The Internet of things (IoT) has been a promising communication paradigm that involves sensors, microcontrollers, and transceivers for an efficient communication and computation system. The infrastructure and the applications shall enable and improve the intelligent management of our city service, workspace, and daily life. This paper aims at the future 6G vision of IoT, and discusses the convergence with radio over fiber (RoF) system. Comparing with the IoT services included in the 5G deployment, 6G IoT exploits high-density heterogeneous devices involving extremely high capacity, supporting much more robust system architecture and artificial intelligence (AI) based smart algorithms. The RoF is one of the most promising enablers for the outstanding characters of flexibility and efficiency of 6G IoT systems. The paper first introduces the IoT envolution roadmap from 5G towards 6G and the potency of optic fiber and RoF technologies. Then we present the rapidly expanding RoF market and compatible technologies related to IoT-RoF convergence with the discussion on the current outstanding works in multiple dimensions. Finally, we investigate the challenges ahead for the future RoF supported 6G IoT system and the emerging technology solutions.

With the evolution of the 5th generation (5G) and the advent of the 6th generation (6G) mobile communication technologies, the consumer Internet of Things (CIoT) has emerged as a critical enabler for ubiquitous information exchange across diverse applications. The integration of sensing and communication (ISAC) offers significant potential to enhance communication performance by leveraging multimodal heterogeneous data collected through sensors in CIoT systems. This paper presents a comprehensive discussion on multimodal heterogeneous data-based environment sensing and its integration with communication in CIoT systems. We provide an overview of multimodal ISAC in CIoT, addressing the need for advanced environment understanding and optimization of network performance. Furthermore, we introduce a novel ISAC framework that integrates semantic environmental mapping, beam management, and resource allocation to improve communication efficiency. The concept of a semantic environment map (SEM) is proposed, which fuses multimodal sensing data for precise environment recognition and adaptive communication strategies. Additionally, we identify key future trends and challenges, including performance evaluation, multimodal data transmission, and security, and then propose a roadmap for future research. This paper aims to bridge the gap between multimodal sensing and communication in CIoT systems, offering insights into integrating advanced technologies such as artificial intelligence (AI) and 6G for smarter, more connected consumer applications.

In this paper, we comprehensively investigate robust beamforming schemes for reconfigurable intelligent surface (RIS)-assisted multi-cell downlink transmissions in the presence of channel estimation error (CEE), where an RIS is deployed at the joint boundary of multiple cells to assist downlink transmissions to cell-edge users. Specifically, we study a unified robust beamforming framework to maximize the minimum achievable rate or the coverage probability of cell-edge users with a bounded or a statistical CEE model. The bounded and statistical CEE models characterize the absolute robustness and statistical robustness demands of RIS-assisted cell-edge users. In particular, we reformulate the non-convex robust beamforming problem into two sub-problems, i.e., the optimal transmit beamforming at the base station (BS) and the optimal phase shifts at the RIS, which are alternately optimized. Moreover, we apply S-Procedure and Bernstein-type inequality to transform non-convex bounded and statistical CEE constraints into convex forms. Consequently, the BS transmit beamforming sub-problem and the RIS phase shifts sub-problem can be solved by a novel penalty-based method. Finally, the two sub-problems are iteratively optimized via a bisection-based strategy to obtain the joint robust beamforming. Simulation results demonstrate that the proposed robust beamforming scheme can converge rapidly and outperform reference schemes in the presence of CEE. Moreover, our proposed penalty-based method achieves better performance than that of the conventional semidefinite relaxation method.

The digitization and intellectualization have been envisioned as the fundamental basis for future Industrial Internet of Things, which integrates sensor technology, industrial control technology, communication technology, and artificial intelligence (AI). Specifically, the collaboration among these above techniques is crucial for the successful implementation of intelligent applications. This article develops an end-to-end visual control framework to accomplish multicrane collaborative sorting in wireless time sensitive networking (TSN) networks. The design primarily incorporates field devices, data transmission, AI, and industrial control. An advanced binocular stereo visual recognition model based on deep learning is investigated to accurately obtain the world coordinates and types. A cooperative control scheduling model that combines a scheduling strategy with an anti-collision strategy is presented to effectively control multiple cranes for sorting tasks. The device data and commands are transmitted through industrial 5G-TSN integrated network for ultrareliable, low-latency, and deterministic transmission. The proposed visual sorting system is further validated through the establishment of an experimental prototype, demonstrating its exceptional real-time performance while enabling flexible intelligent manufacturing.

Person re-identification aims to retrieve the pedestrian across different cameras. It is still a challenging task for the intelligent visual surveillance system because of similar appearances, camera shooting angles, scene illumination, and pedestrian pose. In this paper, we propose a novel two-stream network named Spatial Segmentation Network that learns both the global and local features in a unified framework for nonaligned person re-identification. One stream focuses on spatial feature learning using global adaptive average pooling in deep convolutional neural networks. Another stream is utilized to learn the fine local features by adopting horizontal average pooling without division that depends on the pose predictor. To assess the importance ranking of all features, we also obtain the performance of every part feature and global features. Our evaluation of the proposed method on Market-1501 acquires 94.51% Rank-1 and 90.78% mAP, that on DukeMTMC-re-ID acquires 87.52% Rank-1 and 84.82% mAP, and that on CHUK03-detected acquires 69.71% Rank-1 and 71.67% mAP; these findings verify the state-of-art performance of the proposed method.