arXiv subjects

Israat Haque

Publications and source records attributed to Israat Haque.

18 recordsLinked to original sources

arXiv2026

Effective Pivot Attack Detection via System and Network Information

Perimeter-based security appliances, such as firewalls or Intrusion Detection Systems, are ineffective against modern attacks that use pivoting, wherein attackers "pivot" traffic through compromised hosts to gain access to additional targets that would otherwise be inaccessible. Due to the legitimate appearance of the relayed traffic, pivoting is extremely difficult to detect. Although the consequences of these attacks are known to be severe, existing defenses suffer from drawbacks such as high processing delays, low accuracy, or reliance on network-wide participation, making them inconvenient or even ineffective. This work presents Stitch, a host-based system that uses the programmable kernel to detect pivoting in real time. By observing host-traversing flows, Stitch uses process tracing to effectively combine system and network-level information, connecting incoming and outgoing communications and identifying pivoting characteristics between them. Showing 31% gains in accuracy over state-of-the-art pivoting defenses and a maximum false positive rate of 0.006% over two separate real-world deployments, Stitch covers the gap in current pivot detection solutions by providing accurate, lightweight, and independent coverage for vulnerable hosts in a network.

cs.CR↗

arXiv2026

Performance Characterization of dApps in Open Radio Access Networks

Despite recommendations to deploy real-time Open Radio Access Network (O-RAN) applications (dApps) in containerized environments, existing approaches predominantly rely on bare-metal servers. Moreover, current dApp deployments offer limited visibility into the resource usage patterns of both intelligent and non-intelligent dApps, hindering informed deployment decisions. This work addresses these gaps by implementing and evaluating representative dApps across multiple deployment scenarios (bare-metal and containers) to characterize the trade-offs in latency, scalability, and resource utilization. Additionally, we identify key performance bottlenecks and demonstrate how offloading dApps to emerging hardware accelerators, such as smart Network Interface Cards (NICs), can alleviate these limitations and improve real-time responsiveness in O-RAN systems.

cs.NI↗

arXiv2026

Characterizing Performance-Energy Trade-offs of Large Language Models in Multi-Request Workflows

Large language models (LLMs) are increasingly used in applications forming multi-request workflows like document summarization, search-based copilots, and multi-agent programming. While these workflows unlock richer functionality, they also amplify latency and energy demand during inference. Existing measurement and benchmarking efforts either focus on assessing LLM inference systems or consider single-request evaluations, overlooking workflow dependencies and cross-request interactions unique to multi-request workflows. Moreover, the energy usage of such interdependent LLM calls remains underexplored. To address these gaps, this paper presents the first systematic characterization of performance-energy trade-offs in multi-request LLM inference. We develop four representative workloads capturing sequential, interactive, agentic, and composite patterns common in modern deployments. Using an NVIDIA A100 testbed with state-of-the-art serving systems (vLLM and Parrot), we analyze how key energy knobs affect latency, throughput, and component-level energy use. Our findings reveal batch size as the most impactful lever, though benefits are workload dependent. While optimal batching benefits workloads with large shared prompts, it is ineffective for sequential summarization and only partially effective for multi-agent coding. GPU power capping provides modest but predictable savings, while output length induces linear energy scaling with limited efficiency gains. We further show that engine-level optimizations in vLLM maintain higher GPU utilization and efficiency, especially for decode-heavy workloads, while Parrot's workflow-aware scheduling achieves lower energy consumption under strict power constraints. These findings offer actionable guidelines for developers and system operators designing performance- and energy-aware LLM serving systems in emerging multi-request workflows.

cs.DC↗

arXiv2026

An Explainable Failure Prediction Framework for Neural Networks in Radio Access Networks

As 5G networks continue to evolve to deliver high speed, low latency, and reliable communications, ensuring uninterrupted service has become increasingly critical. While millimeter wave (mmWave) frequencies enable gigabit data rates, they are highly susceptible to environmental factors, often leading to radio link failures (RLF). Predictive models leveraging radio and weather data have been proposed to address this issue; however, many operate as black boxes, offering limited transparency for operational deployment. This work bridges that gap by introducing a framework that combines explainability based feature pruning with model refinement. Our framework can be integrated into state of the art predictors such as GNN Transformer and LSTM based architectures for RLF prediction, enabling the development of accurate and explainability guided models in 5G networks. It provides insights into the contribution of input features and the decision making logic of neural networks, leading to lighter and more scalable models. When applied to RLF prediction, our framework unveils that weather data contributes minimally to the forecast in extensive real world datasets, which informs the design of a leaner model with 50 percent fewer parameters and improved F1 scores with respect to the state of the art solution. Ultimately, this work empowers network providers to evaluate and refine their neural network based prediction models for better interpretability, scalability, and performance.

cs.NI↗

arXiv2025

A Comprehensive Survey on Smart Home IoT Fingerprinting: From Detection to Prevention and Practical Deployment

Smart homes are increasingly populated with heterogeneous Internet of Things (IoT) devices that interact continuously with users and the environment. This diversity introduces critical challenges in device identification, authentication, and security, where fingerprinting techniques have emerged as a key approach. In this survey, we provide a comprehensive analysis of IoT fingerprinting specifically in the context of smart homes, examining methods for device and their event detection, classification, and intrusion prevention. We review existing techniques, e.g., network traffic analysis or machine learning-based schemes, highlighting their applicability and limitations in home environments characterized by resource-constrained devices, dynamic usage patterns, and privacy requirements. Furthermore, we discuss fingerprinting system deployment challenges like scalability, interoperability, and energy efficiency, as well as emerging opportunities enabled by generative AI and federated learning. Finally, we outline open research directions that can advance reliable and privacy-preserving fingerprinting for next-generation smart home ecosystems.

cs.CR↗

arXiv2025

On the Sustainability of AI Inferences in the Edge

The proliferation of the Internet of Things (IoT) and its cutting-edge AI-enabled applications (e.g., autonomous vehicles and smart industries) combine two paradigms: data-driven systems and their deployment on the edge. Usually, edge devices perform inferences to support latency-critical applications. In addition to the performance of these resource-constrained edge devices, their energy usage is a critical factor in adopting and deploying edge applications. Examples of such devices include Raspberry Pi (RPi), Intel Neural Compute Stick (INCS), NVIDIA Jetson nano (NJn), and Google Coral USB (GCU). Despite their adoption in edge deployment for AI inferences, there is no study on their performance and energy usage for informed decision-making on the device and model selection to meet the demands of applications. This study fills the gap by rigorously characterizing the performance of traditional, neural networks, and large language models on the above-edge devices. Specifically, we analyze trade-offs among model F1 score, inference time, inference power, and memory usage. Hardware and framework optimization, along with external parameter tuning of AI models, can balance between model performance and resource usage to realize practical edge AI deployments.

cs.LG↗

arXiv2025

Performance Analysis of Decentralized Federated Learning Deployments

The widespread adoption of smartphones and smart wearable devices has led to the widespread use of Centralized Federated Learning (CFL) for training powerful machine learning models while preserving data privacy. However, CFL faces limitations due to its overreliance on a central server, which impacts latency and system robustness. Decentralized Federated Learning (DFL) is introduced to address these challenges. It facilitates direct collaboration among participating devices without relying on a central server. Each device can independently connect with other devices and share model parameters. This work explores crucial factors influencing the convergence and generalization capacity of DFL models, emphasizing network topologies, non-IID data distribution, and training strategies. We first derive the convergence rate of different DFL model deployment strategies. Then, we comprehensively analyze various network topologies (e.g., linear, ring, star, and mesh) with different degrees of non-IID data and evaluate them over widely adopted machine learning models (e.g., classical, deep neural networks, and Large Language Models) and real-world datasets. The results reveal that models converge to the optimal one for IID data. However, the convergence rate is inversely proportional to the degree of non-IID data distribution. Our findings will serve as valuable guidelines for designing effective DFL model deployments in practical applications.

cs.LG↗

arXiv2024

Fast Prototyping of Distributed Stream Processing Applications with stream2gym

Stream processing applications have been widely adopted due to real-time data analytics demands, e.g., fraud detection, video analytics, IoT applications. Unfortunately, prototyping and testing these applications is still a cumbersome process for developers that usually requires an expensive testbed and deep multi-disciplinary expertise, including in areas such as networking, distributed systems, and data engineering. As a result, it takes a long time to deploy stream processing applications into production and yet users face several correctness and performance issues. In this paper, we present stream2gym, a tool for the fast prototyping of large-scale distributed stream processing applications. stream2gym builds on Mininet, a widely adopted network emulation platform, and provides a high-level interface to enable developers to easily test their applications under various operating conditions. We demonstrate the benefits of stream2gym by prototyping and testing several applications as well as reproducing key findings from prior research work in video analytics and network traffic monitoring. Moreover, we show stream2gym presents accurate results compared to a hardware testbed while consuming a small amount of resources (enough to be supported in a single commodity laptop even when emulating a dozen of processing nodes).

cs.DC↗

arXiv2024

A Generalized Transformer-based Radio Link Failure Prediction Framework in 5G RANs

Radio link failure (RLF) prediction system in Radio Access Networks (RANs) is critical for ensuring seamless communication and meeting the stringent requirements of high data rates, low latency, and improved reliability in 5G networks. However, weather conditions such as precipitation, humidity, temperature, and wind impact these communication links. Usually, historical radio link Key Performance Indicators (KPIs) and their surrounding weather station observations are utilized for building learning-based RLF prediction models. However, such models must be capable of learning the spatial weather context in a dynamic RAN and effectively encoding time series KPIs with the weather observation data. Existing works fail to incorporate both of these essential design aspects of the prediction models. This paper fills the gap by proposing GenTrap, a novel RLF prediction framework that introduces a graph neural network (GNN)-based learnable weather effect aggregation module and employs state-of-the-art time series transformer as the temporal feature extractor for radio link failure prediction. The proposed aggregation method of GenTrap can be integrated into any existing prediction model to achieve better performance and generalizability. We evaluate GenTrap on two real-world datasets (rural and urban) with 2.6 million KPI data points and show that GenTrap offers a significantly higher F1-score (0.93 for rural and 0.79 for urban) compared to its counterparts while possessing generalization capability.

cs.AI↗

arXiv2024

Root Cause Analysis of Anomalies in 5G RAN Using Graph Neural Network and Transformer

The emergence of 5G technology marks a significant milestone in developing telecommunication networks, enabling exciting new applications such as augmented reality and self-driving vehicles. However, these improvements bring an increased management complexity and a special concern in dealing with failures, as the applications 5G intends to support heavily rely on high network performance and low latency. Thus, automatic self-healing solutions have become effective in dealing with this requirement, allowing a learning-based system to automatically detect anomalies and perform Root Cause Analysis (RCA). However, there are inherent challenges to the implementation of such intelligent systems. First, there is a lack of suitable data for anomaly detection and RCA, as labelled data for failure scenarios is uncommon. Secondly, current intelligent solutions are tailored to LTE networks and do not fully capture the spatio-temporal characteristics present in the data. Considering this, we utilize a calibrated simulator, Simu5G, and generate open-source data for normal and failure scenarios. Using this data, we propose Simba, a state-of-the-art approach for anomaly detection and root cause analysis in 5G Radio Access Networks (RANs). We leverage Graph Neural Networks to capture spatial relationships while a Transformer model is used to learn the temporal dependencies of the data. We implement a prototype of Simba and evaluate it over multiple failures. The outcomes are compared against existing solutions to confirm the superiority of Simba.

cs.NI↗

arXiv2024

STX-Vote: Improving Reliability with Bit Voting in Synchronous Transmission-based IoT Networks

Industrial Internet of Things (IIoT) networks must meet strict reliability, latency, and low energy consumption requirements. However, traditional low-power wireless protocols are ineffective in finding a sweet spot for balancing these performance metrics. Recently, network flooding protocols based on Synchronous Transmissions (STX) have been proposed for better performance in reliability-critical IIoT, where simultaneous transmissions are possible without packet collisions. STX-based protocols can offer a competitive edge over routing-based protocols, particularly in dependability. However, they notably suffer from the beating effect, a physical layer phenomenon that results in sinusoidal interference across a packet and, consequently, packet loss. Thus, we introduce STX-Vote, an error correction scheme that can handle errors caused by beating effects. Importantly, we utilize transmission redundancy already inherent within STX protocols so do not incur additional on-air overhead. Through simulation, we demonstrate STX-Vote can provide a 40% increase in reliability. We subsequently implement STX-Vote on nRF52840-DK devices and perform extensive experiments. The results confirm that STX-Vote improves reliability by 25-28% for BLE 5 PHYs and 8% for IEEE 802.15.4; thus, it can complement existing error correction schemes.

cs.NI↗

arXiv2024

Araucaria: Simplifying INC Fault Tolerance with High-Level Intents

Network programmability allows modification of fine-grain data plane functionality. The performance benefits of data plane programmability have motivated many researchers to offload computation that previously operated only on servers to the network, creating the notion of in-network computing (INC). Because failures can occur in the data plane, fault tolerance mechanisms are essential for INC. However, INC operators and developers must manually set fault tolerance requirements using domain knowledge to change the source code. These manually set requirements may take time and lead to errors in case of misconfiguration. In this work, we present Araucaria, a system that aims to simplify the definition and implementation of fault tolerance requirements for INC. The system allows requirements specification using an intent language, which enables the expression of consistency and availability requirements in a constrained natural language. A refinement process translates the intent and incorporates the essential building blocks and configurations into the INC code. We present a prototype of Araucaria and analyze the end-to-end system behavior. Experiments demonstrate that the refinement scales to multiple intents and that the system provides fault tolerance with negligible overhead in failure scenarios.

cs.NI↗

arXiv2024

A Calibrated and Automated Simulator for Innovations in 5G

The rise of 5G deployments has created the environment for many emerging technologies to flourish. Self-driving vehicles, Augmented and Virtual Reality, and remote operations are examples of applications that leverage 5G networks' support for extremely low latency, high bandwidth, and increased throughput. However, the complex architecture of 5G hinders innovation due to the lack of accessibility to testbeds or realistic simulators with adequate 5G functionalities. Also, configuring and managing simulators are complex and time consuming. Finally, the lack of adequate representative data hinders the data-driven designs in 5G campaigns. Thus, we calibrated a system-level open-source simulator, Simu5G, following 3GPP guidelines to enable faster innovation in the 5G domain. Furthermore, we developed an API for automatic simulator configuration without knowing the underlying architectural details. Finally, we demonstrate the usage of the calibrated and automated simulator by developing an ML-based anomaly detection in a 5G Radio Access Network (RAN).

cs.NI↗

arXiv2023

Learn to Compress (LtC): Efficient Learning-based Streaming Video Analytics

Video analytics are often performed as cloud services in edge settings, mainly to offload computation, and also in situations where the results are not directly consumed at the video sensors. Sending high-quality video data from the edge devices can be expensive both in terms of bandwidth and power use. In order to build a streaming video analytics pipeline that makes efficient use of these resources, it is therefore imperative to reduce the size of the video stream. Traditional video compression algorithms are unaware of the semantics of the video, and can be both inefficient and harmful for the analytics performance. In this paper, we introduce LtC, a collaborative framework between the video source and the analytics server, that efficiently learns to reduce the video streams within an analytics pipeline. Specifically, LtC uses the full-fledged analytics algorithm at the server as a teacher to train a lightweight student neural network, which is then deployed at the video source. The student network is trained to comprehend the semantic significance of various regions within the videos, which is used to differentially preserve the crucial regions in high quality while the remaining regions undergo aggressive compression. Furthermore, LtC also incorporates a novel temporal filtering algorithm based on feature-differencing to omit transmitting frames that do not contribute new information. Overall, LtC is able to use 28-35% less bandwidth and has up to 45% shorter response delay compared to recently published state of the art streaming frameworks while achieving similar analytics performance.

eess.IV↗

arXiv2023

Preventing Discriminatory Decision-making in Evolving Data Streams

Bias in machine learning has rightly received significant attention over the last decade. However, most fair machine learning (fair-ML) work to address bias in decision-making systems has focused solely on the offline setting. Despite the wide prevalence of online systems in the real world, work on identifying and correcting bias in the online setting is severely lacking. The unique challenges of the online environment make addressing bias more difficult than in the offline setting. First, Streaming Machine Learning (SML) algorithms must deal with the constantly evolving real-time data stream. Second, they need to adapt to changing data distributions (concept drift) to make accurate predictions on new incoming data. Adding fairness constraints to this already complicated task is not straightforward. In this work, we focus on the challenges of achieving fairness in biased data streams while accounting for the presence of concept drift, accessing one sample at a time. We present Fair Sampling over Stream ($FS^2$), a novel fair rebalancing approach capable of being integrated with SML classification algorithms. Furthermore, we devise the first unified performance-fairness metric, Fairness Bonded Utility (FBU), to evaluate and compare the trade-off between performance and fairness of different bias mitigation methods efficiently. FBU simplifies the comparison of fairness-performance trade-offs of multiple techniques through one unified and intuitive evaluation, allowing model designers to easily choose a technique. Overall, extensive evaluations show our measures surpass those of other fair online techniques previously reported in the literature.

cs.LG↗

arXiv2022

BlueTiSCH: A Multi-PHY Simulation of Low-Power 6TiSCH IoT Networks

Low-power wireless IoT networks have traditionally operated over a single physical layer (PHY) -- many based on the IEEE 802.15.4 standard. However, recent low-power wireless chipsets offer both the IEEE 802.15.4 and all four PHYs of the Bluetooth 5 (BT 5) standard. This introduces the intriguing possibility that IoT solutions might not necessarily be bound by the limits of a single PHY, and could actively or proactively adapt their PHY depending on RF or networking conditions (e.g., to offer a higher throughput or a longer radio range). Several recent studies have explored such use-cases. However, these studies lack comprehensive evaluation over various metrics (such as reliability, latency, and energy) with regards to scalability and the Radio Frequency (RF) environment. In this work we evaluate the performance of IEEE 802.15.4 and the four BT 5 2.4GHz PHY options for the recently completed IETF 6TiSCH low-power wireless standard. To the best of our knowledge, this is the first work to directly compare these PHYs in identical settings. Specifically, we use a recently released 6TiSCH simulator, TSCH-Sim, to compare these PHY options in networks of up to 250 nodes over different RF environments (home, industrial, and outdoor), and highlight from these results how different PHY options might be better suited to particular application use-cases.

cs.NI↗

arXiv2021

ACE: an Accurate and Cost-Effective Measurement System in SDN

Packet-level traffic measurement is essential in applications like QoS, traffic engineering, or anomaly detection. Software-Defined Networking (SDN) enables efficient and dynamic network configuration that we can deploy for fine-grained network state measurement. As the state changes dynamically, the sampling frequency must cope with that for accurate measurement. At the same time, we must consider the measurement cost, e.g., nodes' resource utilization. Existing works fall short in offering an optimal model to balance both the measurement accuracy and cost. We fill that gap by proposing ACE, an accurate and cost-effective measurement system that builds on a multi-objective optimization problem. As the optimization problem is NP-hard, we develop a heuristic. We solve the model using CPLEX; then implement a prototype of ACE in Mininet over real-world network topologies. The results confirm that ACE outperforms its counterparts in balancing both accuracy and cost.

cs.NI↗

arXiv2021

Embedded vs. External Controllers in Software-Defined IoT Networks

The flexible and programmable architectural model offered by Software-Defined Networking (SDN) has re-imagined modern networks. Supported by powerful hardware and high-speed communications between devices and the controller, SDN provides a means to virtualize control functionality and enable rapid network reconfiguration in response to dynamic application requirements. However, recent efforts to apply SDN's centralized control model to the Internet of Things (IoT) have identified significant challenges due to the constraints faced by embedded low-power devices and networks that reside at the IoT edge. In particular, reliance on external SDN controllers on the backbone network introduces a performance bottleneck (e.g., latency). To this end, we advocate a case for supporting Software-Defined IoT networks through the introduction of lightweight SDN controllers directly on the embedded hardware. We firstly explore the performance of two popular SDN implementations for IoT mesh networks, $μ$SDN and SDN-WISE, showing the former demonstrates considerable gains over the latter. We consequently employ $μ$SDN to conduct a study of embedded vs. external SDN controller performance. We highlight how the advantage of an embedded controller is reduced as the network scales, and quantify a point at which an external controller should be used for larger networks.

cs.NI↗