Low Complexity Joint Power and Bandwidth Allocation for 3D Video SoftCast

Abstract

SoftCast transmission achieves a linear video quality transition commensurate with the wireless channel conditions, allowing it to avoid the cliff effect, unlike the digital video transmission system. When adopting SoftCast for three-dimensional video transmission, several issues arise. 1) Optimally allocate the power budget to texture and depth to minimize transmission distortion. 2) Select the appropriate number of texture and depth chunks to meet the bandwidth constraints. 3) Reduce the 3D video computation complexity. This paper aims to optimally solve the joint power and bandwidth allocation with low complexity. We derive a closed-form based on the power-distortion optimization problem to estimate the optimal power allocation ratio between the texture and depth map. Then to adapt to the bandwidth constraints. We first set the available bandwidth to be shared equally between texture and depth, this joint bandwidth allocation technique reduces the complexity, and to improve the overall 3D video quality, the joint power allocation is estimated at different bandwidth constraints. With low complexity, the proposed method achieves a graceful video quality transition with the improvement of channel conditions under bandwidth constraints and better performance than its counterpart default fixed-ratio power allocation between texture and depth.