Ad Hoc Networks最新文献_第7页

RELTO: A reliability-oriented DRL approach with context-aware adaptive reward weighting for multi-objective task offloading in MEC 基于上下文感知自适应奖励加权的MEC多目标任务卸载可靠性导向DRL方法

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-10-30 DOI: 10.1016/j.adhoc.2025.104065

Anam Nasir, Xiang He, Teng Wang, Haomai Shi, Zhongjie Wang

Task offloading in Mobile Edge Computing (MEC) enables resource-constrained IoT devices to reduce latency and energy consumption while enhancing computational performance. However, designing effective offloading strategies presents a multi-objective optimization challenge, particularly in ensuring task reliability while optimizing energy efficiency and latency under dynamic conditions with unpredictable task failures caused by fluctuating computation demands and unstable communication links that severely degrade Quality of Service (QoS). Existing Deep Reinforcement Learning (DRL) approaches struggle to address these reliability-centered challenges, primarily due to their limited adaptability to dynamic reliability requirements, inadequate hybrid action space management, and insufficient handling of complex system state representations. To address these limitations, this work formulates a reliability-aware task offloading problem that explicitly integrates communication and computation reliability with latency and energy consumption into a multi-objective optimization formulation. To solve this optimization problem, the proposed Reliability Energy Latency Task Offloading (RELTO) algorithm employs Proximal Policy Optimization (PPO) within hybrid action spaces and incorporates a context-aware adaptive reward weighting mechanism driven by dual-attention architecture. The mechanism dynamically adjusts objective prioritization particularly emphasizing reliability based on real-time conditions, while attention-based state representation enables proactive decision-making through temporal pattern recognition. Extensive experiments in simulated MEC environments demonstrate that RELTO achieves optimal trade-offs across the key performance metrics, providing a more adaptive and robust solution for multi-objective task offloading.

移动边缘计算（MEC）中的任务卸载使资源受限的物联网设备能够减少延迟和能耗，同时提高计算性能。然而，设计有效的卸载策略是一个多目标优化挑战，特别是在动态条件下，由于计算需求波动和通信链路不稳定导致的不可预测的任务失败严重降低了服务质量（QoS），在保证任务可靠性的同时优化能效和延迟。现有的深度强化学习（DRL）方法难以解决这些以可靠性为中心的挑战，主要是因为它们对动态可靠性要求的适应性有限，混合动作空间管理不足，以及对复杂系统状态表示的处理不足。为了解决这些限制，本工作制定了一个可靠性感知任务卸载问题，该问题明确地将通信和计算可靠性与延迟和能耗集成到一个多目标优化公式中。为了解决这一优化问题，本文提出的可靠性能量延迟任务卸载（RELTO）算法在混合动作空间中采用了近端策略优化（PPO），并结合了由双注意力架构驱动的上下文感知自适应奖励加权机制。该机制动态调整客观优先级，特别强调基于实时条件的可靠性，而基于注意力的状态表示通过时间模式识别实现主动决策。在模拟MEC环境中进行的大量实验表明，RELTO在关键性能指标之间实现了最佳权衡，为多目标任务卸载提供了更具适应性和鲁棒性的解决方案。

{"title":"RELTO: A reliability-oriented DRL approach with context-aware adaptive reward weighting for multi-objective task offloading in MEC","authors":"Anam Nasir, Xiang He, Teng Wang, Haomai Shi, Zhongjie Wang","doi":"10.1016/j.adhoc.2025.104065","DOIUrl":"10.1016/j.adhoc.2025.104065","url":null,"abstract":"<div><div>Task offloading in Mobile Edge Computing (MEC) enables resource-constrained IoT devices to reduce latency and energy consumption while enhancing computational performance. However, designing effective offloading strategies presents a multi-objective optimization challenge, particularly in ensuring task reliability while optimizing energy efficiency and latency under dynamic conditions with unpredictable task failures caused by fluctuating computation demands and unstable communication links that severely degrade Quality of Service (QoS). Existing Deep Reinforcement Learning (DRL) approaches struggle to address these reliability-centered challenges, primarily due to their limited adaptability to dynamic reliability requirements, inadequate hybrid action space management, and insufficient handling of complex system state representations. To address these limitations, this work formulates a reliability-aware task offloading problem that explicitly integrates communication and computation reliability with latency and energy consumption into a multi-objective optimization formulation. To solve this optimization problem, the proposed Reliability Energy Latency Task Offloading (RELTO) algorithm employs Proximal Policy Optimization (PPO) within hybrid action spaces and incorporates a context-aware adaptive reward weighting mechanism driven by dual-attention architecture. The mechanism dynamically adjusts objective prioritization particularly emphasizing reliability based on real-time conditions, while attention-based state representation enables proactive decision-making through temporal pattern recognition. Extensive experiments in simulated MEC environments demonstrate that RELTO achieves optimal trade-offs across the key performance metrics, providing a more adaptive and robust solution for multi-objective task offloading.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104065"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145419849","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Location-based early detection and prevention of DDoS attacks in mMTC networks mMTC网络中基于位置的DDoS攻击早期检测与防范

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-11-06 DOI: 10.1016/j.adhoc.2025.104090

Zeinab Rezaeifar , Zahra Alavikia , Changhee Hahn

The Random Access (RA) procedure of the current 3GPP cellular network has been adopted for small data packet transmissions by massive Machine Type Communication Devices (MTCDs). However, the initial steps of the RA procedure lack an authentication mechanism, making it susceptible to Distributed Denial of Service (DDoS) attacks, particularly in massive access scenarios. In these cases, attackers can hide among a large number of legitimate stationary devices with limited processing capabilities, such as installed sensors or smart meters. To address this issue, this paper proposes an early DDoS attack detection and prevention method that leverages the Timing Advance (TA) information from stationary MTCDs. The proposed method detects the approximate location of malicious devices sending consecutive preamble codes and blocks them by withholding Resource Blocks (RBs) during the RA procedure. Numerical results from a simulated Physical Random Access Channel (PRACH) for Machine Type Communications (MTC), considering noise and multipath effects, demonstrate the effectiveness of the proposed method in detecting and mitigating DDoS attacks. Under intense attack scenarios, the proposed method effectively identifies network attackers while reducing RA delay and RB consumption for MTCDs by approximately 50% compared to the baseline. This improvement enhances overall network performance and sustainability.

当前3GPP蜂窝网络采用随机存取（RA）程序进行大规模机型通信设备（mtcd）的小数据包传输。然而，RA过程的初始步骤缺乏身份验证机制，使其容易受到分布式拒绝服务（DDoS）攻击，特别是在大规模访问场景中。在这些情况下，攻击者可以隐藏在大量处理能力有限的合法固定设备中，例如安装的传感器或智能电表。为了解决这一问题，本文提出了一种利用静止mtcd的时序提前（TA）信息的早期DDoS攻击检测和预防方法。提出的方法检测发送连续前导码的恶意设备的大致位置，并在RA过程中通过扣留资源块（Resource block, RBs）来阻止它们。在考虑噪声和多径影响的情况下，模拟机器类型通信（MTC）的物理随机接入信道（PRACH）的数值结果证明了该方法在检测和减轻DDoS攻击方面的有效性。在激烈的攻击场景下，该方法有效地识别网络攻击者，同时将mtcd的RA延迟和RB消耗与基线相比减少了约50%。这种改进提高了整体网络性能和可持续性。

{"title":"Location-based early detection and prevention of DDoS attacks in mMTC networks","authors":"Zeinab Rezaeifar , Zahra Alavikia , Changhee Hahn","doi":"10.1016/j.adhoc.2025.104090","DOIUrl":"10.1016/j.adhoc.2025.104090","url":null,"abstract":"<div><div>The Random Access (RA) procedure of the current 3GPP cellular network has been adopted for small data packet transmissions by massive Machine Type Communication Devices (MTCDs). However, the initial steps of the RA procedure lack an authentication mechanism, making it susceptible to Distributed Denial of Service (DDoS) attacks, particularly in massive access scenarios. In these cases, attackers can hide among a large number of legitimate stationary devices with limited processing capabilities, such as installed sensors or smart meters. To address this issue, this paper proposes an early DDoS attack detection and prevention method that leverages the Timing Advance (TA) information from stationary MTCDs. The proposed method detects the approximate location of malicious devices sending consecutive preamble codes and blocks them by withholding Resource Blocks (RBs) during the RA procedure. Numerical results from a simulated Physical Random Access Channel (PRACH) for Machine Type Communications (MTC), considering noise and multipath effects, demonstrate the effectiveness of the proposed method in detecting and mitigating DDoS attacks. Under intense attack scenarios, the proposed method effectively identifies network attackers while reducing RA delay and RB consumption for MTCDs by approximately 50% compared to the baseline. This improvement enhances overall network performance and sustainability.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104090"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145468553","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

TOP: A forward and reverse offloading strategy in MEC-enabled Cooperative Vehicle–Infrastructure System TOP：基于mec的协同车辆基础设施系统中的正向和反向卸载策略

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-11-01 DOI: 10.1016/j.adhoc.2025.104058

Dun Cao , Weijia Xiao , Dan Cai , Yifan Yang , Fayez Alqahtani , Jin Wang

Enabled by Mobile Edge Computing (MEC) equipped on Base Station (BS), Collaborative Vehicle–Infrastructure Systems (CVIS) can provide efficient and reliable computing services for mobile vehicles. Vehicles can achieve intelligent applications such as autonomous driving by forward offloading tasks to base stations. However, most existing studies focus on the BS merely as a task receiver and integrator in CVIS, and neglecting its role as a task generator for information processing. When the BS is overloaded, CVIS will deteriorate drastically. Reverse offloading from BS to idle vehicles can relieve the pressure. Nonetheless, with the huge volume of tasks generated by both some task vehicles and the BS, how to select appropriate offloading and resource allocation strategies is a challenge. The situation will become more complex when facing heterogeneous nodes, i.e., task vehicles, the BS, and the idle vehicles in the range of the task vehicle or in the range of the BS but out of the task vehicle in dynamic scenarios. Thus, we propose an optimization problem joint multiple task offloading and resource partitioning to maximize the average task satisfaction of the system. To address the above proposed optimization problem, we propose Two-way Offloading & Partitioning (TOP) strategy, where a Two-way Collaborative Edge Node Dividing and Offloading Algorithm determines the cooperative edge nodes for different tasks and obtains the offloading strategy for each task set. Furthermore, we optimize the resource partitioning using the Genetic Algorithm to avoid resource wastage while enhancing the overall satisfaction of the system. Extensive experimental results show that our proposed TOP strategy improves average system satisfaction by up to 33% compared to other baseline strategies.

通过基站（BS）上的移动边缘计算（MEC），协同车辆基础设施系统（CVIS）可以为移动车辆提供高效可靠的计算服务。车辆可以通过将任务转发给基站来实现自动驾驶等智能应用。然而，现有的研究大多只关注脑电信号在CVIS中的任务接收和集成商作用，而忽视了脑电信号在信息加工中的任务生成作用。当BS过载时，CVIS将急剧恶化。倒车卸至怠速车辆可减轻压力。然而，由于某些任务车和BS都产生了大量的任务，如何选择合适的卸载和资源分配策略是一个挑战。当面对异构节点，即任务车辆、BS、在任务车辆范围内的空闲车辆或在BS范围内但不在任务车辆的动态场景时，情况会变得更加复杂。因此，我们提出了一个多任务卸载和资源分配相结合的优化问题，以最大限度地提高系统的平均任务满意度。为了解决上述优化问题，我们提出了双向卸载和卸载（TOP）策略，其中双向协作边缘节点划分和卸载算法确定不同任务的合作边缘节点，并获得每个任务集的卸载策略。此外，我们利用遗传算法优化资源分配，避免资源浪费，同时提高系统的整体满意度。广泛的实验结果表明，与其他基线策略相比，我们提出的TOP策略将平均系统满意度提高了33%。

{"title":"TOP: A forward and reverse offloading strategy in MEC-enabled Cooperative Vehicle–Infrastructure System","authors":"Dun Cao , Weijia Xiao , Dan Cai , Yifan Yang , Fayez Alqahtani , Jin Wang","doi":"10.1016/j.adhoc.2025.104058","DOIUrl":"10.1016/j.adhoc.2025.104058","url":null,"abstract":"<div><div>Enabled by Mobile Edge Computing (MEC) equipped on Base Station (BS), Collaborative Vehicle–Infrastructure Systems (CVIS) can provide efficient and reliable computing services for mobile vehicles. Vehicles can achieve intelligent applications such as autonomous driving by forward offloading tasks to base stations. However, most existing studies focus on the BS merely as a task receiver and integrator in CVIS, and neglecting its role as a task generator for information processing. When the BS is overloaded, CVIS will deteriorate drastically. Reverse offloading from BS to idle vehicles can relieve the pressure. Nonetheless, with the huge volume of tasks generated by both some task vehicles and the BS, how to select appropriate offloading and resource allocation strategies is a challenge. The situation will become more complex when facing heterogeneous nodes, i.e., task vehicles, the BS, and the idle vehicles in the range of the task vehicle or in the range of the BS but out of the task vehicle in dynamic scenarios. Thus, we propose an optimization problem joint multiple task offloading and resource partitioning to maximize the average task satisfaction of the system. To address the above proposed optimization problem, we propose Two-way Offloading & Partitioning (TOP) strategy, where a Two-way Collaborative Edge Node Dividing and Offloading Algorithm determines the cooperative edge nodes for different tasks and obtains the offloading strategy for each task set. Furthermore, we optimize the resource partitioning using the Genetic Algorithm to avoid resource wastage while enhancing the overall satisfaction of the system. Extensive experimental results show that our proposed TOP strategy improves average system satisfaction by up to 33% compared to other baseline strategies.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104058"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145468469","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

CNN-MHBiGRU: A two-stage deep learning framework with multi-attention mechanisms for IoT intrusion detection CNN-MHBiGRU：基于多关注机制的两阶段深度学习框架物联网入侵检测

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-11-04 DOI: 10.1016/j.adhoc.2025.104082

Yufeng Zhang , Yulong Wang , Liting Gao

The growing complexity of cyberattacks in the Internet of Things (IoT) highlights the need for intrusion detection systems that are both accurate and efficient. We present CNN-MHBiGRU, a two-stage framework that integrates enhanced convolutional feature extraction with multi-head recurrent temporal modeling. The CNN stage employs residual blocks, SE attention, SoftPool, and attention pooling to obtain multi-scale spatial representations, while the second stage uses a multi-head bidirectional GRU with intra-head attention to capture diverse temporal dependencies. Data preprocessing follows a leakage-free protocol in which standardization and SMOTE are applied only to the training set, combined with Focal Loss to improve minority-class recognition. Experiments on four public benchmarks (NF-BoT-IoT v1/v2 and NF-ToN-IoT v1/v2) demonstrate that CNN-MHBiGRU consistently surpasses recent methods. On NF-BoT-IoT-v2 (binary), the model achieves an F1-score of 0.9997, while on NF-ToN-IoT-v2 (binary) it reaches 0.9907, both exceeding strong graph-based baselines. For multiclass tasks, weighted recall and F1-scores of 0.9901/0.9901 and 0.9601/0.9595 are obtained on the two v2 datasets. Temporal–spatial efficiency evaluations further show a compact footprint of fewer than 300K parameters and

< 1.2

MB memory, with only moderate inference overhead. Unlike prior CNN–GRU or graph-based IDSs, CNN-MHBiGRU achieves state-of-the-art detection performance while maintaining this lightweight design, offering a distinctive accuracy–efficiency trade-off suitable for realistic IoT-edge deployment.

物联网（IoT）中日益复杂的网络攻击凸显了对准确高效的入侵检测系统的需求。我们提出了CNN-MHBiGRU，这是一个两阶段框架，集成了增强的卷积特征提取和多头递归时间建模。CNN阶段使用残差块、SE注意力、SoftPool和注意力池来获得多尺度空间表示，而第二阶段使用带有头部内注意力的多头双向GRU来捕获不同的时间依赖性。数据预处理遵循无泄漏协议，其中标准化和SMOTE仅应用于训练集，并结合Focal Loss来提高少数类识别。在四个公共基准测试（NF-BoT-IoT v1/v2和NF-ToN-IoT v1/v2）上的实验表明，CNN-MHBiGRU始终优于最近的方法。在NF-BoT-IoT-v2（二进制）上，模型的f1得分为0.9997，在NF-ToN-IoT-v2（二进制）上达到0.9907，两者都超过了基于强图的基线。对于多类任务，两个v2数据集的加权召回率和f1得分分别为0.9901/0.9901和0.9601/0.9595。时空效率评估进一步显示，紧凑的内存占用少于300K个参数和1.2 MB内存，只有适度的推理开销。与之前的CNN-GRU或基于图形的ids不同，CNN-MHBiGRU在保持轻量级设计的同时实现了最先进的检测性能，提供了适合现实物联网边缘部署的独特精度和效率权衡。

{"title":"CNN-MHBiGRU: A two-stage deep learning framework with multi-attention mechanisms for IoT intrusion detection","authors":"Yufeng Zhang , Yulong Wang , Liting Gao","doi":"10.1016/j.adhoc.2025.104082","DOIUrl":"10.1016/j.adhoc.2025.104082","url":null,"abstract":"<div><div>The growing complexity of cyberattacks in the Internet of Things (IoT) highlights the need for intrusion detection systems that are both accurate and efficient. We present CNN-MHBiGRU, a two-stage framework that integrates enhanced convolutional feature extraction with multi-head recurrent temporal modeling. The CNN stage employs residual blocks, SE attention, SoftPool, and attention pooling to obtain multi-scale spatial representations, while the second stage uses a multi-head bidirectional GRU with intra-head attention to capture diverse temporal dependencies. Data preprocessing follows a leakage-free protocol in which standardization and SMOTE are applied only to the training set, combined with Focal Loss to improve minority-class recognition. Experiments on four public benchmarks (NF-BoT-IoT v1/v2 and NF-ToN-IoT v1/v2) demonstrate that CNN-MHBiGRU consistently surpasses recent methods. On NF-BoT-IoT-v2 (binary), the model achieves an F1-score of 0.9997, while on NF-ToN-IoT-v2 (binary) it reaches 0.9907, both exceeding strong graph-based baselines. For multiclass tasks, weighted recall and F1-scores of 0.9901/0.9901 and 0.9601/0.9595 are obtained on the two v2 datasets. Temporal–spatial efficiency evaluations further show a compact footprint of fewer than 300K parameters and <span><math><mrow><mo><</mo><mn>1</mn><mo>.</mo><mn>2</mn></mrow></math></span> MB memory, with only moderate inference overhead. Unlike prior CNN–GRU or graph-based IDSs, CNN-MHBiGRU achieves state-of-the-art detection performance while maintaining this lightweight design, offering a distinctive accuracy–efficiency trade-off suitable for realistic IoT-edge deployment.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104082"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145468467","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

ADMZ: Adaptive Dynamic Mix Zone pseudonym change strategy for location privacy in Vehicular Ad-hoc Networks ADMZ：车辆自组织网络中位置隐私的自适应动态混合区域假名更改策略

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-11-05 DOI: 10.1016/j.adhoc.2025.104066

Liyan Wang, Yingxu Lai, Congai Zeng

Vehicular Ad-Hoc Networks (VANETs) enable efficient traffic coordination through real-time broadcasting of Basic Safety Messages (BSMs), yet they also introduce significant location privacy concerns. Existing dynamic mix zone strategies trigger pseudonym changes based on vehicle density awareness. However, their reliance on single-time-slot verification often results in residual ”ghost vehicles” and false mix zone activation, which not only wastes pseudonym resources but also compromises privacy protection. To overcome these limitations, we propose ADMZ, an Adaptive Dynamic Mix Zone pseudonym change strategy. ADMZ includes a time slot-based pseudonym cleanup mechanism carried out by Roadside Units (RSUs) to periodically remove expired pseudonyms. It also incorporates a dynamic boundary adjustment method based on trajectory prediction region overlap to counter semantic linking attacks, as well as a density-distance weighted priority model for mix zone selection. Simulation results show that ADMZ significantly improves location privacy protection, reducing the false trigger rate from 40.5% to 17.6% and reducing the adversary’s trajectory tracking success rate to 14.3% in scenarios with 300 vehicles. Furthermore, ADMZ reduces pseudonym consumption and achieves a better balance between privacy preservation and system stability.

车载自组织网络（VANETs）通过实时广播基本安全信息（BSMs）实现高效的交通协调，但它们也带来了严重的位置隐私问题。现有的动态混合区域策略基于车辆密度感知触发假名变化。然而，它们依赖于单一时隙验证，往往会导致残留的“鬼车”和虚假的混合区激活，这不仅浪费了假名资源，而且损害了隐私保护。为了克服这些限制，我们提出了自适应动态混合区假名更改策略ADMZ。ADMZ包括一个基于时隙的假名清理机制，由路边单位（rsu）执行，定期清除过期的假名。采用基于轨迹预测区域重叠的动态边界调整方法对抗语义链接攻击，采用密度距离加权优先级模型选择混合区域。仿真结果表明，ADMZ显著提高了位置隐私保护，在300辆车辆的场景下，将误触发率从40.5%降低到17.6%，将对手的轨迹跟踪成功率降低到14.3%。此外，ADMZ减少了假名消耗，并在隐私保护和系统稳定性之间取得了更好的平衡。

{"title":"ADMZ: Adaptive Dynamic Mix Zone pseudonym change strategy for location privacy in Vehicular Ad-hoc Networks","authors":"Liyan Wang, Yingxu Lai, Congai Zeng","doi":"10.1016/j.adhoc.2025.104066","DOIUrl":"10.1016/j.adhoc.2025.104066","url":null,"abstract":"<div><div>Vehicular Ad-Hoc Networks (VANETs) enable efficient traffic coordination through real-time broadcasting of Basic Safety Messages (BSMs), yet they also introduce significant location privacy concerns. Existing dynamic mix zone strategies trigger pseudonym changes based on vehicle density awareness. However, their reliance on single-time-slot verification often results in residual ”ghost vehicles” and false mix zone activation, which not only wastes pseudonym resources but also compromises privacy protection. To overcome these limitations, we propose ADMZ, an Adaptive Dynamic Mix Zone pseudonym change strategy. ADMZ includes a time slot-based pseudonym cleanup mechanism carried out by Roadside Units (RSUs) to periodically remove expired pseudonyms. It also incorporates a dynamic boundary adjustment method based on trajectory prediction region overlap to counter semantic linking attacks, as well as a density-distance weighted priority model for mix zone selection. Simulation results show that ADMZ significantly improves location privacy protection, reducing the false trigger rate from 40.5% to 17.6% and reducing the adversary’s trajectory tracking success rate to 14.3% in scenarios with 300 vehicles. Furthermore, ADMZ reduces pseudonym consumption and achieves a better balance between privacy preservation and system stability.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104066"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145468470","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Bi-functional glucose Sensing-Transmission (Sens-Tra) sensor for IoT-based applications 基于物联网应用的双功能葡萄糖传感传输（Sens-Tra）传感器

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-11-04 DOI: 10.1016/j.adhoc.2025.104086

Swati Todi, Shankul Patel, Poonam Agarwal

In the era of the Internet of Things (IoT) Ecosystem, multi-functional components have the potential to meet the needs of the exponential rise in the number of sensors/devices. Hence, in the future, the feasible complexity of the systems in IoT could be maintained using multi-functional components. This research aims to lower the hardware complexity at the edge devices in IoT architecture by incorporating both the sensing and data transmission capabilities into a single device, thereby eliminating the necessity of an external transmission device. Here, a state-of-the-art technique is developed, where a real-time, label-free, contact-based bi-functional Sens-Tra (Sensing-Transmission) operates as a sensor for glucose detection in aqueous solution as well as wireless transmission. To materialize this, a Circular Interdigital Capacitor Sensor (CIDCS) fed by a coplanar waveguide transmission line is proposed. The sensor operates over the Wi-Fi network, integrating leading-edge wireless technology with contact-based glucose sensing. The Wi-Fi network is established through ESP8266 Wi-Fi modules, and the Received Signal Strength Indicator (RSSI) is utilized as the sensing parameter for glucose detection. In addition to the sensing purpose, the bi-functional sensor is demonstrated for IoT-based applications, where RSSI is wirelessly transmitted from Sens-Tra to the cloud server, ThingSpeak, through mobile data. The data from the cloud can be processed and utilized through any web/mobile application, providing desired flexibility to the system. The experimental findings exhibit the potential of the bi-functional CIDCS for sensing as well as data transmission.

在物联网（IoT）生态系统时代，多功能组件有可能满足传感器/设备数量呈指数级增长的需求。因此，在未来，物联网系统的可行复杂性可以使用多功能组件来维持。本研究旨在通过将传感和数据传输功能整合到单个设备中，从而消除外部传输设备的必要性，从而降低物联网架构中边缘设备的硬件复杂性。在这里，开发了一种最先进的技术，其中实时，无标签，基于接触的双功能Sens-Tra（传感传输）作为水溶液中葡萄糖检测和无线传输的传感器。为了实现这一目标，提出了一种由共面波导传输线馈电的圆形数字电容传感器（CIDCS）。该传感器通过Wi-Fi网络运行，将先进的无线技术与接触式葡萄糖传感技术相结合。通过ESP8266 Wi-Fi模块建立Wi-Fi网络，利用接收信号强度指标（Received Signal Strength Indicator， RSSI）作为葡萄糖检测的传感参数。除了传感用途外，双功能传感器还用于基于物联网的应用，其中RSSI通过移动数据从Sens-Tra无线传输到云服务器ThingSpeak。来自云的数据可以通过任何web/移动应用程序进行处理和利用，从而为系统提供所需的灵活性。实验结果显示了双功能CIDCS在传感和数据传输方面的潜力。

{"title":"Bi-functional glucose Sensing-Transmission (Sens-Tra) sensor for IoT-based applications","authors":"Swati Todi, Shankul Patel, Poonam Agarwal","doi":"10.1016/j.adhoc.2025.104086","DOIUrl":"10.1016/j.adhoc.2025.104086","url":null,"abstract":"<div><div>In the era of the Internet of Things (IoT) Ecosystem, multi-functional components have the potential to meet the needs of the exponential rise in the number of sensors/devices. Hence, in the future, the feasible complexity of the systems in IoT could be maintained using multi-functional components. This research aims to lower the hardware complexity at the edge devices in IoT architecture by incorporating both the sensing and data transmission capabilities into a single device, thereby eliminating the necessity of an external transmission device. Here, a state-of-the-art technique is developed, where a real-time, label-free, contact-based bi-functional Sens-Tra (Sensing-Transmission) operates as a sensor for glucose detection in aqueous solution as well as wireless transmission. To materialize this, a Circular Interdigital Capacitor Sensor (CIDCS) fed by a coplanar waveguide transmission line is proposed. The sensor operates over the Wi-Fi network, integrating leading-edge wireless technology with contact-based glucose sensing. The Wi-Fi network is established through ESP8266 Wi-Fi modules, and the Received Signal Strength Indicator (RSSI) is utilized as the sensing parameter for glucose detection. In addition to the sensing purpose, the bi-functional sensor is demonstrated for IoT-based applications, where RSSI is wirelessly transmitted from Sens-Tra to the cloud server, <em>ThingSpeak</em>, through mobile data. The data from the cloud can be processed and utilized through any web/mobile application, providing desired flexibility to the system. The experimental findings exhibit the potential of the bi-functional CIDCS for sensing as well as data transmission.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104086"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145468468","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Transfer function-guided mixed-variable optimization for joint mining decisions and resource allocation in mobile edge computing-integrated blockchain networks 移动边缘计算集成区块链网络联合挖掘决策与资源分配的传递函数引导混合变量优化

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-11-07 DOI: 10.1016/j.adhoc.2025.104088

Mohamed Abdel-Basset , Reda Mohamed , Karam M. Sallam , Saber Elsayed

Recently, mobile edge computing (MEC) technology has been integrated with wireless blockchain networks to improve the computational capabilities of Internet of Things devices during the mining process. Jointly, optimizing miner selection (discrete) and resource allocation (continuous) in MEC-integrated blockchain networks is a challenging mixed-variable, NP-hard problem. Although several algorithms have been presented in the literature to solve it, they still suffer from low-quality results due to either slow convergence speed, local optima stagnation, or both, especially for small or medium problem sizes. To address this, we propose a transfer-function-guided encoding (TFE) framework that introduces a principled link between continuous metaheuristic search and discrete miner-operator control. Specifically, each individual maintains one discrete control variable determining insertion, deletion, or replacement of a miner and two continuous controls representing transmission power and computing resource allocation. Continuous metaheuristic outputs are converted to discrete decisions through families of S-shaped and V-shaped transfer functions, providing tunable exploration–exploitation balance and probabilistic control over operator selection. This mechanism is integrated with several state-of-the-art algorithms. Extensive experiments on MEC-blockchain networks with

m \in [50, 1000]

miners demonstrate that TFE consistently accelerates convergence and improves system profit for small–medium scales, with HNOA-TFE achieving the best overall performance. The numerical results show that the hybrid nutcracker optimization algorithm with the TFE mechanism is effective across most problem instances. Also, the comparative study with recent MEC/blockchain resource-allocation and vehicular-edge benchmarks shows the robustness and scalability of the proposed method.

最近，移动边缘计算（MEC）技术与无线区块链网络相结合，以提高物联网设备在采矿过程中的计算能力。在mec集成区块链网络中，优化矿工选择（离散）和资源分配（连续）是一个具有挑战性的混合变量np困难问题。虽然文献中已经提出了几种算法来解决它，但由于收敛速度慢，局部最优停滞或两者兼而有之，它们仍然遭受低质量结果的困扰，特别是对于中小型问题规模。为了解决这个问题，我们提出了一个传递函数引导编码（TFE）框架，该框架在连续元启发式搜索和离散矿工算子控制之间引入了原则性的联系。具体来说，每个个体维护一个离散控制变量，决定插入、删除或替换一个矿工，以及两个连续控制变量，代表传输功率和计算资源分配。连续的元启发式输出通过s形和v形传递函数族转换为离散决策，提供可调的勘探-开采平衡和对操作员选择的概率控制。该机制集成了几种最先进的算法。在m∈[50,1000]矿工的mec -区块链网络上进行的大量实验表明，TFE持续加速收敛并提高中小型规模的系统利润，其中HNOA-TFE实现了最佳的整体性能。数值结果表明，结合TFE机制的混合胡桃夹子优化算法在大多数情况下都是有效的。此外，与最近的MEC/区块链资源分配和车辆边缘基准测试的比较研究表明了该方法的鲁棒性和可扩展性。

{"title":"Transfer function-guided mixed-variable optimization for joint mining decisions and resource allocation in mobile edge computing-integrated blockchain networks","authors":"Mohamed Abdel-Basset , Reda Mohamed , Karam M. Sallam , Saber Elsayed","doi":"10.1016/j.adhoc.2025.104088","DOIUrl":"10.1016/j.adhoc.2025.104088","url":null,"abstract":"<div><div>Recently, mobile edge computing (MEC) technology has been integrated with wireless blockchain networks to improve the computational capabilities of Internet of Things devices during the mining process. Jointly, optimizing miner selection (discrete) and resource allocation (continuous) in MEC-integrated blockchain networks is a challenging mixed-variable, NP-hard problem. Although several algorithms have been presented in the literature to solve it, they still suffer from low-quality results due to either slow convergence speed, local optima stagnation, or both, especially for small or medium problem sizes. To address this, we propose a transfer-function-guided encoding (TFE) framework that introduces a principled link between continuous metaheuristic search and discrete miner-operator control. Specifically, each individual maintains one discrete control variable determining insertion, deletion, or replacement of a miner and two continuous controls representing transmission power and computing resource allocation. Continuous metaheuristic outputs are converted to discrete decisions through families of S-shaped and V-shaped transfer functions, providing tunable exploration–exploitation balance and probabilistic control over operator selection. This mechanism is integrated with several state-of-the-art algorithms. Extensive experiments on MEC-blockchain networks with <span><math><mrow><mi>m</mi><mspace></mspace><mo>∈</mo><mspace></mspace><mrow><mo>[</mo><mn>50</mn><mo>,</mo><mn>1000</mn><mo>]</mo></mrow></mrow></math></span> miners demonstrate that TFE consistently accelerates convergence and improves system profit for small–medium scales, with HNOA-TFE achieving the best overall performance. The numerical results show that the hybrid nutcracker optimization algorithm with the TFE mechanism is effective across most problem instances. Also, the comparative study with recent MEC/blockchain resource-allocation and vehicular-edge benchmarks shows the robustness and scalability of the proposed method.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104088"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145468556","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A blockchain-integrated PUF framework for secure authentication and communication 一个区块链集成的PUF框架，用于安全认证和通信

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-10-24 DOI: 10.1016/j.adhoc.2025.104059

Koustav Kumar Mondal , Debasis Das , Arpit Khandelwal

Prior work on Internet-of-Things (IoT) security often splits between hardware roots of trust and decentralized key management: Physically Unclonable Function (PUF) schemes frequently depend on centralized helper infrastructures, whereas blockchain systems typically lack a hardware seed. We present a unified, protocol-enforced framework that integrates PUFs, blockchain-backed Shamir’s Secret Sharing (SSS), and elliptic-curve cryptography (ECC) to remove this gap. Concretely: (i) a Static Monostable PUF with error correction derives device keys without storage, achieving

> 95 %

reconstruction success at

\approx 10 %

noise; (ii) SSS shares are posted on-chain but re-wrapped every epoch under fresh IND-CPA encryption, eliminating ciphertext staleness and bounding ledger-scraping advantage; and (iii) ECC (Elliptic Curve Cryptography) – including ephemeral Elliptic Curve Diffie–Hellman (ECDH) for forward secrecy and Elliptic Curve Digital Signature Algorithm (ECDSA) for authentication – must rotate each epoch from high-min-entropy PUF material. Our analysis proves hardware-anchored uniqueness, information-theoretic threshold secrecy, and forward secrecy under bounded leakage with non-compounding attacker advantage. We also derive a unified impersonation bound that composes the PUF/ML front-end with the ECC back-end (acceptance probability

\leq α + α_{ML} + negl (κ)

). A duty-cycle/connectivity model shows constant device-resident state (independent of

(t, n)

), microjoule–millijoule energy per epoch, and graceful PUF-only operation during outages with automatic re-incorporation of on-chain shares upon reconnection. Containerized experiments across domain-specific deployments – secure supply-chain identification, critical-infrastructure control, and healthcare telemetry – demonstrate sub-second end-to-end handshakes under load, with SSS and ECC costs scaling linearly in

t

and

Θ (log p)

, respectively. These results indicate that enforced rotation plus on-chain ephemerality yields an efficient, scalable, and formally validated tamper-resistant root of trust for next-generation IoT networks.

先前在物联网（IoT）安全方面的工作通常分为信任的硬件根和分散的密钥管理：物理不可克隆功能（PUF）方案经常依赖于集中式辅助基础设施，而区块链系统通常缺乏硬件种子。我们提出了一个统一的、协议强制的框架，该框架集成了puf、区块链支持的Shamir秘密共享（SSS）和椭圆曲线密码学（ECC）来消除这一差距。具体而言：(i)带纠错的静态单稳态PUF无需存储即可导出设备密钥，在≈10%噪声下实现95%重构成功率；（ii） SSS股票在链上发布，但在新的IND-CPA加密下重新包装每个epoch，消除密文的过时性和边界分类账抓取优势；（iii） ECC（椭圆曲线密码学）-包括用于前向保密的短暂椭圆曲线Diffie-Hellman （ECDH）和用于身份验证的椭圆曲线数字签名算法(ECDSA) -必须从高最小熵PUF材料旋转每个历元。我们的分析证明了硬件锚定唯一性、信息论阈值保密和有界泄漏下的前向保密具有非复合攻击者的优势。我们还推导了由PUF/ML前端与ECC后端组成的统一模拟界（接受概率≤α+αML+negl(κ)）。占空比/连接模型显示恒定的设备驻留状态（独立于（t,n）），每个历元的微焦耳-毫焦耳能量，以及停机期间仅puf的优雅运行，并在重新连接时自动重新合并链上份额。跨特定领域部署的容器化实验（安全供应链识别、关键基础设施控制和医疗遥测）演示了负载下的亚秒级端到端握手，其中SSS和ECC成本分别在t和Θ（logp）中呈线性扩展。这些结果表明，强制旋转加上链上短暂性为下一代物联网网络提供了高效、可扩展且经过正式验证的防篡改信任根。

{"title":"A blockchain-integrated PUF framework for secure authentication and communication","authors":"Koustav Kumar Mondal , Debasis Das , Arpit Khandelwal","doi":"10.1016/j.adhoc.2025.104059","DOIUrl":"10.1016/j.adhoc.2025.104059","url":null,"abstract":"<div><div>Prior work on Internet-of-Things (IoT) security often splits between hardware roots of trust and decentralized key management: Physically Unclonable Function (PUF) schemes frequently depend on centralized helper infrastructures, whereas blockchain systems typically lack a hardware seed. We present a unified, protocol-enforced framework that integrates PUFs, blockchain-backed Shamir’s Secret Sharing (SSS), and elliptic-curve cryptography (ECC) to remove this gap. Concretely: (i) a Static Monostable PUF with error correction derives device keys without storage, achieving <span><math><mrow><mo>></mo><mn>95</mn><mtext>%</mtext></mrow></math></span> reconstruction success at <span><math><mrow><mo>≈</mo><mn>10</mn><mtext>%</mtext></mrow></math></span> noise; (ii) SSS shares are posted on-chain but re-wrapped every epoch under fresh IND-CPA encryption, eliminating ciphertext staleness and bounding ledger-scraping advantage; and (iii) ECC (Elliptic Curve Cryptography) – including ephemeral Elliptic Curve Diffie–Hellman (ECDH) for forward secrecy and Elliptic Curve Digital Signature Algorithm (ECDSA) for authentication – must rotate each epoch from high-min-entropy PUF material. Our analysis proves hardware-anchored uniqueness, information-theoretic threshold secrecy, and forward secrecy under bounded leakage with non-compounding attacker advantage. We also derive a unified impersonation bound that composes the PUF/ML front-end with the ECC back-end (acceptance probability <span><math><mrow><mo>≤</mo><mi>α</mi><mo>+</mo><msub><mrow><mi>α</mi></mrow><mrow><mi>ML</mi></mrow></msub><mo>+</mo><mi>negl</mi><mrow><mo>(</mo><mi>κ</mi><mo>)</mo></mrow></mrow></math></span>). A duty-cycle/connectivity model shows constant device-resident state (independent of <span><math><mrow><mo>(</mo><mi>t</mi><mo>,</mo><mi>n</mi><mo>)</mo></mrow></math></span>), microjoule–millijoule energy per epoch, and graceful PUF-only operation during outages with automatic re-incorporation of on-chain shares upon reconnection. Containerized experiments across domain-specific deployments – secure supply-chain identification, critical-infrastructure control, and healthcare telemetry – demonstrate sub-second end-to-end handshakes under load, with SSS and ECC costs scaling linearly in <span><math><mi>t</mi></math></span> and <span><math><mrow><mi>Θ</mi><mrow><mo>(</mo><mo>log</mo><mi>p</mi><mo>)</mo></mrow></mrow></math></span>, respectively. These results indicate that enforced rotation plus on-chain ephemerality yields an efficient, scalable, and formally validated tamper-resistant root of trust for next-generation IoT networks.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104059"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145365861","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Optimizing energy-efficient routing in Mobile Internet of Things (MIoT) networks using Grey Wolf Optimization and Recurrent Neural Networks 利用灰狼优化和递归神经网络优化移动物联网（MIoT）网络中的节能路由

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-10-25 DOI: 10.1016/j.adhoc.2025.104047

Seyed Salar Sefati , Sanda Osiceanu Maiduc , Bahman Arasteh , Winfred Ofoe Larkotey , Asgarali Bouyer , Wali Ullah Khan

The Mobile Internet of Things (MIoT) represents a significant evolution of traditional IoT by enabling seamless connectivity for mobile devices and sensors in dynamic environments. Given the resource constraints and mobility challenges in MIoT networks, developing adaptive and energy-efficient routing strategies is important. This paper proposes a novel routing protocol that integrates Grey Wolf Optimization (GWO) and Recurrent Neural Networks (RNNs) to enhance energy efficiency, reliability, and responsiveness in MIoT systems. The protocol features dynamic clustering, predictive traffic load balancing, and multi-objective optimization for Cluster Head (CH) selection, where RNNs forecast traffic trends and GWO optimizes routing paths. Simulation results demonstrate that the proposed method reduces energy consumption, lowers end-to-end delays, and improves packet delivery ratio (PDR) and network reliability under both static and mobile conditions. Compared to existing methods such as the Krill Herd (KH) algorithm, Dynamic Multi-Sink Routing Protocol (DMS-RP), and Evolutionary Fuzzy Rule-based (EFR) models, the proposed solution exhibits superior performance, validating its scalability and effectiveness for real-world MIoT applications.

移动物联网（MIoT）通过在动态环境中实现移动设备和传感器的无缝连接，代表了传统物联网的重大演变。考虑到物联网网络中的资源限制和移动性挑战，开发自适应和节能的路由策略非常重要。本文提出了一种集成灰狼优化（GWO）和循环神经网络（rnn）的新型路由协议，以提高物联网系统的能源效率、可靠性和响应性。该协议具有动态聚类、预测流量负载均衡和多目标簇头选择的特点，其中rnn预测流量趋势，GWO优化路由路径。仿真结果表明，该方法在静态和移动两种情况下都能降低能耗，降低端到端时延，提高分组分发率（PDR）和网络可靠性。与Krill Herd （KH）算法、动态多汇路由协议（DMS-RP）和进化模糊规则（EFR）模型等现有方法相比，该解决方案表现出卓越的性能，验证了其在实际工业物联网应用中的可扩展性和有效性。

{"title":"Optimizing energy-efficient routing in Mobile Internet of Things (MIoT) networks using Grey Wolf Optimization and Recurrent Neural Networks","authors":"Seyed Salar Sefati , Sanda Osiceanu Maiduc , Bahman Arasteh , Winfred Ofoe Larkotey , Asgarali Bouyer , Wali Ullah Khan","doi":"10.1016/j.adhoc.2025.104047","DOIUrl":"10.1016/j.adhoc.2025.104047","url":null,"abstract":"<div><div>The Mobile Internet of Things (MIoT) represents a significant evolution of traditional IoT by enabling seamless connectivity for mobile devices and sensors in dynamic environments. Given the resource constraints and mobility challenges in MIoT networks, developing adaptive and energy-efficient routing strategies is important. This paper proposes a novel routing protocol that integrates Grey Wolf Optimization (GWO) and Recurrent Neural Networks (RNNs) to enhance energy efficiency, reliability, and responsiveness in MIoT systems. The protocol features dynamic clustering, predictive traffic load balancing, and multi-objective optimization for Cluster Head (CH) selection, where RNNs forecast traffic trends and GWO optimizes routing paths. Simulation results demonstrate that the proposed method reduces energy consumption, lowers end-to-end delays, and improves packet delivery ratio (PDR) and network reliability under both static and mobile conditions. Compared to existing methods such as the Krill Herd (KH) algorithm, Dynamic Multi-Sink Routing Protocol (DMS-RP), and Evolutionary Fuzzy Rule-based (EFR) models, the proposed solution exhibits superior performance, validating its scalability and effectiveness for real-world MIoT applications.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104047"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145419848","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Authentication protocol for the Internet of Drones with fog computing based on aggregate signatures for forest inventory 基于集合签名的无人机互联网雾计算森林清查认证协议

IF 4.8 3区计算机科学 Q1 COMPUTER SCIENCE, INFORMATION SYSTEMS

Ad Hoc Networks

Pub Date : 2026-02-01 Epub Date: 2025-10-27 DOI: 10.1016/j.adhoc.2025.104034

Manuela de Jesus Sousa , Paulo Roberto L. Gondim , Sandra Sendra , Jaime Lloret

The Internet of Drones (IoD) has become increasingly important in applications such as forest inventory, leveraging advanced sensors and internet connectivity to enable efficient data collection. Compared to traditional methods, IoD offers superior cost-effectiveness. However, its reliance on public channels, unreliable connectivity, and dynamic environments poses significant security and privacy challenges. Safeguarding forest inventory data is essential to maintaining accuracy, preventing unauthorized access, and mitigating the risk of data manipulation, which can lead to suboptimal management decisions. To address these concerns, it is essential to design a lightweight authentication protocol that secures IoD communication, accounts for network bandwidth limitations and scalability, and supports integration with emerging technologies. This manuscript introduces a new Authentication and Key Agreement (AKA) protocol specifically designed for the Internet of Drones (IoD), leveraging asymmetric cryptography and aggregate signatures to enhance security and privacy in forest inventories with fog computing. Its robustness was confirmed through informal and formal security analyses by the AVISPA tool and the ROR model, demonstrating resistance to known attacks and superior communication, computational, and energy performance compared to existing protocols.

无人机互联网（IoD）在森林清查等应用中变得越来越重要，利用先进的传感器和互联网连接来实现高效的数据收集。与传统方法相比，IoD具有更高的成本效益。然而，它对公共通道的依赖、不可靠的连接和动态环境构成了重大的安全和隐私挑战。保护森林清查数据对于保持准确性、防止未经授权的访问和减轻可能导致次优管理决策的数据操纵风险至关重要。为了解决这些问题，必须设计一个轻量级的身份验证协议，以保护IoD通信，考虑网络带宽限制和可扩展性，并支持与新兴技术的集成。本文介绍了一种专门为无人机互联网（IoD）设计的新的身份验证和密钥协议（AKA）协议，利用非对称加密和聚合签名来增强雾计算森林清单的安全性和隐私性。AVISPA工具和ROR模型通过非正式和正式的安全分析证实了它的鲁棒性，与现有协议相比，它展示了对已知攻击的抵抗力，以及更好的通信、计算和能源性能。

{"title":"Authentication protocol for the Internet of Drones with fog computing based on aggregate signatures for forest inventory","authors":"Manuela de Jesus Sousa , Paulo Roberto L. Gondim , Sandra Sendra , Jaime Lloret","doi":"10.1016/j.adhoc.2025.104034","DOIUrl":"10.1016/j.adhoc.2025.104034","url":null,"abstract":"<div><div>The Internet of Drones (IoD) has become increasingly important in applications such as forest inventory, leveraging advanced sensors and internet connectivity to enable efficient data collection. Compared to traditional methods, IoD offers superior cost-effectiveness. However, its reliance on public channels, unreliable connectivity, and dynamic environments poses significant security and privacy challenges. Safeguarding forest inventory data is essential to maintaining accuracy, preventing unauthorized access, and mitigating the risk of data manipulation, which can lead to suboptimal management decisions. To address these concerns, it is essential to design a lightweight authentication protocol that secures IoD communication, accounts for network bandwidth limitations and scalability, and supports integration with emerging technologies. This manuscript introduces a new Authentication and Key Agreement (AKA) protocol specifically designed for the Internet of Drones (IoD), leveraging asymmetric cryptography and aggregate signatures to enhance security and privacy in forest inventories with fog computing. Its robustness was confirmed through informal and formal security analyses by the AVISPA tool and the ROR model, demonstrating resistance to known attacks and superior communication, computational, and energy performance compared to existing protocols.</div></div>","PeriodicalId":55555,"journal":{"name":"Ad Hoc Networks","volume":"181 ","pages":"Article 104034"},"PeriodicalIF":4.8,"publicationDate":"2026-02-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145419844","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0