Packet Size Research Articles

Entropy Weighted TOPSIS Based Cluster Head Selection in Wireless Sensor Networks under Uncertainty

In recent decades, wireless sensor networks (WSNs) have become a popular ambient sensing and model-based solution for various applications. WSNs are now achievable due to the developments of micro electro mechanical and semiconductors logic circuits with rising computational power and wireless communication technology. The most difficult issues concerning WSNs are related to their energy consumption. Since communication typically requires a significant amount of energy, there are some techniques/ways to reduce energy consumption during the operation of the sensor’s communication systems. The topology control technique is one such effective method for reducing WSNs’ energy usage. A cluster head (CH) is usually selected using a topology control technique known as clustering to control the entire network. A single factor is inadequate for CH selection. Additionally, with the traditional clustering method, each round exhibits a new batch of head nodes. As a result, when using conventional techniques, nodes decay faster and require more energy. Furthermore, the inceptive energy of nodes, the range between sensor nodes and base stations, the size of data packets, voltage and transmission energy measurements, and other factors linked to sensor nodes are also completely unexpected due to irregular or hazardous natural circumstances. Here, unpredictability represented by Triangular Fuzzy Numbers (TFNs). The associated parameters of nodes were converted into crisp ones via the defuzzification of fuzzy numbers. The fuzzy number has been defuzzified using the well-known signed distance approach. Here, we have employed a multi-criteria decision-making (MCDM) approach to choosing the CHs depending on a bunch of characteristics of each node (i) residual energy, (ii) the number of neighbors, (iii) distance from the sink, (iv) average distance of cluster node, (v) distance ratio, and (vi) reliability. This study used the entropy-weighted Technique for Order Preference by Similarity to the Ideal Solution (TOPSIS) approach to select the CH in WSNs. For experiments, we have used the NSG2.1 simulator, and based on six characteristics comprising residual energy, number of neighbor nodes, distance from the sink or base station (BS), average distance of cluster nodes, distance ratio, and reliability, optimal CHs have been selected. Finally, experimental results have been presented and compared graphically with the existing literature. A statistical hypothesis test has also been conducted to verify the results that have been provided.

Packet Length Covert Channel Detection: An Ensemble Machine Learning Approach

New undetectable attack methods made possible by advances in covert channel techniques. Due to their nonstandard data transfer methods, common countermeasures are ineffective. The detection, mitigation, and elimination of hidden channels that operate on the packet length level are extremely challenging problems. Differences in packet size can be used as part of a covert communication strategy for the delivery of sensitive information across networks. The use of machine learning techniques for detecting covert channel attacks has been praised in recent academic research. Researchers in this work developed an effective ensemble classification method for spotting such breaches. Our approach relies on an ensemble model, which is a combination of three different machine learning methods. There are several applications for classifiers including the Support Vector Machine (SVM), Random Forest (RF), and Naive Bayes (NB). The logistic regression (LR) classifier served as a meta classifier to aggregate the results of the component classifiers into an ensemble classifier. According to the findings, the proposed ensemble model is effective. Among single-classification algorithms, it is unrivalled in its ability to detect such covert channel attacks.

Dynamic Bandwidth Reallocation for Enhanced Data Slice Distribution in Multi-Cloud Storage Environments

This article examines the advantages of employing many cloud platforms for data storage, alongside providing a comprehensive description of cloud computing. The significance of this matter is indisputable, considering the substantial duration required for the consolidation of such a vast amount of data into a singular cloud platform. When faced with a complex problem requiring significant time for resolution, it is advantageous to transfer a substantial amount of data across different cloud platforms. In addition to variations in transmission rates between different cloud providers, the size of Internet data packets employed by service providers may also play a factor in the transmission process. This study proposes an optimization approach to enhance data transfer and facilitate the optimal distribution of data packets when compared to a method that evenly distributed data across various clouds without considering optimal bandwidth allocation.

WFP-Collector: Automated dataset collection framework for website fingerprinting evaluations on Tor Browser

Website Fingerprinting (WFP) is a technique that analyses browsing network traffic to infer a webpage that a user browsed in Tor Browser. A sufficiently large and clean dataset is essential for quality and accurate WFP experiments. Thus, there is a corresponding need to automate the dataset collection, filtering, and validation processes. This work introduces a new paradigm, WFP-Collector, an automatic dataset collection framework for WFP experiments. WFP-Collector enables researchers to automatically (1) create a visit database for webpage crawling, (2) collect various data and log for in-depth analysis, (3) webpage visits in tablet and mobile browsing modes, (4) throttle network bandwidth and latency performance, (5) validate and filter the collected data, (6) compress and upload the collected data to cloud storage, and (7) completion notification using Telegram and email. We developed a proof-of-concept of the WFP framework for simulation and comparison. We found that the WFP-Collector framework collects nine data items and produces over 55% larger collected data size than existing approaches. The captured packet size in tablet and mobile browsing modes is up to 57.5% smaller than in desktop mode. Moreover, the file compression in WFP-Collector can reduce up to 39.9% of the storage space required for data collection.

The effect of high-temperature ECAP on dynamic recrystallization behavior and material strength of 42CrMo steel

With the development of mechanical equipment towards long life and high reliability, the traditional heat treatment method is challenging to meet the high strength requirements of 42CrMo steel for mechanical equipment. High-temperature equal channel angular pressing (ECAP) of 42CrMo steel was carried out above austenite transformation temperature, and the microstructure and material strength of the specimen at different radial positions were characterized and tested. The dynamic recrystallization (DRX) kinetic model of 42CrMo steel was established. The distribution of strain, strain rate, temperature and DRX volume fraction in high-temperature ECAP was studied by coupling finite element software and DRX kinetic model, and the formation mechanism of microstructure and material strength difference was revealed. The results show that the different degrees of DRX caused by the uneven distribution of strain and strain rate on the specimen is the main reason for the gradient distribution of microstructure and material strength in the radial direction of the specimen. From the top to the bottom of the specimen, with the decrease of strain and strain rate, the degree of DRX increases continuously, the microstructure inhomogeneity decreases, the size of the martensite packets and blocks increases gradually, and the strength decreases gradually. The grain boundary strengthening caused by the decreased size of martensite packets and blocks is the main reason for the radial difference in material yield strength. Grain boundary strengthening and the coordinated deformation of geometrically necessary dislocations at the soft-hard grain interface of heterogeneous materials are the main mechanisms leading to the tensile strength gradient distribution. This study reveals the dynamic recrystallization behavior and material strength variation of high-temperature ECAP, which contributes to the qualification of the ECAP process and promotes the development of severe plastic deformation on the further strength enhancement of high-strength steel.

On performance of short-packet ultra-reliable and low-latency communications NOMA MIMO relay networks

This paper proposes and analyzes the performance of a multi-antenna NOMA relay system over the Nakagami-m fading channel, in which the size of signal packets is finite. Based on the application of the Gaussian–Chebyshev quadrature, Riemann integral and incomplete Gamma function, the exact and asymptotic closed-form expressions of the block error rate (BLER), throughput, goodput of the proposed finite block-length multi-antenna NOMA relay system are derived. Moreover, in order to improve the BLER, throughput, and goodput of the proposed system, maximal-ratio combining and maximal-ratio transmission are applied to receivers and transmitters, respectively. The provided analysis and simulation results indicate that the proposed system meets the requirements of ultra-reliable and low-latency communications (URLLC). Especially, the reliability is higher than 99.99%, i.e., the BLER is less than 10−4. The accuracy of analysis results is verified by simulation results. Furthermore, the number of transmission bits and power allocation coefficients are optimized to maximize the throughput, and goodput and minimize the BLER.

Random Segmentation: New Traffic Obfuscation against Packet-Size-Based Side-Channel Attacks

Despite encryption, the packet size is still visible, enabling observers to infer private information in the Internet of Things (IoT) environment (e.g., IoT device identification). Packet padding obfuscates packet-length characteristics with a high data overhead because it relies on adding noise to the data. This paper proposes a more data-efficient approach that randomizes packet sizes without adding noise. We achieve this by splitting large TCP segments into random-sized chunks; hence, the packet length distribution is obfuscated without adding noise data. Our client–server implementation using TCP sockets demonstrates the feasibility of our approach at the application level. We realize our packet size control by adjusting two local socket-programming parameters. First, we enable the TCP_NODELAY option to send out each packet with our specified length. Second, we downsize the sending buffer to prevent the sender from pushing out more data than can be received, which could disable our control of the packet sizes. We simulate our defense on a network trace of four IoT devices and show a reduction in device classification accuracy from 98% to 63%, close to random guessing. Meanwhile, the real-world data transmission experiments show that the added latency is reasonable, less than 21%, while the added packet header overhead is only about 5%.

Neural-Network-Assisted Packet Accelerators for Internet of Things Network Systems

Major device nodes within the internet of things (IoT) system collects and store information in bit forms of 0’s and 1’s regardless of its repetition. The nodes do not possess the capability of processing redundant data information except for outright rejection/replacement of packets of similar sizes. This becomes a research problem since high volumes of packet redundancy are prevalent owing to repetitive information. Many optimal solutions have been provided to reprocess redundant packets and to store them in edge server for accessibility by other connected IoT systems and networks servers. To do so, major IoT platforms implements tier-based network layers which primarily aid seamless communication among nodes. These network layers perform near-similar tasks of guaranteeing packet sensing and exchange although at often-higher energy requirement. To mitigate the energy concerns, packets are clustered and compressed, allowing exchange of essential information only. But the approach continues to present heavy packet-losses and/or redundancies. In this paper, two-tier layered network | where packet exchange is conducted at the top layer in order to lower energy consumption and promote system-reliability is investigated. All packets are first segregated into multiple clusters using the Voronoi cell-based correlation cluster formation (VC3F) technique. Cluster heads (CHs) are identified by their multi-path (M-Score) value, thus, assuming sole responsibility of redundant packet removal within each cluster. The redundant packets are then moved to the edge-tier layer using optimized multi-objective flower pollination (MO-FPO) routing, and finally processed using hybrid models of novel fast-fully connected neural network (F2CNN) accelerator and Lempel-Ziv-Welch (LZW) data compression. The F2CNN and LZW models are harmonized to further collectively explore potential benefits of both models. These benefits include the neural capability to work on the sensitivity level of the packets determined by the packet classification and validation approaches. The system is evaluated with detailed experimental investigations where higher system throughput, packet delivery ratio (PDR), end-to-end delay and system reliability are corroborated.

Resilient Edge: Building an Adaptive and Resilient Multi-Communication Network for IoT Edge Using LPWAN and WiFi

Edge computing has gained attention in recent years due to the adoption of many Internet of Things (IoT) applications in domestic, industrial and wild settings. The resiliency and reliability requirements of these applications vary from noncritical (best delivery efforts) to safety-critical with time-bounded guarantees. The network connectivity of IoT edge devices remains the central critical component that needs to meet the timebounded Quality of Service (QoS) and fault-tolerance guarantees of the applications. Therefore, in this work, we systematically investigate how to meet IoT applications mixed-criticality QoS requirements in multi-communication networks. We (i) present the network resiliency requirements of IoT applications by defining a system model (ii) analyse and evaluate the bandwidth, latency, throughput, maximum packet size of many state-of-theart LPWAN technologies, such as Sigfox, LoRa, and LTE (CATM1/ NB-IoT) and Wi-Fi, (iii) implement and evaluate an adaptive system Resilient Edge and Criticality-Aware Best Fit (CABF) resource allocation algorithm to meet the application resiliency requirements using Raspberry Pi 4 and Pycom FiPy development board having five multi-communication networks.We present our findings on how to achieve 100% of the best-effort high criticality level message delivery using multi-communication networks.

Delay and size‐dependent priority‐aware scheduling for IoT‐based healthcare traffic using heterogeneous multi‐server priority queueing system

AbstractInternet of Things (IoT) based healthcare applications are time‐sensitive and any delay can cause alarming situations, including death of patients. The Earliest Deadline First (EDF) scheduling scheme has been proposed for use in IoT‐based healthcare applications. However, the EDF scheme performs poorly under overloaded conditions due to giving highest priority to packets that are close to missing their deadlines. Some studies have proposed the use of Priority EDF to overcome the challenges of EDF; however, Priority EDF still favours higher priority queues which increases the waiting times of lower priority queues. In order to overcome the limitation of EDF and its variants, this paper proposes a system model for a prioritized scheduling (PS) scheme. The PS scheme is an improvement of the Earliest Deadline First (EDF) scheme and its variants for IoT‐based healthcare applications. The PS scheme uses a heterogeneous multi‐server priority queuing system to provide service differentiation by prioritizing short packets over large packets and delay sensitive packets are serviced before delay tolerant packets. Numerical results demonstrate that the PS scheme minimizes the mean slowdown for both delay sensitive short and large packets at low and high load values. Additionally, the PS scheme performs better than the EDF and Priority EDF schemes in terms of reducing mean slowdown of packets and the PS scheme performs better than the EDF in terms of throughput for all packet sizes at both low and high load values. The performance improvement in terms of throughput is more pronounced at high load values. This addresses the challenge of the EDF scheme which performs poorly under overloaded conditions and the challenge of the Priority EDF scheme which favours higher priority queues at the expense of low priority queues.

Analysis of the Efficiency of Various Receipting Multiple Access Methods with Acknowledgement in IoT Networks

An Internet of things (IoT) network is a distributed set of “smart” sensors, interconnected via a radio channel. The basic method of accessing the radio channels for these networks is Carrier Sense Multiple Access/Collision Avoidance (CSMA/CA), in which access is carried out on the basis of contention, and confirmation of the correct reception of the packet is achieved using a receipt. If the sizes of information packets are small and comparable to the sizes of receipts, then the transmission of receipts requires a significant bandwidth of the channel, which reduces the efficiency of the network. This problem exists not only for IoT networks but also for monitoring systems, operational management of fast processes, telemetry, short messaging and many other applications. Therefore, an urgent task is to develop effective methods of multiple random access in the transmission of short information packets, the size of which is comparable to the size of receipts. To solve this problem, the authors proposed modifications of CSMA/CA random access in which, when packet collisions are detected, a diagnostic message (DM) is generated and transmitted in the broadcast mode. Based on simulation modeling, it is shown that in a wide range of network loads, the proposed random access options provide an increase in network capacity (the number of connected subscribers) of 1.5–2 times compared to the basic CSMA/CA access method when the size of the information packet is an order of magnitude larger than the size of receipts. The variant of access without acknowledgment is also considered, in which, as shown by the simulation results, at sufficiently large loads, the network can go into an unstable state.

Effect of Microstructural Evolution on Mechanical Properties of Secondary Hardening Low‐Carbon High‐Alloy Bearing Steel

The hierarchical martensite structure of secondary hardening low‐carbon high‐alloy bearing steel at different solution temperatures is characterized in detail, and the effect of microstructural evolution on mechanical properties of the experimental steel is studied. The results show there are a large number of micron‐M6C at the boundary of the prior austenite grain when the solution temperature is 1,000–1,025 °C, which is conducive to hinder the growth of the prior austenite grain and produces smaller packets that improve the yield strength of the material. At the same time, such micron‐M6C is prone to stress concentration when the material is subjected to an external force load, which can severely reduce the material toughness. The M6C completely dissolves into the matrix when the solution temperature exceeds 1,050 °C, and the packets size grows rapidly with the prior austenite grain size. At this point, the effect of packets on yield strength no longer dominates, and the rate of decrease in yield strength tends to be gradual. Under the comprehensive function of the micron‐M6C‐remelting, the martensitic blocks refinement, the increased number of the Σ3 grain boundaries, and the growth of cracks can be hindered, and the toughness of material can be improved accordingly.

Tracking the Effect of Human Activity on MeSO-net Noise Using Seismic Data Traffic—Did Seismic Noise in Tokyo Truly Decrease during the COVID-19 State of Emergency?

Abstract Human activities cause seismic noise over 1 Hz (cultural noise), and the recent articles have reported that the curtailing of socioeconomic activities during the novel coronavirus (COVID-19) pandemic in 2020 appeared to reduce high-frequency seismic noise amplitudes in cities. The Tokyo metropolitan area in Japan, where seismic stations are densely distributed and various anthropogenic activities have been closely monitored, is an ideal study area to investigate the effect of human activity on high-frequency seismic noise during the pandemic. We demonstrated that the magnitude of seismic data traffic (SDT), indexed by the packet size of continuous seismic data in WIN32 format, is a good indicator for monitoring time-dependent changes in high-frequency noise levels. The SDT of 169 Metropolitan Seismic Observation network (MeSO-net) stations—a continuous accelerometer network that is mostly located at schools in the Tokyo metropolitan area—decreased by approximately 1%–3% from March to June 2020, when a state of emergency in Japan was first declared, compared with that in the previous year. We revealed that the SDT decrease was prominent only at stations near school buildings, and the SDT trend was uncorrelated with the temporal changes in the population and vehicular traffic volume near the seismic stations. We also found strong correlations between the SDT reduction and school size (classified by the number of students enrolled), implying that the noise decrease at the MeSO-net stations during the pandemic was strongly influenced by school-based activities. Thus, the noise reduction observed at MeSO-net stations during the COVID-19 pandemic in 2020 did not provide strong evidence of quieting in the Tokyo metropolitan area.

Performance analysis and optimization of UAV-assisted NOMA short packet communication systems

In this paper, we propose and analyze an unmanned aerial vehicle (UAV)-assisted non-orthogonal multiple access (NOMA) short packet communication (SPC) system, where finite-size data packets are transmitted from a UAV to multiple users. The block error rate (BLER), throughput, and energy efficiency (EE) are used as metrics to evaluate the proposed system performance. We adopt the Gaussian-Chebyshev quadrature and the first-order Riemann integral to derive the approximate expressions of BLERs. The accuracy of the derived expressions is verified by presenting that numerical results are perfectly matched with analysis results. We also formulate optimization problems to maximize the throughput with respect to the number of transmitted bits and the altitude of the UAV. Numerical results in this paper provide useful knowledge about the relationships among BLER, throughput, EE, transmitted packet size, the altitude of the UAV, and the Rician factor.

Adaptive Algorithms for Batteryless LoRa-Based Sensors.

Ambient energy-powered sensors are becoming increasingly crucial for the sustainability of the Internet-of-Things (IoT). In particular, batteryless sensors are a cost-effective solution that require no battery maintenance, last longer and have greater weatherproofing properties due to the lack of a battery access panel. In this work, we study adaptive transmission algorithms to improve the performance of batteryless IoT sensors based on the LoRa protocol. First, we characterize the device power consumption during sensor measurement and/or transmission events. Then, we consider different scenarios and dynamically tune the most critical network parameters, such as inter-packet transmission time, data redundancy and packet size, to optimize the operation of the device. We design appropriate capacity-based storage, considering a renewable energy source (e.g., photovoltaic panel), and we analyze the probability of energy failures by exploiting both theoretical models and real energy traces. The results can be used as feedback to re-design the device to have an appropriate amount energy storage and meet certain reliability constraints. Finally, a cost analysis is also provided for the energy characteristics of our system, taking into account the dimensioning of both the capacitor and solar panel.

An Advanced Energy-Efficient Environmental Monitoring in Precision Agriculture Using LoRa-Based Wireless Sensor Networks.

Sensor networks, as a special subtype of wireless networks, consist of sets of wirelessly connected sensor nodes often placed in hard-to-reach environments. Therefore, it is expected that sensor nodes will not be powered from the power grid. Instead, sensor nodes have their own power sources, the replacement of which is often impractical and requires additional costs, so it is necessary to ensure minimum energy consumption. For that reason, the energy efficiency of wireless sensor networks used for monitoring environmental parameters is essential, especially in remote networking scenarios. In this paper, an overview of the latest research progress on wireless sensor networks based on LoRa was provided. Furthermore, analyses of energy consumption of sensor nodes used in agriculture to observe environmental parameters were conducted using the results of real measurements in the field, as well as simulations carried out based on collected data about real equipment. Optimization methods of energy consumption, in terms of choosing the appropriate data collection processes from the conducted field measurements, as well as the settings of network radio parameters imitating real conditions used in conducted simulations were highlighted. In the analyses, special emphasis was placed on choosing the optimal packet size. Unlike in other papers analyzing energy efficiency of LoRa communication, in this paper, it was proven that the adjustment of the transmission speed to the actual size of the packet is important for better energy efficiency of communication and that it can reduce energy consumption considerably. Moreover, in the paper, the contents of a packet that can be used in precision agriculture is suggested in order to prove that the 6-bit packet is sufficient for energy-efficient collection of parameters from the environment, in contrast to the 11-bit packets used in standard commercially available equipment.

Unleashing the Power of Multi-Agent Deep Learning: Cyber-Attack Detection in IoT

Detecting botnet and malware cyber-attacks is a critical task in ensuring the security of computer networks. Traditional methods for identifying such attacks often involve static rules and signatures, which can be easily evaded by attackers. Dl is a subdivision of ML, has shown promise in enhancing the accuracy of detecting botnets and malware by analyzing large amounts of network traffic data and identifying patterns that are difficult to detect with traditional methods. In order to identify abnormal traffic patterns that can be a sign of botnet or malware activity, deep learning models can be taught to learn the intricate interactions and correlations between various network traffic parameters, such as packet size, time intervals, and protocol headers. The models can also be trained to detect anomalies in network traffic, which could indicate the presence of unknown malware. The threat of malware and botnet assaults has increased in frequency with the growth of the IoT. In this research, we offer a unique LSTM and GAN-based method for identifying such attacks. We utilise our model to categorise incoming traffic as either benign or malicious using a dataset of network traffic data from various IoT devices. Our findings show how well our method works by attaining high accuracy in identifying botnet and malware cyberattacks in IoT networks. This study makes a contribution to the creation of stronger and more effective security systems for shielding IoT devices from online dangers. One of the major advantages of using deep learning for botnet and malware detection is its ability to adapt to new and previously unknown attack patterns, making it a useful tool in the fight against constantly evolving cyber threats. However, DL models require large quantity of labeled data for training, and their performance can be affected by the quality and quantity of the data used. Deep learning holds great potential for improving the accuracy and effectiveness of botnet and malware detection, and its continued development and application could lead to significant advancements in the field of cybersecurity.

Influence of Martensite/Bainite Dual Phase-Content on the Mechanical Properties of EA4T High-Speed Axle Steel.

In this work, we have investigated the effect of martensite/bainite dual phase content on the mechanical properties of EA4T high-speed axle steel. For evaluation and control of the strength, ductility, and toughness of steel, the microstructure of lath martensite (LM) and granular bainite (GB) was clarified through an optical microscope (OM), electron backscatter diffraction (EBSD), and transmission electron microscopy (TEM). Besides, the tensile fracture morphology was studied by scanning electron microscopy (SEM). For this purpose, this study conducted a quantitative analysis of the LM and GB fractions using the Pro Imaging software-2018 of OM. The remarkable effect of the LM/GB structure on mechanical properties is discussed. The results have shown that by increasing the volume fraction of the GB structure, the LM structure is refined and its microhardness and strength are improved. Meanwhile, the micro strength of LM follows the Hall-Petch relationship with the lath martensite packet size. Subsequently, the mechanical property prediction model of EA4T steel based on the LM/GB content was established by regression analysis of all experiment dates. When the LM fraction in the steel is about 40-70%, a superior combination of strength, ductility, and toughness can be obtained in EA4T steel.

A Novel Energy-Efficient Reservation System for Edge Computing in 6G Vehicular Ad Hoc Network.

The roadside unit (RSU) is one of the fundamental components in a vehicular ad hoc network (VANET), where a vehicle communicates in infrastructure mode. The RSU has multiple functions, including the sharing of emergency messages and the updating of vehicles about the traffic situation. Deploying and managing a static RSU (sRSU) requires considerable capital and operating expenditures (CAPEX and OPEX), leading to RSUs that are sparsely distributed, continuous handovers amongst RSUs, and, more importantly, frequent RSU interruptions. At present, researchers remain focused on multiple parameters in the sRSU to improve the vehicle-to-infrastructure (V2I) communication; however, in this research, the mobile RSU (mRSU), an emerging concept for sixth-generation (6G) edge computing vehicular ad hoc networks (VANETs), is proposed to improve the connectivity and efficiency of communication among V2I. In addition to this, the mRSU can serve as a computing resource for edge computing applications. This paper proposes a novel energy-efficient reservation technique for edge computing in 6G VANETs that provides an energy-efficient, reservation-based, cost-effective solution by introducing the concept of the mRSU. The simulation outcomes demonstrate that the mRSU exhibits superior performance compared to the sRSU in multiple aspects. The mRSU surpasses the sRSU with a packet delivery ratio improvement of 7.7%, a throughput increase of 5.1%, a reduction in end-to-end delay by 4.4%, and a decrease in hop count by 8.7%. The results are generated across diverse propagation models, employing realistic urban scenarios with varying packet sizes and numbers of vehicles. However, it is important to note that the enhanced performance parameters and improved connectivity with more nodes lead to a significant increase in energy consumption by 2%.

Station Grouping Mechanism using Machine Learning Approach for IEEE 802.11ah

IEEE 802.11ah, also known as Wi-Fi HaLow, improves the scalability of the Internet of Things (IoT) by connecting a large number of low-power devices. IEEE 802.11ah has developed a novel feature called Restricted Access Window (RAW) that aims to address one of the main problems of the IoT: excessive channel contention in large sensor networks. All stations can access the channel from the same group when using RAW, which allows the Access Point (AP) to split stations into different groups. Existing station grouping schemes support only the homogeneous traffic scenarios, having the same data transmission interval, packet size, and Modulation and Coding Scheme (MCS). In this paper, we highlight two contributions to solving this problem. The first is a Machine Learning (ML) based model that predicts RAW configuration parameters and performance under specific network conditions. It predicts accurately and is very quick to train. Second, a technique for a channel access mechanism that uses an ML model to determine the optimal RAW station grouping configuration for dynamic traffic conditions and heterogeneous stations. We compare our ML-based model with the existing state-of-the-art solutions. We calculated the performance of our proposed model and compared it to traditional 802.11ah and the Traffic Adaptive RAW Optimization Algorithm (TAROA). The results show that our proposed model outperforms a significant improvement in terms of throughput, packet received ratio, delay, and execution time.