{
 "doc_id": "io05",
 "title": "Energy-aware lightweight intrusion detection for industrial iot networks: an empirical study",
 "sections": {
  "abstract": "This paper proposes a energy-aware lightweight intrusion detection framework for industrial IoT networks. We address botnet infiltration by introducing federated anomaly detection built on edge gateway accelerators. Experiments on 300 cases show a 62% improvement in attack detection latency. The framework demonstrates an F1 score of 0.93 with p = 0.014. Our findings establish lightweight intrusion detection as a practical remedy for botnet infiltration.",
  "introduction": "The rapid growth of sensor telemetry streams has made botnet infiltration a central concern for industrial IoT networks [2]. Recent advances in lightweight intrusion detection offer a promising direction for industrial IoT networks. However, existing deployments rarely integrate federated anomaly detection in an end-to-end manner. In this paper we design, implement, and evaluate a energy-aware pipeline that couples lightweight intrusion detection with federated anomaly detection. Our contributions include a reference architecture on edge gateway accelerators and an evaluation across 300 cases. The remainder of this paper reviews prior work, details the methodology, and reports results.",
  "literature_review": "Early studies of lightweight intrusion detection focused primarily on feasibility within industrial IoT networks [2]. Subsequent work between 2015 and 2022 shifted attention toward federated anomaly detection [4]. Several authors report that botnet infiltration remains the dominant failure mode in production systems [8]. Benchmark efforts using edge gateway accelerators demonstrate significant gains in attack detection latency [4]. Surveys of industrial IoT networks consistently identify sensor telemetry streams as the most vulnerable asset class [2]. Despite this progress, no prior work unifies lightweight intrusion detection and federated anomaly detection under realistic workloads.",
  "methodology": "Our research design follows a controlled experimental protocol over 300 cases collected from 2015 to 2022. The core of the system is federated anomaly detection implemented on edge gateway accelerators. Each instance of sensor telemetry streams is normalized, segmented, and assigned a cryptographic provenance tag. We configure the lightweight intrusion detection layer with a three-stage validation pipeline to suppress botnet infiltration. Statistical significance is assessed with paired tests at alpha = 0.05. All experiments are repeated five times and we report the mean with confidence intervals.",
  "results": "The proposed framework improves attack detection latency by 62% relative to the strongest baseline. Across 300 cases the system attains an F1 score of 0.93 with p = 0.014. Ablation shows that removing federated anomaly detection degrades performance by 21%. Latency overhead introduced by the lightweight intrusion detection layer remains below acceptable operational thresholds. These results confirm that energy-aware integration of edge gateway accelerators is feasible at scale. Error analysis attributes most residual failures to noisy sensor telemetry streams.",
  "conclusion": "We presented a energy-aware lightweight intrusion detection framework that mitigates botnet infiltration in industrial IoT networks. Evaluation over 300 cases demonstrated a 62% improvement in attack detection latency. Future work will extend federated anomaly detection to cross-organizational settings. We will also study the long-term governance of sensor telemetry streams under this architecture."
 },
 "references": [
  "O. Okafor, \"Federated anomaly detection: opportunities and challenges,\" Future Generation Computer Systems, 2020",
  "M. Martinez, \"On the limits of lightweight intrusion detection for sensor telemetry streams,\" IEEE Internet of Things Journal, 2021",
  "P. Patel, \"Benchmarking attack detection latency under lightweight intrusion detection,\" IEEE Transactions on Industrial Informatics, 2018",
  "N. Nguyen, \"Adaptive federated anomaly detection for dynamic environments,\" ACM Computing Surveys, 2017",
  "K. Kowalski, \"Edge gateway accelerators in practice,\" Journal of Network and Computer Applications, 2019",
  "J. Johansson, \"Mitigating botnet infiltration using federated anomaly detection,\" Future Generation Computer Systems, 2019",
  "T. Tanaka, \"Towards energy-aware industrial iot networks,\" Future Generation Computer Systems, 2021",
  "C. Chen, \"A survey of lightweight intrusion detection in industrial iot networks,\" Sensors, 2023",
  "C. Chen, \"Deep evaluation of federated anomaly detection at scale,\" Journal of Network and Computer Applications, 2020",
  "N. Nguyen, \"Benchmarking average commute delay under spatio-temporal graph networks,\" ACM Computing Surveys, 2016",
  "K. Kowalski, \"A framework for city-scale deployment of adaptive signal controllers,\" Expert Systems with Applications, 2015"
 ],
 "tier": 2,
 "year": 2019,
 "topic": "io",
 "vals": {
  "imp": 62,
  "imp2": 21,
  "n": 300,
  "p": 0.014,
  "f1": 0.93,
  "y0": 2015,
  "y1": 2022,
  "r1": 2,
  "r2": 4,
  "r3": 8,
  "r4": 4
 }
}