Satellite communication networks play a critical role in providing connectivity to remote regions and areas with limited infrastructure. However, their inherently open nature and physical exposure make them particularly susceptible to security threats, including replay, impersonation, and man-in- the-middle attacks. The emergence of quantum computing further undermines the robustness of conventional cryptographic schemes that rely on number-theoretic assumptions. To mitigate these challenges, this article proposesHBQS, a lightweight post-quantum authentication framework designed for satellite platforms with limited resources.HBQSintegrates Physically Unclonable Functions (PUFs) with hash-based cryptography, leveraging SPHINCS+ digital signatures and SHA-3 hashing to provide secure mutual authentication and session key establishment. The protocolHBQSwas implemented and evaluated on embedded hardware platforms, including Raspberry Pi 4.0, PYNQ-Z2 FPGA, and a Dell ground control station. The experimental results demonstrate thatHBQSachieves mutual authentication in 0.88 ms, offering an approximately 67% reduction in execution time compared to representative baselines from the prior literature. Entropy analysis confirms that the proposed protocol maintains a high entropy across critical components, withHBQSachieving a signature entropy of 164.63 bits and PUF response entropy of 172.58 bits, indicating strong resistance to statistical and modeling attacks. A formal security analysis conducted within the Random Oracle Model demonstrates semantic security against both classical and quantum adversaries. The protocolHBQSshows a significant improvement over existing methods, achieving approximately 67% reduction in computational execution time, approximately 11.5% faster user-side handshake time, approximately 0.6% improvement on the satellite side and full security coverage across all evaluated security features with only a 21.4% increase in static memory usage as the trade-off. These findings positionHBQSas an efficient, secure, and scalable authentication solution for next-generation satellite communication systems operating in the post-quantum era.

Malicious traffic detection often requires large, labeled datasets, which are challenging due to privacy concerns, labeling costs, and evolving threat patterns. Although recent self-supervised pretraining methods address this issue, they rely on complex transformer-based architectures that are computationally expensive and have high inference times, making them unsuitable for real-time use. In addition, most existing approaches process packets or flows independently, and often rely on per-packet dataset splits that introduce implicit flow-level data leakage, thereby limiting their ability to capture meaningful semantic and behavioral relationships across flows for detecting stealthy encrypted threats. To address these issues, we propose LitCVit, a lightweight self-supervised contrastive Vision Transformer-based framework that captures cross-flow semantic and behavioral patterns to generate robust latent representations of encrypted traffic. Without relying on decryption or manually engineered features, our method enables efficient detection of encrypted malicious flows with low inference time. Extensive evaluations on benchmark datasets demonstrate that the proposed framework achieves an average detection accuracy of 98.10% and F1-score of 98.08%. Compared to the best state-of-the-art model, LitCVit achieves an average improvement of 2.49% in F1-score, 2.12% in precision, and 2.50% in recall, highlighting its superior detection capability in encrypted traffic scenarios. Additionally, LitCVit achieves an 8.7× reduction in inference time compared to the best existing self-supervised approach, making it highly suitable for deployment on resource-constrained devices.

Security and privacy are two primary concerns in critical applications within Internet of Things (IoT) environments. The Long Range Wide Area Network (LoRaWAN) protocol facilitates long-range communication for battery-powered end devices in IoT and has gained widespread adoption among both individuals and industries. To foster trust and facilitate its use, ensuring security and privacy for data collected by end devices is essential. User authentication and key establishment protocols play a pivotal role in this regard. While existing authentication schemes in the literature are unsuitable for LoRaWAN networks, this article proposes an energy-efficient LoRaWAN authentication protocol for privacy preservation in IoT applications. Through formal verification using the Random Oracle Model (ROM) and AVISPA tool, we demonstrate our protocol’s resilience against common attacks including replay, man-in-the-middle, and impersonation threats. Performance evaluations reveal significant advantages over existing schemes: 48% faster computation (0.1953 ms vs. 0.3647 ms baseline), 24% lower communication overhead (2398 bits vs. 3168 bits), and 44% energy savings (1.27 mJ vs. 2.27 mJ) over state-of-the-art protocols. These improvements, combined with enhanced security features, such as resistance to sensor capture and stolen device attacks, make our protocol particularly suitable for practical IoT deployments in smart agriculture and industrial monitoring systems where both security and energy efficiency are paramount. The balance of strong security guarantees and low operational overhead represents a significant advance in LoRaWAN authentication mechanisms.

Industrial intrusion detection systems (IIDS) are crucial in defending digital industrial infrastructures by detecting unauthorized activities within industrial network traffic. As cyber threats continue to evolve, collaborative advancements in IIDS are necessary to ensure robust defenses across future industries. While machine learning (ML) enhances IIDS capabilities, collaborative ML approaches face privacy and regulatory challenges, limiting data sharing across industries. Federated learning (FL) offers a solution by enabling collaborative training without direct data sharing; however, industries remain vulnerable to data reconstruction attacks and FL can impose high overheads. To address these challenges, we propose federated neural-network-based gradient boosting (FNGB) method for collaborative IIDS. FNGB introduces GradProtect, a privacy-preserving mechanism that mitigates data reconstruction, and DynamicLR, an adaptive learning rate method for efficient distributed gradient boosting. Extensive evaluations demonstrate that FNGB delivers enhanced privacy, high performance, and high convergence with low overhead compared to existing methods, making it well-suited for cross-industry deployment.

Summary Data centers today host a number of computational resources to support the increasing demand for computation and storage. Understanding how these physical and virtual machines transition between different states of operation (referred to as machine lifecycle) enables more efficient data center operation management. Furthermore, it helps data center operators define policies on how new computational resources can be added or existing infrastructure decommissioned. Using Google cluster trace data set version 3 collected from approximately 96 k machines, we analyze machine failure and changes in machine lifecycle over time. We observed that there is a 13% chance of another machine failure under the same network switch within 1 min of the previous machine failure. A Markov chain‐based model is proposed, that can predict machine states at any given time. Using the model and estimated probabilities, we predicted the machine state over a span of several days with a high probability. Using the predicted machine state, we reconstructed the active machines trend and compared this with the trend reported in the data set, observing an error of 1.76%.

Content caching has been considered as an effective way to offload contents at network edge in order to alleviate backhaul load. Recently, federated learning (FL) based edge caching has gained a lot of popularity due to its prominent features of data privacy, distributed mode of operation, and scalability. However, these FL-based schemes ignore the behavior of the communication channels during the federated weight averaging procedure. In this paper, we introduce a novel robust federated learning-based content caching approach for fog radio access networks (F-RANs) that mitigates the effect of communication channel noise. In our proposed robust FL approach, each cell employs a deep neural network (DNN)-based model to predict users’ future files rating score based on user and file contextual information and shares its learned weights to the fog server. The fog server is responsible for global weight averaging. Prior to FL weight averaging fog sever feed incoming local model weights to a generative adversarial neural network (GANs) model which differentiates between noisy and actual federated weights, and passes only actual weights based on the distribution of the weight matrices. Extensive simulations have been carried out to validate the performance of our proposed approach. Results show that the GAN-aided federated model yields 23.1% more prediction accuracy as compared to the federated noisy model without GANs based noise mitigation.

With the emergence of Internet-of-Things (IoT) and ever-increasing demand for the newly connected devices, there is a need for more effective storage and processing paradigms to cope with the data generated from these devices. In this study, we have discussed different paradigms for data processing and storage including Cloud, Fog, and Edge computing models and their suitability in integrating with the IoT. Moreover, a detailed discussion on low latency and massive connectivity requirements of future cellular networks in accordance with machine-type communication (MTC) is also presented. Furthermore, the need to bring IoT devices to Internet connectivity and a standardized protocol stack to regulate the data transmission between these devices is also addressed, while keeping in view the resource-constraint nature of IoT devices.

Fog computing is a key paradigm that brings together shared storage, low latency communication, and computation resources closer to users' end-devices. While most IoT services adopt a three-tier computing architecture, where fog nodes are always probed first before reaching a distant Cloud, collaboration across multi-stakeholder, multi-tenant fog providers remains unexplored. In this paper, we quantitatively highlight the gain which may arise if a collaborative fog computing paradigm is deployed. Next, we propose CICO, an incentive based collaborative mechanism for fog computing networks. CICO regulates a multi-stakeholder, multi-tenant cooperation among fog providers. We present a mathematical model and an experimental approach to make the case for such cooperation paradigm.

The popularity of cellular internet of things (IoT) is increasing day by day and billions of IoT devices will be connected to the internet. Many of these devices have limited battery life with constraints on transmit power. High user power consumption in cellular networks restricts the deployment of many IoT devices in 5G. To enable the inclusion of these devices, 5G should be supplemented with strategies and schemes to reduce user power consumption. Therefore, we present a novel joint uplink user association and resource allocation scheme for minimizing user transmit power while meeting the quality of service. We analyze our scheme for two-tier heterogeneous network (HetNet) and show an average transmit power of -2.8 dBm and 8.2 dBm for our algorithms compared to 20 dBm in state-of-the-art Max reference signal received power (RSRP) and channel individual offset (CIO) based association schemes.

The growing subscriber Quality of Experience demands are posing significant challenges to the mobile cellular network operators. One such challenge is the autonomic detection of sleeping cells in cellular networks. Sleeping Cell (SC) is a cell degradation problem, and a special case in Cell Outage Detection (COD) because it does not trigger any alarm due to hardware or software problems in the BS. To minimize the effect of such outages, researchers have proposed autonomous outage detection and compensation solutions. State-of-the-art SC detection depends on drive tests and subscriber complaints to identify the effected cells. However, this approach is quickly becoming unsustainable due to rising operational expenses. To address this particular issue, we employ a Deep Learning based framework which uses Minimization of Drive Tests (MDT) functionality introduced in LTE networks. In our proposed framework, MDT measurements are used to train the deep learning model. Anomalies or cell outages in the network can be then quickly detected and localized, thus significantly reducing the duty cycle of self-healing process in SON. In our simulation setup, we also quantitatively compare and demonstrate superior performance of our proposed approach with state of the art machine learning algorithm such as One Class SVM using multiple performance metrics.

Wireless mesh networks have been considered as a cost effective means of extending coverage which is beyond the capability of a single wireless device. The license free IEEE 802.11 based Wi-Fi is one of the handy available technologies for building a wireless mesh network. The legacy Wi-Fi technology is however not designed for building mesh networks that often require supporting mobility of wireless nodes. The IEEE 802.11s adds the feature of mobile mesh networking to standard Wi-Fi by providing layer-2 routing and better handoff mechanisms. This creates a new set of applications that can be provided with mobile mesh capable Wi-Fi devices. In this work we build a real 802.11s mesh network testbed and performed extensive experimentation with speeds ranging up to 80km/h. The study is aimed at analyzing the effect of speed and other factors on different QoS parameters in a real environment with the objective of finding the feasibility of different applications for mobile 802.11s networks.

Devices in Low Power and Lossy Networks (LLNs) based Internet of Things (IoT) systems have energy, memory and processing constraints. For LLNs, IETF standardized a light-weight routing protocol referred to as RPL. The energy efficiency of RPL routing protocol has been under investigation in prior studies but its stability issue is recently identified. The stability of RPL is critical for its energy efficient and application oriented routing operation. In RPL, child nodes select best parent towards the sink node based on some Objective Function (OF). Current OFs proposed for RPL result in unstable operation of RPL due to frequent route changes. Consequently, high control traffic is generated to re-construct routes which consumes scarce energy resources and thus reduces the network lifetime. In this paper, a novel OF for RPL is proposed which exploits channel adaptability and stability provided by ETX and HOP routing OFs, respectively. The proposed enhancement of RPL is implemented in Contiki operating system and its analysis is carried out in Cooja simulator. The analysis suggests reduction in frequent parent changes, control traffic and energy consumption, thereby improves RPL stability and energy efficiency.

Development of Low-power Low-rate wireless Networks (LLNs) based systems is prodigiously escalated, which helps realizing the notion of Internet of Things (IoT). Routing Protocol for Low-power and Lossy Networks (RPL) is the standard routing protocol for IPv6 LLNs. RPL has gotten tremendous appreciation due to its flexibility and adaptiveness to support dynamic application requirements in IoT. In this paper, we identify multiple issues in RPL route maintenance and recovery mechanism and investigate their drastic effects on the performance of RPL based on a simulation as well as an experimental study using Zolertia Z1 motes. In addition, the paper proposes an enhanced route maintenance and recovery mechanism to improve and rectify RPL operation. Performance analysis of the proposed scheme promise significantly lower reconnection time and high certainty in route recovery.

MCC (Mobile Cloud Computing) performs computation offloading from mobile edge devices to centralized cloud services for achieving the reduction in resource consumption of mobile devices. MBaaS is considered as a new service model within the range of MCC and firebase by Google has been accepted as the most popular MBaaS. However, firebase has not made the efforts for performing computation offloading, as it is expected from an MCC platform. We have contributed in two folds by 1) identifying the limitations of firebase as an MCC platform and by 2) proposing RQL (REST Query Language) as a wrapper over firebase to improve its services for achieving computational offloading. RQL is used for requesting data in the form of JSON instead of using URLs for making REST requests. We have also explained different case studies to evaluate the effectiveness of RQL for converting firebase to an MCC platform.

Seamless mobility in legacy multi-hop IEEE 802.11 Wi-Fi network is difficult to accomplish due to the lack of handoff mechanism. IEEE 802.11s MAC layer protocol adds the feature of mobile mesh networking to standard Wi-Fi by providing layer-2 routing and handoff mechanisms. With ubiquitous connectivity and seamless handoff, many applications requiring continuous data streaming like video, VOIP, file downloading, etc. become possible in Wi-Fi networks that requires mobility support like highways, railways and underground transportation. In this paper we study the handoff mechanism of IEEE 802.11s protocol extensively in a real mobile testbed. Our testbed comprises three mesh nodes, one of which is installed on a vehicle for finding the impact of vehicle speed on 802.11s handoff mechanism. We found that the vehicle speed does not drastically affect the performance metrics during handoff as previously understood. Moreover the results show frequent handoffs during mobility. This is called the ping pong effect, which indicates the absence of hysteresis mechanism in 802.11s protocol. For improving the performance during handoffs we therefore suggest the use of hysteresis mechanism along with the airtime link metric being used in the default routing protocol of 802.11s.

In low-power low-rate wireless networks, IEEE 802.15.4 is a standard protocol for communication. Devices in such networks are usually battery-operated so radio-transceiver component in such networks are typically of low range but most power consuming. Proper antenna orientation, distance between nodes and channel selection are among the ways to achieve reliability in these networks. But these arrangements can affect the link qualities and achieved communication performance in IoT applications. In this work, we have evaluated the performance of IEEE 802.15.4 links using Zolertia Z1 motes experimentally through indoor/outdoor real scenarios. We have found that RSSI decreases with distance and is effected by the height of the motes. The antenna polarization drastically affects the RSSI whereas LQI and packet delivery ratio is not much affected. We also found that the non-interfering channels 26 and channel 15 are effected by Wi-Fi in the same way. Moreover the contiki MAC performs better than XMAC protocol.

The pervasiveness of small mobile devices equipped with multiple wireless interfaces enabled novel communication paradigms: opportunistic data transfer between the mobile devices. In this paper, we consider the heterogeneity in forwarding mechanisms as one of the major reasons which affects the performance of forwarding algorithms. Heterogeneity arises in different contexts, for example in device types (i.e., tablets vs. phones vs wearable), operating systems or node objectives. Each node may have its own objectives e.g. security, privacy, trust, battery life, etc. We specifically address the heterogeneity in nodes objectives and propose two different architectures, cooperative and collaborative. In the cooperative mode nodes can only communicate if they share the same forwarding objective such as saving energy or minimizing delays. In the collaborative mode, all nodes can communicate with other nodes after a negotiation phase within which they find objective trade-offs. We then present three different collaborative algorithms and evaluate all proposed algorithms using real mobility traces to find out the message delivery success rate.

Virtual Local Area Networks (VLANs) provide logical grouping in LANs that are used for many purposes including to limit broadcast traffic and to provide better traffic localization. For better traffic localization and broadcast containment, VLANs could perform well only when the VLAN memberships for nodes are optimized. We propose a Genetic Algorithm-based solution to help the network administrator in identifying node groupings in the LAN prior to VLANs implementation. Identification of such groups allows a network administrator to decide a VLAN topology that provides maximum traffic localization and broadcast containment.We select a fitness function and compare three selection mechanisms namely roulette wheel, ranking and tournament selection for different number of nodes, VLANs, crossover and mutation probabilities. Our experiments show that the ranking selection with a crossover probability of 0.3 and mutation probability of 0.05 produces good result. Moreover, we also show that the proposed GA-based approach is most efficient in comparison with other four heuristics being used to solve the same problem.

Random walk (RW) is simple to implement and has a better termination control. The Markov chain analysis informs that RW eventually visits all vertices of a connected graph. Due to such nice properties, RW is often proposed for information dissemination or collection from all or part of a large scale unstructured network. The random walker, which can be used to disseminate or collect information, visits the nodes while selecting randomly one of the neighbors. The selection of neighbors is effected by the neighbor density or the connectivity degree of the nodes. The connectivity degree in turn depends on the radius of transmission of wireless nodes. In this paper we studied the coverage process of the RW on random geometric graph. The random geometric graphs are often considered as a model for wireless ad hoc and sensor networks. We defined and studied a metric called “attenuation” that indicates how fast a RW can move in the network while disseminating or collecting information. We showed that attenuation depends on the topology, the number of nodes in a network and the transmission radius of the nodes. We then studied the effect of attenuation on the RW coverage process analytically and through simulations and showed that attenuation is the normalized estimated search time of the network. In the end we applied the results obtained to show that the estimated search time in random geometric graphs is proportional to the reciprocal of the number of replicated targets.

This paper is about energy efficient and robust implementation of random walks in mobile wireless networks. While random walk based algorithm are often proposed to solve many problems in wireless networks, their implementation is usually done at the application layer so that many characteristics of the wireless transmissions are not exploited. In this paper we show that we can greatly reduce the energy requirements to perform a walk by better exploiting the broadcast nature of the transmissions. We propose a robust, energy efficient distributed next hop selection algorithm. To evaluate the algorithm we present a simulation study performed with ns-2. We found that in the proposed algorithm energy is reduced to more than 4 times and the selection delay is reduced to more than 8 times as compared to a standard next hop selection implementation.