufficient in the mobile environment. The findings indicate that QoE metrics designed for the desktop environment are not necessarily adequate for the mobile environment, and appropriate metrics should be devised to reflect the mobile web experience.