A memory mechanism has attracted growing popularity in tracking tasks due to the ability of learning long-term-dependent information. However, it is very challenging for existing memory modules to provide the intrinsic attribute information of the target to the tracker in complex scenes. In this article, by considering the biological visual memory mechanisms, we propose the novel online tracking method via an attention-driven memory network, which can mine discriminative memory information and enhance the robustness and reliability of the tracker. First, to reinforce effectiveness of memory content, we design a novel attention-driven memory network. In the network, the long memory module gains property-level memory information by focusing on the state of the target at both the channel and spatial levels. Meanwhile, in reciprocity, we add a short-term memory module to maintain good adaptability when confronting drastic deformation of the target. The attention-driven memory network can adaptively adjust the contribution of short-term and long-term memories to tracking results under the weighted gradient harmonized loss. On this basis, to avoid model performance degradation, an online memory updater (MU) is further proposed. It is designed to mining for target information in tracking results through the Mixer layer and the online head network together. By evaluating the confidence of the tracking results, the memory updater can accurately judge the time of updating the model, which guarantees the effectiveness of online memory updates. Finally, the proposed method performs favorably and has been extensively validated on several benchmark datasets, including object tracking benchmark-50/100 (OTB-50/100), temple color-128 (TC-128), unmanned aerial vehicles-123 (UAV-123), generic object tracking -10k (GOT-10k), visual object tracking-2016 (VOT-2016), and VOT-2018 against several advanced methods.

A user-centric no-cell (UCNC) framework has been proposed which enables different level coordination over a large number of TRPs. This paper presents some key technologies in UCNC framework with the focus on PoC of UL grant- free transmissions in UCNC architecture. Massive connectivity support is one of the three important scenarios in 5G networks. The challenges to support massive connectivity lie in the cost of signaling overhead and transmission latency. An uplink grant- free transmission based on SCMA (sparse code multiple access) design is proposed. The proposed scheme can provide different levels of overloading to efficiently meet the massive connectiv- ity requirements. Extensive lab tests have been conducted to verify the performance matrices. Taken LTE as a baseline for comparison purpose, it is shown that the signaling overhead by using uplink grant-free transmission can be reduced by around 80% and the user plane transmission latency in ECO state can be reduced by around 93%. Moreover, massive connectivity is efficiently supported with SCMA scheme that yields around 230% gain over OFDMA in terms of supported active users.

A 5G mobile network has evolved from a cell centric radio access network to a user centric one. A user-centric no-cell (UCNC) framework has been proposed, which enables physical layer coordination over a large number of transmission and reception points for each mobile user to always experience cell-center-like communications. This paper presents some key technologies in the UCNC framework. The focus is on proof of concept of UL grant-free transmissions in the UCNC architecture. Massive connectivity supporting a huge number of devices is one of the three important scenarios in 5G networks. The challenges to support massive connectivity lie in the cost of signaling overhead and transmission latency. An uplink grant-free transmission based on sparse code multiple access (SCMA) design is proposed. The proposed scheme can provide different levels of overloading to efficiently meet the massive connectivity requirements. Grant-free transmission conducted in RRC connected state can reduce signaling overhead, while in energy conserved operation (ECO) state, it can significantly reduce the user plane transmission latency. Extensive laboratory testing has been conducted in a collaborative project of Huawei and Telefonica to verify the performance matrices. Taking LTE as a baseline for comparison purposes, it is shown that the signaling overhead by using uplink grant-free transmission can be reduced by around 80% and the user plane transmission latency in ECO state can be reduced by around 93%. Moreover, massive connectivity is efficiently supported with the SCMA scheme that yields around 230% gain over OFDMA in terms of supported active users.