Search

Keywords Access control; Multi-domain Authorization; Mutual trust mechanism Abstract The problem of mutual trust between users and edge nodes in an edge computing environment is a sufficient guarantee for the double security of edge computing service users and service providers. How to achieve the user's security and credibility and ensure the trust of the edge service nodes is a serious challenge facing the security of edge computing. Therefore, the research target of this paper is to establish the mutual trust relationship between users and edge service nodes and to formulate dynamic access control strategies. While realizing user security and trustworthiness in the edge environment, it also ensures the trustworthiness of service nodes. The research content of this paper is the access control model of mutual trust between nodes and users in the edge computing environment. This model closely combines trust management with the RBAC (Role Based Access Control) model. Given the multi-domain characteristics of the edge computing environment, the intra-domain access control and cross-domain access control strategies based on mutual trust are implemented, which realizes the improvement and expansion of the RBAC model in the edge computing environment. The performance of EUSMTM is compared and analyzed through simulation experiments, and the validity of the trust model between users and edge servers is verified. Then through the comparative experiment of two-way trust and one-way trust between edge users and edge servers, the advantages of access control based on two-way trust in the EUSMTM model are analyzed. Author Affiliation: (a) School of Computer Science, Hunan University or Technology and Business, Changsha, 410205, China (b) Xiangtan Institute of Technology, Xiangtan University, Xiangtan City, Hunan Province, China * Corresponding author. Article History: Received 28 March 2022; Revised 6 September 2022; Accepted 15 March 2023 Byline: Yongxia Sun (a), Weijin Jiang [jwj3666@163.com] (a,*), Ying Yang (a), Haoran Zhu (a), Yirong Jiang (b)

Keywords Unmanned aerial vehicles (UAV); Flying ad hoc network (FANET); Topology construction; Topology adjustment; Relay transmission Abstract A flying ad hoc network (FANET) is formed by multiple unmanned aerial vehicles (UAVs) in an ad hoc manner in which UAVs can cooperate more effectively to complete complex tasks. In this paper, we mainly investigate the topology construction and topology adjustment for FANETs in which relay UAVs are used for relay transmission between mission UAVs and a ground control station (GCS), so that each mission UAV can communicate with the GCS through a multi-hop communication connection. Firstly, we propose an alternating optimization algorithm to optimize the routing paths and the positions of relay UAVs alternately, so as to construct a feasible FANET topology using the minimal number of relay UAVs. Then, we propose a greedy algorithm to deploy new relay UAVs to the FANET topology constructed by the alternating optimization algorithm, so as to construct a high-performance FANET topology using the given number of relay UAVs. Both the alternating optimization algorithm and the greedy algorithm need to be executed in a centralized manner, so these two algorithms can only be used for FANET topology construction. Furthermore, we propose a distributed algorithm to adjust the existing FANET topology when the positions of mission UAVs change, so as to improve the FANET network performance. This distributed algorithm can be executed in a distributed manner, so this algorithm can be used for FANET topology adjustment. The simulation results show that the algorithms proposed in this paper are superior to other baselines. Author Affiliation: Command & Control Engineering College, Army Engineering University, Nanjing 210007, China * Corresponding author. Article History: Received 19 August 2022; Revised 23 March 2023; Accepted 31 March 2023 Byline: Yaqun Liu, Jun Xie [xiejun@aeu.edu.cn] (*), Changyou Xing, Shengxu Xie

Keywords Infrastructure protection; Network modelling; Multilayer networks; Network analytics; Core and metro networks Abstract The physical assets within a critical infrastructure system are pivotal to its efficient performance and protection and that of other dependent systems. This is particularly the case for communication systems where network protection strategies usually involve asset redundancy. Although such redundancy is well-modelled in the literature, there is a gap in knowledge from a network science perspective in terms of its implications for network modelling and performance assessment. This paper presents a multilayer complex network framework that takes into account the heterogeneity of the redundant infrastructure for realistic network modelling and further analysis, a step change from using a single network model. Key performance indicators (KPIs) for communication networks (i.e., latency and jitter, bandwidth and throughput, queue depth and packet drops) are redefined to evaluate key important features of a long-haul backbone network such as network capacity and average use. In addition, these KPIs are adapted to deal with the aforementioned redundancy and so inform network managers with values defined over a model closer to the real system. The paper analyses the use case of a nationwide core and metro network infrastructure of one of the main UK internet service providers. The results of the analysis of KPIs showcase the advantage of the proposed multilayer complex network framework over the traditional single network model. Critical network elements within different dimensions of a communication network are identified based on their performance for prioritising network management measures. Author Affiliation: (a) Institute for Manufacturing, Department of Engineering, University of Cambridge, 17 Charles Babbage Road, Cambridge CB3 0FS, United Kingdom (b) British Telecomm plc (BT), Adastral Park -- Barrack Square, Martlesham Heath, Martlesham, Ipswich IP5 3RE, United Kingdom * Corresponding author. Article History: Received 14 July 2022; Revised 20 February 2023; Accepted 26 March 2023 (footnote)[white star] This research was funded by the EPSRC, United Kingdom and BT, United Kingdom Prosperity Partnership project: Next Generation Converged Digital Infrastructure, grant number EP/R004935/1. For the purpose of open access, the authors have applied a Creative Commons Attribution (CC BY) licence to any Author Accepted Manuscript version arising. Byline: Manuel Herrera [amh226@cam.ac.uk] (a,*), Manu Sasidharan (a), Stephen Cassidy (b), Ajith Kumar Parlikad (a)

Keywords Virtual computer networks; Automatic network security; Security function selection Abstract The incessant growth of network virtualization determined the proliferation of Virtual Network Functions (VNFs), software programs that can run on general-purpose servers and that can also integrate security controls for protection from cyber-attacks. However, a high availability of VNFs may be counterproductive for the network administrators who have to select the most suitable ones to establish the security configuration of their network. On the one hand, the vendor-dependent technicalities of each VNF may cloud the security controls it can actually perform. On the other hand, VNF selection traditionally occurs before the synthesis of the virtual network graph, so it does not employ any network information and it may outcome unoptimized results. In light of these shortcomings, this paper proposes a novel security configuration workflow, based on new abstractions that we call projections. They represent the security-related operations that VNFs should perform to enforce a security policy. Thanks to these abstractions, the actual selection of the VNFs can be postponed to the moment their deployment in the physical network is actually required. In fact, projections are enough for the synthesis of the virtual security graph. This paper also proposes a two-step algorithm for computing projection chains as candidate solutions for graph synthesis. The proposed approach has been implemented as a Java framework and a set of tests have validated its applicability to real-world VNFs, correctness, scalability and optimization. These tests showed that the new security configuration workflow can achieve a significant reduction for the number of selected VNFs and their deployment cost. Specifically, in the analyzed scenario, the improvement percentages for these two parameters are 79% and 90% with respect to the worst-case strategy, while 68% and 77% with respect to a traditional more optimized configuration strategy. Author Affiliation: Dipartimento di Automatica e Informatica, Politecnico di Torino, Torino, Italy * Corresponding author. Article History: Received 17 January 2023; Revised 17 March 2023; Accepted 25 March 2023 Byline: Daniele Bringhenti [daniele.bringhenti@polito.it] (*), Riccardo Sisto [riccardo.sisto@polito.it], Fulvio Valenza [fulvio.valenza@polito.it]

Keywords Blockchain; Smart contracts; DLT; Internet of Things (IoT) Abstract In recent times, IoT has emerged as a new paradigm for the interconnection of heterogeneous, resource-constrained, and communication-capable smart devices. It has been anticipated as a key enabler for various domains of applications such as health care, automotive, agriculture, industrial operations, automation, energy, and the next generation of living. However, the current IoT applications face significant challenges in terms of the huge amount of collected data, intensive data exchange, security, privacy, centralized processing, and interoperability. To mitigate many of these issues, blockchain has been identified as a promising innovative technology. Blockchain, in conjunction with smart contracts, has received significant attention both from the industry and academia and offers features such as irreversibility, non-repudiation, proof of provenance, fault tolerance, pseudonymity, decentralized operations and decision-making, and distributed ledger. The integration of blockchain with IoT requires essential insights concerning the application areas, scalability, security, privacy, data college and storage, performance, and governance. Thus, this paper intends to expound on the opportunities and key aspects of using blockchain in the IoT landscape. Specifically, this paper surveys the utilization of blockchain for various IoT applications. Besides, the paper distinguishes different technical aspects and presents the associated research challenges. At last, future research directions are discussed depending on the lessons learned. Author Affiliation: (a) Department of Computer and Communication Engineering, Manipal University Jaipur, Jaipur, India (b) Department of Computer Engineering, CGPIT, Uka Tarsadia University, Bardoli, Gujarat, India (c) Institute of Applied Information Technology (InIT), Zurich University of Applied Sciences (ZHAW), Switzerland (d) School of Computer Science, University College Dublin, Ireland * Corresponding authors. Article History: Received 28 April 2022; Revised 25 February 2023; Accepted 16 March 2023 Byline: Shikha Mathur [shikhamathur806@gmail.com] (a), Anshuman Kalla [anshuman.kalla@ieee.org] (b,*), Gürkan Gür [gueu@zhaw.ch] (c), Manoj Kumar Bohra [manojkumar.bohra@jaipur.manipal.edu] (a,*), Madhusanka Liyanage [madhusanka@ucd.ie] (d)

Keywords Resource allocation; MEC; Edge computing; 5G Highlights * New 5G MEC resource allocation problems classified based on problem nature: target, resource type, issue, assumptions. Characterization not only based on existing works but also presents future opportunities. * Proposes a method classification for MEC resource allocation. Categorized by solved resource issues. * The paper summarizes the main findings for each problem category (e.g., target, resource type, resource issue, assumption) and the corresponding methods, and relates them to future research opportunities and open issues. * This is the first study on 5G-MEC resource allocation that discusses the state-of-the-art and future challenges for URLLC services in 5G and 6G. Abstract Innovative services with strict requirements are expected in the fifth generation (5G) of mobile networks and beyond. For example, the Ultra-Reliable Low-Latency Communication (URLLC) requires up to 1 ms latency, end-to-end security, and reliability of up to 99.999%. The Multi-access Edge Computing (MEC) promises to support the delivery of URLLC services by providing computing and storage resources in the proximity of user equipment. The data which previously needed to be processed and stored in the cloud systems can be kept at the edge network, decreasing the total latency and increasing the context-awareness, security, and dependability. Vastly available resources, which are available from cloud to edge, must be appropriately allocated to deliver a service efficiently. The resource allocation problem in MEC for 5G-and-beyond networks can be formulated differently, depending on the nature of the problem. This survey outlines the resource allocation problem as a proper problem formulation, which can be addressed by target, resource type, resource issue, and the considered assumptions. Moreover, this paper also describes the open issues and future directions for MEC resource allocation based on the state of the art on this research topic. Author Affiliation: University of Stavanger, postboks 8600, Stavanger, 4036, Norway * Corresponding author. Article History: Received 17 October 2022; Revised 8 February 2023; Accepted 15 March 2023 (footnote)[white star] This work has been funded by the The Research Council of Norway through the 5G-MODaNeI project (no. 308909). Byline: Annisa Sarah [annisa.sarah@uis.no] (*), Gianfranco Nencioni [gianfranco.nencioni@uis.no], Md. Muhidul I. Khan [md.m.khan@uis.no]

Keywords Reconfigurable intelligent surface; UAV network; Multicast communication; Resource allocation; MP-DQN Abstract This paper investigates an unmanned aerial vehicle (UAV)-enabled multicast network, where the UAV serves as a mobile transmitter to send typical contents to its corresponding ground receivers. A reconfigurable intelligent surface (RIS) is deployed to enhance the service quality with a limited power supply in the UAV-enabled multicast network. It can reconfigure the signal propagation environment and improve the received power of ground receivers by adjusting the reflection coefficients. The sum rate maximization problem is formulated by jointly designing the UAV movement, RIS reflection matrix, and beamforming design from the UAV to users. This paper proposes a Beamforming control and Trajectory design algorithm based on a Multi-Pass Deep Q-Network (BT-MP-DQN). In the proposed algorithm, the UAV acts as an agent for periodically observing the state of the UAV multicast network and takes actions to adapt to the dynamic environment. Specifically, the movement of the UAV is discrete action, and the beamforming design is continuous action. The simulation results show that this proposed algorithm can effectively improve the achievable rate and satisfy the minimum rate of multicast group users. The deployment of the RIS is beneficial to network performance enhancement. In addition, the multicast network with UAV also outperforms the conventional multicast channel with a fixed-location transmitter. Author Affiliation: (a) School of Computer Science and Engineering, Northeastern University, Shenyang, Liaoning, 110819, China (b) Engineering Research Center of Security Technology of Complex Network System, Ministry of Education, China * Corresponding author at: School of Computer Science and Engineering, Northeastern University, Shenyang, Liaoning, 110819, China. Article History: Received 8 October 2022; Revised 22 February 2023; Accepted 9 March 2023 Byline: Pengshuo Ji (a,b), Jie Jia [jiajie@mail.neu.edu.cn] (a,b,*), Jian Chen (a,b), Liang Guo (a), An Du (a), Xingwei Wang (a)

Keywords Sketch; Sliding window sketch; Network measurement Abstract As an important basis for network management, effective network measurement is critical for improving network performance and security. As an efficient tool for network measurement, sketch is a probabilistic data structure which can measure traffic statistics with low overhead. Considering that the most recent data of flows are more significant, sliding window sketch is proposed by combining sliding window model with the sketch. Sliding window sketch focuses on measuring information of the most recent period, e.g., data of the last N items or in the last N time units of measured traffic. By prioritizing the latest data, sliding window sketch can better reflect the current network situation and the future trend while avoiding information loss. However, implementation of sliding window sketch is very challenging considering the memory limitations of network devices and the need to maintain the window content in real-time. This paper conducts a comprehensive survey on the latest research works and provides insights into sliding window sketch. First, we briefly review the fundamentals of network measurement and sketch, and then we thoroughly go through and analyze the existing works on sliding window sketch for three different types. Afterwards, we provide a comparative analysis on the design of sliding window sketch in terms of data structures, supported operations, measurement tasks, implementation platforms and performance. Finally, we summarize this survey paper and highlight some future research directions. Author Affiliation: Department of Computer Science, Jinan University, Gangzhou, China * Corresponding author. Article History: Received 16 November 2022; Revised 3 February 2023; Accepted 8 March 2023 Byline: Zijie Zeng, Lin Cui [tcuilin@jnu.edu.cn] (*), Mimi Qian, Zhen Zhang, Kaimin Wei

Keywords Status update systems; Age of information; Peak age of information; Proportional fairness; Temporal fairness Abstract Proportional Fair (PF) scheduling with successful deployments in various cellular wireless networks and wireless LANs, aims at maximizing the sum of the logarithms of user throughputs. PF scheduling is known to strike an appropriate balance between fairness and throughput, for conventional data traffic. On the other hand, there has recently been a surge of interest in status update networks carrying age-sensitive traffic for which information freshness is crucial and therefore network performance metrics driven by Age of Information (AoI) are instrumental, as opposed to conventional performance metrics such as delay, loss, or throughput, used for conventional data traffic. This paper studies the scheduling problem for the downlink of a cellular wireless network with a transmitter sending age-sensitive status update packets from multiple information sources to users with the goal of keeping the information as fresh as possible for the users. For this purpose, under the generate-at-will scenario, an age-agnostic model-free scheduler is proposed with the goal of minimizing the weighted sum peak AoI of the network, which is the performance metric used in this paper for quantifying information freshness. With numerical examples, the proposed scheduler is compared and contrasted with weighted PF scheduling in terms of implementation and performance, in both non-opportunistic and opportunistic scenarios. Author Affiliation: Bilkent University, Electrical and Electronics Engineering Department, Bilkent, 06800, Ankara, Turkey * Corresponding author. Article History: Received 9 October 2022; Revised 10 February 2023; Accepted 28 February 2023 Byline: Nail Akar [akar@ee.bilkent.edu.tr] (*), Ezhan Karasan

Keywords Social networks; Competitive propagation; Target user; Influence maximization Highlights * To make the dissemination targeted and the revenue maximized, we propose the lowest cost problem to achieve dominant coverage for target users under competitive propagation(LC-CTU problem), using target user profiles and setting the objective to dominate coverage. * To better model this scene, this paper proposes the IC-CRA model, which adds competing, delay and attribute to the basic cascade model. After proving that the objective function is monotonic and submodular, the LCH-TU problem can be solved by a greedy algorithm. * To improve the efficiency of the algorithm, we propose the LCH-TU algorithm. It simplifies the calculation by restricting the influence of nodes in the maximum activation probability path and calculating the local influence recursively using the linear relationship of the path. Abstract Multiple pieces of information disseminated in social networks can create a competitive environment where the influence of each message will be reduced. Thus, how to increase the influence under competitive propagation is significant for many applications such as product marketing and rumor suppression. In this paper, to make the dissemination targeted and maximize revenue, we propose the lowest cost problem to achieve dominant coverage for target users under competitive propagation(LC-CTU problem), using target user profiles and setting the objective to dominate coverage. To better model this scene, this paper proposes a competitive restricted propagation independent cascade model based on weights of attribute-generating edges(IC-CRA model), which adds competition, delay, and attributes to the basic independent cascade model. After proving that the LC-CTU problem under the IC-CRA model is NP-hard and the objective function is monotonic and submodular, the problem can be solved by a greedy algorithm. However, the greedy algorithm is time-consuming for large social networks. Thus, to improve the efficiency of the algorithm, we propose the lowest cost heuristic algorithm based on target users and restricted competition (LCH-TU algorithm), which uses the local influence update of each node for selecting the seed node. Finally, this paper demonstrates the effectiveness, efficiency, and scalability of the LCH-TU algorithm through extensive experiments on seven datasets, including real and artificial networks. Author Affiliation: Institute of Cyber Science and Technology, Shanghai Jiao Tong University, Shanghai 200240, China Shanghai Key Laboratory of Integrated Administration Technologies for Information Security, Shanghai, China * Correspondence to: School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China. Article History: Received 31 December 2022; Revised 5 March 2023; Accepted 8 March 2023 (footnote)[white star] This work is supported by the National Natural Science Foundation of China under Grant 62172278. Byline: Jiaxing Zheng, Li Pan [panli@sjtu.edu.cn] (*)

Keywords Internet of Underwater Things; Opportunistic routing; Route metric; Forwarder selection; Priority scheduling; Packet holding time Abstract Compared with traditional ocean monitoring technologies, such as cable seabed observation networks, the Internet of Underwater Things (IoUT) can provide real-time underwater environment monitoring and improve efficiency of data collection. Data collection is one of the most essential functions of IoUT, while it is very challenging to achieve efficient data transmission in IoUT. Opportunistic routing is considered to have great advantages and potential in mitigating the drawbacks of underwater communication and improving the performance of data transmission. Therefore, in this paper, the opportunistic routing protocols for IoUT are reviewed and summarized. Taking advantage of the broadcast characteristic of wireless underwater communication, opportunistic routing selects a group of suitable neighbors as potential forwarders instead of selecting only one or a few specific next-hop nodes as in conventional routing, thus improving the success rate of data packet delivery. Critical to routing performance is which nodes participate in data forwarding as forwarders. The nodes act as forwarders are determined by user-defined routing parameters, which are referred to as route metrics in this paper. First, route metrics of opportunistic routing are analyzed and a comprehensive classification of opportunistic routing protocols is discussed based on the route metrics. Then, opportunistic routing protocols for IoUT are further investigated and compared in terms of the adopted communication technique, data delivery ratio, transmission delay efficiency, energy efficiency, buffer efficiency, reliability, etc. Based on the comprehensive classification and comparison, existing problems and future research issues regarding opportunistic routing in IoUT are finally presented. Author Affiliation: (a) College of Internet of Things Engineering, Hohai University, 200 North Jinling Road, Changzhou 213022, China (b) Northeastern University, China * Corresponding author. Article History: Received 18 May 2022; Accepted 20 February 2023 Byline: Jinfang Jiang [jiangjinfang@hhu.edu.cn] (a), Guangjie Han [hanguangjie@gmail.com] (a,*), Chuan Lin [chuanlin1988@gmail.com] (b)

Keywords LEO routing; LEO SLA; Demand Island Abstract Low Earth Orbit (LEO) constellations create a network that includes the satellites (as routing nodes) connected by Inter-Satellite Links (ISLs) and the terminals dynamically connected to one or more satellites. The combination of transient high rate changes in the network topology due to terminals handover, and the end-to-end high propagation time between the routing nodes presents a unique challenge for designing a guaranteed-bandwidth routing protocol that can support the frequent changes. In addition, terminals connected to multiple satellites can balance their traffic between multiple paths in a frame granularity based on the links conditions, and require real-time multipath guaranteed-bandwidth. Prior work focuses on end-to-end routing between multiple gateways and terminals and does not provide a multipath guaranteed-bandwidth service. This paper addresses the problem of sending traffic from a source terminal to a destination terminal connected through multiple satellites, while guaranteeing and enabling planning of the multipath service metrics (bandwidth and latency) and handling satellite handovers. A novel paradigm called Demand Islands is presented. This paradigm takes advantage of the geographical properties of the demand. The service area is split into geographical rectangles (Demand Islands), with a gateway (GW) located at the center of each island. Each Demand Island is dynamically associated with a set of satellites that cover it. We model the set of satellites as a grid. The routing algorithm uses the grid properties to provide real-time multipath guaranteed-bandwidth routing for the demand allocated to each terminal. Each satellite executes the algorithm to calculate the next hop for a received message in the time complexity of O(1). The analysis and simulation demonstrate the unique offering of the algorithm to route real-time guaranteed-bandwidth to terminals connected to multiple satellites. Author Affiliation: School of Electrical and Computer Engineering, Ben-Gurion University of the Negev, Beer-Sheva, Israel * Corresponding author. Article History: Received 21 April 2022; Revised 9 February 2023; Accepted 20 February 2023 (footnote)[white star] Preliminary version of this paper was published in IEEE ICC 2021. Byline: Oren Markovitz [orenmar@post.bgu.ac.il], Michael Segal [segal@bgu.ac.il] (*)

Keywords Edge computing; DDoS attacks; Collaborative prediction; Distributed SDN; LSTM Abstract Edge computing (EC) has greatly facilitated the deployment of networked services with fast responses and low bandwidth, by deploying computing and storage at the network edge which is closer to the data sources. However, it is challenging to have the EC servers gain security protections like those in centralized data centers, making them more vulnerable to security attacks, especially distributed denial-of-service (DDoS) attacks. Existing detection approaches relying on the feedback from EC servers under the attacks can incur high bandwidth costs and service performance degradation. In this paper, we propose a new framework CoWatch for collaborative prediction and detection of DDoS Attacks in EC scenarios. Based on the distributed software-defined networking (SDN) architecture, CoWatch can collaboratively predict the DDoS attacks towards the EC servers and detect the attack flows in time. To efficiently filter the suspicious flows in distributed SDN, we devise an optimal threshold model by balancing the trade-off between collaboration efficiency and prediction effectiveness. We also explore and build on the LSTM model to design an algorithm for collaborative prediction and detection of DDoS Attacks. Experiment results on a number of datasets demonstrate the promising performance of CoWatch in effectiveness and efficiency. Author Affiliation: (a) School of Computer Science and Technology, Harbin Institute of Technology, Shenzhen, China (b) Department of Computer Science, City University of Hong Kong, Hong Kong, China (c) Cyberspace Security Research Center, Peng Cheng Laboratory, Shenzhen, China * Corresponding author. Article History: Received 2 August 2022; Revised 8 January 2023; Accepted 15 February 2023 (footnote)[white star] A preliminary version (Zhou et al., 2021) of this paper was presented at The IEEE Global Communications Conference (GLOBECOM), 2021. Byline: Hongliang Zhou [hlzhou@hit.edu.cn] (a), Yifeng Zheng [yifeng.zheng@hit.edu.cn] (a,*), Xiaohua Jia [csjia@cityu.edu.hk] (a,b), Jiangang Shu [shujg@pcl.ac.cn] (c)

Keywords 5 G, 6 G; Healthcare; Heterogeneous networks; Wireless networks; Wireless body area network Abstract Digital healthcare applications are part of the Industry 4.0 which is based on smart devices and trustworthy communications technology. The concept of Digital-care is based on providing a permanent connection between the patient and the medical department (medical staff), and the ability to analyze the patient's data automatically. With the existence of modern wireless networks, which support, for example, the Internet of Things (IoT), cloud computing, virtual reality, virtual machine, and low-power wireless networks it is possible to achieve this permanent connection while maintaining confidentiality and privacy of patient data. Digital-care applications rely on essential features, such as ensuring availability, almost zero delays, adequate storage capacity, and data protection. To achieve this term, many challenges must be solved. This paper presents a comprehensive concept of Digital-care applications, the wireless network infrastructure required by those applications, the role of current and future wireless networks in their prospects, and the challenges they face. Moreover, this paper also provides a comprehensive discussion of the proposed solutions that can be used on modern wireless networks for adoption. Author Affiliation: (a) College of Computer Information Technology, American University in the Emirates, Emirates 503000, United Arab Emirates (b) Institute of Networked and Embedded Systems, University of Klagenfurt, 9020 Klagenfurt, Austria (c) Ubiquitous Sensing Systems Lab, University of Klagenfurt-Silicon Austria Labs, 9020 Klagenfurt, Austria (d) Centre for Wireless Communications, PO Box 4500, 90014 Oulu University of Oulu, Finland (e) College of Computing and Informatics, Saudi Electronic University, Riyadh 11673, Saudi Arabia * Corresponding author. Article History: Received 27 January 2022; Revised 24 October 2022; Accepted 27 January 2023 Byline: Rami Ahmad [rami.alshwaiyat@aue.ae] (a,b,c,*), Matti Hämäläinen [matti.hamalainen@oulu.fi] (d), Raniyah Wazirali [r.wazirali@seu.edu.sa] (e), Tarik Abu-Ain [t.aboain@seu.edu.sa] (e)

Keywords Privacy; Anonymity protocol; Relationship anonymity; Internet Abstract Lightweight anonymity protocols in the network layer are promising in terms of high throughput and low latency. The realistic and weak assumption of local adversaries deletes the need for hop-by-hop encryption of onion routing, and this contributes to high throughput and low latency. Among them, PHI and its successor dPHI are promising due to the fact that path-setup packets for anonymous paths are forwarded according to IP routing. This feature enables easy deployment of the protocol on the Internet infrastructure. However, PHI and dPHI are vulnerable to adversaries such as malicious ASes who compromise servers for path setup, called helpers, and malicious ASes who leverage topological information and routing policies of the IP network. The paper resolves the vulnerabilities in the two steps. In the first step, we design attacks that leverage the vulnerabilities to break anonymity. In the second step, we extend dPHI to mitigate such attacks by adding a new server called a guard. The extended protocol is called gPHI, and gPHI is resilient against adversaries who perform such attacks. This paper designs gPHI and validates its resilience through formal proof and simulation. Author Affiliation: Graduate School of Information Science and Technology, Osaka University, Japan * Corresponding author. Article History: Received 26 August 2022; Revised 23 December 2022; Accepted 2 January 2023 (footnote)[white star] This work has been supported by NICT in Japan (Contract No.224). Byline: Yutaro Yoshinaka [y-yoshinaka@ist.osaka-u.ac.jp] (*), Junji Takemasa, Yuki Koizumi, Toru Hasegawa

Keywords Internet of Things (IoT); Bluetooth Low Energy (BLE); Generic ATTribute (GATT); Internet Gateway; Dynamic network; Smart agriculture Abstract Bluetooth Low Energy (BLE) is a wireless communication technology for power-constrained Internet of Things (IoT) applications. BLE data can be transmitted via either the IPv6 or the Generic ATTribute (GATT) Profile protocol, with the former supporting dynamic IoT structures and the latter being application-friendly. In fact, GATT requires the data layout to be known in advance by peer devices, in order to properly interpret the received data. In this paper, we introduce DynGATT, a protocol that achieves the benefits of both IPv6 and GATT, by extending GATT in a seamless fashion to support dynamic IoT structures. The key idea of DynGATT is to use GATT descriptors, originally intended to specify data in static IoT scenarios, to also specify IoT systems whose structures may dynamically evolve. Peer devices reading these descriptors will know how to interpret the data of GATT characteristics provided by devices joining the IoT network. Because no additional data have to be transmitted, the connection time is then reduced with respect to classical BLE. DynGATT has been implemented and tested in an agricultural IoT application, with different types of sensor nodes. Our experimental evaluation shows that DynGATT is very power-efficient, despite its added flexibility. Its worst-case power consumption is only around 19.37 [micro]A per data transmission and around 41.37 [micro]A overall. This consumption can be further reduced by using the methods discussed in this paper. To the best of our knowledge, this work is the first to support dynamic IoT structures in a GATT-based setting. Author Affiliation: (a) Independent Researcher, Wien, Austria (b) Internet of Things (IoT) Laboratory, Department of Engineering and Architecture, University of Parma, Parma, Italy (c) Institute of Computer Engineering, Vienna University of Technology, Wien, Austria * Corresponding author. Article History: Received 17 July 2022; Revised 3 December 2022; Accepted 3 January 2023 (footnote)1 He was with the Institute of Computer Engineering of the Vienna University of Technology when contributing to this work. Byline: Christian Hirsch [christian@hirsch.zone] (a,1), Luca Davoli [luca.davoli@unipr.it] (b,*), Radu Grosu [radu.grosu@tuwien.ac.at] (c), Gianluigi Ferrari [gianluigi.ferrari@unipr.it] (b)

Keywords Deep learning; 6G; 5G; Federated learning; Reinforcement learning; O-RAN; Zero-trust architecture Abstract In this position paper, we discuss the critical need for integrating zero trust (ZT) principles into next-generation communication networks (5G/6G). We highlight the challenges and introduce the concept of an intelligent zero trust architecture (i-ZTA) as a security framework in 5G/6G networks with untrusted components. While network virtualization, software-defined networking (SDN), and service-based architectures (SBA) are key enablers of 5G networks, operating in an untrusted environment has also become a key feature of the networks. Further, seamless connectivity to a high volume of devices has broadened the attack surface on information infrastructure. Network assurance in a dynamic untrusted environment calls for revolutionary architectures beyond existing static security frameworks. To the best of our knowledge, this is the first position paper that presents the architectural concept design of an i-ZTA upon which modern artificial intelligence (AI) algorithms can be developed to provide information security in untrusted networks. We introduce key ZT principles as real-time Monitoring of the security state of network assets, Evaluating the risk of individual access requests, and Deciding on access authorization using a dynamic trust algorithm, called MED components. To ensure ease of integration, the envisioned architecture adopts an SBA-based design, similar to the 3GPP specification of 5G networks, by leveraging the open radio access network (O-RAN) architecture with appropriate real-time engines and network interfaces for collecting necessary machine learning data. Therefore, this work provides novel research directions to design machine learning based components that contribute towards i-ZTA for the future 5G/6G networks. Author Affiliation: (a) Marconi-Rosenblatt AI/ML Innovation Laboratory, ANDRO Computational Solutions, LLC, Rome, 13440, NY, USA (b) Department of Electrical Engineering, University at Buffalo, The State University of New York, Buffalo, 14260, NY, USA * Corresponding author at: Marconi-Rosenblatt AI/ML Innovation Laboratory, ANDRO Computational Solutions, LLC, Rome, 13440, NY, USA. Article History: Received 19 February 2022; Revised 26 July 2022; Accepted 3 September 2022 Byline: Keyvan Ramezanpour (a), Jithin Jagannath [jjagannath@androcs.com] (a,b,*)

Keywords Network function virtualization; Virtual network function allocation; Service function chain; Network resilience; Redundancy; Network diversity Abstract This paper proposes an optimization model to derive a resilient virtual network function allocation in service function chains aiming to reduce the end-to-end (E2E) latency during the migrations from the primary functions to backup functions. The model considers k-fault tolerance assurance and the satisfactions of service requirements under different error patterns in this model. The allocation provided by the proposed model ensures that the processing ability satisfies the requirements even though there are k failed nodes in the network. Diversity splits a single virtual network function (VNF) into a pool of replicas with different specifications. The diversity of both primary and backup functions is considered. Redundancy is used for recovering the failed functions. We formulate the proposed model as a mixed integer linear programming problem to select suitable replicas from the pools of replicas and decide the locations of these replicas for both primary and backup functions. The objective of the proposed model is to minimize the sum of the maximum E2E latencies among functions under all possible failure patterns which have k node failures. The numerical results show that the proposed model reduces the E2E latency between the pair of primary and backup VNFs with ensuring the resiliency of the functions compared with baseline models in the examined cases. Two approximate approaches are developed to reduce the computation time of solving the proposed model with a limited performance penalty. We derive theorems to give the bounds of maximum resiliency in the proposed model. Author Affiliation: Graduate School of Informatics, Kyoto University, Yoshida-honmachi, Sakyo-ku, Kyoto 606-8501, Japan * Corresponding author. Article History: Received 23 February 2022; Revised 10 August 2022; Accepted 11 August 2022 (footnote)[white star] Funding for this research was provided by: Japan Society for the Promotion of Science (21H03426, 21J21552). A part of this paper was presented in Kang et al. (2021) at IEEE International Conference on Communications 2021 (ICC 2021), June 2021. Byline: Rui Kang [kang.rui.74z@st.kyoto-u.ac.jp] (*), Fujun He [fujunhe@ieee.org], Eiji Oki [oki@i.kyoto-u.ac.jp]

Keywords Wireless networks; Proactive caching; Multipath routing; Energy saving; Integer linear programming; Simulated annealing Abstract The ever-continuing explosive growth of on-demand content distribution has imposed great pressure on mobile/wireless network infrastructures. To ease congestion in the network and to increase perceived user experience, caching of popular content closer to the end-users can play a significant role and as such this issue has received significant attention over the last few years. Additionally, energy saving is treated as a fundamental requirement in the design of next-generation mobile networks. However, there has been little attention to the overlapping area between energy saving and network caching especially when considering multipath routing. To this end, this paper proposes an energy-efficient caching with multipath routing support. The proposed scheme provides a joint anchoring of popular content into a set of potential caching nodes with optimized multipath support whilst ensuring a balance between transmission and caching energy cost. The proposed model also considers different content delivery modes, such as multicast and unicast. Two separated Integer-Linear Programming (ILP) models are formulated for each delivery mode. To tackle the curse of dimensionality we then provide a greedy simulated annealing algorithm, which not only reduces the time complexity but also provides a competitive performance. A wide set of numerical investigations reveal that the proposed scheme reduces the energy consumption up to 75% compared with other widely used caching approaches under the premise of network resource limitation. Sensitivity analysis to different parameters is also meticulously discussed in this paper. Author Affiliation: (a) School of Information Science and Engineering, Shandong Normal University, Changqing District, 250358, Ji'nan, China (b) Center of Telecommunication Research, Department of Engineering, King's College London, Strand, WC2R 2LS, London, UK * Corresponding author. Article History: Received 8 January 2022; Revised 30 July 2022; Accepted 6 August 2022 Byline: Yantong Wang [yantong@sdnu.edu.cn] (a,*), Vasilis Friderikos [vasilis.friderikos@kcl.ac.uk] (b)

Keywords DDoS attack; Cloud; DEHO; SVM; KPCA; Databases; Data samples; Cyber security Highlights * Proposing SVM based DEHO classifier to detect DDoS attack for the cloud computing environment. * Utilizing KPCA algorithm to obtain optimal features for further detection of normal and attack or malicious data. * Performing experiments for four different databases namely NSL-KDD, UNSW-NB15, ISCX IDand CIC-IDS2017 databases to determine the effectiveness of the proposed approach. Abstract Distributed denial of service (DDoS) attacks is rising exponentially and creates a severe threat to security. Generally, the DDoS attack may appear uncomplicated but they are hard to prevent and considered as one of the most significant cybersecurity issues. Hence, addressing against DDoS attacks turns out to be imperative. The major goal of this paper involves the optimal detection of data samples as normal data samples and malicious/attacked data samples. This paper proposes a security algorithm against DDoS attacks by employing four significant phases namely the Database training phase, Data pre-processing phase, Feature selection phase and Classification phase. Initially, the data samples are to be trained before using it for attack detection. Later, the sample group is created for every file and the data samples are pre-processed in the data pre-processing phase. Secondly, in feature selection phase, the selected features are optimized by employing kernel principal component analysis (KPCA) to obtain optimal features. Later, in the classification phase, a support vector machine-based discrete elephant herding optimization (SVM-DEHO) classifier is utilized to detect the data sample as normal data and attacked or malicious data. Finally, the proposed approach is examined for four different databases namely NSL-KDD, UNSW-NB15, ISCX ID and CIC-IDS2017 databases respectively. The experimental analyses are performed for various simulation metrics and the outcome reveals that the detection system performances are high using SVM-DEHO approach than other approaches. Author Affiliation: (a) Department of Computer Science and Engineering, Presidency University, Bengaluru, India (b) School of Computing Science and Engineering, Galgotias University, Greater Noida, India (c) Department of Computer Science and Engineering, Vardhaman College of Engineering, Hyderabad, India * Corresponding author. Article History: Received 12 May 2021; Revised 23 April 2022; Accepted 28 June 2022 Byline: Gowthul Alam MM [gowthulalam2005@gmail.com] (a,*), Jerald Nirmal Kumar S (b), Uma Mageswari R (c), Michael Raj TF (b)

Keywords Mobile crowdsourcing; Privacy protection; Blockchain; Edge computing; Federated learning; Localized Differential Privacy Abstract With the rapid popularization and development of the Internet of Things (IoT) and 5G networks, mobile crowdsourcing (MCS) has become an indispensable part in today's society. However, when task participants submit tasks, they are likely to expose their data privacy and location privacy. These privacy will be maliciously attacked and exploited by attackers (external attackers and untrusted third party). With the rapid increase of MCS data throughput, traditional cloud platforms can no longer meet the huge data processing needs. To solve these problems, this paper proposes an MCS federated learning system based on Blockchain and edge computing. This paper uses federated learning as the framework of the MCS system. The system protects data privacy and location privacy by using the Double local disturbance Localized Differential Privacy (DLD-LDP) proposed in this paper. Because the sensed data exists in multiple modalities (text, video, audio, etc.), this paper uses the Multi-modal Transformer (MulT) method to merge the multi-modal data before subsequent operations. To solve the problem that the third party is untrusted, we utilize Blockchain to distribute tasks and collect models in a distributed way. A reputation calculation method (Sig-RCU) is proposed to calculate the real-time reputation of task participants. Through conducting experiments on real data sets, the effectiveness and adaptation of the proposed DLD-LDP algorithm and Sig-RCU algorithm are verified. Author Affiliation: (a) School of Computer and Control Engineering, Yantai University, 264005, China (b) The Yantai Key Laborary of High-end Ocean Engineering Equipment and Intelligent Technology, Yantai, 264005, China (c) Department of Software Enportraitgineering & Game Development at Kennesaw State University (KSU), 1100 South Marietta Pkwy, Marietta, GA 30060, USA (d) Department of Computer Science, Georgia State University, Atlanta, GA 30303, USA * Corresponding author at: School of Computer and Control Engineering, Yantai University, 264005, China. Article History: Received 14 February 2022; Revised 3 July 2022; Accepted 17 July 2022 Byline: Weilong Wang (a,b), Yingjie Wang [wangyingjie@ytu.edu.cn] (a,b,*), Yan Huang [yhuang24@kennesaw.edu] (c), Chunxiao Mu (a,b), Zice Sun (a,b), Xiangrong Tong (a,b), Zhipeng Cai [zcai@gsu.edu] (d)

Keywords Fog computing; Task offloading; Delay & latency; Energy & power consumption; Cost; Machine learning Abstract The exponential growth in Internet of Things (IoT) devices and the limitations of cloud computing in terms of latency and quality of service for time-sensitive applications have led to the unfolding of the efficient middleware technology called fog. Fog computing circumvents the limitations of the cloud by creating a seamless continuum between the things/IoT/end-user devices and the cloud and reducing the effective distance. However, fog computing faces challenges for offloading tasks for their remote computation at some level. Hence, the optimality of task offloading is the primary research area in fog computing. Several contemporary papers exist on this important subject. The research gap in reviewing all these task offloading algorithms has motivated us for their presentation in the form of a detailed survey in this paper. There exist some survey papers which deal with the task offloading. However, none of them has covered the basics of optimization techniques and their solution approaches. The primary objective of this paper is to provide the readers with a complete overview of the journey from a task offloading idea to its mathematical problem formulation and finally to its solution with all details of optimization techniques. We begin by introducing fog computing, and task offloading process followed by several task offloading factors governing decision-making process and their surveys. A section is fully dedicated to the survey of offloading objectives with examples. We also present several optimization approaches used in task offloading. Finally, the last section dedicates to the challenges and future direction in fog computing. The outcomes of the survey will benefit readers in learning the optimization used in task offloading, and it will also provide them a systematic design of offloading scheme with specific objectives. Author Affiliation: Department of Computer Science and Engineering, Indian Institute of Technology (ISM) Dhanbad, Jharkhand 826004, India * Corresponding author. Article History: Received 21 February 2022; Revised 2 June 2022; Accepted 28 June 2022 Byline: Nidhi Kumari [nidhi3rdjan1995@gmail.com] (*), Anirudh Yadav [yadavanirudh1997@gmail.com], Prasanta K. Jana [prasantajana@iitism.ac.in]

Keywords Matching theory; Human-in-the-loop; Stochastic network calculus Abstract Upcoming sixth-generation networks have to be carefully designed to offer support to a wide range of human-centric applications, typically computationally intensive and heavily dependent on human perception. This paper proposes a human-in-the-loop-based resource allocation approach found on the joint use of matching theory and stochastic network calculus principles to efficiently address tasks offloading issues within a sixth-generation network environment. The considered matching game approach is characterized by externalities, whose stability is properly discussed here. In particular, by focusing on virtual reality services, the paper aims at maximizing the average service reliability in terms of ability in guaranteeing a proper quality of experience target, i.e., keeping the end-to-end delay below a suitable threshold, by the proper exploitation of the human cognitive limitations. Finally, the effectiveness of the proposed human-in-the-loop-based resource allocation approach is verified by proving comparisons with simulation results and conventional brain-agnostic strategies. Author Affiliation: University of Florence, 50139 Firenze, Italy * Corresponding author. Article History: Received 28 September 2021; Revised 30 June 2022; Accepted 1 July 2022 (footnote)1 Member, IEEE. (footnote)2 Fellow, IEEE. Byline: Benedetta Picano [benedetta.picano@unifi.it] (*,1), Romano Fantacci [romano.fantacci@unifi.it] (2)

Keywords 5G; 6G; SDN; NFV; Network slicing; Cloud/edge computing; Network softwarization; Artificial intelligence; Multimedia services; QoE; Network management Abstract The concept for developing the future mobile communication systems has been explored in recent years with a vision towards 5G and 6G systems. Meeting performance targets such as higher data rate transmission, lower end-to-end latency, higher capacity, lower cost, and satisfied Quality of Experience (QoE) for delivered services is the key for the success of 5G and 6G networks. Within these performance targets and associated enabling technologies, QoE and the overall management of multimedia services is the one which has been less well addressed for 5G in the past. This paper provides a comprehensive and ground-breaking discussion regarding QoE management in the context of future softwarized 5G/6G networks. We introduce a generalized QoE provisioning ecosystem for emerging multimedia streaming services in softwarized 5G/6G and beyond networks. We explore potential network softwarization and communication technologies that will enable the end-user's QoE in 5G/6G. We further provide promising use case scenarios regarding (a) QoE-driven management of multimedia services over multi-access cloud/edge softwarized and virtualized 5G networks, (b) QoE monitoring for large traffic variations in business, residential and public areas in 5G networks, along with a Machine Learning (ML)-based softwarized controller for QoE provisioning over 5G/6G networks. We build on the achievements of 5G networks to provide the roadmap towards the development of 6G and beyond networks in terms of requirements, use case scenarios and service classes. We also present potential technologies that will be dominant through 2025--2030 in shaping the vision of 6G and beyond networks including (a) self-driving 3D network architecture, (b) network management, automation and orchestration, and (c) pervasive artificial intelligence. Moreover, we provide challenges and research directions in 6G and beyond wireless communication systems. The paper provides the first roadmap towards the management, orchestration and monitoring of multimedia 3D services in 6G and beyond networks. The comprehensive goal of this paper is to encourage researchers from the industry and academia to work together towards the realization of softwarized 6G and beyond networks while tackling the critical research challenges regarding the management of emerging 3D multimedia services and applications. The comments made by the authors in this paper are expectations on 6G, and the standardization of 6G where these aspects will be decided in 6G networks and beyond, will be started soon by different standardization bodies and fora. Author Affiliation: School of Computing, Dublin City University, Ireland * Corresponding author. Article History: Received 14 January 2021; Revised 20 May 2022; Accepted 27 June 2022 Byline: Alcardo Alex Barakabitze [alcardo.barakabitze@adaptcentre.ie] (*), Ray Walshe

Keywords DNS; Measurement; Resolvers; TTL; Hosting infrastructures; DNS query failures; New gTLDs Abstract The Domain Name System (DNS) is fundamental to the operation of the Internet. Providing an up-to-date view of DNS behavior in-the-wild is thus important for various Internet stakeholders. Among the behavioral characteristics, failure is one of the most import aspects, because failures within DNS can have a dramatic impact on the wider Internet, most notably preventing access to any services dependent on domain names (e.g. web, mobile apps). In this paper, we perform a large study into DNS activity, covering 3B queries in total. We first examine DNS behavior from different perspectives including DNS query types, use of recursive resolvers, TTLs and hosting infrastructures, aiming at providing an up-to-date view of DNS. We then focus on DNS query failures. We find that 13.5% of DNS queries fail, and this leads us to explore the root causes. We observe significant differences between IPv4 and IPv6 lookups, biased failure distribution across domains, the great impact of recursive resolvers and malicious domains on query failures. We also discuss the implications of our findings and provide a handful of recommendations. Author Affiliation: (a) Institute of Computing Technology, Chinese Academy of Sciences, China (b) University of Chinese Academy of Sciences, China (c) Queen Mary University of London, United Kingdom (d) China Internet Network Information Center, China (e) Computer Network Information Center, Chinese Academy of Sciences, China * Corresponding author. Article History: Received 4 January 2022; Revised 3 April 2022; Accepted 26 June 2022 (footnote)[white star] The preliminary shorter version of this paper appeared at USENIX ATC 2020 as a short paper (6 pages). Byline: Donghui Yang [yangdonghui@ict.ac.cn] (a,b), Zhenyu Li [zyli@ict.ac.cn] (a,b), Haiyang Jiang [jianghaiyang@ict.ac.cn] (a,b), Gareth Tyson [gareth.tyson@qmul.ac.uk] (c), Hongtao Li [lihongtao@cnnic.cn] (d), Gaogang Xie [xie@cnic.cn] (b,e), Yu Zeng [zengyu@cnnic.cn] (d,*)

Keywords Forward security; Group signature; Lattice, Message authentication; Vehicular ad hoc networks (VANETs); Traceability; Applied cryptography; Information security Abstract Message authentication has been a research hotspot in current vehicular ad hoc networks (VANETs). Many researchers adopt group signatures based on number-theoretic assumptions to authenticate the vehicular users' identities. Nevertheless, the classical group signature is vulnerable to quantum computing attacks and without considering the negative consequences of secret key disclosure. In this paper, to address these problems, we propose a novel group signature protocol for authentication in VANETs, which is based on lattice cryptography to achieve quantum-resistance and Bonsai-tree signature architecture to achieve forward security. Our scheme is proven secure in terms of traceability, anonymity, and forward-security under the Short Integer Solution (SIS) and Learning With Errors (LWE) hardness. Through comprehensive performance evaluation, we demonstrate that the storage overhead of our scheme is relatively diminutive and the computation cost of the sign and verify algorithms are efficient and practical compared with other existing schemes. Author Affiliation: (a) School of Information Engineering, North China University of Technology, China (b) School of Cyberspace Security, Beijing University of Posts and Telecommunications, China (c) Department of Computer Science, The University of Hong Kong, China (d) Department of Computing, The Hong Kong Polytechnic University, China (e) Institute of Information Engineering, Chinese Academy Sciences, China * Corresponding authors. Article History: Received 29 March 2022; Revised 6 June 2022; Accepted 30 June 2022 (footnote)1 Yibo Cao and Shiyuan Xu have the same contribution to this paper. Byline: Yibo Cao (a,b,1), Shiyuan Xu [13501199447@163.com] (a,c,1,*), Xue Chen (a,d), Yunhua He (a,e,*), Shuo Jiang (a)

Keywords Networ Intrusion detection; AI adversarial robustness; Adversarial attack; Defense technique; NSL-KDD; UNSW-NB15 Abstract Intrusion detection is a key topic in cybersecurity. It aims to protect computer systems and networks from intruders and malicious attacks. Traditional intrusion detection systems (IDS) follow a signature-based approach, but in the last two decades, various machine learning (ML) techniques have been strongly proposed and proven to be effective. However, ML faces several challenges, one of the most interesting being the emergence of adversarial attacks to fool the classifiers. Addressing this vulnerability is critical to prevent cybercriminals from exploiting ML flaws to bypass IDS and damage data and systems. Some research papers have studied the vulnerability of ML based IDS to adversarial attacks, however most of them focused on deep learning based classifiers. Unlike them, this paper pays more attention to shallow classifiers that are still widely used in ML-based IDS due to their maturity and simplicity of implementation. In more detail, we evaluate the robustness of 7 shallow ML-based NIDS including Adaboost, Bagging, Gradient boosting (GB), Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Support Vector Classifier (SVC) and also a Deep Learning Network, against several adversarial attacks widely used in the state of the art (SOA). In addition, we apply a Gaussian data augmentation defense technique and measure its contribution to improving classifier robustness . We conduct extensive experiments in different scenarios using the NSL-KDD benchmark dataset and the UNSW-NB 15 dataset . The results show that attacks do not have the same impact on all classifiers and that the robustness of a classifier depends on the attack and that a trade-off between performance and robustness must be considered depending on the network intrusion detection scenario. Author Affiliation: (a) SAMOVAR, Télécom SudParis, Institut Polytechnique de Paris, Palaiseau, France (b) IRT-SystemX, 2 Boulevard Thomas Gobert , 91120 Palaiseau, France * Corresponding author. Article History: Received 6 October 2021; Revised 26 April 2022; Accepted 23 May 2022 Byline: Houda Jmila (a), Mohamed Ibn Khedher [ibnkhedhermohamed@hotmail.com] (b,*)

Keywords Internet of Things; Edge Networks; Time-Expanded Graph; Utility Maximization; Task Offloading Abstract This paper comprehensively investigates spatio-temporal dynamics for task offloading in the Internet of Things (IoT) Edge Network (iTEN) in order to maximize utility. Different from the previous works in the literature that only consider partially dynamic factors, this paper takes into account the time-varying wireless link quality, communication power, wireless interference on task offloading, and the spatiotemporal dynamics of energy harvested by terminals and their charging efficiency. Our goal is to maximize utility during the task offloading by considering the above-mentioned factors, which are relatively complex but closer to reality. This paper designs the Time-Expanded Graph (TEG) to transfer network dynamics and wireless interference into some static weight in the graph so as to devise the algorithm easily. This paper firstly devises the Single Terminal (ST) utility maximization algorithm on the basis of TEG when there is only one terminal. In the case of multiple terminals, it is very complicated to directly solve the utility maximization of the task offloading. This paper adopts the framework of Garg and Könemann and devises a multi-terminal algorithm (MT) to maximize the total utility of all terminals. MT is a fast approximate algorithm and its approximate ratio is 1-3[final sigma], where 0 Author Affiliation: (a) School of Computer Science and Technology, Hangzhou Dianzi University, Hangzhou, 310018, China (b) Hangzhou Hikvision Digital Technology Co., Ltd., Hangzhou, 310052, China * Corresponding author. Article History: Received 22 December 2021; Revised 27 June 2022; Accepted 4 July 2022 (footnote)1 Equal contribution and shared co-first authorship. Byline: Jiacheng Wang [jcwang@hdu.edu.cn] (a,1), Jianhui Zhang [jh_zhang@hdu.edu.cn] (a,1), Liming Liu [limingliu@hdu.edu.cn] (a), Xuzhao Zheng [zhengxuzhao66@163.com] (b), Hanxiang Wang [hx_wang@hdu.edu.cn] (a), Zhigang Gao [gaozhigang@hdu.edu.cn] (a,*)

Keywords LEO; Stepping beam; Tracking beam; Satellite; User beams Abstract A stepping (or tracking) beam LEO constellation provides service to predefined areas instead of full coverage. As a result, satellites using stepping beams are more efficient, as power is used only for populated areas. When the constellation shifts, each satellite adjusts its beams locations and sizes, such that each area is serviced by one of the satellites that has the area in its FoV. Current algorithms do not scale well and require considerable satellite compute resources. This paper proposes two novel algorithms to manage the beam allocation. The Dynamic Service Cell Beam Allocation algorithm (DSBA) allocates the beam dynamically in O (m.sup.2) time complexity, where m is the number of service cells in the satellite FoV. A novel Pre-planning Service Cell Beam Allocation algorithm (PSBA) is the first to use pre-planning of the user beams. The algorithm is executed once (before the dynamic allocations are made) requiring O (n.sup.2[middle dot]k.sup.3) time, where n is the number of service cells served by the constellation and k is the number of satellite beams. PSBA has demonstrated similar dynamic beam allocation efficiency as DSBA with performance improvements of up to 100%. Author Affiliation: School of Electrical and Computer Engineering, Ben-Gurion University of the Negev, Beer-Sheva, Israel * Corresponding author. Article History: Received 29 December 2021; Revised 1 July 2022; Accepted 2 July 2022 (footnote)[white star] Preliminary version of this paper appeared in IEEE WiMob 2021. Byline: Oren Markovitz [orenmar@post.bgu.ac.il], Michael Segal [segal@bgu.ac.il] (*)

Keywords Service Function Chaining; Network Function Virtualization; Multi-cloud; Latency; Routing; Linear programming Abstract The advent of virtualization-based infrastructure needed for 5G brought new challenges in utilizing the existing physical network infrastructure effectively. Designing physical routes between virtual network functions in a multi-cloud environment is usually handled as a maximum flow problem, separately considering latency requirements. This paper presents an algorithm designed for low latency edge-cloud environments, where results meet both bandwidth and latency requirements of the planned traffic in one step. The proposed algorithm is much faster compared to the wide-spread linear program based flow maximization approach as the incorporated filtering of route candidates by latency allows more efficient search. The paper demonstrates the operation of the algorithm by a detailed example. Author Affiliation: H-1083 Bókay János Street 36-42., Budapest, Hungary Article History: Received 22 July 2021; Revised 28 January 2022; Accepted 13 April 2022 Byline: Péter Hegyi [peter.hegyi@nokia-bell-labs.com]

Keywords Network traffic classification; Encrypted traffic; Network traffic analysis; Application service; Intra-application service; Granular Abstract Granular traffic classification categorizes traffic into detailed classes like application names and services. Application names represent parent applications, such as Facebook, while application services are the individual actions within the parent application, such as Facebook-comment. These granular classes are still insufficient to keep pace with modern applications that offer various services. Accordingly, this paper further divides the application service class into inter-application and intra-application services to provide more insights. Inter-application service refers to a similar service between different parent applications, such as Facebook-comment and Youtube-comment, whereas intra-application service differentiates services within the same parent application, such as Facebook-comment and Facebook-post. Most studies focus on classification at the application name and inter-application service levels. In contrast, classification at the intra-application service level receives far less attention due to its complexity despite providing the highest flexibility. Therefore, this paper presents GRAIN, a granular multi-label approach to classify encrypted traffic at all three levels of granular classification: application name, inter-application and intra-application service levels using a classifier chain. GRAIN chains two random forest classifiers to produce a multi-label classification using seven novel statistical features based on packet payload length. The utilized features are independent of the packet payload content, thus unaffected by packet encryption and preserving user privacy. Our performance evaluation showed that GRAIN achieved an average F-measure of 99% at the application name level, 93% at the inter-application service level and 88% at the intra-application service level. To test for robustness, we compared GRAIN against four baseline classifiers and the ISCX VPN-nonVPN public dataset in which GRAIN maintained its comparable performance across all tests. Author Affiliation: (a) Department of Computer System and Technology, Faculty of Computer Science and Information Technology, Universiti Malaya, 50603 Kuala Lumpur, Malaysia (b) Faculty of Ocean Engineering Technology and Informatics, Universiti Malaysia Terengganu, Kuala Nerus, Terengganu 21030, Malaysia (c) Faculty of Engineering, School of Computing, Universiti Teknologi Malaysia (UTM), Skudai 81310, Malaysia (d) Faculty of Computing and Informatics, University Malaysia Sabah, International Campus Labuan, Sabah, Malaysia * Corresponding author. Article History: Received 24 July 2021; Revised 30 March 2022; Accepted 30 May 2022 Byline: Faiz Zaki (a), Firdaus Afifi (b), Shukor Abd Razak (c), Abdullah Gani (d), Nor Badrul Anuar [badrul@um.edu.my] (a,*)

Keywords DNS; Security; DoH; Measurements Abstract In this paper we present a large measurement study of the impact on the performance of the adoption of HTTPS as a transport for the DNS protocol (DoH) with public resolvers compared to the existent approach of using non-encrypted transport of DNS queries with the resolver services locally provided by ISPs. Using on web-ads as the mean to execute our tests, we perform over 42 million measurements from more than 4 million vantage points distributed in 32 countries and served by over 2,500 ISPs. We find that, the median resolution time increased 17 ms when using DoH with Cloudflare, 41 ms when using DoH with Quad9, 68 ms when using DoH with Google and 170 ms when using DoH with DNS.SB, compared to using Do53 with the local resolver for a non-cached name. We find similar increases even when using caching. The results presented in the paper contribute to the ongoing discussion of the tradeoffs involved in the combined adoption of public resolvers and DoH. Author Affiliation: (a) Universidad Carlos III de Madrid, Spain (b) Telefónica, Spain * Corresponding author. Article History: Received 10 December 2021; Revised 8 April 2022; Accepted 11 May 2022 Byline: Patricia Callejo [pcallejo@inst.uc3m.es] (a), Marcelo Bagnulo [marcelo@it.uc3m.es] (a,*), Jaime González Ruiz (a), Andra Lutu [andra.lutu@telefonica.com] (b), Alberto García-Martínez [alberto@it.uc3m.es] (a), Rubén Cuevas [rcuevas@it.uc3m.es] (a)

Keywords TCP; P4; Programmable data plane; Congestion control; AQM; SmartNICs; Network diagnosis Abstract The increasing performance requirements of today's Internet applications demand a reliable mechanism to transfer data. Many applications rely on the Transmission Control Protocol (TCP) as the transport protocol, due to its ability to adapt to properties of the network and to be robust in the face of many kinds of failures. However, improving the performance of applications that rely on TCP has been limited by the closed nature of legacy switches, which do not provide accurate visibility of network events. With the emergence of P4-programmable devices, developers can rapidly implement and test customized solutions that use fine-grained telemetry, provide sub round-trip time feedback to end devices to enhance congestion control, precisely isolate traffic to offer better Quality of Service (QoS), quickly detect congestion and re-route traffic via alternate paths, and optimize server resources by offloading protocols. This paper first surveys recent works on P4-programmable devices, focusing on schemes aimed at enhancing TCP performance. It provides a taxonomy classifying the aspects that impact TCP's behavior, such as congestion control, Active Queue Management (AQM) algorithms, TCP offloading, and network measurement schemes. Then, it compares the P4-based solutions, and contrasts those solutions with legacy implementations. Lastly, the paper presents challenges and future trends. Author Affiliation: (a) College of Engineering and Computing, University of South Carolina, Columbia, USA (b) Department of Mathematics and Computer Science, Brandon University, Canada * Corresponding author. Article History: Received 3 September 2021; Revised 3 April 2022; Accepted 6 May 2022 Byline: Jose Gomez [gomezgaj@email.sc.edu] (a), Elie F. Kfoury [ekfoury@email.sc.edu] (a), Jorge Crichigno [jcrichigno@cec.sc.edu] (a,*), Gautam Srivastava [srivastavag@brandonu.ca] (b)

Index Terms Indoor-Outdoor Localization; IoT; Localization Technologies; Localization Techniques; Localization Metrics Abstract The number of connected mobile devices and Internet of Things (IoT) is growing around us, rapidly. Since most of people's daily activities are relying on these connected things or devices. Specifically, this past year (with COVID-19) changed daily life in abroad and this is increased the use of IoT-enabled technologies in the health sector, work, and play. Further, the most common service via using these technologies is the localization/positioning service for different applications including: geo-tagging, billing, contact tracing, health-care system, point-of-interest recommendations, social networking, security, and more. Despite the availability of a large number of localization solutions in the literature, the precision of localization cannot meet the needs of consumers. For that reason, this paper provides an in-depth investigation of the existing technologies and techniques in the localization field, within the IoT era. Furthermore, the benefits and drawbacks of each technique with enabled technologies are illustrated and a comparison between the utilized technologies in the localization is made. The paper as a guideline is also going through all of the metrics that may be used to assess the localization solutions. Finally, the state-of-the-art solutions are examined, with challenges and perspectives regarding indoors/outdoors environments are demonstrated. Author Affiliation: (a) Department of Technical Information Systems Engineering, Erbil Technical Engineering College, Erbil Polytechnic University, Erbil, Kurdistan Region of Iraq (b) Department of Software Engineering, Faculty of Engineering, Koya University, Kurdistan Region of Iraq * Corresponding author at: Department of Sotware Engineering, Koya University, 40 Zanko, Koya, Please Select, Iraq. Article History: Received 29 August 2021; Revised 26 March 2022; Accepted 9 May 2022 (footnote) This paragraph of the first footnote will contain the date on which you submitted your paper for review. It will also contain support information, including sponsor and financial support acknowledgment. For example, 'This work was supported in part by the U.S. Department of Commerce under Grant BS123456.' (footnote)1 Member IEEE. Byline: Safar M. Asaad [safar.maghdid@koyauniversity.org] (a,b), Halgurd S Maghdid [halgurd.maghdid@koyauniversity.org] (b,*,1)

Keywords Edge computing; Network function virtualization; Network security Abstract Edge computing emerges as a new paradigm to provide low-latency network services in close proximity to end users. Based on the network function virtualization (NFV) technology, network services can be flexibly and scalably provisioned as virtual network function (VNF) chains deployed at edge servers. With such advantages, both the industry and research communities have done extensive studies on deploying VNF chains at network edge. The existing works mainly take an ideal assumption that the network is totally safe and there are no malicious users. Therefore, they leverage all available resources to serve their users. However, such an assumption is impractical in real networks. Security services, such as firewall, deep packet detection, intrusion detection, are always required for production networks. The existing service deployment methods fail to consider the co-existence of security services and functionality services. In this paper, we present the topic of joint deployment of both security and functionality services, wherein the security services are responsible to check the data flows before being processed by the functionality services. To solve this problem, we propose the Secure Deployment Pattern, which aims to simultaneously satisfy the security protection and QoS requirements at network edge. It divides the services into two kinds, i.e., the user-oriented functionality services, and the service provider-oriented security services. In this case, it is very challenging to jointly deploy the security services and functional services with respect to the resource and latency constraints. We formulate this problem as an integer programming model, and propose the heuristic algorithms to solve it. As far as we know, this paper is the first step, which targets at a proper orchestration of security and functionality services in edge computing. Extensive evaluations show that the proposed algorithms are effective and efficient, in terms of the execution time and the average number of served requests. Author Affiliation: (a) College of Computer Science and Technology, National University of Defense Technology, Changsha, China (b) Science and Technology Laboratory on Information Systems Engineering, National University of Defense Technology, Changsha, China (c) State Key Laboratory of High Performance Computing, National University of Defense Technology, Changsha, China * Corresponding author. Article History: Received 25 November 2021; Revised 3 March 2022; Accepted 4 April 2022 Byline: Yudong Qin [qinyudong12@nudt.edu.cn] (a), Deke Guo [dekeguo@nudt.edu.cn] (b,c,*), Lailong Luo [luolailong09@nudt.edu.cn] (a,b), Ming Xu [xuming@nudt.edu.cn] (a)

Keywords Intrusion detection systems; Security; Fractals; Machine learning; Communication Abstract Distributed Denial of Service (DDOS) attacks are important threats to network services and applications. Studies in literature have proposed various approaches including Intrusion Detection Systems (IDS) based on the application of machine learning and deep learning, but their computational cost can be significant. For this reason, other studies have proposed efficient IDS algorithms based on the online real-time analysis of the network traffic with a sliding window and entropy or other statistical measures. This paper proposes an online algorithm based on a sliding window with the novel application of the Morphological Fractal Dimension (MFD) to this problem. The results presented in this study show that the application of MFD to the recent CICIDS2017 public data set can obtain a significant improvement in the detection of the DDoS attack in comparison to entropy based approaches. In addition, this paper proposes a novel algorithm for the automatic definition of the sliding window size. This paper reports the impact of the different hyper-parameters, including the parameters present in the definition of MFD and the evaluation of the distance measures, where the Chebyschev distance provides the optimal detection accuracy. The results show a detection accuracy of 99.30%, which performs better than similar approaches on the same data set. Author Affiliation: (a) European Commission, Joint Research Centre, Ispra, Italy (b) Sapienza University of Rome, Department of Computer, Control and Management Engineering Rome A.Ruberti, Italy * Corresponding author. Article History: Received 16 February 2021; Revised 3 January 2022; Accepted 15 March 2022 Byline: Gianmarco Baldini [gianmarco.baldini@ec.europa.eu] (a,*), Irene Amerini (b)

Keywords 5G and 6G; Wireless Channel Estimation (WCE); Machine Learning (ML); Federated Learning (FL); Stochastic Gradient Descent (SGD) Abstract In this paper, we propose a Machine Learning (ML) based approach to address Wireless Channel Estimation (WCE) problem for 5G and beyond wireless networks. Accurate wireless channel estimation plays a crucial role for provision of high quality of service to billions of wireless devices in current and future wireless networks. Our proposed approach uses contemporary ML technique, Federated Learning (FL), in conjunction with the Stochastic Gradient Descent (SGD) algorithm to optimise the WCE problem. Our proposed approach leverages the local user wireless channel information to locally optimise the objective at users and then uses this locally optimised information at the server to optimise the global objective function. The proposed approach is referred to as joint Federated Server Learning and Federated Client Learning (j-FSL-FCL) in the paper. We formulate the WCE problem with a novel loss function to be used for the optimisation problem. To evaluate the performance of our proposed j-FSL-FCL approach (with and without SGD), we consider a Down Link (DL) wireless channel model with Multiple-Input Multiple-Output (MIMO) setting that mimic closely to the wireless channel for 5G and beyond wireless channel models. The performance measuring parameters for j-FSL-FCL are to minimise the difference between actual and estimated wireless channel parameters (channel strength and direction). We compare the results of our proposed approach with the other techniques in the literature based on Least Squares (LS), Linear Regression (LR) and Mean Square Error (MSE). It is shown that the proposed algorithm converges to the optimal solution quickly when used with SGD compared to other existing techniques. It is also shown that the efficiency of the proposed approach for WCE problem is much higher compared to other LS and MSE based techniques. Finally, we present some interesting futuristic applications of FL in the context of 5G and beyond wireless networks. Author Affiliation: School of Computing, Mathematics and Engineering, Charles Sturt University, NSW 2678, Australia * Corresponding author. Article History: Received 1 October 2021; Revised 17 January 2022; Accepted 11 March 2022 Byline: Jasneet Kaur [jkaur@csu.edu.au], M. Arif Khan [mkhan@csu.edu.au] (*)

Keywords Mobile edge networks; Cooperative caching; Multi-agent deep reinforcement learning; Partially observable Markov decision process; Multi-agent deep deterministic policy gradient Abstract Caching the most likely to be requested content at the base stations in a cooperative manner can facilitate direct content delivery without fetching content from the remote content server and thus alleviate the user-perceived latency, reduce the burden on backhaul and minimize the duplicated content transmissions. Content popularity plays a vital role, and it drives caching on edge. In the literature, earlier works considered the content popularity either known earlier or obtained on prediction. However, the content popularity is time-varying and unknown in reality, so the above assumption makes it less practical. Therefore, this paper considers the cooperative cache replacement problem in a realistic scenario where the edge nodes are unaware of the content popularity in mobile edge networks. To address this problem, the main contribution of this paper is to design an intelligent content update mechanism using multi-agent deep reinforcement learning in dynamic environments. With the goal of maximizing the saved delay with deadline and capacity constraints, we formulate the cache replacement problem as Integer linear programming problem. Considering the dynamic nature of the content popularity, high dimensional parameters, and for an intelligent caching decision, we model the problem as a partially observable Markov decision process and present an efficient deep reinforcement learning algorithm by embedding the long short-term memory network (LSTM) into a multi-agent deep deterministic policy gradient formalism. The LSTM inclusion reduces the instability produced by partial observability of the environment. Extensive simulation results demonstrate that the proposed cooperative caching mechanism significantly improves the performance in terms of reward, acceleration ratio and hit ratio compared with existing mechanisms. Author Affiliation: (a) Department of Computer Science and Engineering, National Institute of Technology Warangal, Telangana, 506004, India (b) Indian Institute of Information Technology Design and Manufacturing (IIITDM) Kurnool, Andhra Pradesh, 518002, India * Corresponding author. Article History: Received 10 August 2021; Revised 22 December 2021; Accepted 28 February 2022 Byline: Manoj Kumar Somesula [smanoj@student.nitw.ac.in] (a,*), Rashmi Ranjan Rout [rashrr@nitw.ac.in] (a), D.V.L.N. Somayajulu [soma@nitw.ac.in] (a,b)

Keywords Virtual machine migration; Optimization; Memory migration process; Managing dirty pages; Moving target defense Abstract Enterprises are increasingly outsourcing their computing needs to cloud computing. This dependency on cloud services requires uninterrupted and high availability of services. However, achieving the availability of services is not always possible because cloud services may need to be temporarily suspended for several reasons. These include real-time maintenance of cloud infrastructure, load balancing of services, deployment of countermeasures against cyber attacks, power management, proactive fault tolerance, or resource utilization, etc. The unavailability of critical cloud services may negatively impact the business model of cloud providers because clients would likely be unhappy with frequent service interruptions. One solution to this service interruption is real-time Live Machine Migration (LVM) -- a virtualization technique in network-based systems involving migrating virtual machines (VMs) from one host to another. The key objective of LVM is to avoid or minimize the interruption of application services running on a VM during the migration from a source to a destination host server. To achieve this objective, some challenges need to be tackled. An optimal LVM approach first needs to address the challenge of simultaneously minimizing downtime, network resource usage for migration, and total migration time. Several optimization techniques have been proposed for enhancing the performance of LVM. This paper provides a comprehensive review of the LVM challenges particularly focusing on memory content migration and the proposed solutions to overcome the identified challenges and their limitations. The paper also points out some open research issues and future directions towards optimizing LVM. Author Affiliation: Department of Computer Science and Engineering, Qatar University, Qatar * Corresponding author. Article History: Received 29 September 2021; Revised 5 February 2022; Accepted 9 March 2022 Byline: Raseena M. Haris [raseena@qu.edu.qa] (*), Khaled M. Khan, Armstrong Nhlabatsi

Keywords Traffic engineering; Rural broadband; MOEA; SPEA-II Highlights * This paper proposes the use of a multi-objective evolutionary algorithm for traffic engineering purposes. It provides the following features. * The ability to determine a (near) optimal downstream communication path in a dynamic environment, for example, taking into account weather conditions. * The optimality of the path considers conflicting requirements of End-to-End speed, residual battery life and energy drain. The choices are presented as a Pareto front, so various solutions are presented, allowing the operator to make flexible, personalised decisions. * The computationally efficient scheme can be readily extended to consider other parameters of interest. Abstract Providing adequate and cost-effective access to the Internet for rural communities remains a challenge. In this paper, we explore the use of a traffic engineering framework based on a Multi-Objective Evolutionary Algorithm (MOEA) for both path and transmission power selection across a network of Free-Space Optical relay nodes. The configuration takes into account the conflicting requirements of End-to-End speed and the battery state of the relay nodes, as well as the current weather conditions. As such, the system is able to maintain near-optimal transport of data in a changing environment. Furthermore, the use of an MOEA yields a Pareto Front enabling the operator to select from a set of (near) optimal configurations based on the particular balance of requirements that are desired. Author Affiliation: EECS, Queen Mary University of London, London, UK * Corresponding author. Article History: Received 31 May 2021; Revised 27 February 2022; Accepted 2 March 2022 Byline: Nazrul Islam [nazrul.islam@qmul.ac.uk] (*), Chris Phillips [chris.i.phillips@qmul.ac.uk]

Keywords LAA; CQI; RSRQ; Hidden node problem; Machine learning; Unlicensed spectrum Abstract This paper analyzes and evaluates the use of Machine Learning (ML) techniques as an alternative to heuristic methods to identify UEs in Licensed-Assisted Access (LAA) networks that are being affected by collisions from hidden nodes sources. The paper analyzes different models based on supervised ML methods to classify UEs into two possible classes: UEs affected by collisions and UEs free of collisions. To implement them, we employ a set of typical parameters of the LTE standard, e.g., Channel Quality Indicator (CQI), Reference Signal Received Quality (RSRQ) and number of Physical Resource Blocks (PRB). These parameters are fed into different supervised ML models which, once optimized, obtain the best models to predict when an UE is facing interference from hidden nodes in a dynamic channel environment. Results show that classification methods based on ML techniques are capable of predicting with a high level of accuracy when a UE is affected by collisions and, consequently, if this node is located in a hidden area. Nevertheless, and despite the growing interest in applying ML techniques in an increasing number of fields, we can conclude that, in this case, the performance of the ML approach is not necessarily better than heuristic methods. For instance, it does not outperform the Dynamic Collision Detection (DCD) previously presented by the authors. In addition, the ML algorithms exhibit a higher prediction time and computational complexity, with respect to those obtained with DCD algorithm. This should be considered a drawback in some LTE-Advanced and 5G wireless network scenarios that prioritize lower latencies. Author Affiliation: Aragon Institute for Engineering Research (I3A), University of Zaragoza, 50018 Zaragoza (Spain) * Corresponding author. Article History: Received 16 April 2021; Revised 5 August 2021; Accepted 18 February 2022 Byline: Pablo Campos [pcampos@unizar.es] (*), Ángela Hernández-Solana [anhersol@unizar.es], Antonio Valdovinos-Bardají [toni@unizar.es]

Keywords P4 language; Programmable data plane; P4 security applications and implications; STRIDE model; Challenges and solutions in P4 Abstract The emergence of the IoT, cloud systems, data centers, and 5G networks is increasing the demand for a rapid development of new applications and protocols at all levels of the protocol stack. However, traditional fixed-function data planes have been characterized by a lengthy and costly development process at the hand of few chip manufacturers. Recently, data plane programmability has attracted significant attention, permitting network owners to run customized packet processing functions using P4, the de facto data plane programming language. Network security is one of the key research areas exploiting the capabilities of programmable switches. Examples include new encapsulations and secure tunnels implemented in short times, mitigation techniques for DDoS attacks that occur at terabit rates, customized firewalls that track hundreds of thousands of connections per second, and traffic anonymization systems that operate at line rate. Moreover, applications can be reconfigured in the field without additional hardware upgrades, facilitating the deployment of new defenses against unforeseen attacks and vulnerabilities. Furthermore, these security applications are designed by network owners who can meet their specific requirements, rather than by chip manufacturers. Despite the impressive advantages of programmable data plane switches, the literature has been missing a comprehensive survey on security applications. To this end, this paper provides a concise background on programmable switches and their main features that are relevant to security. It then presents a taxonomy that surveys, classifies, and analyzes articles related to security applications developed with P4. Additionally, the paper employs a STRIDE analysis to examine vulnerabilities related to general P4 applications (e.g., congestion control, load balancing, in-network cache) and proposes plausible remediation approaches. Furthermore, challenges associated with programmable data planes, the impact of these challenges on security implementations, and schemes to eliminate or mitigate them are discussed. Finally, the paper discusses future endeavors and open research problems. Author Affiliation: (a) College of Engineering and Computing, University of South Carolina, Columbia, USA (b) The Cyber Center For Security and Analytics, University of Texas at San Antonio, USA * Corresponding author. Article History: Received 5 November 2021; Accepted 17 January 2022 Byline: Ali AlSabeh [aalsabeh@email.sc.edu] (a,*), Joseph Khoury [joseph.khoury@utsa.edu] (b), Elie Kfoury [ekfoury@email.sc.edu] (a), Jorge Crichigno [jcrichigno@cec.sc.edu] (a), Elias Bou-Harb [elias.bouharb@utsa.edu] (b)

Keywords Machine Learning; Distributed machine learning; Distributed inference; Latency; 5G networks Abstract 5G is the fifth generation of cellular networks. It enables billions of connected devices to gather and share information in real time; a key facilitator in Industrial Internet of Things (IoT) applications. It has more capabilities in terms of bandwidth, latency/delay, processing powers and flexibility to utilize either edge or cloud resources. Furthermore, 6G is expected to be equipped with the new capability to converge ubiquitous communication, computation, sensing and controlling for a variety of sectors, which heightens the complexity in a more heterogeneous environment This increased complexity, combined with energy efficiency and Service Level Agreement (SLA) requirements makes application of Machine Learning (ML) and distributed ML necessary. A decentralized approach stemming from distributed learning is a very attractive option compared with a centralized architecture for model learning and inference. Distributed ML exploits recent Artificial Intelligence (AI) technology advancements to allow collaborated ML, whilst safeguarding private data, minimizing both communication and computation overhead along with addressing ultra-low latency requirements. In this paper, we review a number of distributed ML architectures and designs, that focus on optimizing communication, computation and resource distribution. Privacy, information security and compute frameworks, are also analyzed and compared with respect to different distributed ML approaches. We summarize the major contributions and trends in this area and highlight the potential of distributed ML to help researchers and practitioners make informed decisions on selecting the right ML approach for 5G and Beyond related AI applications. To enable distributed ML for 5G and Beyond, communication, security, and computing platform often counter balance each other, thus, consideration and optimization of these aspects at an overall system level is crucial to realize the full potential of AI for 5G and Beyond. These different aspects do not only pertain to 5G, but will also enable careful design of distributed machine learning architectures to circumvent the same hurdles that will inevitably burden 5G and Beyond network generations. This is the first survey paper that brings together all these aspects for distributed ML. Author Affiliation: (a) Centre of Telecommunication Research, Kings College London, United Kingdom (b) Ericsson * Corresponding author. Article History: Received 7 July 2021; Revised 25 January 2022; Accepted 27 January 2022 Byline: Omar Nassef [omar.nassef@kcl.ac.uk] (a,*), Wenting Sun [wenting.sun@ericsson.com] (b), Hakimeh Purmehdi [hakimeh.purmehdi@ericsson.com] (b), Mallik Tatipamula [mallik.tatipamula@ericsson.com] (b), Toktam Mahmoodi [toktam.mahmoodi@kcl.ac.uk] (a)

Keywords Threat situation awareness; IoT edge; Network traffic; FPGA Abstract Threat situation awareness is one of the new major technologies to avoid network attacks and ensure equipment security. Facing the current IoT network architecture which is characterized by end equipments' complex services, huge traffic and computing marginalization, real time threat situation awareness based on network traffic can effectively warn and clean latent threat. However, the existing threat situation awareness methods are mostly unitary and dependent on the central node for collection, detection and cleaning. First, it takes too much bandwidth and is not suitable for high-speed scenes. Second, the transmission of traffic or log leads to poor privacy and risk of leakage. Most of all, the perception time is too long, which leads to the performance degradation. This paper proposes a threat situation awareness architecture based on IoT edge and network traffic. Firstly, this paper designs an edge computing device SIE based on CPU and FPGA, the FPGA pipeline is used to analyze the traffic and summarize it in real time. A fast threat situation detection method deployed on SIE's CPU is proposed which uses flow entropy algorithm to generate situation information. Secondly, this paper introduces the threat situation understanding method based on machine learning. It improves the AdaBoost algorithm and uses uploaded situation information to judge the threat in the traffic. Finally, the method obtains the defensive measure according to the threat intelligence. It can issue the SIE for situation projection and completes threat situation awareness closed loop. Experimental results on KDD99, UNSW-NB15 show that under the premise of ensuring the normal business of IoT equipment and the second level early warning ability, the proposed method can still show good performance under the recognition recall rate, success rate of cleaning threat and other indicators. Author Affiliation: (a) School of Cyber Science and Engineering, Southeast University, Nanjing, China (b) Key Laboratory of Computer Network and Information Integration, Ministry of Education, Nanjing, China (c) Jiangsu Provincial Key Laboratory of Computer Network Technology, Southeast University, Nanjing, China (d) Hunan Huaxintong Network, Changsha, China * Correspondence to: 2 Southeast University Road, Jiangning District, Nanjing, Jiangsu Province, China. Article History: Received 23 February 2021; Revised 18 August 2021; Accepted 4 October 2021 Byline: Yuyu Zhao [yyzhao@njnet.edu.cn] (a,b,c), Guang Cheng [gcheng@njnet.edu.cn] (a,b,c,*), Yu Duan [yduan@njnet.edu.cn] (a,b,c), Zhouchao Gu [zcgu@njnet.edu.cn] (a,b,c), Yuyang Zhou [yyzhou@njnet.edu.cn] (a,b,c), Lu Tang [lutang@c2comm.cn] (d)

Keywords Friendship selection; Relationship management; Network navigability; Social internet of things; Systematic literature review; Smart objects Abstract The Social Internet of Things (SIoT) paradigm integrates the Internet of Things (IoT) and human social networks. SIoT network plays an essential role in establishing social friendships between smart devices. Any physical object deployed on the SIoT environment will access the required service utilizing its social friendships. Discovering desirable and efficient services among SIoT smart devices is possible only if we create, select, and manage social friendships intelligently and reliably. However, despite the significant importance of the thrust area of social friendship selection and management in the SIoT network, there is no systematic and comprehensive research study about exclusively analyzing and reviewing all the key strategies of this research field. Therefore, this survey paper investigates the state-of-the-art SIoT friendship selection and management strategies based on the Systematic Literature Review (SLR) methodology. It achieves this target by looking at the journal and conference papers published on social friendship selection and Relationship Management (RM) scope between 2013 and December 2020. We propose a novel technical taxonomy to categorize the state-of-the-art studies into the five categories in which social relationship selection and management play a significant role: structure-based, community-based, ontology-based, recommendation-based, and other strategies. In each category, the relevant research studies are presented and described. Also, a summary of the benefits and drawbacks of the reviewed studies is provided, and a comprehensive comparison of studies based on some evaluation metrics is made. Finally, we outline new emerging challenges and discuss and identify future research directions and open issues. Author Affiliation: (a) Department of Computer Engineering, Science and Research Branch, Islamic Azad University, Tehran, Iran (b) Future Technology Research Center, National Yunlin University of Science and Technology, 123 University Road, Section 3, Douliou, Yunlin 64002, Taiwan (c) Department of Computer Engineering, Central Tehran Branch, Islamic Azad University, Tehran, Iran (d) Mental Health Research Center, Psychosocial Health Research Institute, Iran University of Medical Sciences, Tehran, Iran (e) Computer Science, University of Human Development, Sulaymaniyah, Iraq * Corresponding author. Article History: Received 18 June 2021; Revised 29 September 2021; Accepted 19 October 2021 Byline: Babak Farhadi (a), Amir Masoud Rahmani [rahmania@yuntech.edu.tw] (b,*), Parvaneh Asghari (c), Mehdi Hosseinzadeh (d,e)

Keywords Malware classification; Malware attacks; Machine classifier; Convolutional neural network; Data processing; Honeypot Abstract Modern malware threats demand a robust and scalable detection system. This paper presents a novel proactive monitoring and analysis architecture called malware threat intelligence system (MTIS) to collect and classify real-world samples of modern malware families. Microsoft Window's portable executable (PE) files are systematically labeled using clustering and AVClass engine. These labeled malware samples are visualized into grayscale images, which are further utilized for the extraction of textural features. This paper uses local descriptors (LBP, DSIFT, GLCM) and global descriptors(GIST) to extract local and global textural features from a grayscale image. Malware images are distinguished using a hybrid approach of machine learning methods and a proposed convolutional neural network (CNN) employed with early stopping configurations. Results have demonstrated that the proposed architecture detects modern and real-world malware samples with better classification accuracy without code reversing and domain expertise. Four machine learning algorithms, namely, K-nearest neighbors (k-NN), Support vector machine(SVM), Naive Bayes(NB), and Random forest, are compared systematically, with different image resolutions and data split, enabling efficient model selection. Deep learning of MTIS (DL-MTIS) is compared with former methods and with CNN+GIST, CNN with the whole image as inputs, and k-NN+GIST method, and it is observed to be superior in performance. The results also reveal that the proposed method is resilient to packed and encrypted malware samples. Author Affiliation: Department of Computer Application, National Institute of Technology, Tiruchirappalli, India * Corresponding author. Article History: Received 11 July 2021; Revised 8 October 2021; Accepted 31 October 2021 Byline: Sanjeev Kumar [405918054@nitt.edu] (*), B. Janet

Keywords Blockchain; Internet of Things; Efficiency; Security; Integrity Abstract The IoT ecosystem envisions a new world fully integrated by smart devices exchanging data to enable more automated, accurate, and timely decisions. Nonetheless, this idealized scenario may be only implemented if the system database allow smart devices' transactions to be efficiently processed and security is guaranteed. To tackle this challenge, this paper then analyzes three Blockchain integrated architectures, covering an assemblage of different popular IoT-application scenarios. The performance evaluation is based on analytical queueing models, simulations, and theoretical discussion. The results show that efficiency and security requirements may be met when the Blockchain system is suitably configured, mainly observing a trade-off between data integrity and response time. In this sense, our main contribution is to yield valuable subsidies which can act as guidance for the development of real IoT-ecosystem projects. At last, general conclusions and directions for further research close this paper. Author Affiliation: Center for Mathematics, Computing and Cognition -- CMCC, Federal University of ABC -- UFABC, Avenida dos Estados, 5001, CEP 09210-580, Santo André, SP, Brazil Article History: Received 25 June 2021; Revised 29 September 2021; Accepted 5 November 2021 Byline: Carlo Kleber da Silva Rodrigues [carlo.kleber@ufabc.edu.br]

Keywords Congestion control; Deep Learning (DL); Network coding; Transport Layer; UAV Networks Abstract This paper focuses on the congestion control issues in Unmanned Aerial Vehicle (UAV) Swarm Networks (USNs). In a USN, many network factors can cause segment loss, including dynamic swarming, high mobility, and link fading loss. With traditional transport layer protocols such as Transmission Control Protocol (TCP), these losses are interpreted as congestion events and will cause the data sending rate being decreased dramatically, therefore impacting throughput. In this paper, a learning-based adaptive network coding scheme is proposed to handle segment loss. In this scheme, a certain amount of redundancy is attached to the original data. If the segment loss is caused by random factors (such as radio interference), the lost segments are retrieved by decoding. However, if the loss is caused by congestion, the sender will retransmit the lost segments and decrease the sending rate. The coding rate is a critical factor, which should guarantee that the random loss can be retrieved by decoding while the congestion loss triggers retransmission and sending rate deduction. To achieve this goal, a Deep Learning (DL) algorithm is proposed, which comprehensively considers the wireless network conditions and dynamically optimizes the coding rate. Our experimental results show that the DL-based network coding scheme provides improved throughput and end-to-end delay compared to the TCP and general network coding schemes. Author Affiliation: (a) Mathematics & Computer Science Department, Whitworth University, Spokane, USA (b) Electrical and Computer Engineering, University of Alabama, Tuscaloosa, AL, USA (c) Air Force Research Laboratory, Rome, NY, USA (d) Electrical and Computer Engineering, San Diego State University, San Diego, CA, USA * Corresponding author. Article History: Received 10 February 2021; Revised 25 August 2021; Accepted 25 September 2021 (footnote)[white star] This work was supported by the U.S. Air Force Research Laboratory, under agreement No. FA8750-18-1-0023. DISTRIBUTION STATEMENT A: Approved for Public Release; distribution unlimited 88ABW-2020-3119 on 08 Oct 2020. Byline: Qian Mao (a), Lin Zhang (b), Fei Hu [fei@eng.ua.edu] (b,*), Elizabeth Serena Bentley (c), Sunil Kumar (d)

Keywords Elastic optical network; Sliceable bandwidth variable transponder; Super-channel Abstract In recent days, Elastic Optical Networks (EONs) have arisen as a viable solution to improve spectrum utilization by elastically assigning spectral resources customized for diverse bandwidth requirements. The key facilitators for EONs are flexible grid technology and sliceable bandwidth variable transponders (SBVTs). SBVT in EONs has recently been proposed as a next-generation bandwidth variable transponder (BVT), which supports high data-rate super-channels (i.e., coherent optical subcarriers are aggregated to build a channel of desired capacity which is provisioned in a single operational cycle) and sliceability (i.e., optical flows are flexibly related to the arriving traffic demands, and apart from forming a super-channel, they could be transmitted towards distinct destinations simultaneously). In this paper, we present an explication on SBVT technology enabling distance adaptive hybrid super-channels. This paper begins by proposing a novel SBVT architecture, including a distance adaptive sub-carrier estimator module (DASEM) and transmission and modulation module (T&MM). DASEM is a programmable controller which operates on the concept of distance adaptive modulation technique and accomplishes the transmission of the data-streams/main-streams via distance adaptive hybrid super-channels by configuring the T&MM dynamically. Further, we implement the proposed distance adaptive hybrid super-channels using the two most promising transmission techniques: Coherent optical orthogonal frequency division multiplexing (CO-OFDM) and Nyquist wavelength-division multiplexing (NWDM). It has been exhibited that distance adaptive hybrid super-channels outperform conventional super-channels by achieving an increment in spectral efficiency of approximately 55%. In addition, the system transmission performance (NWDM super-channel) with an acceptable bit error rate (BER) of 10.sup.3 is analyzed over 240 km of standard single-mode fiber (SSMF). The proposed SBVT architecture with a programmable controller enables cost reduction and provides a significant saving in spectrum consumption. Author Affiliation: Indian Institute of Technology (Indian School of Mines), Dhanbad 826004, India * Corresponding author. Article History: Received 13 August 2020; Revised 30 July 2021; Accepted 5 October 2021 Byline: Ujjwal, Neha Kumari, Jaisingh Thangaraj [jaisingh@iitism.ac.in] (*), Chhandita Roy

Keywords Unmanned aerial vehicles; Task offloading; Edge computing; Q-learning Abstract Unmanned aerial vehicle (UAV) have been deployed in many applications, such as Power Grid inspection, forest fire prevention, and pollution surveillance. They often cruise along a fixed route above the target area. Due to the cost of remote communication and local computationally intensive tasks, resource-constrained drones tend to offload tasks to edge servers. In most cases, drones do not know the prior knowledge of user nodes and edge servers, and must reduce the altitude to provide services. Therefore, it is necessary to carefully decide when and where to collect and offload tasks to avoid unnecessary energy consumption and time delays. In this paper, we propose the benefit maximization problem under constraints such as time sensitivity, and propose an optimized task offloading strategy based on the reinforcement learning algorithm. We strive to directly solve the deficiencies in the profit maximization problem with modified Q-Learning algorithm. We test the performance under practical application scenarios with different environmental parameters. The experimental results prove that the solution proposed in this paper has better convergence and performance, as well as better reusability in similar application scenarios. Author Affiliation: (a) School of Computer Science and Technology, Hangzhou Dianzi University, Hangzhou 310018, China (b) Key Laboratory of Electromagnetic Wave Information Technology and Metrology of Zhejiang Province, College of Information Engineering, China Jiliang University, Hangzhou 310018, China (c) College of Computer and Information Sciences and Research Chair of Smart Technologies, King Saud University, Riyadh 11543, Saudi Arabia * Corresponding author. Article History: Received 30 March 2021; Revised 5 July 2021; Accepted 2 August 2021 Byline: Peng Liu (a,b), Han He [15306507997@163.com] (a,*), Tingting Fu (a), Huijuan Lu (b), Abdulhameed Alelaiwi (c), Md Wasif Islam Wasi (a)