A novel clustering and optimal resource scheduling in vehicular cloud networks using MKMA and the CM‐CSO algorithm. (3rd January 2023)

• Record Type: Journal Article
• Authors: Ruhin Kouser, R.; Manikandan, T.
• Is Part Of: International journal of communication systems. Volume 36:Number 5(2023)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2023-01-03
• Subjects: cloud computing -- intelligent transportation system (ITS) -- resource management -- vehicular Adhoc networks -- vehicular cloud computing (VCC); Telecommunication systems -- Periodicals; 621.382
• DOI: 10.1002/dac.5424
• Languages: English
• ISSNs: 1074-5351

Abstract

Summary: Various opportunities aimed at the employment of delay‐sensitive applications in the vehicular environment are presented by the vehicular cloud (VC). Contrary to diverse wireless networks, VC networks possess exceptional features among others, namely, shorter transmission time along with a higher dynamic topology. Although integration with the cloud offers higher storage along with computation capabilities, it as well entails restricted resource availability. The restrictions on the number of resources serve as a challenge in servicing the applications with their necessary quality of service (QoS) guarantees as the number of service requests for applications keeps on augmenting with diverse circumstances. Thus, the need for an effective scheduling methodology arises to decide the sequence of servicing application requests and successful utilization of a broadcast medium, along with data transmission. To do efficient resource scheduling on VC networks, an optimization algorithm, namely, the crossover and mutation (CM)‐centered chicken swarm optimization (CSO) is proposed and implemented with the help of a publicly available dataset. Initially, the VC infrastructure is initialized and some vehicle information is extracted as features. Next, the Brownian motion‐centered bacteria foraging optimization (BM‐BFO) algorithm chooses the essential features. Centered on the chosen features, the vehicles are clustered using the modified K‐means algorithm. Next, as for the cloud server's virtual machines (VMs), the resource information is extracted. Lastly, the CM‐CSO algorithm carries out the optimal scheduling in the VC by means of the clustered features of vehicles and features of the VM. The proposed techniques' findings are scrutinized and analogized to the other prevailing methodologies to confirm that the proposed work performs effectively and gives optimal resource allocation (RA) to the VC. Abstract : VCC can deliver several on board data services, such as easy navigation, road safety, traffic efficiency, infotainment, and comfort driving. It is a facilitating technology aimed at intelligent transportation system (ITS), Smart cities, and autonomous driving. The resources are allotted to vehicles on demand that could well be performed by means of appropriate scheduling of resources aimed at cooperation and safety betwixt vehicles in VCC. Here, an efficient and optimal resource scheduling of VC is performed by means of proposing the crossover and mutation (CM)‐centered chicken swarm optimization (CSO) algorithm. The proposed work begins by initializing the VC infrastructure. After that, feature extraction (FE) of vehicles utilizing the silhouette, feature selection (FS), clustering of vehicles, and extraction of resource information, together with optimal resource scheduling is carried out. Initially, as of the silhouette, the vehicle's information or features, such as circularity, distance circularity, radius ratio, and scatter ratio, are extracted. Next, the Brownian motion‐centered bacteria foraging optimization (BM‐BFO) algorithm selects the necessary features. Subsequent to the FS, centered on the chosen feature values, the vehicles are clustered. After that, the resource information of the VM is extracted as of the cloud server (CS) like speed, memory, and cost. Lastly, the CM‐CSO algorithm performs the optimal resource scheduling of the VC. Centered on the retrieved information of the VM, this technique efficiently allows the resources to the clustered vehicles.