Institution
China Mobile Research Institute
About: China Mobile Research Institute is a based out in . It is known for research contribution in the topics: MIMO & Wireless network. The organization has 579 authors who have published 542 publications receiving 13897 citations.
Topics: MIMO, Wireless network, Cellular network, Wireless, Spectral efficiency
Papers
More filters
••
10 Oct 2011TL;DR: Experimental results demonstrate the proposed low complexity rate control method outperforms the method using in reference software, which can achieve better performance in terms of increasing coding quality, depressing quality fluctuations and reducing buffer overflow/underflow.
Abstract: Rate control algorithm is one of important factors influencing the coding performance. In this paper, a low complexity rate control (LCRC) method is proposed. Instead of using RQ model and MAD model, LCRC determine the QP for P frames according to the allocated bits for current frame and the encoded bits of previous frame. Moreover, LCRC exploits a simple model to improve the initial QP determination. Also, considering the trade-off between similar reconstructed image quality and the fast constringency, some rate control parameters are changed in LCRC, including limitation of GOP's QP variation range. Five sequences are coded at different bit rate to verify the proposed method. Experimental results demonstrate the proposed method outperforms the method using in reference software, which can achieve better performance in terms of increasing coding quality, depressing quality fluctuations and reducing buffer overflow/underflow.
1 citations
••
TL;DR: It is found that users behave differently not only inside different type of channels but also different programs, which has important implications on not only the design and development of P2P streaming system and IPTV systems, but also the existing general TV system.
Abstract: understanding the characteristics of user activities make sense to various application designs, and consequently has impact on business benefit directly. In this paper, we present a study of a peer to peer (P2P) live streaming system based on measurement. We classify those channels into three types and study the statistical characteristics on user behaviors, such as user arriving and leaving, especially we conduct an in-depth analysis of their relations with program time point and time duration. We observe that users behave differently not only inside different type of channels but also different programs. We find that 1) peers watching time duration has no necessary relationship with peer interest; 2) peers watching time duration is enormously affected by program arrangement; 3) peers behavior is substantially active in the first 10min of a program. We also come up with a heuristic model about the number of online peers during a program and the watching time duration of one online user. Our study can be used as a reference for arranging the program resources and channel resources so as to attract more viewers to stay longer in the streaming system. This has important implications on not only the design and development of P2P streaming system and IPTV systems, but also the existing general TV system.
••
29 Oct 2021TL;DR: Wang et al. as discussed by the authors proposed a novel attention mechanism, named category-shared and category-specific feature extraction module (CSS-FEM), which first extracts the category shared features based on the intra-class semantic relationship, then focuses on the discriminative parts.
Abstract: The attention mechanism is one of the most vital branches to solve fine-grained image classification (FGIC) tasks, while most existing attention-based methods only focus on inter-class variance and barely model the intra-class similarity. They perform the classification tasks by enhancing inter-class variance, which narrows down the intra-class similarity indirectly. In this paper, we intend to utilize the intra-class similarity as assistance to improve the classification performance of the obtained attention feature maps. To obtain and utilize the intra-class information, a novel attention mechanism, named category-shared and category-specific feature extraction module (CSS-FEM) is proposed in this paper. CSS-FEM firstly extracts the category-shared features based on the intra-class semantic relationship, then focuses on the discriminative parts. CSS-FEM is assembled by two parts: 1) The category-shared feature extraction module extracts category-shared features that contain high intra-class semantic similarity, to reduce the large intra-class variances. 2) The category-specific feature extraction module performs spatial-attention mechanism in category-shared features to find the discriminative information as category-specific features to decrease the high inter-class similarity. Compared with the state-of-the-art methods, the experimental results on three commonly used FGIC datasets show that the effectiveness and competitiveness of the proposed CSS-FEM. Ablation experiments and visualizations are also provided for further demonstrations.
••
21 Sep 2020TL;DR: This paper presents a solution for API admission control, which comprises of a function entity and a related procedure, by which location information and identity of mobile devices are acquired by network.
Abstract: Quality of Services (QoS) capability exposure enables mobile cellular network operators to enhance their revenue mode. It offers QoS differentiating and prioritizing service for individual subscribers and Internet Service Providers by Application Programming Interfaces (APIs). The admission control for API has two purposes: Firstly, it limits the frequency and times of API calling to protect network resource; Secondly, it ensures every paid calling can satisfy caller's expectation because network status is not stable. This paper presents a solution for API admission control, which comprises of a function entity and a related procedure, by which location information and identity of mobile devices are acquired by network. Network accepts API requests conditionally based on cell status and predefined policies. Some cases are shown on the application of this solution.
••
01 Jan 2017TL;DR: The feasibility of general purpose processor (GPP) platform adoption in baseband processing with optimized virtualization implementation is functionally demonstrated and initially verified in terms of interruption time through prototype development implemented with a commercial LTE protocol stack.
Abstract: This chapter discusses one of the key design principles for 5G systems: “No More Cells” (NMC) [1]. NMC transfers the traditional cell-centric network design to a user-centric design principle. It is pointed out that NMC realization could be facilitated by the Cloud RAN (C-RAN) architecture which pools the processing resources and virtualizes “soft” BBUs and various applications on demand. The major challenges for C-RAN, including the transport networks to connect the resource pool and the remote sites as well as virtualization with potential solutions, are analyzed in detail. Various fronthaul solutions, including Common Public Radio Interface (CPRI) compression, single-fiber bi-direction, as well as wavelength division multiplexing (WDM) technology, are demonstrated and verified through our extensive field trials. In addition, the feasibility of general purpose processor (GPP) platform adoption in baseband processing with optimized virtualization implementation is functionally demonstrated and initially verified in terms of interruption time through prototype development implemented with a commercial LTE protocol stack.
Authors
Showing all 579 results
Name | H-index | Papers | Citations |
---|---|---|---|
Chih-Lin I | 54 | 206 | 14480 |
Yifei Yuan | 49 | 277 | 9760 |
Shuangfeng Han | 29 | 55 | 7360 |
Lei Lei | 27 | 107 | 3715 |
Corbett Rowell | 22 | 63 | 4661 |
Zhikun Xu | 19 | 43 | 3213 |
Zhengang Pan | 16 | 44 | 1886 |
Qi Sun | 13 | 19 | 2346 |
Zhen Cao | 10 | 29 | 332 |
Dawei Ge | 9 | 53 | 254 |
Xueying Hou | 8 | 18 | 274 |
Xuefei Cao | 8 | 15 | 542 |
Yang Li | 8 | 18 | 538 |
Jian Qiu | 7 | 12 | 208 |
Yami Chen | 7 | 21 | 255 |