scispace - formally typeset
Search or ask a question
Journal Article•DOI•

Survey and taxonomy of clustering algorithms in 5G

Muhammad Fahad Khan1, Muhammad Fahad Khan2, Kok-Lim Alvin Yau2, Rafidah Md Noor3, Muhammad Imran4 •
COMSATS Institute of Information Technology1, Sunway University2, University of Malaya3, University of Glasgow4
15 Mar 2020-Journal of Network and Computer Applications (Academic Press)-Vol. 154, pp 102539
TL;DR: A comprehensive survey of the research work of clustering schemes proposed for various scenarios in 5G networks is presented and various aspects of clustered schemes, including objectives, challenges, metrics, characteristics, performance measures are highlighted.
About: This article is published in Journal of Network and Computer Applications.The article was published on 2020-03-15 and is currently open access. It has received 21 citations till now. The article focuses on the topics: Cluster analysis & Network topology.

Summary (8 min read)

Jump to: [1. Introduction] – [1.1. Our contributions] – [1.2. Organization of this paper] – [2. Background and Motivating the Need for Clustering in 5G Networks] – [2.1. What will the next generation wireless mobile networks be like?] – [2.2. What is 5G?] – [2.3. How can 5G benefit from clustering?] – [2.4. What are the costs of clustering?] – [3. Taxonomy of Clustering Attributes in 5G Networks] – [3.1. Clustering objectives] – [3.2. Clustering challenges] – [3.3. Clustering metrics] – [3.4. Clustering characteristics] – [3.5. Performance measures] – [4. Clustering Framework] – [5. Clustering Algorithms in 5G Networks] – [5.1.1. Reducing CH election time based on energy consumption] – [5.1.4. Electing backup CHs] – [5.1.5. Merging clusters based the number of handovers of UEs between two SCs] – [5.2.1. Adjusting cluster size to transfer traffic load among SCs] – [5.3.1. Joining node based on social interests] – [5.4. Clustering Algorithms for Enhancing Fairness] – [6.1. Performing cluster size adjustment] – [6.2. Performing cluster maintenance] – [6.3. Ensuring node connectivity] – [6.4. Adopting self-organization] – [6.5. Enhancing quality of service] and [7. Conclusion]

1. Introduction

  • The tremendous growth of user equipment (UE) expecting to reach up to billions in number [1, 2] , along with bandwidthstarving applications (e.g., video streaming, multimedia sharing, and online gaming), has contributed to 74% increment in data traffic over the years [3] .
  • Hence, there is a colossal demand for significantly higher network capacity and lower delay to support higher mobility UEs, leading to the introduction of the nextgeneration mobile wireless network, namely fifth generation (5G).
  • Hence, D2D can support the deployment of SCs through spatial reuse of frequency bands to cater for ultra-densification.
  • Of particular interest is clustering in the access networks, which is different from cell clustering performed in the core network [52] .
  • Cluster member, which is associated with the cluster head, performs intra-cluster communication.

1.1. Our contributions

  • A taxonomy for clustering attributes, covering clustering objectives, challenges, metrics, characteristics, and performance measures, as well as a clustering framework, are presented.
  • The clustering algorithms are classified, analyzed, and discussed based on the taxonomy and clustering framework.
  • Open issues of this research topic are also outlined.
  • While a survey paper that focuses on clustering using coordinated multi-point (CoMP) [30] , which is one of the main technologies in 5G, has been presented, this paper focuses on clustering using various technologies in 5G.

1.2. Organization of this paper

  • Section 2 presents background and the motivation for clustering in 5G.
  • Section 5 presents various clustering algorithms in 5G based on the taxonomy and clustering framework.

2. Background and Motivating the Need for Clustering in 5G Networks

  • With a massive amount of highly variable data generated by the tremendous growth of a diverse range of UEs, clustering has been proposed to organize network topology and summarize data in order to improve network performance (e.g., higher network scalability, spectral efficiency, data availability, and load balancing) [23, 31] based on new and traditional clustering metrics.
  • Clustering schemes can use multiple clustering metrics, such as weighted clustering that uses four kinds of clustering metrics, namely the node degree, transmission power, mobility, and residual energy of a node [53] .
  • Section 2.1 presents three main characteristics of the next generation wireless mobile networks.
  • Section 2.3 presents how clustering can cater for the three main characteristics in the 5G context.
  • Finally, Section 2.4 presents the cost of clustering.

2.1. What will the next generation wireless mobile networks be like?

  • The next generation wireless mobile networks are foreseen to possess three main characteristics.
  • Ultra-densification raises the challenge to provide up to 10× higher network capacity than that in 4G networks, while providing greater mobility support and efficiency in resource allocation.
  • Secondly, network heterogeneity whereby there is a diverse range of UEs (e.g, laptops, smart gadgets, unmanned aerial vehicles (UAVs), and sensors), network cells (e.g., macrocells, picocells, and femtocells), and networks (e.g., Internet of everything (IoE), Internet of humans (IoH), and IoT) [21, 61] .
  • As an example, the deployment of macrocells caters for long-range transmission, while the deployment of SCs caters for short-range transmission that enables efficient resource utilization via spatial reuse of frequency bands [63] .
  • High variability raises the challenge to achieve scalability, agility, and reconfigurability [67, 68] in order to cater for the scarcity and overabundance of network capacity caused by high variability in traffic and spectrum availability [69, 70] .

2.2. What is 5G?

  • Generally speaking, mean data rate and peak data rate are expected to increase up to 10× and 20×, respectively; traffic capacity and energy efficiency are expected to increase up to 100×; and delay is expected to reduce to 1 ms, and so on, with the comparison of 4G.
  • Data plane contains network infrastructure, such as routers and switches, that performs tasks following instructions, logics, or rules, given by the control plane [71] .
  • So, there is a trade-off between the level of inter-cell interference and the number of network cells participating in a collaboration (or the amount of overhead).
  • This enables BSs with directional antennas to coordinate resource usage among BSs and UEs, particularly at overlapping areas, in order to improve CoMP gain (e.g., which is based on spectral ef-ficiency, energy efficiency, load balancing, and fairness of resource distribution) [81] [82, 83] and reduce overhead [30] , otherwise a large-scale collaboration can congest the network.
  • Clustering has been proposed to handle handover and user association with BS in order to optimize load balancing [89] .

2.3. How can 5G benefit from clustering?

  • Traditionally, clustering has been proposed to achieve network stability and scalability in order to improve network performance, such as throughput, the fairness of resource distribution, a better load balancing, as well as the lifetime of a CH, a cluster, or a network, while providing support for routing [29] .
  • Since network disconnections can be caused by high mobility UEs [109] , clustering segregates nodes in a network with similar behavior (i.e., similar speed) into logical groups, and each cluster may connect to different BSs in order to achieve network stability [110, 56] .
  • Secondly, in order to cater for network heterogeneity, clustering segregates network entities into logical groups based on common characteristics (e.g., relative speed, degree centrality, and social relation) or metrics (e.g., fairness index), so that heterogeneous network entities can prolong their respective connections.
  • As another example, SCs are overlaid in macrocells to increase network capacity, and clustering segregates heterogeneous network cells into logical groups in order to offload computation tasks from macrocells.
  • As an example, in [118] , clustering segregates nodes with similar behavior (i.e., traffic characteristics and network resource requirements) into logical groups to enable spatial reuse of frequency bands in order to improve spectral efficiency and reduce network congestion [21, 118] .

2.4. What are the costs of clustering?

  • Cluster maintenance (or re-clustering [119] ) requires message exchange [40] , and so it incurs clustering cost.
  • During re-clustering, a new CH is re-elected, and each nonclustered node associates itself with the new or an existing CH to join the cluster.
  • Not only does frequent re-clustering reduce network scalability and stability, as well as the capability of clustering to cater for the three main characteristics (i.e., ultra-densification, network heterogeneity, and high variability) (see Section 2.3), but also increases clustering cost and reduces network performance.
  • There are three main types of clustering costs.
  • Firstly, bandwidth wastage due to explicit clustering message exchange among nodes (or clustering overhead) in order to exchange information for clustering purpose.

3. Taxonomy of Clustering Attributes in 5G Networks

  • The rest of this subsection explains the taxonomy.
  • Cluster stability is affected by: a) ultra-densification whereby higher node density increases the probability of a connected network remaining connected [122] , and b) heterogeneity whereby a diverse range of network entities (i.e., UEs, network cells, and networks), which possess a diverse range of capabilities, overlap and increase interference [123] .
  • Both (c) and (d) are incurred during re-clustering.
  • As an example, D2D reduces access to network core leading to reduced network congestion and overhead, and so it increases cluster lifetime, and hence cluster stability. [125] [126] .

3.1. Clustering objectives

  • In 5G networks, load balancing improves with efficient distribution of network traffic and lower traffic variability.
  • Load balancing is affected by: a) ultra-densification whereby higher node density increases the amount of data generated, b) heterogeneity whereby a diverse range of network cells increases the number of handovers, handover overhead, and computational cost, and c) high variability whereby higher variability causes a sudden demand for significantly higher network capacity.
  • Social awareness is affected by: a) ultradensification whereby higher node density provides a higher value of social relation and more options in selecting influential nodes, and b) heterogeneity whereby a diverse range of interests (e.g., using the same set of services) and behaviors pose a great challenge to segregate alike network entities into logical groups.
  • In 5G networks, QoS increases with cluster lifetime, fair distribution of network traffic and resources, as well as reduces with computational cost and the occurrence of re-clustering.
  • D2D node pairs must compete with each other to adjust its transmission power in order to minimize its own price for a given reuse price vector broadcast by the BS, while achieving a better QoS [139] .

3.2. Clustering challenges

  • There are three main clustering challenges that must be addressed during cluster formation and maintenance (or reclustering) in 5G networks: X.1 High heterogeneity, which is an intrinsic characteristic of 5G, poses difficulties in segregating nodes with different nature of heterogeneity, including interests or behaviors, in a network into logical groups [140, 141] .
  • High heterogeneity can be addressed by deploying SCs and promoting collaboration among them to reduce interference among SCs in order to maximize throughput [116] .
  • High cost, which is incurred when re-clusterings become more frequent, poses difficulties in improving the efficiency of clustering and resource utilization.
  • More descriptions about cost are presented in Section 2.4.
  • High dynamicity, which is caused by nodal mobility, poses difficulties in maintaining a cluster structure, as well as maintaining operation during network failure (or being fault tolerance).

3.3. Clustering metrics

  • Firstly, it achieves the objective of enhancing social awareness (O.3) because alike nodes form clusters; for instance, nodes with interests towards a specific multimedia content form a cluster and share the content.
  • Using this clustering metric helps to achieve two main objectives in 5G.
  • Firstly, it achieves the objective of enhancing social awareness (O.3) because mobile nodes with similar mobility pat-terns (e.g., similar AoA) form a cluster to prolong the lifetime of a CH and a cluster.
  • M.5 Residual energy represents the remaining energy level of a node.

3.4. Clustering characteristics

  • There are three main clustering characteristics in 5G networks: C.1 Number of hops (or intra-cluster distance) represents the maximum number of hops between a CH and its CMs.
  • There are two possible options as follows: C.1.
  • These clusters are more difficult to maintain as compared to those with low mobility.

3.5. Performance measures

  • There are six main performance measures for clustering in 5G networks.
  • Table 2 shows the details of performance measures and performance metrics.
  • Cluster lifetime has been investigated with respect to energy efficiency and cluster stability as shown in the.

4. Clustering Framework

  • During the first stage, non-clustered nodes exchange messages among themselves, or receive broadcasts from BS, to gather clustering information about: a) neighboring nodes, and b) network topology.
  • The proposed scheme has shown to increase cluster lifetime.
  • In [124] , a CH is elected using a residual energy metric (M.5) so that it has a higher residual energy among non-clustered nodes, subsequently it broadcasts an energy threshold and non-clustered nodes that fulfill the threshold joins the cluster as CMs [154, 155] .
  • In [127] , the threshold is based on a social relation metric (M.3) whereby a maximum number of CMs with similar interests (i.e., multimedia content) can join a cluster.
  • During the fourth stage, cluster maintenance (or reclustering) occurs when there are major changes to the underlying cluster structure which are affecting the entire cluster, such as a large number of non-clustered nodes joining a cluster, or a large number of clustered nodes leaving a cluster.

5. Clustering Algorithms in 5G Networks

  • The clustering algorithms are segregated into four categories based on the clustering objectives (see Section 3.1), namely enhancing cluster stability (O.1), enhancing load balancing (O.2), enhancing social awareness (O.3), and enhancing fairness (O.4).
  • In addition to achieving the aforementioned clustering objectives, some of these clustering algorithms achieve the objective of enhancing QoS (O.5).
  • Table 4 shows the summary of the clustering schemes presented in this section.

5.1.1. Reducing CH election time based on energy consumption

  • Lina et al. [124, 156] propose a clustering scheme that minimizes energy consumption during CH election to prolong network lifetime in order to increase cluster stability.
  • During data communication, a CH, which is equipped with multiple interfaces and uses MIMO, select the right interfaces for communication (e.g., video transmission uses high-bandwidth interface, and data transmission uses low-bandwidth interface, to reduce energy consumption), and can operate at several channels simultaneously.
  • Smart-BEE(M) has shown to improve spectral efficiency (P.3), energy efficiency (P.4), reduces the number of clusters in the network (P.6), and improve cluster stability (P.7).
  • Khan et al. [122] propose a clustering scheme that elects reliable CH to prolong network connectivity in order to increase cluster stability.
  • During the first stage, non-clustered UEs generate NS using physical parameters to develop a network topology consisting the UEs in the network.

5.1.4. Electing backup CHs

  • Duan et al. [126] propose a clustering scheme that elects backup CHs to prolong network connectivity in order to increase cluster stability and network lifetime.
  • The CHs can aggregate and forward traffic pattern information to BSs in a single hop, and subsequently the CHs forward the information to a centralized controller.
  • The backup CH becomes the CH of a cluster when the existing CH leaves the cluster.
  • During the third stage, a non-clustered UE joins a cluster if both are located at the same region and are moving in the same direction based on a geographical location (M.2) and a mobility (M.4) metrics, respectively, which increases the connection time of a link in a node pair.
  • This scheme has shown to improve QoS performance (P.1) (i.e., lower packet loss), reduce the number of clusters in the network (P.6), and improve cluster stability (P.7).

5.1.5. Merging clusters based the number of handovers of UEs between two SCs

  • Ying et al. [127] propose a clustering scheme that merges two clusters to prolong network connectivity.
  • Each cluster consists of a centralized controller as the CH, and UEs as the CMs.
  • The main objective is to enhance cluster stability (O.1) among network entities.
  • The clustering scheme addresses the challenge of high cost (X.2) caused by clustering overhead and energy consumption.
  • This is because, as the number of SCs (i.e., femtocells) increases, the inter-cell interference level and the number of handovers of mobile UEs from one femtocell to another increases [171] .

5.2.1. Adjusting cluster size to transfer traffic load among SCs

  • Ali et al. [128] propose a clustering scheme that adjusts its cluster size (i.e., the number of nodes in a cluster) to self-organize traffic load in order to achieve load balancing in 5G networks.
  • The main objective is to enhance load balancing (O.2).
  • When the traffic load exceeds the pre-defined threshold, the cluster size increases in order to increase the CoMP gain leading to more available resources.
  • During the fourth stage, when a cluster becomes too large causing high clustering overhead and energy consumption, cluster maintenance (or re-clustering) is initiated to form new clusters and the traffic load of SCs is equally distributed among the newly formed clusters in order to increase cluster lifetime.

5.3.1. Joining node based on social interests

  • In these schemes, UEs form clusters among themselves and connect to macrocell BSs.
  • The trust value increases with the number of message exchanges (i.e., higher number of message exchanges, or communications, increases the effectiveness in improving privacy and security, so it leads to a higher trust value) [132, 131] .

5.4. Clustering Algorithms for Enhancing Fairness

  • This section presents two clustering algorithms for achieving the objective of enhancing fairness. ).
  • Huang et al. [134] propose a clustering scheme that allocate resources (i.e., channels) among clusters in a SC (i.e., femtocell) fairly to minimize inter-cluster interference, and hence inter-cell interference [181] .
  • High-quality channels are allocated to clusters with higher priority to achieve a fair distribution of resources among clusters in a femtocell.

6.1. Performing cluster size adjustment

  • A) increases network scalability by increasing the number of CMs in a cluster (or reduces the number of clusters in a network) leading to reduced clustering overhead (or clustering message exchanges) which helps to address the challenge of high cost, and b) increases cluster stability by reducing the number of re-clusterings [184], also known as Large cluster size.
  • In other words, higher number of nodes increases the number of clusters and message exchanges in the network, and hence requiring more network resources and reducing network performance [187] .
  • Cluster size adjustment is necessary to adjust the number of clusters and message exchanges in the network.
  • Mobility pattern can also be applied to predict new positions of nodes for the node joining procedure.

6.2. Performing cluster maintenance

  • Cluster maintenance or re-clustering, which is the fourth stage of the clustering framework (see Section 4), forms new cluster(s) after the dissolution of an existing cluster in order to maintain network performance as time goes by [190, 192] .
  • In short, cluster maintenance ensures that network resources are efficiently utilized in order to enhance QoS (P.1), spectral efficiency (P.3), energy efficiency (P.4), as well as to reduce cost (P.5), and the number of clusters in the network (P.6).

6.3. Ensuring node connectivity

  • Intra-and inter-cluster connectivities among nodes reduce network partitions.
  • Long-term connectivity among nodes in a cluster increases network lifetime leading to cluster stability [194] ; while short-term connectivity reduces network lifetime causing cluster instability [195] .
  • This can be performed using particle swarm optimization [197] , artificial bees colony [198] , and ant colony optimization, that extracts social behaviors of bird flocks, bees, and ant colonies, respectively [199] .

6.4. Adopting self-organization

  • Traditionally, artificial intelligence approaches have been used to perform selforganization, whereby macro-learning is applied at the centralized controller and micro-learning is applied at the distributed entities.
  • Clustering messages, generated by a large number of network entities (e.g., sensors, meters, and tracking devices), must be minimized as local information, such as the congestion level, mobility pattern, and data rate of the node [26] , must be gathered from the network entities by the centralized controller [204] .
  • 5G networks pose three main challenges to self-organization: a) high heterogeneity (X.1) whereby there are a diverse range of nodes with different capabilities, b) high cost (X.2), particularly clustering overhead, is incurred for selecting new CHs and node joining under ultra-densification, and c) high dynamicity (X.3) whereby nodes with high mobility introduces unpredictability in selforganization.
  • In addition to performing continuous monitoring, unexpected events, such as those that occur during disaster and emergency, should be detected.

6.5. Enhancing quality of service

  • Due to ultra-densification in 5G networks, clustering a large number of nodes, with some using real-time applications, demands a stringent level of user requirements and QoS [206] .
  • In order to provide more reliable communication and larger coverage, control message transmission uses licensed channel, while data transmission uses either licensed or unlicensed (e.g., IEEE 802.11p) channel.
  • 5G networks pose three main challenges to QoS enhancement, namely high heterogeneity (X.1), high cost (X.2), and c) high dynamicity (X.3), all of which reduce QoS.
  • Addressing these challenges can enhance QoS (P.1) (e.g., higher throughput and lower end-to-end delay), as well as reduce cost (P.5) and the number of clusters in the network (P.6).
  • Further studies can be pursued in this topic.

7. Conclusion

  • Clustering schemes for the next-generation mobile wireless network, namely 5G, are presented.
  • This article refreshes the topic of clustering, its motivation and background in 5G through a review of the limited research works in this topic.
  • The clustering schemes discussed in this paper, which are mainly based on five types of objectives (i.e., enhancing cluster stability, load balancing, social awareness, fairness, and quality of service (QoS)) address three main challenges, namely high heterogeneity, high cost, and high dynamicity.
  • These schemes are based on three main characteristics, namely the number of hops (i.e., single-hop and multihop), node type (i.e., homogeneous and heterogeneous), and mobility rate (i.e., quasi static and mobile/ nomadic).
  • The paper is expected to support and motivate researchers for further exploration and investigation in this research area.

Did you find this useful? Give us your feedback

Figures (13)
Citations
More filters
Millimeter wave beamforming for wireless backhaul and access in small cell networks and practical approaches in software-defined radio

[...]

James V. Krogmeier1, David J. Love1, Sooyoung Hur1•
Purdue University1
01 Jan 2013
TL;DR: In this paper, an efficient beam alignment technique using adaptive subspace sampling and hierarchical beam codebooks was proposed to solve the problem of spectrum reusability and flexible prototyping radio platform using software-defined radio (SDR).
Abstract: Mobile data traffic will continue its tremendous growth in some markets, and has already resulted in an apparent radio spectrum scarcity. There is a strong need for more efficient methods to use spectrum resources, leading to extensive research on increasing spectrum reusability on flexible radio platforms. This study solves this problem in two sub topics, millimeter wave communication on wireless backhaul for spectrum reusability, and flexible prototyping radio platform using software-defined radio (SDR). Wireless backhaul has received significant attention as a key technology affecting the development of future wireless cellular networks because it helps to easily deploy many small size cells, an essential part of a high capacity system. Millimeter wave is considered a possible candidate for cost-effective wireless backhaul. In the outdoor deployment using a millimeter wave, beamforming methods are key techniques to establish wireless links in the 60 GHz to 80 GHz to overcome pathloss constraints (i.e., rainfall effect and oxygen absorption). The millimeter wave communication system cannot directly access the channel knowledge. To overcome this, a beamforming method based on codebook search is considered. The millimeter wave communication cannot access channel knowledge, therefore alternatively a beamforming method based on a codebook search is considered. In the first part, we propose an efficient beam alignment technique using adaptive subspace sampling and hierarchical beam codebooks. A wind sway analysis is presented to establish a notion of beam coherence time. This highlights a previously unexplored tradeoff between array size and wind-induced movement. Generally, it is not possible to use larger arrays without risking a performance loss from wind-induced beam misalignment. The performance of the proposed alignment technique is analyzed and compared with other search and alignment methods. Results show significant performance improvement with reduced search time. In the second part of this study, SDR is discussed as an approach toward flexible wireless communication systems. Most layers of SDR are implemented by software. Therefore, only a software change is needed to transform the type of radio system. The translation of the signal processing into software performed by a regular computer opens up a huge number of possibilities at a reasonable price and effort. SDR systems are widely used to build prototypes, saving time and money. In this project, a robust wireless communication system in high interference environment was developed. For the physical layer (PHY) of the system, we implemented a channel sub-bandding method that utilizes frequency division multiplexing to avoid interference. Then, to overcome a further interfered channel, Direct Spread Spectrum System (DSSS) was considered and implemented. These prototyped testbeds were evaluated for system performance in the interference environment.

103 citations

Journal Article•DOI•
Optimized Node Clustering in VANETs by Using Meta-Heuristic Algorithms

[...]

Waleed Ahsan, Muhammad Fahad Khan, Farhan Aadil, Muazzam Maqsood, Staish Ashraf, Yunyoung Nam, Seungmin Rho 
27 Feb 2020-Electronics
TL;DR: The grasshoppers’ optimization-based node clustering algorithm for VANETs (GOA) for optimal cluster head selection reduced network overhead in unpredictable node density scenarios and indicated that GOA outperformed existing methodologies.
Abstract: In a vehicular ad-hoc network (VANET), the vehicles are the nodes, and these nodes communicate with each other. On the road, vehicles are continuously in motion, and it causes a dynamic change in the network topology. It is more challenging when there is a higher node density. These conditions create many difficulties for network scalability and optimal route-finding in VANETs. Clustering protocols are being used frequently to solve such type of problems. In this paper, we proposed the grasshoppers’ optimization-based node clustering algorithm for VANETs (GOA) for optimal cluster head selection. The proposed algorithm reduced network overhead in unpredictable node density scenarios. To do so, different experiments were performed for comparative analysis of GOA with other state-of-the-art techniques like dragonfly algorithm, grey wolf optimizer (GWO), and ant colony optimization (ACO). Plentiful parameters, such as the number of clusters, network area, node density, and transmission range, were used in various experiments. The outcome of these results indicated that GOA outperformed existing methodologies. Lastly, the application of GOA in the flying ad-hoc network (FANET) domain was also proposed for next-generation networks.

49 citations

Cites methods from "Survey and taxonomy of clustering a..."

  • ...A clustering-based reliable low-latency routing scheme using the ACO method for vehicular networks [36] is used to enhance VANETs QoS....

    [...]

Journal Article•DOI•
K-means clustering algorithms: A comprehensive review, variants analysis, and advances in the era of big data

[...]

Belal Abuhaija1, Namig Isazade•
Wenzhou-Kean University1
01 Apr 2023-Information Sciences
TL;DR: An overview and taxonomy of the K-means clustering algorithm and its variants can be found in this article , where the authors present an overview of the current trends, open issues and challenges, and recommended future research perspectives.

32 citations

Journal Article•DOI•
SAMNET: Self-adaptative multi-kernel clustering algorithm for urban VANETs

[...]

Lamaa Sellami1, Bechir Alaya2, Bechir Alaya1•
Qassim University1, University of Gabès2
01 Jun 2021-Vehicular Communications
TL;DR: This work takes into account the random and continuous evolution of traffic in the VANET environment and adopts a system to model the mode of evolution based on commutation, a self-adapting clustering algorithm that consists of modeling each sub-model based on a linear regression function.

18 citations

Proceedings Article•DOI•
A comprehensive survey on clustering in vehicular networks: Current solutions and future challenges

[...]

01 Jan 2022
TL;DR: A comprehensive survey on the recent advancements in the clustering schemes for vehicular networks is presented in this article , where the authors take a holistic approach to classify the algorithms by focusing on, (i) the objective of clustering mechanisms (i.e., reliability, scalability, stability, routing overhead, and delay), (ii) general-purpose clustering algorithms, (iii) application-based (i., QoS, MAC, security, etc.) clustering, and iv) technology-based clustering (machine learning-based, nature-inspired, fuzzy logic-based and software-defined networking-based).
Abstract: Vehicular networks are on the verge of deployment, thanks to the advancements in computation and communication technologies. This breed of ad hoc networks leverages vehicles as nodes with Vehicle-to-anything (V2X) communication paradigm. Clustering is considered one of the most important techniques used to enhance network stability, reliability, and scalability. Furthermore, clustering employs bandwidth optimization by reducing the overhead and transmission delay and helps in mitigating the hidden node problem. To date, extensive research has been done to address clustering issues in vehicular networks, and several surveys have also been published in the literature. However, a holistic approach towards clustering in vehicular networks is still lacking. In this regard, we conduct a comprehensive survey on the recent advancements in the clustering schemes for vehicular networks. We take a holistic approach to classify the algorithms by focusing on, (i) the objective of clustering mechanisms (i.e., reliability, scalability, stability, routing overhead, and delay), (ii) general-purpose clustering algorithms, (iii) application-based (i.e., QoS, MAC, security, etc.) clustering, and iv) technology-based clustering (machine learning-based, nature-inspired, fuzzy logic-based and software-defined networking-based clustering). We investigate the existing clustering mechanisms keeping in mind the factors such as cluster formation, maintenance, and management. Additionally, we present a comprehensive set of parameters for selecting cluster heads and the role of enabling technologies for cluster maintenance. Finally, we identify future research trends in clustering techniques for vehicular networks and their various breeds. This survey will act as a one-stop shop for the researchers, practitioners, and system designers to select the right clustering mechanism for their applications, services, or for their research. As a result of this survey, we can see that clustering is heavily dependent on the underlying application, context, environment, and communication paradigm. Furthermore, clustering in vehicular networks can greatly benefit from enabling technologies such as artificial intelligence.

16 citations

References
More filters
Journal Article•DOI•
Millimeter Wave Mobile Communications for 5G Cellular: It Will Work!

[...]

Theodore S. Rappaport1, Shu Sun1, Rimma Mayzus1, Hang Zhao1, Yaniv Azar1, Kevin H. Wang1, George N. Wong1, Jocelyn K. Schulz1, Mathew K. Samimi1, Felix Gutierrez1 •
New York University1
10 May 2013-IEEE Access
TL;DR: The motivation for new mm-wave cellular systems, methodology, and hardware for measurements are presented and a variety of measurement results are offered that show 28 and 38 GHz frequencies can be used when employing steerable directional antennas at base stations and mobile devices.
Abstract: The global bandwidth shortage facing wireless carriers has motivated the exploration of the underutilized millimeter wave (mm-wave) frequency spectrum for future broadband cellular communication networks. There is, however, little knowledge about cellular mm-wave propagation in densely populated indoor and outdoor environments. Obtaining this information is vital for the design and operation of future fifth generation cellular networks that use the mm-wave spectrum. In this paper, we present the motivation for new mm-wave cellular systems, methodology, and hardware for measurements and offer a variety of measurement results that show 28 and 38 GHz frequencies can be used when employing steerable directional antennas at base stations and mobile devices.

6,708 citations

Journal Article•DOI•
Internet of Things: A Survey on Enabling Technologies, Protocols, and Applications

[...]

Ala Al-Fuqaha1, Mohsen Guizani2, Mehdi Mohammadi1, Mohammed Aledhari1, Moussa Ayyash3 •
Western Michigan University1, Qatar University2, Chicago State University3
15 Jun 2015-IEEE Communications Surveys and Tutorials
TL;DR: An overview of the Internet of Things with emphasis on enabling technologies, protocols, and application issues, and some of the key IoT challenges presented in the recent literature are provided and a summary of related research work is provided.
Abstract: This paper provides an overview of the Internet of Things (IoT) with emphasis on enabling technologies, protocols, and application issues. The IoT is enabled by the latest developments in RFID, smart sensors, communication technologies, and Internet protocols. The basic premise is to have smart sensors collaborate directly without human involvement to deliver a new class of applications. The current revolution in Internet, mobile, and machine-to-machine (M2M) technologies can be seen as the first phase of the IoT. In the coming years, the IoT is expected to bridge diverse technologies to enable new applications by connecting physical objects together in support of intelligent decision making. This paper starts by providing a horizontal overview of the IoT. Then, we give an overview of some technical details that pertain to the IoT enabling technologies, protocols, and applications. Compared to other survey papers in the field, our objective is to provide a more thorough summary of the most relevant protocols and application issues to enable researchers and application developers to get up to speed quickly on how the different protocols fit together to deliver desired functionalities without having to go through RFCs and the standards specifications. We also provide an overview of some of the key IoT challenges presented in the recent literature and provide a summary of related research work. Moreover, we explore the relation between the IoT and other emerging technologies including big data analytics and cloud and fog computing. We also present the need for better horizontal integration among IoT services. Finally, we present detailed service use-cases to illustrate how the different protocols presented in the paper fit together to deliver desired IoT services.

6,131 citations

Journal Article•DOI•
HEED: a hybrid, energy-efficient, distributed clustering approach for ad hoc sensor networks

[...]

O. Younis1, Sonia Fahmy1•
Purdue University1
01 Oct 2004-IEEE Transactions on Mobile Computing
TL;DR: It is proved that, with appropriate bounds on node density and intracluster and intercluster transmission ranges, HEED can asymptotically almost surely guarantee connectivity of clustered networks.
Abstract: Topology control in a sensor network balances load on sensor nodes and increases network scalability and lifetime. Clustering sensor nodes is an effective topology control approach. We propose a novel distributed clustering approach for long-lived ad hoc sensor networks. Our proposed approach does not make any assumptions about the presence of infrastructure or about node capabilities, other than the availability of multiple power levels in sensor nodes. We present a protocol, HEED (Hybrid Energy-Efficient Distributed clustering), that periodically selects cluster heads according to a hybrid of the node residual energy and a secondary parameter, such as node proximity to its neighbors or node degree. HEED terminates in O(1) iterations, incurs low message overhead, and achieves fairly uniform cluster head distribution across the network. We prove that, with appropriate bounds on node density and intracluster and intercluster transmission ranges, HEED can asymptotically almost surely guarantee connectivity of clustered networks. Simulation results demonstrate that our proposed approach is effective in prolonging the network lifetime and supporting scalable data aggregation.

4,889 citations

Journal Article•DOI•
Five disruptive technology directions for 5G

[...]

Federico Boccardi1, Robert W. Heath2, Angel Lozano3, Thomas L. Marzetta4, Petar Popovski5 •
Vodafone1, University of Texas at Austin2, Pompeu Fabra University3, Bell Labs4, Aalborg University5
19 Feb 2014-IEEE Communications Magazine
TL;DR: In this article, the authors describe five technologies that could lead to both architectural and component disruptive design changes: device-centric architectures, millimeter wave, massive MIMO, smarter devices, and native support for machine-to-machine communications.
Abstract: New research directions will lead to fundamental changes in the design of future fifth generation (5G) cellular networks. This article describes five technologies that could lead to both architectural and component disruptive design changes: device-centric architectures, millimeter wave, massive MIMO, smarter devices, and native support for machine-to-machine communications. The key ideas for each technology are described, along with their potential impact on 5G and the research challenges that remain.

3,711 citations

Journal Article•DOI•
Software-Defined Networking: A Comprehensive Survey

[...]

Diego Kreutz1, Fernando M. V. Ramos2, Paulo Veríssimo1, Christian Esteve Rothenberg3, Siamak Azodolmolky, Steve Uhlig4 •
University of Luxembourg1, University of Lisbon2, State University of Campinas3, Queen Mary University of London4
01 Jan 2015
TL;DR: This paper presents an in-depth analysis of the hardware infrastructure, southbound and northbound application programming interfaces (APIs), network virtualization layers, network operating systems (SDN controllers), network programming languages, and network applications, and presents the key building blocks of an SDN infrastructure using a bottom-up, layered approach.
Abstract: The Internet has led to the creation of a digital society, where (almost) everything is connected and is accessible from anywhere. However, despite their widespread adoption, traditional IP networks are complex and very hard to manage. It is both difficult to configure the network according to predefined policies, and to reconfigure it to respond to faults, load, and changes. To make matters even more difficult, current networks are also vertically integrated: the control and data planes are bundled together. Software-defined networking (SDN) is an emerging paradigm that promises to change this state of affairs, by breaking vertical integration, separating the network's control logic from the underlying routers and switches, promoting (logical) centralization of network control, and introducing the ability to program the network. The separation of concerns, introduced between the definition of network policies, their implementation in switching hardware, and the forwarding of traffic, is key to the desired flexibility: by breaking the network control problem into tractable pieces, SDN makes it easier to create and introduce new abstractions in networking, simplifying network management and facilitating network evolution. In this paper, we present a comprehensive survey on SDN. We start by introducing the motivation for SDN, explain its main concepts and how it differs from traditional networking, its roots, and the standardization activities regarding this novel paradigm. Next, we present the key building blocks of an SDN infrastructure using a bottom-up, layered approach. We provide an in-depth analysis of the hardware infrastructure, southbound and northbound application programming interfaces (APIs), network virtualization layers, network operating systems (SDN controllers), network programming languages, and network applications. We also look at cross-layer problems such as debugging and troubleshooting. In an effort to anticipate the future evolution of this new paradigm, we discuss the main ongoing research efforts and challenges of SDN. In particular, we address the design of switches and control platforms—with a focus on aspects such as resiliency, scalability, performance, security, and dependability—as well as new opportunities for carrier transport networks and cloud providers. Last but not least, we analyze the position of SDN as a key enabler of a software-defined environment.

3,589 citations

Related Papers (5)
SDN Enabled 5G-VANET: Adaptive Vehicle Clustering and Beamformed Transmission for Aggregated Traffic

[...]

14 Jul 2017-IEEE Communications Magazine
Xiaoyu Duan, Yanan Liu, Xianbin Wang
Cluster-based location service schemes in VANETs: current state, challenges and future directions

[...]

01 Mar 2021-Telecommunication Systems
Muhammad Akram Mujahid, Kamalrulnizam Abu Bakar, Tasneem Darwish, Fatima Tul Zuhra 
Network Slicing Based 5G and Future Mobile Networks: Mobility, Resource Management, and Challenges

[...]

24 Apr 2017-arXiv: Information Theory
Haijun Zhang, Na Liu, Xiaoli Chu, Keping Long, Abdol Hamid Aghvami, Victor C. M. Leung 
LiM-AHP-G-C: Life Time Maximizing based on Analytical Hierarchal Process and Genetic Clustering protocol for the Internet of Things environment

[...]

20 Jul 2020-Computer Networks
Khalid A. Darabkh, Wafa'a K. Kassab, Ala' F. Khalifeh
Content sharing in Internet of Vehicles: Two matching-based user-association approaches

[...]

01 Apr 2017-Vehicular Communications
Francesco Chiti, Romano Fantacci, Yunan Gu, Zhu Han 
Frequently Asked Questions (22)
Q1. What are the contributions in "Survey and taxonomy of clustering algorithms in 5g" ?

In this paper, the authors focus on clustering in access networks, rather than cell clustering, in order to achieve a better Quality of Service ( QoS ) of up to 10× higher data rate and up to 1000× lower delay. 

The paper is expected to support and motivate researchers for further exploration and investigation in this research area. 

Load balancing is affected by: a) ultra-densification whereby higher node density increases the amount of data generated, b) heterogeneity whereby a diverse range of network cells increases the number of handovers, handover overhead, and computational cost, and c) high variability whereby higher variability causes a sudden demand for significantly higher network capacity. 

The effects of a lower cluster stability are that it: a) increases packet loss rate, b) reduces cluster lifetime [124], c) increases clustering overhead or message exchange, and hence network congestion [103], and d) increases computational cost or time complexity [122]. 

Higher energy efficiency reduces node failure, and so it: a) increases the lifetime of a cluster and a CH, and reduces the occurrence of re-clustering (i.e., CH election and node joining), and b) minimizes network partition (or network topology disconnection) [152]. 

5G networks pose three main challenges to QoS enhancement, namely high heterogeneity (X.1), high cost (X.2), and c) high dynamicity (X.3), all of which reduce QoS. 

The clustering schemes mainly use five types of clustering metrics, namely node density, geographical location, mobility metrics, social relation, and residual energy, for the clusterhead election and node joining procedures. 

5G networks pose three main challenges to self-organization: a) high heterogeneity (X.1) whereby there are a diverse range of nodes with different capabilities, b) high cost (X.2), particularly clustering overhead, is incurred for selecting new CHs and node joining under ultra-densification, and c) high dynamicity (X.3) whereby nodes with high mobility introduces unpredictability in selforganization. 

There are three main clustering challenges that must be addressed during cluster formation and maintenance (or reclustering) in 5G networks: X.1 High heterogeneity, which is an intrinsic characteris-tic of 5G, poses difficulties in segregating nodes with different nature of heterogeneity, including interests or behaviors, in a network into logical groups [140, 141]. 

A CM can maintain its status quo, join another cluster, or leave its existing cluster in a highly dynamic network with frequent network topological variations, resulting in shorter cluster lifetime and frequent re-clustering. 

5G networks pose three main challenges to achieving long-term connectivity, including: a) high heterogeneity (X.1) whereby there are a diverse range of nodes (i.e., with different interests and purposes) with different capabilities (i.e., mobility, transmission power, and social relations), b) high cost (X.2), particularly clustering overhead, is incurred due to ultra-densification, and c) high dynamicity (X.3) whereby nodes’ or UEs’ with highmobility increase the changes of the underlying network topology (e.g., nodes move out of the coverage of a cluster and cause dis-connectivity). 

In addition, insufficient of resources can cause the dissolution of a cluster as a results of increased packet loss and reduced QoS of the entire cluster [193, 26]. 

The tremendous growth of user equipment (UE) expecting to reach up to billions in number [1, 2], along with bandwidthstarving applications (e.g., video streaming, multimedia sharing, and online gaming), has contributed to 74% increment in data traffic over the years [3]. 

The centralized controller coordinates with clusters to execute the first to third stages again whenever important UEs (e.g., CH) leave the cluster. 

Ali et al. [128] propose a clustering scheme that adjusts its cluster size (i.e., the number of nodes in a cluster) to self-organize traffic load in order to achieve load balancing in 5G networks. 

computational cost (or time complexity) due to the time incurred for re-clustering, specifically from the dissolution of a cluster until all non-clustered nodes are clustered again. 

As an example, in [112], there are heterogeneous UEs with different requirements for networkresources [113], and clustering segregates SCs into logical groups in order to reduce interference. 

Due to ultra-densification in 5G networks, clustering a large number of nodes, with some using real-time applications, demands a stringent level of user requirements and QoS [206]. 

Duan et al. [126] propose a clustering scheme that elects backup CHs to prolong network connectivity in order to increase cluster stability and network lifetime. 

This is because, as the number of SCs (i.e., femtocells) increases, the inter-cell interference level and the number of handovers of mobile UEs from one femtocell to another increases [171]. 

Clustering has been investigated to support mobile user equipment (UE) in access networks, whereby UEs form clusters themselves and may connect to BSs. 

The clustering scheme addresses three challenges, namely high heterogeneity (X.1) because different UEs have different transmission ranges, high cost (X.2) whereby there is a high energy consumption and overhead due to re-transmission [163, 164], and high dynamicity (X.3) due to high mobility of UEs.