Abstract
Cellular vehicle-to-everything (C-V2X) communications have elicited significant scrutiny in recent years owing to amenities such as comfort, efficiency, and an enhanced line-of-sight through the exchange of statistics between vehicles and other entities. In general, achieving less interference in terms of resource allocation and a higher quality of forwarding (QoF) during routing is a major problem owing to the highly mobile environment used in such communications. To overcome this bottleneck, the present study focuses on emperor-based resource allocation for device-to-device (D2D) communications and QoF routing (ERA-D(2)Q) in C-V2X. Our ERA-D(2)Q offers three sub-sequential processes: (1) coalition-based routing, (2) D2D communications and SignRank oriented pedestrian selection, and (3) emperor based resource allocation. To achieve QoF routing, ERA-D(2)Q introduces coalition game theory (CGT), which elects an optimum relay and reduces the amount of time for a packet transmission. Our D2D communication is applied in two different cases: (1) devices discovered within the coverage of the discoverer and (2) devices discovered out of the coverage of the discoverer. The second case is conducted with the aid of the best pedestrian, who is elected using the SignRank algorithm. The best pedestrian sends a relay request to the 5G base station (BS) to discover a destination for D2D communications. For this purpose, 5G BS utilizes the Mamdani interval type 2 fuzzy algorithm. To optimize the resource blocks assigned to the D2D communications, an emperor penguin colony algorithm is established for the proposed ERA-D(2)Q, which is employed in an OMNET++ simulator. Finally, we validate the performance achieved in our study using five metrics, namely, the packet delivery ratio (PDR), average transmission delay (ATD), throughput, mean opinion score (MoS), and jitter. The evaluation results prove that our ERA-D(2)Q enhances the PDR and throughput by up to 35%, and decreases the ATD by up to 48%, the jitter by up to 33%, and the MoS by 40% compared to the existing methods QFRG, CF, SMRS, SRA, and DTDMD.