Period
2020.1-2024.12
Introduction
As a bridge between the physical world and the digital world, passive sensing network technology has the characteristics of ultra-low monitoring cost, unlimited duration, diverse sensing capabilities and rapid information transfer, which has an irreplaceable and important position in the research of informationization of modern logistics. This project focuses on the basic theories and key technologies of high performance IoT, researching the transmission and networking of large-scale passive networks, the localization and tracking of massive and highly dynamic logistics nodes, and the intelligent sensing and autonomous decision making in complex scenarios, solving the four major problems of single network structure, inefficient location services, separation of transmission and sensing, and lack of scenario capabilities in large-scale industrial applications of passive sensing networks. . We will explore the multi-hop
network theoretical framework and key technologies of large-scale passive sensing networks, study the high-precision, high-reliability and high-availability positioning and tracking methods, build a new generation of intelligent sensing system with integrated transmission and perception, build an unmanned supermarket intelligent logistics demonstration platform with a daily processing capacity of more than 100,000 passive nodes, break through the barriers to the industrial application of passive sensing networks, and provide a basic platform for seamlessly connecting modern logistics with intelligent manufacturing. Provide a basic platform for seamless connection between modern logistics and intelligent manufacturing, thus providing a comprehensive and reliable network infrastructure support for China's intelligent manufacturing upgrade.
Target
To solve a series of urgent contradictions such as single network structure, inefficient location services, transmission-perception separation, and lack of scenario capability of high performance passive sensing network in large-scale industrial applications. To build a high performance passive sensing network system platform for smart logistics, and provide scientific basis for monitoring and management of the whole logistics cycle. Strive to realize the demonstration platform of unmanned supermarket smart logistics with a daily processing capacity of more than 100,000 passive nodes in demonstration applications, achieving passive multi-hops of more than 10 hops and transmission rates of more than 1 Mbps, and providing theoretical and technical support for the innovative development of IoT and large-scale application of smart logistics in China.
Methodology
I. Multi-protocol Passive Transmission and Multi-hop Networking
In view of the existing logistics-oriented IoT system with a single network structure, inflexible network configuration, and inability to adapt to highly dynamic network changes, this project intends to study low-latency passive multipoint multi-hop transmission as a breakthrough, through the combination of the physical layer and routing layer, innovative passive multi-hop transmission mechanism to achieve network routing and transmission path diversification, greatly enhancing the robustness of the network; at the same time, through the study of At the same time, by studying the efficient cross-protocol passive communication, we realize the coexistence and co-integration of multi-protocols and enhance the compatibility of communication systems; furthermore, we build a passive transmission mechanism based on multiple input and output spatial flows to improve the transmission efficiency of passive networks; at the same time, we establish an asymmetric passive link model to realize the adaptive transmission mechanism of passive network channels.
II. High-capacity, high-precision and high-availability location awareness
In view of the high capacity, high accuracy and round-the-clock service characteristics of large-scale smart logistics, the project intends to first study the location capacity estimation under large-scale passive nodes. Capacity estimation is the basic index of location service and network transmission, and fast and accurate capacity estimation will significantly reduce location service delay and improve network transmission efficiency. Immediately after that, a real-time positioning and tracking system with millimeter-level accuracy and below is constructed for the characteristics of large-scale logistics. Meanwhile, for the weak signal strength and insufficient beacon positioning signal in some special harsh scenarios, the dynamic positioning mechanism with high robustness and the single AP positioning scheme with high availability are studied respectively, so as to provide complete and reliable location services for the whole cycle of smart logistics.
III. The intelligent fusion of sensing in complex scenes
Combined with the application of large-scale IoT in logistics, the project intends to firstly study the intelligent wireless sensing of deviceless objects, by deploying several passive nodes in the environment, the identification of the location, trajectory and identity of people (not carrying any IoT devices) can be realized; secondly, study the intelligent detection and identification of abnormal nodes in logistics applications to realize the rapid classification and monitoring of abnormal vibrations and guarantee the safety of the whole logistics chain nodes;. Again, constructing intelligent behavior recognition based on deep learning, decomposition of multipath effect and deep learning network construction have the opportunity to realize fast and accurate recognition of human behavior; finally, researching the adversarial learning intelligent recognition that can overcome the complexity of multiple scenes and realize highly robust recognition under multiple scenes, multiple targets and multiple movements.
IV. High-performance IoT platform for smart logistics
To establish an efficient smart logistics verification platform for unmanned supermarkets, and to verify and evaluate the correctness, effectiveness, and specific performance indexes of the proposed theories and technical methods. The goal is to build a high-performance passive network system that can handle more than 100,000 passive nodes per day, support more than 10 hops of passive multi-hop transmission, and sense more than 10 categories of complex application scenarios.