Large Language Models (LLMs) have demonstrated remarkable general capabilities, yet they falter in domains requiring deep strategic reasoning.A primary obstacle is the need to navigate a game tree that grows exponentially with search depth, a task for which their generative nature is ill-suited.To address this, we introduce Generative Gamer (GenGamer), a framework that trains LLMs to reason like an expert player.Instead of attempting an exhaustive search, GenGamer learns to generate a compact, pruned reasoning trajectory termed as a Dynamic Deduction.This is achieved by integrating three key strategies: action pruning based on policy confidence, state pruning via value estimation, and branch pruning inspired by alpha-beta principles.Furthermore, to train the model effectively, we propose the Deduction Tree Reward (DTR), a process-oriented mechanism that provides step-by-step feedback on the quality of the reasoning process, rather than relying solely on the final game outcome.Experiments on complex games such as Tic-Tac-Toe and Leduc Poker demonstrate that GenGamer significantly enhances the strategic capabilities of LLMs, enabling them to achieve performance that surpasses current state-of-theart language models.

The security of wireless challenge-response Physical Layer Authentication (PLA) based on Orthogonal Frequency Division Multiplexing (OFDM) relies on a sufficiently random fading channel condition, which is commonly assumed in existing studies. However, in practical scenarios, such a condition is not always guaranteed and the responses of OFDM subchannels may exhibit correlation. Consequently, ensuring the security of such PLA systems remains an unsolved problem. In this paper, we propose a novel adversary model, called Maximum Differential Likelihood Generator (MDLG), which exploits the weak correlation property in practical wireless channel to launch effective attacks against PLA. Based on this model, we create a measurable guideline using randomness testing to decide when we can in fact use PLA in a practical wireless channel condition. Extensive real-world experiments validate the effectiveness of the MDLG attack and demonstrate how the proposed guideline can help protect the security of PLA.

Accurate drug-target interaction prediction is vital for drug discovery and optimization. Traditional experimental methods, while effective, are time-intensive and costly. HyperGCN-DTI, a novel framework that explicitly advances beyond existing models such as CHL-DTI and HHDTI by leveraging hypergraph neural networks with a multimodal feature fusion strategy. While exsiting methods primarily focuses on low-order graph representations and fixed heterogeneous network structures, HyperGCN-DTI incorporates richer multimodal fused features including embeddings from pretrained language models and diverse biological networks and build robust hypergraphs that capture high-order multi-entity relationships within drug-target pairs. This dual-channel architecture effectively captures both local topological connections and higher-order structural dependencies. HyperGCN-DTI outperforms state-of-the-art DTI prediction models across multiple datasets and remains robust under imbalanced and large-scale real-world datasets, demonstrating its superior predictive power. The model demonstrates significant improvements when using multimodal features and hypergraph-based message passing, with sensitivity analysis confirming stability across hyperparameter variations. Top-ranked predictions are validated through biomedical literature and molecular docking, underscoring the reliability and practical relevance of our approach. HyperGCN-DTI is the first DTI prediction model to jointly integrate such a wide range of heterogeneous information sources with hypergraph representation, significantly enhancing accuracy and robustness, particularly in sparse or noisy settings. The proposed model offers a powerful and generalizable tool for accelerating drug development and target identification.

Argument mining has garnered increasing attention over the years, with the recent advancement of Large Language Models (LLMs) further propelling this trend. However, current argument relations remain relatively simplistic and foundational, struggling to capture the full scope of argument information. To address this limitation, we propose a systematic framework comprising 14 fine-grained relation types from the perspectives of vertical argument relations and horizontal discourse relations, thereby capturing the intricate interplay between argument components for a thorough understanding of argument structure. On this basis, we conducted extensive experiments on three tasks: argument component prediction, relation prediction, and automated essay grading. Additionally, we explored the impact of writing quality on argument component prediction and relation prediction, as well as the connections between discourse relations and argumentative features. The findings highlight the importance of fine-grained argumentative annotations for argumentative writing assessment and encourage multi-dimensional argument analysis.

Accurate and non-invasive monitoring of hematopoietic stem cell transplantation (HSCT) patients is crucial but challenging due to factors including the limitations of traditional biopsies and the intensive manual analysis required for emerging comprehensive imaging techniques like FLT PET/CT. Furthermore, low-dose CT resolution in this vulnerable patient population often hinders precise vertebral body segmentation, a critical step for comprehensive marrow compartment assessment. This paper presents a novel approach for per-vertebral body instance segmentation in HSCT FLT PET/CT scans, addressing the inherent difficulties of low-resolution data and limited annotated training cases. Our method leverages an attention-gated U-Net architecture, significantly enhanced by a novel data augmentation strategy involving downsampled high-resolution VerSe dataset images. We demonstrate, for the first time, accurate vertebral body segmentation on this challenging low-resolution dataset. Our approach integrates an attention-based U-Net model and is compared against TotalSegmentator as a baseline, showing superior segmentation performance, particularly in the anatomically complex upper spine where TotalSegmentator exhibits suboptimal results. To the best of our knowledge, this work reports fully automated high-quality instance segmentation results for individual vertebral bodies in CT volumes of HSCT FLT PET/CT patients for the first time, promising to facilitate automation for critical quantitative assessments like SUV measurement and ultimately improve long-term patient management and outcomes.

Large Language Models (LLMs) have demonstrated impressive capabilities across a wide range of tasks. However, their proficiency and reliability in the specialized domain of financial data analysis, particularly focusing on data-driven thinking, remain uncertain. To bridge this gap, we introduce FinDABench, a comprehensive benchmark designed to evaluate the financial data analysis capabilities of LLMs within this context. The benchmark comprises 15,200 training instances and 8,900 test instances, all meticulously crafted by human experts. FinDABench assesses LLMs across three dimensions: 1) Core Ability, evaluating the models’ ability to perform financial indicator calculation and corporate sentiment risk assessment; 2) Analytical Ability, determining the models’ ability to quickly comprehend textual information and analyze abnormal financial reports; and 3) Technical Ability, examining the models’ use of technical knowledge to address real-world data analysis challenges involving analysis generation and charts visualization from multiple perspectives. We will release FinDABench, and the evaluation scripts at https://github.com/xxx. FinDABench aims to provide a measure for in-depth analysis of LLM abilities and foster the advancement of LLMs in the field of financial data analysis.

Recently, adversarial machine learning attacks have posed serious security threats against practical audio signal classification systems, including speech recognition, speaker recognition, music copyright detection. Most existing studies have mainly focused on ensuring the effectiveness of attacking an audio signal classifier via creating a noise-like perturbation on the original signal, which remains a gap in preserving the human perception of adversarial audios. This paper presents a novel perspective to create adversarial audios by integrating the human perception model into the attack formulation to generate well-perceived adversarial examples. Different from conventional approaches which primarily focused on using$L_{p}$norm to preserve the audio quality, we adopt a human study to understand how human participants react to different types of music perturbations, build a Siamese Neural Network (SNN) based model to characterize the human perception. The new findings of the human perception study guide us to formulate a new computationally efficient, multiple-feature-based perception-aware (CEMF-PA) attack, which manipulates different audio signal features to find an optimal perturbed music signal against music copyright detection. This novel attack vector opens a new door to generating highly effective, well-perceived adversarial audio signals via manipulating the auditory features. Experimental results show that the proposed attack is effective against YouTube’s copyright detection. Finally, we propose the defense strategy design to make the copyright detection more robust to adversarial music signals generated by the CEMF-PA attack.

Large-scale dual-stream Vision-Language Pre-training (VLP) models provide an efficient solution for text-image retrieval tasks. Despite this, their performance often falls short of the most current single-stream models, primarily due to limited fine-grained text-image interactions. Recent trends indicate a union of these two types of networks. Some methods adopt a retrieve and rerank strategy, their performance improvements largely hinge on the single-stream encoder during inference. Other approaches utilize knowledge distillation to strengthen either the single-stream encoder or the dual-stream encoder, surpassing their previous capabilities. However, existing distillation techniques typically focus on a single knowledge type, neglecting the richer insights available in the teacher model. To bridge this gap, we introduce a Lightweight and Effective Multi-View Knowledge Distillation approach, named LEMKD, for text-image retrieval. This method effectively utilizes response-based, feature-based and relation-based knowledge, transferring the knowledge from the single-stream encoder to the dual-stream encoder. Our approach is executed on the widely used MS-COCO and Flickr30K datasets. Results demonstrate that LEMKD not only matches the exceptional performance of the most advanced single-stream models but also excels in dual-stream encoder performance amidst the recent integration of single-stream and dual-stream models.

Machine learning has been adopted for efficient cooperative spectrum sensing. However, it incurs an additional security risk due to attacks leveraging adversarial machine learning to create malicious spectrum sensing values to deceive the fusion center, called adversarial spectrum attacks. In this paper, we propose an efficient framework for detecting adversarial spectrum attacks. Our design leverages the concept of the distance to the decision boundary (DDB) observed at the fusion center and compares the training and testing DDB distributions to identify adversarial spectrum attacks. We create a computationally efficient way to compute the DDB for machine learning based spectrum sensing systems. Experimental results based on realistic spectrum data show that our method, under typical settings, achieves a high detection rate of up to 99% and maintains a low false alarm rate of less than 1%. In addition, our method to compute the DDB based on spectrum data achieves 54%–64% improvements in computational efficiency over existing distance calculation methods. The proposed DDB-based detection framework offers a practical and efficient solution for identifying malicious sensing values created by adversarial spectrum attacks.

The channel state information (CSI) has been extensively studied in the literature to facilitate authentication in wireless networks. The less focused is a systematic attack model to evaluate CSI-based authentication. Existing studies generally adopt either a random attack model that existing designs are resilient to or a specific-knowledge model that assumes certain inside knowledge for the attacker. This paper proposes a new, realistic attack model against CSI-based authentication. In this model, an attacker Eve tries to actively guess a user Alice’s CSI, and precode her signals to impersonate Alice to the verifier Bob who uses CSI to authenticate users. To make the CSI guessing effective and low-cost, we use theoretical analysis and CSI dataset validation to show that there is no need to guess CSI values in all signal propagation paths. Specifically, Eve can adopt a Dominant Path Construction (DomPathCon) strategy that only focuses on guessing the CSI values on the first few paths with the highest channel response amplitude (called dominant paths). Comprehensive experimental results show that DomPathCon is effective and achieves up to 61% attack success rates under different wireless network settings, which exposes new limitations of CSI-based authentication. We also propose designs to mitigate the adverse impact of DomPathCon.

Location spoofing attack deceiving a Wi-Fi positioning system has been studied for over a decade. However, it has been challenging to construct a practical spoofing attack in urban areas with dense coverage of legitimate Wi-Fi APs. This paper identifies the vulnerability of the Google Geolocation API, which returns the location of a mobile device based on the information of the Wi-Fi access points that the device can detect. We show that this vulnerability can be exploited by the attacker to reveal the black-box localization algorithms adopted by the Google Wi-Fi positioning system and easily launch the location spoofing attack in dense urban areas with a high success rate. Furthermore, we find that this vulnerability can also lead to severe consequences that hurt user privacy, including the leakage of sensitive information like precise locations, daily activities, and demographics. Ultimately, we discuss the potential countermeasures that may be used to mitigate this vulnerability and location spoofing attack.

Breathing rates and crowd counting can be used to verify the human presence, especially the former one can disclose a person’s physiological status. Many studies have demonstrated success in applying channel state information (CSI) to estimate the breathing rates of stationary individuals and count the number of people in motion. Due to the invisibility of radio signals, the ubiquitous deployment of wireless infrastructures, and the elimination of the line-of-sight (LOS) requirement, such wireless inference techniques can surreptitiously work and violate user privacy. However, little research has been conducted specifically in mitigating misuse of those techniques. This paper proposes new proactive countermeasures against all existing CSI-based vital signs and crowd counting inference methods. Specifically, we set up ambush locations with carefully designed wireless signals, allowing eavesdroppers to infer a false breathing rate or person count specified by the transmitter. The true breathing rate or person count is thus protected. Experimental results on software-defined radio platforms with 5 participants demonstrate the effectiveness of the proposed defenses. An eavesdropper can be misled into believing any desired breathing rate with an error of less than 1.2 bpm when the user lies on a bed in a bedroom, and 0.9 bpm when the user sits in a chair in an office room. Additionally, our proposed defense mechanisms can deceive an attacker into believing there are moving individuals in an empty room with a 100% success rate, using both Support Vector Machine (SVM) and Decision Tree (DT) classifiers.

To enable ultra-high throughputs while addressing the potential blockage problem, maintaining an adaptive access point (AP) planning is critical to mmWave networking. By investigating the hidden interaction between the environment map and the placement of mmWave APs, we develop an adaptive AP planning (E-app) approach that can accurately sense the environment dynamics, reconstruct the obstacle map, and then predict the placements of mmWave APs adaptively. Specifically, our solution leverages mmWave radio itself to sniff the unacceptable performance degradation through sensing only a small fraction of observation points that are identified by a sparsity-aware analytical model, thereby accurately triggering a prediction module for AP positioning when necessary. Extensive evaluations show a very high prediction accuracy for our solution, which can provide around 25% improvement on user throughput performance in mmWave WLANs. This intelligent AP-planning framework well handles the environment dynamics that affect the average-case network performance, which is of utmost interest for network deployers because of its usage convenience and adaptivity.

New network architectures, such as the Internet of Things (IoT), 5G, and next-generation (NextG) cellular systems, put forward emerging challenges to the design of future wireless networks toward ultra-high data rate, massive data processing, smart designs, low-cost deployment, reliability and security in dynamic environments. As one of the most promising techniques today, artificial intelligence (AI) is advocated to enable a data-driven paradigm for wireless network design. In this paper, we are motivated to review existing AI techniques and their applications for the full wireless network protocol stack toward improving network performance and security. Our goal is to summarize the current motivation, challenges, and methodology of using AI to enhance wireless networking from the physical to the application layer, and shed light on creating new AI-enabled algorithms, mechanisms, protocols, and system designs for future data-driven wireless networking.

Due to the open nature of the wireless medium, wireless communications are especially vulnerable to eavesdropping attacks. This article designs a new wireless communication system to deal with eavesdropping attacks. The proposed system can enable a legitimate receiver to get desired messages and meanwhile an eavesdropper to hear “fake” but meaningful messages by combining confidentiality and deception, thereby confusing the eavesdropper and achieving additional concealment that further protects exchanged messages. Towards this goal, we propose techniques that can conceal exchanged messages by utilizing wireless channel characteristics between the transmitter and the receiver, as well as techniques that can attract an eavesdropper to gradually approach a trap region, where the eavesdropper can get fake messages. We also provide both theoretical and empirical analysis of the established secure channel between the transmitter and the receiver. We develop a prototype system using Universal Software Defined Radio Peripherals (USRPs). Experimental results show that an eavesdropper at a trap location can receive fake information with a bit error rate (BER) close to 0, and the transmitter with multiple antennas can successfully deploy a trap area.

Federated learning (FL) provides a high efficient decentralized machine learning framework, where the training data remains distributed at remote clients in a network. Though FL enables a privacy-preserving mobile edge computing framework using IoT devices, recent studies have shown that this approach is susceptible to poisoning attacks from the side of remote clients. To address the poisoning attacks on FL, we provide atwo-phasedefense algorithm called${\underline{Lo}cal\ \underline{Ma}licious\ Facto\underline{r}}$(LoMar). In phase I, LoMar scores model updates from each remote client by measuring the relative distribution over their neighbors using a kernel density estimation method. In phase II, an optimal threshold is approximated to distinguish malicious and clean updates from a statistical perspective. Comprehensive experiments on four real-world datasets have been conducted, and the experimental results show that our defense strategy can effectively protect the FL system. Specifically, the defense performance on Amazon dataset under a label-flipping attack indicates that, compared with FG+Krum, LoMar increases the target label testing accuracy from$96.0\%$to$98.8\%$, and the overall averaged testing accuracy from$90.1\%$to$97.0\%$.

Previous adversarial audio attacks have mainly focused on ensuring the effectiveness of attacking an audio signal classifier via creating a small noise-like perturbation on the original signal. It is still unclear if an attacker is able to create audio signal perturbations that can be well perceived by human beings in addition to its attack effectiveness. In this work, we formulate the adversarial attack against music signals as a new perception-aware attack framework, which integrates human study into adversarial attack design. Specifically, we invite human participants to rate their perceived deviation based on pairs of original and perturbed music signals, and reverse-engineer the human perception process by regression analysis to predict the human-perceived deviation given a perturbed signal. The perception-aware attack is then formulated as an optimization problem that finds an optimal perturbation signal to minimize the prediction of perceived deviation from the regressed human perception model. Experiments show that the attack produces adversarial music with significantly better perceptual quality than prior work against YouTube's copyright detector.

Defense strategies have been well studied to combat Byzantine attacks that aim to disrupt cooperative spectrum sensing by sending falsified versions of spectrum sensing data to a fusion center. However, existing studies usually assume network or attackers as passive entities, e.g., assuming the prior knowledge of attacks is known or fixed. In practice, attackers can actively adopt arbitrary behaviors and avoid pre-assumed patterns or assumptions used by defense strategies. In this paper, we revisit this security vulnerability as an adversarial machine learning problem and propose a novel learning-empowered attack framework named Learning-Evaluation-Beating (LEB) to mislead the fusion center. Based on the black-box nature of the fusion center in cooperative spectrum sensing, our new perspective is to make the adversarial use of machine learning to construct a surrogate model of the fusion center's decision model. We propose a generic algorithm to create malicious sensing data using this surrogate model. Our real-world experiments show that the LEB attack is effective to beat a wide range of existing defense strategies with an up to 82 percent of success ratio. Given the gap between the proposed LEB attack and existing defenses, we introduce a non-invasive method named as influence-limiting defense, which can coexist with existing defenses to defend against LEB attack or other similar attacks. We show that this defense is highly effective and reduces the overall disruption ratio of LEB attack by up to 80 percent.

Existing research work has identified a new class of attacks that can eavesdrop on the keystrokes in a non-invasive way without infecting the target computer to install malware. The common idea is that pressing a key of a keyboard can cause a unique and subtle environmental change, which can be captured and analyzed by the eavesdropper to learn the keystrokes. For these attacks, however, a training phase must be accomplished to establish the relationship between an observed environmental change and the action of pressing a specific key. This significantly limits the impact and practicality of these attacks. In this paper, we discover that it is possible to design keystroke eavesdropping attacks without requiring the training phase. We create this attack based on the channel state information extracted from the wireless signal. To eavesdrop on keystrokes, we establish a mapping between typing each letter and its respective environmental change by exploiting the correlation among observed changes and known structures of dictionary words. To defend against this attack, we propose a reactive jamming mechanism that launches the jamming only during the typing period. Experimental results on software-defined radio platforms validate the impact of the attack and the performance of the defense.

Network tomography is an important tool to estimate link metrics from end-to-end network measurements. An implicit assumption in network tomography is that observed measurements indeed reflect the aggregate of link performance (i.e.,seeing is believing). However, it is not guaranteed today that there exists no anomaly (e.g., malicious autonomous systems and insider threats) in large-scale networks. Malicious nodes can intentionally manipulate link metrics via delaying or dropping packets to affect measurements. Will such an assumption render a vulnerability when facing attackers? The problem is of essential importance in that network tomography is developed towards effective network diagnostics and failure recovery. In this article, we demonstrate that the vulnerability is real and propose a new attack strategy, calledmeasurement integrity attack, in which malicious nodes can substantially damage a network (e.g., delaying packets) and at the same time maliciously manipulate end-to-end measurement results such that a legitimate node is misleadingly identified as the root cause of the damage (thereby becoming a scapegoat) under network tomography. We formulate three basic attack approaches and show under what conditions attacks can be successful. We also reveal conditions to detect and locate such attacks in a network. Our theoretical and experimental results show that simply trusting measurements leads to measurement integrity vulnerabilities. Thus, existing methods should be revisited accordingly for security in various applications.

We revisit the traditional framework of wireless secret key generation, where two parties leverage the wireless channel randomness to establish a secret key. The essence in the framework is to quantify channel randomness into bit sequences for key generation. Conducting randomness tests on such bit sequences has been a common practice to provide the confidence to validate whether they are random. Interestingly, despite different settings in the tests, existing studies interpret the results the same: passing tests means that the bit sequences are indeed random. In this paper, we investigate how to properly test the wireless channel randomness to ensure enough security strength and key generation efficiency. In particular, we define an adversary model that leverages the imperfect randomness of the wireless channel to search the generated key, and create a guideline to set up randomness testing and privacy amplification to eliminate security loss and achieve efficient key generation rate. We use theoretical analysis and comprehensive experiments to reveal that common practice misuses randomness testing and privacy amplification: (i) no security insurance of key strength, (ii) low efficiency of key generation rate. After revision by our guideline, security loss can be eliminated and key generation rate can be increased significantly.

Preventing the source-destination network flow information from being disclosed is pivotal for anonymous wireless network applications. However, the advance of network inference, which is able to obtain the flow information without directly measuring it, poses severe challenges towards this goal. Randomized routing is capable of hiding the flow information by injecting substantial errors to the network inference process. In this paper, we systematically study the behavior of randomized routing protocols, and categorize them into three templates, k -random-relay, k -random-neighbor and k -random-path based on their routing behaviors. We propose technical models to characterize these templates in terms of their induced inference errors and their delay costs. We also use simulations to validate the theoretical results. Our work provides the first systematic study on understanding both the benefit and the cost of using randomized routing to hide the flow information in wireless networks.

This paper investigates the joint effects of in-phase and quadrature-phase imbalance (IQI) and imperfect successive interference cancellation (ipSIC) on the cooperative Internet of Things (IoT) nonorthogonal multiple access (NOMA) networks where the Nakagami-m fading channel is taken into account. The closed-form expressions of outage probability for the far and near IoT devices are derived to evaluate the outage behaviors. For deeper insights of the performance of the considered system, the approximate outage probability and diversity order in high signal-to-noise ratio (SNR) regime are obtained. In addition, we also analyze the throughput and energy efficiency to characterize the performance of the considered system. The simulation results demonstrate that, compared with IQI, ipSIC has a greater impact on the outage performance for the near-IoT-device of the considered system. Furthermore, we also find that the outage probabilities of IoT devices can be minimized by selecting a specific power allocation scheme.

Due to the open nature of wireless medium, wireless communications are especially vulnerable to eavesdropping attacks. This paper designs a new wireless communication system to deal with eavesdropping attacks. The proposed system can enable a legitimate receiver to get desired messages and meanwhile an eavesdropper to hear “fake” but meaningful messages, thereby confusing the eavesdropper and achieving additional concealment that further protects exchanged messages. Towards this goal, we propose techniques that can conceal exchanged messages by utilizing wireless channel characteristics between the transmitter and the receiver, as well as techniques that can attract an eavesdropper to gradually approach a trap region, where the eavesdropper can get fake messages. We also implement and evaluate the proposed system on top of Universal Software Defined Radio Peripherals (USRPs). Experimental results show that an eavesdropper at a trap location can receive fake information with a bit error rate (BER) that is close to 0, and the transmitter with multiple antennas can successfully deploy a trap area.

Wireless jamming remains as one of the primary threats towards wireless security. Traditionally, jamming is able to disrupt wireless signals within, but not beyond, its covered bandwidth. In this paper, we propose a novel attack strategy, called orthogonality-sabotaging attack, against orthogonal frequency division multiple access (OFDMA) that has been widely adopted in today's wireless network standards (e.g., 4G/5G and 802.11ax). The attack intentionally introduces an unaligned narrowband jamming signal to an OFDMA network so as to destroy the orthogonality among all subcarriers in broadband signals. We theoretically formulate and optimize the attack strategies, and then use real-world experiments to show that orthogonality sabotaging is very efficient and can take down an 802.11ax network with only 1/5-1/4 of the full network bandwidth. Finally, we propose an attack identification and localization method to identify and localize orthogonality-sabotaging attacks in the fullband spectrum with 92% overall accuracy and localization errors within about 0.4 subcarrier spacing in experiments.

Traditional methods to eavesdrop keystrokes leverage some malware installed in a target computer to record the keystrokes for an adversary. Existing research work has identified a new class of attacks that can eavesdrop the keystrokes in a non-invasive way without infecting the target computer to install a malware. The common idea is that pressing a key of a keyboard can cause a unique and subtle environmental change, which can be captured and analyzed by the eavesdropper to learn the keystrokes. For these attacks, however, a training phase must be accomplished to establish the relationship between an observed environmental change and the action of pressing a specific key. This significantly limits the impact and practicality of these attacks. In this paper, we discover that it is possible to design keystroke eavesdropping attacks without requiring the training phase. We create this attack based on the channel state information extracted from wireless signal. To eavesdrop keystrokes, we establish a mapping between typing each letter and its respective environmental change by exploiting the correlation among observed changes and known structures of dictionary words. We implement this attack on software-defined radio platforms and conduct a suite of experiments to validate the impact of this attack. We point out that this paper does not propose to use wireless signal for inferring keystrokes, since such work already exists. Instead, the main goal of this paper is to propose new techniques to remove the training process, which can make existing work unpractical.

Network tomography is a vital tool to estimate link qualities from end-to-end network measurements. An implicit assumption in network tomography is that observed measurements indeed reflect the aggregate of link performance (i.e., seeing is believing). However, it is not guaranteed today that there exists no anomaly (e.g., malicious autonomous systems and insider threats) in large-scale networks. Malicious nodes can intentionally manipulate link metrics via delaying or dropping packets to affect measurements. Will such an assumption render a vulnerability when facing attackers? The problem is of essential importance in that network tomography is developed towards effective network diagnostics and failure recovery. In this paper, we demonstrate that the vulnerability is real and propose a new attack strategy, called scapegoating, in which malicious nodes can substantially damage a network (e.g., delaying packets) and at the same time maliciously manipulate end-to-end measurement results such that a legitimate node is misleadingly identified as the root cause of the damage (thereby becoming a scapegoat) under network tomography. We formulate three basic scapegoating approaches and show under what conditions attacks can be successful. We also reveal conditions to detect such attacks. Our theoretical and experimental results show that simply trusting measurements leads to scapegoating vulnerabilities. Thus, existing methods should be revisited accordingly for security in various applications.

Traditional signal processing design (e.g., frequency offset and channel estimation) at a receiver treats each packet arrival as an independent process to facilitate decoding and interpreting packet data. In this paper, we enhance the performance of this process in the wireless network domain. We propose STAteful inter-Packet signaL procEssing (STAPLE), a framework of stateful signal processing residing between the physical and link layers. STAPLE transforms the signal processing procedure into a lightweight stateful process that caches in a small-sized memory table physical and link layer header fields as packet state information. The similarity of such information among packets serves as prior knowledge to further enhance the reliability of signal processing and thus improve the wireless network performance. We implement STAPLE on USRP X300-series devices with adapted configurations for 802.11a/b/g/n/ac and 802.15.4. The STAPLE prototype is of low processing complexity and does not change any wireless standard specification. Comprehensive experimental results show that the benefit from STAPLE is universal in various wireless networks.