The rapid growth of Internet-of-Things (IoT) in the current decade has led to the development of a multitude of new access technologies targeted at low-power, wide area networks (LP-WANs). However, this has also created another challenge pertaining to technology selection. This paper reviews the performance of LP-WAN technologies for IoT, including design choices and their implications. We consider Sigfox, LoRaWAN, WavIoT, random phase multiple access (RPMA), narrow band IoT (NB-IoT) as well as LTE-M and assess their performance in terms of signal propagation, coverage and energy conservation. The comparative analyses presented in this paper are based on available data sheets and simulation results. A sensitivity analysis is also conducted to evaluate network performance in response to variations in system design parameters. Results show that each of RPMA, NB-IoT and LTE-M incurs at least 9 dB additional path loss relative to Sigfox and LoRaWAN. This study further reveals that with a 10% improvement in receiver sensitivity, NB-IoT 882 MHz and LoRaWAN can increase coverage by up to 398% and 142% respectively, without adverse effects on the energy requirements. Finally, extreme weather conditions can significantly reduce the active network life of LP-WANs. In particular, the results indicate that operating an IoT device in a temperature of-20 • C can shorten its life by about half; 53% (WavIoT, LoRaWAN, Sigfox, NB-IoT, RPMA) and 48% in LTE-M compared with environmental temperature of 40 • C.

In recent years, IoT has developed into many areas of life including smart homes, smart cities, agriculture, offices, and workplaces. Everyday physical items such as lights, locks and industrial machineries can now be part of the IoT ecosystem. IoT has redefined the management of critical and non-critical systems with the aim of making our lives more safe, efficient and comfortable. As a result, IoT technology is having a huge positive impact on our lives. However, in addition to these positives, IoT systems have also attracted negative attention from malicious users who aim to infiltrate weaknesses within IoT systems for their own gain, referred to as cyber security attacks. By creating an introduction to IoT, this paper seeks to highlight IoT cyber security vulnerabilities and mitigation techniques to the reader. The paper is suitable for developers, practitioners, and academics, particularly from fields such as computer networking, information or communication technology or electronics. The paper begins by introducing IoT as the culmination of two hundred years of evolution within communication technologies. Around 2014, IoT reached consumers, early products were mostly small closed IoT networks, followed by large networks such as smart cities, and continuing to evolve into Next Generation Internet; internet systems which incorporate human values. Following this evolutionary introduction, IoT architectures are compared and some of the technologies that are part of each architectural layer are introduced. Security threats within each architectural layer and some mitigation strategies are discussed, finally, the paper concludes with some future developments.

Deep Learning (DL) has been widely proposed for botnet attack detection in Internet of Things (IoT) networks. However, the traditional Centralized DL (CDL) method cannot be used to detect previously unknown (zero-day) botnet attack without breaching the data privacy rights of the users. In this paper, we propose Federated Deep Learning (FDL) method for zero-day botnet attack detection to avoid data privacy leakage in IoT edge devices. In this method, an optimal Deep Neural Network (DNN) architecture is employed for network traffic classification. A model parameter server remotely coordinates the independent training of the DNN models in multiple IoT edge devices, while Federated Averaging (FedAvg) algorithm is used to aggregate local model updates. A global DNN model is produced after a number of communication rounds between the model parameter server and the IoT edge devices. Zero-day botnet attack scenarios in IoT edge devices is simulated with the Bot-IoT and N-BaIoT data sets. Experiment results show that FDL model: (a) detects zero-day botnet attacks with high classification performance; (b) guarantees data privacy and security; (c) has low communication overhead (d) requires low memory space for the storage of training data; and (e) has low network latency. Therefore, FDL method outperformed CDL, Localized DL, and Distributed DL methods in this application scenario. Index Terms-Cybersecurity, botnet detection, federated learning, deep learning, deep neural network, Internet of Things. I. INTRODUCTION B OTNET attack is a serious cyber security challenge facing the Internet of Things (IoT) [1]-[3]. In our context, a botnet is a network of compromised devices that is used to launch cyber attack against critical infrastructures [4]. This cyber attack may be in form of Denial of Service (DoS), Distributed DoS (DDoS), reconnaissance, or data theft [5].

Nowadays, hackers take illegal advantage of distributed resources in a network of computing devices (i.e., botnet) to launch cyberattacks against the Internet of Things (IoT). Recently, diverse Machine Learning (ML) and Deep Learning (DL) methods were proposed to detect botnet attacks in IoT networks. However, highly imbalanced network traffic data in the training set often degrade the classification performance of state-of-the-art ML and DL models, especially in classes with relatively few samples. In this paper, we propose an efficient DL-based botnet attack detection algorithm that can handle highly imbalanced network traffic data. Specifically, Synthetic Minority Oversampling Technique (SMOTE) generates additional minority samples to achieve class balance, while Deep Recurrent Neural Network (DRNN) learns hierarchical feature representations from the balanced network traffic data to perform discriminative classification. We develop DRNN and SMOTE-DRNN models with the Bot-IoT dataset, and the simulation results show that high-class imbalance in the training data adversely affects the precision, recall, F1 score, area under the receiver operating characteristic curve (AUC), geometric mean (GM) and Matthews correlation coefficient (MCC) of the DRNN model. On the other hand, the SMOTE-DRNN model achieved better classification performance with 99.50% precision, 99.75% recall, 99.62% F1 score, 99.87% AUC, 99.74% GM and 99.62% MCC. Additionally, the SMOTE-DRNN model outperformed state-of-the-art ML and DL models.

Corporations and end users are nding it hard to keep their devices safe from the ever evolving and complicated threat of cyber a acks. Currently, with the widespread adoption of the Internet of ings (IoT), cyber threat is becoming an even greater challenge for both technology providers and consumers. is paper presents a review of the recent and signi cant cyber security issues a ecting many areas of digital technology. From IoT devices and smart automobiles to commonly used computers and typical corporate servers, we focus our analysis on current a ack trends and the e ects of intrusion on Small and Medium sized Enterprises(SMEs). is paper helps to build awareness among non-technical experts, practitioners and researchers about a ack and defense strategies in the current digital market. We have created a guide with input from our in-house security researchers and information gathered from the literature to help the reader understand the challenges faced by the IT industry in the future. CCS CONCEPTS •Security and privacy → security services; Intrusion/anomaly detection and malware mitigation; •Computer systems organization → Dependable and fault-tolerant systems and networks; •Networks → Network reliability;