INFOCOM 2023 Papers

Contemporary society survives on several network infrastructures, such as communication and transportation. These network infrastructures are required to keep all nodes connected, although these nodes are occasionally disconnected due to failures. Thus, the expected number of connected node pairs (ECP) during an operation period is a reasonable reliability measure in network design. However, no work has studied ECP due to its computational hardness; we have to solve the reliability evaluation problem, which is a computationally tough problem, for O(n2) times where n is the number of nodes in a network. This paper proposes an efficient method that exactly computes ECP. Our method performs dynamic programming just once without explicit repetition for each node pair and obtains an exact ECP value weighted by the number of users at each node. A thorough complexity analysis reveals that our method is faster than an existing reliability evaluation method, which can be transferred to ECP computation, by O(n). Numerical experiments using real topologies show great efficiency; e.g., our method computes the ECP of an 821-link network in ten seconds; the existing method cannot complete it in an hour. This paper also presents two applications: critical link identification and optimal resource (e.g., a server) placement.

Age of Information (AoI) is a metric that can be used to measure the freshness of information. Since its inception, there have been active research efforts on designing scheduling algorithms to various AoI-related optimization problems. For each problem, typically a custom-designed scheduler was developed. Instead of following the (custom-design) path, we pursue a general framework that can be applied to design a wide range of schedulers to solve AoI-related optimization problems. As a first step toward this vision, we present a general framework—Eywa, that can be applied to construct high-performance schedulers for a family of AoI-related optimization problems, all sharing a common setting of an IoT data collection network. We show how to apply Eywa to solve two important AoI-related problems: to minimize the weighted sum of AoIs and to minimize the bandwidth requirement under AoI constraints. We show that for each problem, Eywa can either offer a stronger performance guarantee than the state-of-the-art algorithms or provide new results that are not available in the literature.

Differential privacy (DP) enables model training with a guaranteed bound on privacy leakage, therefore is widely adopted in federated learning (FL) to protect the model update. However, each DP-enhanced FL job accumulates privacy leakage, which necessitates a unified platform to enforce a global privacy budget for each dataset owned by users. In this work, we present a novel DP-enhanced FL platform that treats privacy as a resource and schedules multiple FL jobs across sensitive data. It first introduces a novel notion of device-time blocks for distributed data streams. Such data abstraction enables fine-grained privacy consumption composition across multiple FL jobs. Regarding the non-replenishable nature of the privacy resource (that differs it from traditional hardware resources like CPU and memory), it further employs an allocation-then-recycle scheduling algorithm. Its key idea is to first allocate an estimated upper-bound privacy budget for each arrived FL job, and then progressively recycle the unused budget as training goes on to serve further FL jobs. Extensive experiments show that our platform is able to deliver up to 2.1× as many completed jobs while reducing the violation rate by up to 55.2% under limited privacy budget constraint.

With the popularity of machine learning, it has been a growing concern on the trained model revealing the private information of the training data. Membership inference attack (MIA) poses one of the threats by inferring whether a given sample participates in the training of the target model. Although MIA has been widely studied for discriminative models, for generative models, neither it nor its defense is extensively investigated. In this work, we propose a mixup training method for generative adversarial networks (GANs) as a defense against MIAs. Specifically, the original training data is replaced with their interpolations so that GANs would never overfit the original data. The intriguing part is an analysis from the hypothesis test perspective to theoretically prove our method could mitigate the AUC of the strongest likelihood ratio attack. Experimental results support that mixup training successfully defends the state-of-the-art MIAs for generative models, yet without model performance degradation or any additional training efforts, showing great promise to be deployed in practice.

We consider the design of distributed scheduling algorithms that minimize age of information in single-hop wireless networks. The centralized max-weight policy is known to be nearly optimal in this setting; hence, our goal is to design a distributed CSMA scheme that can mimic its performance. To that end, we propose a distributed protocol called Fresh-CSMA and show that in an idealized setting, Fresh-CSMA can match the scheduling decisions of the max-weight policy with high probability in each frame, and also match the theoretical performance guarantees of the max-weight policy over the entire time horizon. We then consider a more realistic setting and study the impact of protocol parameters on the probability of collisions and the overhead caused by the distributed nature of the protocol. Finally, we provide simulations that support our theoretical results and show that the performance gap between the ideal and realistic versions of Fresh-CSMA is small.

Federated learning (FL) trains a shared global model by periodically aggregating gradients from local devices. Communication overhead becomes a principal bottleneck in FL since participating devices usually suffer from limited bandwidth and unreliable connections in uplink transmission. To address this problem, the gradient compression methods based on compressed sensing (CS) theory have been put forward recently. However, most existing CS-based works compress gradients independently, ignoring the gradient correlations between participants or adjacent communication rounds, which constrains the achievement of higher compression rates. In view of the above observation, we propose a novel gradient compression scheme named FedDCS, guided by distributed compressed sensing (DCS) theory. Following the design philosophy of separate encoding and joint decoding in DCS, FedDCS compresses gradients for participants in each round separately while reconstructing them at the central server jointly via fully exploiting correlated gradients from the previous round, which are known as side information (SI). Benefiting from this design, reconstruction performance is significantly improved with fewer decoding errors also iterations under the identical compression rate, and the total uploading bits to achieve model convergence are considerably reduced. Theoretical analysis and extensive experiments conducted on MNIST and Fashion-MNIST both verify the effectiveness of our approach.

Sharding is an effective technique to improve the scalability of blockchain. It splits nodes into multiple groups so that they can process transactions in parallel. To achieve higher parallelism and concurrency at large scales, it is desirable to maintain a large number of small shards. However, simply configuring small shards easily results in a higher fraction of malicious nodes inside shards, causing shard corruption and compromising system security. Existing sharding techniques hence demand large shards, at the expense of limited concurrency. To address this limitation, we propose CoChain: a blockchain sharding system that can securely configure small shards for enhanced concurrency. CoChain allows some shards to be corrupted. For security, each shard is monitored by multiple other shards. The latter reach a cross-shard Consensus on the Consensus results of their monitored shard. Once a corrupted shard is found, its subsequent consensus will be taken over by another shard, hence recovering the system. Via Consensus on Consensus, CoChain allows the existence of shards with more fraction of malicious nodes (<2/3) while securing the system, thus reducing the shard size safely. We implement CoChain based on Harmony and conduct extensive experiments. Compared with Harmony, CoChain achieves 35x throughput gain with 6,000+ nodes.

The social trust assessment can spur extensive applications such as social recommendations, shopping, financial investment strategies, etc, but remain a challenging problem having limited exploration. Such explorations mainly limit their studies to static network topology or simplified dynamic networks, toward the social trust relationship prediction. In contrast, in this paper, we explore the social trust by taking into account the time-varying online social networks whereas the social trust relationship may vary over time. The DTrust, a dynamic graph neural network-based solution, will be proposed for accurate social trust prediction. In particular, DTrust is composed of a static aggregation unit and a dynamic unit, respectively responsible for capturing both the spatial dependence features and temporal dependence features. In the former unit, we stack multiple NNConv layers derived from the edge-conditioned convolution network for capturing the spatial dependence features correlated to the network topology and the observed social relationships. In the latter unit, a gated recurrent unit (GRU) is employed for learning the evolution law of social interaction and social trust relationships. Based on the extracted spatial and temporal features, we then employ a fully connected neural network for learning, able to predict the social trust relationships for both current and future time slots. Extensive experimental results exhibit that our DTrust can outperform the benchmark counterparts on two real-world datasets.

User-plane machine learning facilitates low-latency, high-throughput inference at line rate. Yet, user planes are highly constrained environments, and restrictions are especially marked in programmable switches with limited memory and minimum support for mathematical operations or data types. Thus, current solutions for in-switch inference that are compatible with production-level hardware lack support for complex features or suffer from limited scalability, and hit performance barriers in complex tasks involving large decision spaces. To address this limitation, we present Flowrest, a first complete Random Forest (RF) model implementation that operates at the level of individual flows in commercial switches. Our solution builds on (i) an original framework to embed flow-level machine learning models into programmable switch ASICs, and (ii) novel guidelines for tailoring RF models to operations in programmable switches already at the design stage. We implement Flowrest as an open-source software using the P4 language, and assess its performance in an experimental platform based on Intel Tofino switches. Tests with tasks of unprecedented complexity show how our model can improve accuracy by up to 39% over previous approaches to implement RF models in real-world equipment.

Mobile CrowdSensing (MCS) is a mobile computing paradigm, through which a platform can coordinate a crowd of workers to accomplish large-scale data collection tasks using their mobile devices. Information freshness has attracted much focus on MCS research worldwide. In this paper, we investigate the incentive mechanism design in MCS systems that take the freshness of collected data and social benefits into concerns. First, we introduce the Age of Information (AoI) metric to measure the freshness of data. Then, we model the incentive mechanism design with AoI guarantees as a novel incomplete information two-stage Stackelberg game with multiple constraints. Next, we derive the optimal strategies of this game so as to determine the optimal reward paid by the platform and the optimal data update frequency for each worker. Moreover, we prove that these optimal strategies form a unique Stackelberg equilibrium. Based on the optimal strategies, we propose an AoI-Aware Incentive (AIAI) mechanism for the MCS system, whereby the platform and all workers can maximize their utilities simultaneously. Meanwhile, the system can ensure that the AoI values of all data uploaded to the platform are not larger than a given threshold to achieve high data freshness. Extensive simulations on real-world traces are conducted to demonstrate the significant performance of AIAI.

The Age-of-Information (AoI) metric has been widely studied in the theoretical communication networks and queuing systems literature. However, experimental evaluation of its applicability to complex real-world time-sensitive systems is largely lacking. In this work, we develop, implement, and evaluate an AoI-based application layer middleware that enables the customization of WiFi networks to the needs of time-sensitive applications. By controlling the storage and flow of information in the underlying WiFi network, our middleware can: (i) prevent packet collisions; (ii) discard stale packets that are no longer useful; and (iii) dynamically prioritize the transmission of the most relevant information. To demonstrate the benefits of our middleware, we implement a mobility tracking application using a swarm of UAVs communicating with a central controller via WiFi. Our experimental results show that, when compared to WiFi-UDP/WiFi-TCP, the middleware can improve information freshness by a factor of 109x/48x and tracking accuracy by a factor of 4x/6x, respectively. Most importantly, our results also show that the performance gains of our approach increase as the system scales and/or the traffic load increases.

Low-Earth-orbit (LEO) satellite networking is a promising way of providing low-latency and high-throughput global Internet access. Unlike the static terrestrial network infrastructure, LEO satellites constantly revolve around the Earth and thus bring instability to their networks. Understanding the dynamics and properties of a LEO satellite network and developing mechanisms to address the dynamics become crucial. In this work, we first introduce a high-fidelity and highly configurable real-time emulator called LeoEM to capture detailed dynamics of LEO satellite networks. We then present SaTCP, a cross-layer solution that enables TCP to avoid overly conservative congestion control and improve its performance under high LEO link dynamics. As an upgrade to CUBIC TCP, SaTCP forecasts the time of disruptive events (i.e., satellite handovers or route updates) by tactfully utilizing the predictability of satellite locations, taking into account the prediction inaccuracy, and informs TCP to adapt its decision accordingly. Experiments across various scenarios show SaTCP increases the goodput by multi-folds compared with state-of-the-art protocols while preserving fairness.

To make machine learning (ML) sustainable and apt to run on the diverse devices where relevant data is, it is essential to compress ML models as needed, while still meeting the required learning quality and time performance. However, how much and when an ML model should be compressed, and {\em where} its training should be executed, are hard decisions to make, as they depend on the model itself, the resources of the available nodes, and the data such nodes own. Existing studies focus on each of those aspects individually, however, they do not account for how such decisions can be made jointly and adapted to one another. In this work, we model the network system focusing on the training of DNNs, formalize the above multi-dimensional problem, and, given its NP-hardness, formulate an approximate dynamic programming problem that we solve through the PACT algorithmic framework. Importantly, PACT leverages a time-expanded graph representing the learning process, and a data-driven and theoretical approach for the prediction of the loss evolution to be expected as a consequence of training decisions. We prove that PACT's solutions can get as close to the optimum as desired, at the cost of an increased time complexity, and that, in any case, such complexity is polynomial. Numerical results also show that, even under the most disadvantageous settings, PACT outperforms state-of-the-art alternatives and closely matches the optimal energy cost.

Low-power wide-area networks (LPWAN) evolve rapidly with advanced communication primitives (e.g., coding, modulation) being continuously invented. This rapid iteration on LPWAN, however, forms a communication barrier between legacy wireless sensor nodes deployed years ago (e.g., ZigBee-based sensor node) with their latest competitor running a different communication protocol (e.g., LoRa-based IoT node): they work on the same frequency band but share different MAC- and PHY-layer regulations and thus cannot talk to each other directly. To break this barrier, we propose LigBee, a cross-technology communication (CTC) solution that enables symbol-level communication from the latest LPWAN LoRa node to legacy ZIGBEE node. We have implemented LigBee on both software-defined radios and commercial-off-the-shelf (COTS) LoRa and ZigBee nodes, and demonstrated that LigBee builds a reliable CTC link from LoRa node to ZigBee node on both platforms. Our experimental results show that i) LigBee achieves a bit error rate (BER) in the order of 10−3with 70 ∼ 80% frame reception ratio (FRR), ii) the range of LigBee link is over 300m, which is 6 ∼ 7.5× the typical range of legacy ZigBee and state-of-the-art solution, and iii) the throughput of LigBee link is maintained on the order of kbps, which is close to the LoRa’s throughput.

Human-machine interactions (HMIs), e.g., touchscreens, are essential for users to interact with mobile devices. They are also beneficial in resisting emerging active attacks, which aim at maliciously controlling mobile devices, e.g., smartphones and tablets. With touchscreen-like HMIs, users can notice and interrupt malicious actions conducted by the attackers timely and perform necessary countermeasures, e.g., tapping the ‘Quit’ button on the touchscreen. However, the effect of HMI-oriented active attacks has not been investigated yet. In this paper, we present a practical attack towards touch-based devices, namely Expelliarmus. It reveals a new attack surface of active attacks for hijacking users’ operations and thus taking full control over victim devices. Expelliarmus neutralizes users’ touch commands by producing a reverse current via electromagnetic interference (EMI). Since the reverse current offsets the current change caused by a touch, the touchscreen detects no current change and thus ignores users’ commands. Besides this basic denial-of-service attack, we also realize a target cancellation attack, which can neutralize target commands, e.g., ‘Quit’ without interference in irrelevant operations. Thus, the active attack can be completely performed without interruption from users, even if they are alerted by the abnormal events. Extensive evaluations demonstrate the effectiveness of Expelliarmus on 29 off-the-shelf devices.