Complexity Scalable Motion Estimation Algorithm Based On QoE Control

With the large-scale construction of3G/4G networks by telecom operators, mobile real-time multimedia communication has increasingly become a core business of the next-generation mobile communications network. Because of the limitations of power consumption and processing power of mobile phones and other mobile devices, coupled with the high bit error rate of the mobile communications network and its limited available bandwidth, it is still very difficult and challenging in order to achieve a stable and reliable real-time video communication in the mobile communications network while ensuring that the user can enjoy the quality of service (QoS) and Quality of Experience (QoE).Based on the characteristics of real-time mobile video communication, combined with the limited nature of the cell phone battery, against the video encoding for mobile terminal, this paper constructs a QoE calculation model with two user experience dimensions, so that the user can select and adjust between video quality and duration as needed. The QoE calculation model constructed in this paper does not contain the parameters whose evaluation criteria will change over time, choosing instead the relative parameters that will better reflect the nature of the user experience, that mobile computation resource utilization and charge status. The formula used in the QoE calculation model constructed in this paper is the ratio of mobile computation resource utilization and charge status, at the same time develop a balance between video quality and duration, and get its reasonable explanation through the relationship between money and spending. The concept used in the QoE calculation model constructed in this paper that integrate multiple experience dimensions in a QoE calculated model and let the user to select and adjust as needed is of great theoretical value and practical value in how to construct and use the QoE calculation model.In addition, in order to achieve real-time video communication on mobile phones, mobile phone side encoder algorithm must have low complexity and complexity scalable. To this end, this paper proposes a new complexity scalable motion estimation algorithm using a single pass scanning process, which can take advantage of the relevant information of the neighboring blocks. The algorithm includes two processes:the allocation of the frame-level computation resource and the allocation of the block-level computation resource. The allocation of the frame-level computation resource is used in a manner like rate control, in order to facilitate the comparison with other programs, a fixed amount of computation is allocated for each frame in the experiment. The allocation of the block-level computation resource comprises two parts:the allocation of the block base layer computation resource and the allocation of the block gain layer computation resource. The allocation of the block base layer computation resource is performed in accordance with the initial matching error of each block, the allocation of the block gain layer computation resource is performed based on the distortion gain information of the previous frame. In order to reduce the impact of the non-validity of the distortion gain information, the algorithm adopts adaptive adjustment strategy and the early termination of the detection.The QoE calculation model constructed in this paper is used in the proposed complexity scalable motion estimation algorithm, and verify the practicality of the QoE calculation model and the effectiveness of the proposed complexity scalable motion estimation algorithm through experimental simulation.