A novel cross-layer based priority aware scheduling scheme for QoE guaranteed video transmission over wireless networks

The rapid proliferation of wireless networks increases the demand for video transmission in surveillance, online gaming, video streaming, and other applications. Still, assurance of Quality of Experience (QoE) in video transmission is restricted by multiple constraints, including wireless medium characteristics, multipath routing limitations, and so on. QoE is measured at the source in many prior works, which is not suitable for QoE assurance. This paper addresses all these issues in video transmission over wireless networks with a novel cross-layer design. The proposed cross-layer approach initially measures video quality at a destination based on video quality score (VQS) which is obtained from past destination by application layer, and the feedback is given to the source to guarantee QoE. For improving the quality of video transmission, Quality-based Adaptive Scalable Video Coding (QA-SVC) based video coding is performed in the source node to increase the transmission efficiency. The encoded video is transmitted over multiple paths which are scheduled using the Enriched Particle Swarm Optimization with Multiple Solutions (EPSO-MS) algorithm by considering numerous metrics to reduce the data loss. Improved Artificial Neural Network (IANN) and Deficit Weighted Round Robin (DWRR) are jointly used to schedule the video in intermediate nodes based on priority level which reduces the waiting delay with efficient video quality for transmitting the video before deadline. Video scheduling and priority classification are supported by the Modified Real-time Transport protocol (M-RTP) for fast priority provisioning. QoE of the video is assured and enhanced by performing video quality estimation based on VQS and it is carried out at the destination node using Type-2 Fuzzy Logic (T2FL). Finally, the proposed cross-layer design is modeled in NS-3.26 and evaluated based on throughput (2mbps (high)), jitter (20 ms (low)), goodput (1 mbps(high)), delay (25 ms (low)), Peak Signal-to-Noise Ratio (PSNR) (30 dB (high)), packet drop (7% (low)), bandwidth utilization (20%(high)), and mean opinion score (MOS) (2 (high)).