End-to-End Distortion-Based Multiuser Bandwidth Allocation for Real-Time Video Transmission Over LTE Network

Yuan, Hui; Fu, Huayong; Liu, Ju; Xiao, Jimin

doi:10.1109/tbc.2016.2630267

Cited by 8 publications

(2 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As shown in FIGURE 1, the source coding (e.g., HEVC [9]) is used to eliminate the redundancy within the original video signal, while the channel coding (e.g., Turbo code, and convolutional code) is utilized to deal with the transmission errors over the wireless channel. This kind of video transmission strategy could obtain the optimal performance with the assumption that the channel condition is known before the source and channel coding to choose the appropriate coding rate [10]. Apparently, it is not suitable for the wireless scenarios where the channel condition may vary drastically.…”

Section: Introductionmentioning

confidence: 99%

Block DCT Based Optimization for Wireless SoftCast of Depth Map

et al. 2019

View full text Add to dashboard Cite

With the recent emergence of naked-eye 3D mobile devices, the 3D video-based applications on the mobile terminal are showing great potential. However, the conventional transmission scheme based on separate source coding and channel coding may fail in the broadcast and mobile scenarios due to the drastic changes in wireless channel condition. Recently, an uncoded transmission scheme called SoftCast was introduced, which shows efficient performance for wireless 2D video transmission. However, it may not work well for 3D video wireless transmission, especially for the depth maps. Based on the piecewise smoothness of depth maps, we propose an uncoded wireless depth map transmission scheme, which incorporates three customized components of optimization into the SoftCast framework. First, compared with the framelevel discrete cosine transform (DCT) used in 2D video SoftCast, the mean removed block-based DCT is preferred as the decorrelation transform for depth maps to accommodate its piecewise smooth property, which also facilitates the following subsequent optimization. Second, view synthesis distortion-based chunk dropping is utilized for more efficient power allocation in the case of limited bandwidth, where the block DCT provides a better estimation for synthesis distortion. Third, inter-block DCT coefficients are rearranged to form chunks in terms of minimizing the intra-chunk variance, further enhancing the chunk-level power allocation efficiency. The experimental results demonstrate that the proposed scheme can provide superior performance than the state-of-the-art SoftCast methods while achieving graceful synthesis quality transition within a wide channel signal-to-noise ratio range.

show abstract

Section: Introductionmentioning

confidence: 99%

Block DCT Based Optimization for Wireless SoftCast of Depth Map

et al. 2019

View full text Add to dashboard Cite

show abstract

“…Alternatively, [2] designs an optimal rate allocation and description distribution for high performance video streaming over 5G networks. Reference [3] proposes an end-to-end distortion-based multiuser bandwidth allocation for real-time video transmission over LTE network. The works in [4] and [5] are particularly engaged in the rate adaptation and bandwidth allocation for dynamic video streaming over HTTP.…”

Section: Introductionmentioning

confidence: 99%

A Lossless Recompression Approach for Video Streaming Transmission

Wang

Hou

et al. 2019

IEEE Access

View full text Add to dashboard Cite

With the increasing amount of HD or even UHD video, the video streaming transmission over the mobile network is confronted with various quality of experience issues. To facilitate the delivery of compressed video over the network, it is extremely helpful to further compress the coded video bit stream without any information loss, so-called lossless recompression. However, the existing lossless video/image compression algorithms aim to remove the pixel-domain (e.g., lossless coding mode of HEVC) or the coefficient-domain (e.g., entropy coding) redundancy rather than the bit-stream-domain redundancy. To this end, we propose a novel lossless recompression approach to eliminate statistical redundancy in the coded video bit stream. The core idea is to increase the probability of this repetitive pattern of continuous ''0'' or ''1'' symbol strings in the binary symbol sequence by rearranging the sequence, including value mapping and position aggregation. In particular, Fibonacci-based mapping rule is used to convert the original sequence (those binary symbols in Fibonacci positions) into a new one with more continuous ''0'' or ''1.'' Besides, we aggregate as many identical symbols as possible together to further increase the probability of the occurrence of repetitive patterns. Finally, the lossless compression algorithm and the asymptotic lossless compression scheme are designed to achieve a compact representation of the rearranged symbol sequence. We also regulate formal syntactic structures for the proposed mapping rule, aggregation algorithm, and recompression scheme so as to deterministically revert from the recoded version to the original bit stream. The experimental results on compressed H.264/AVC and H.265/HEVC video data show that our approach can considerably reduce the bit rate of streaming video. Video streaming, video coding, lossless compression, recompression, Fibonacci. INDEX TERMS

show abstract