Oprfid Technology:Internet of things is a new connection mode based on information network and mobile network. It grew up in the soil of the Internet and is based on the boundary expansion and connotation extension of the Internet. At present, the world Internet of things industry chain has formed a closed-loop development, but there are still problems such as application demand, fragmentation of standards and insufficient in-depth application, and the profitability of enterprises needs to be improved. Overview of the development of Internet of things Concept definition of Internet of things Internet of things is a new connection mode based on information and mobile Internet Internet of things is a network that connects daily necessities, facilities, equipment, vehicles and other items on the basis of Internet and communication network. As a broad concept, the Internet of things uses sensors, communication networks, software, control systems and other things to connect and interact with the network and other things, so as to realize the digitization and automation of the real world. The Internet of things has changed the situation that all information in the Internet is obtained and created by people, and all goods need human instructions and operations. In the future, it will have a far-reaching impact on every aspect of production and life. In the future, the scale of the interconnection between things and things in the world will far exceed that of the interconnection between people and people. This exponential growth mainly comes from the diversified connection and independent operation between goods. picture Internet of things, cloud computing and artificial intelligence The Internet of things leads the way, and multiple technologies complement each other and coexist Further disassembly of the IOT architecture can be divided into perception layer, network layer, platform layer, data analysis layer and application layer. The perception layer of the Internet of things generates and collects data, which is successively transported to the edge side and the platform center through 5g communication technology. Edge computing and cloud computing work together to provide computing power support. The data analysis layer mainly uses big data technology to complete data preprocessing and analysis. Artificial intelligence supports optimization algorithms based on cloud computing, big data data data and computing power, and finally feeds the scene application of the Internet of things. The Internet of things is like the main vein of a leaf, continuously importing data nutrients, while artificial intelligence, cloud computing and other technologies are like the side vein, which accept data nourishment and assist in the value release of data elements and the better implementation of Internet of things applications. Various technical elements coexist and depend on each other, running through the flow and application path of data. In the future, the Internet of things, 5g, cloud computing, big data, artificial intelligence and other technologies will be more closely linked, boosting the implementation of Internet of things applications into industrial upgrading and scene intelligence. picture Ecological analysis of Internet of things industry World Internet of things industry chain map With many participants, the role boundaries began to blur, and the number of players on the platform layer increased significantly picture World Internet of things industry chain map Many vertical enterprises integrate Internet of things technology, innovate and launch solutions picture Problems in the world Internet of things industry: data and equipment security The control ability of information security lags behind, and the number of devices attacked and controlled has doubled Internet of things devices are extremely valuable. After being attacked, they may have a large-scale direct impact on the real world, such as traffic paralysis, stagnation of public facilities (water, electricity, gas and heating), remote control, environmental pollution and even casualties. The sensing layer is located at the bottom of the overall architecture of the Internet of things and is the most vulnerable part. In its main application of RFID and WSN technology, WSN routing protocol has inherent defects. When using RFID, readers and electromagnetic waves are easy to copy, and information is easy to be stolen on the way of remote transmission; The network layer is vulnerable to DOS attack, counterfeiting attack, man in the middle attack, etc; The main value of the platform layer is information processing. When the amount of data is too large to be processed in time, it will increase the probability of equipment failure, resulting in security vulnerabilities. At the same time, Internet of things devices are numerous and diverse, and will also become part of the botnet controlled by hackers. Since 2016, the number of smart devices attacked by botnets has been increasing all over the world. Botnets have even been rated as “2017 top ten breakthrough technologies in the world” by MIT Technology Review. At present, the control ability of the world Internet of things for information security lags behind the development speed of the whole Internet of things, and reliable and stable transmission has not been realized, which hinders the overall development rhythm of the Internet of things. picture picture World Internet of things industry: scale and customization Fragmented application scenarios and weak network foundation restrict the large-scale implementation of start-ups If Internet of things enterprises want to realize large-scale application in more vertical industries, they must build a convenient and low-cost Internet of things application ecology, and control the proportion of customized projects or customized proportion of single projects, so as to form economies of scale. According to the survey results of IOT enterprises in iots, about 1 / 3 of IOT projects fail to pass the proof of concept (POC) stage, which is usually due to the high scale cost of the project (respondent data: 32%). At the same time, according to the interview and Research Report released by Oracle, 64% of the pioneers in the field of Internet of things prefer to purchase ready-made solutions (COTS), with shorter project cycle and lower cost. However, at present, start-ups are hampered by both internal and external obstacles in building benchmark cases and improving the reuse rate of project modules. picture Industrial problems of the world Internet of things: Technology Application The sinking of bottom layer technology is insufficient, resulting in insufficient intelligent penetration speed and depth of application layer At present, the accumulation of Internet of things technology in the world is relatively weak, and the limitation of technical level limits the application ability to a great extent. First, the overall underlying technology is not sinking enough to support the data incubation of the platform layer. Finally, it is fed back to the application layer. For example, in terms of chips, most chips have poor anti network attack ability and lack of security of Internet of things devices; At the same time, its internal application processor does not form a unified operating system, which is not open enough; The demand of Internet of things scenario is complex, and the product needs to inherit multiple functions. At present, the chip integration is insufficient, and multi chip cooperation is often required. In terms of application scenarios, in addition to the technical support of network communication and sensing equipment in the field of life, the deepening degree of AI technology also determines the ceiling of scene intelligence. In the production field, due to the particularity of production facilities and environment, whether the equipment can simultaneously have low power consumption and stable transmission has become the key, and the real-time processing and analysis ability has high requirements for WSN, sensors, edge computing and other technologies. In the application of Internet of things in the public domain, the whole process from front-end collection to back-end analysis is faced with the collection, processing and application of massive data, which largely depends on the development of RFID, 5g and other technologies. picture “Object”: data entry Perception layer infrastructure Entering the explosive period of multi technology integration and innovation, accurate and efficient data acquisition is the goal The perception layer is the entrance to the influx of massive “things” data in the Internet of things, and a large number of sensors and identification devices are the physical infrastructure of the perception layer. Based on the above definition of “things” in the Internet of things, the technology of perception layer consists of sensing and identification technology and network communication technology. The network communication technology will be discussed in detail in Chapter 5 and will not be repeated here. Identification technology is a technology that identifies and identifies objects or people through RFID tags, bar codes, two-dimensional codes, biometrics and other means. It has been developed and widely used. The development of sensing technology, especially intelligent sensors, should be guided by the market demand of “intelligence, integration and high performance”, and will be in the development stage of multi technology integration and exploration for a long time. At present, multi-sensor fusion, mems-cmos compatible technology and intelligent sensors integrated with MCU are the technical hotspots and difficulties of the sensing layer of the Internet of things. picture Comparison of perceived layer technology maturity: from the perspective of supply chain There are shortcomings in the iterative cycle of R & D and design and the industrialization process of manufacturing and packaging Affected by market competition, domestic MEMS manufacturers share high supply chain costs and R & D expenses, and cannot transmit them to the downstream. On the contrary, they need to use price as a stepping stone. The pricing of similar products with single specification of domestic manufacturers is about 80-90% of that of the leading manufacturers. MEMS industry chain has high concentration and limited production capacity. Most domestic manufacturers are fabless business model. Compared with the head manufacturers, IDM model does not have scale effect and bargaining advantage in wafer manufacturing and packaging, and the link cost is about 20%. The degree of technical maturity is relatively high, accounting for 2% of the market. In terms of R & D cost allocation, domestic manufacturers expect the product life cycle to be three to five years, and the renewal and iteration of technical process is relatively slow. In addition, in order to explore the market, domestic manufacturers have a high degree of cooperation with terminal customers. The additional software adaptation cost will further increase the R & D and design cost and reduce the profit space, with a floating proportion of about 8% – 10%. It can be seen that domestic MEMS manufacturers still have great room for improvement in the iteration rate of design and R & D, wafer manufacturing and packaging industrialization. picture Development trend of perception layer I: Multi-sensor Fusion It is difficult to remove the false and preserve the true under the multi-source data stack, law extraction and multi parameter balance Multisensor fusion has a double-layer meaning. The surface meaning refers to the physical integration of two into one, integrating multiple sensors in a compact sensor device, typically IMU inertial unit, while the deeper meaning refers to multisensor data fusion. Multisensor data fusion can be likened to a process in which the human brain comprehensively processes the information detected by various functional organs, so as to judge the environment and situation. In consumer electronics, autonomous driving, robotics and other scenarios, through the configuration and management of a large number of and multiple kinds of sensor nodes, multi-source data redundancy and complementarity to make up for the errors and defects of a single sensor signal, solve the problems of data heterogeneity and data conflict through data model and fusion algorithm, and finally give consistent conclusions or provide effective decision support, which is the technical barrier that manufacturers need to establish to overcome the tight encirclement. This requires repeated debugging and trade-offs between sensor combination scheme, cost, computing power, communication and other resource allocation. Taking automatic driving as an example, its sources include radar, infrared, image, etc. by mining the internal relationship between redundant and complementary data, high-precision environmental perception image and positioning results are constructed, so as to guide the vehicle to perform driving tasks such as automatic obstacle avoidance and constant speed cruise. picture Development trend of perception layer II: intelligent sensor Edge computing and edge AI innovate terminals to fill market vacancies Driven by the development of 5g communication and Internet of things, the number of terminals and data continue to accumulate, and the centralized processing architecture has a bottleneck. However, the advantages of distributed local processing in reducing the burden of communication and storage, reducing delay and data security appear. Intelligent sensor integrates the traditional sensing unit with computing unit and AI algorithm, so that the sensor has the ability of information processing other than measurement. Through the decentralization of computing power algorithm from the center to the edge side, the intelligent sensor can independently complete the inspection, diagnosis and calibration of real-time metadata, optimize data quality, independently complete data analysis and execute decision feedback. In emerging application scenarios such as industrial control, medical services and face recognition, there are extreme needs such as real-time response, minimal error and citizen privacy protection. Intelligent sensors with computing power and algorithm support can make up for the limitations of current sensing technology, meet the needs of efficient and secure processing of a large amount of real-time data in the above scenarios, and reduce the cost and resource consumption related to cloud analysis. picture Perception layer development trend 3: high end market The decline of unit price drives down the growth of sales, seizes the high-end market or promotes profit growth The world sensor industry started late, so it lags behind western countries in the medium and high-end sensor market. About 80% of the domestic market depends on imports, and the sensor chip market is weaker, with only 10% of self-produced products. With the increasing competition in the middle and low-end market and the superposition of the trend of sensor integration, the unit price of middle and low-end sensors continues to decline, which not only reduces the market growth rate, but also compresses the profit space of manufacturers, and small and medium-sized manufacturers struggle to survive. The company will try to break through the “low-cost” strategy and focus on the development of high-end sensors in the domestic market. In the segment track, Weier shares in CIS image sensor field, goer micro shares in acoustic sensor field and Minxin shares in pressure sensor field have achieved a certain degree of breakthrough, seizing the original market share of foreign manufacturers in the market. picture picture “Connection”: Data Unicom Analysis on the application status of Internet of things communication technology The infrastructure construction of mobile Internet of things continues to advance, and WLAN technology is eager to make a breakthrough The data perceived by IOT terminal equipment is transmitted through the network. The transmission network carrying IOT equipment mainly includes wired transmission and wireless transmission, of which wireless transmission is the main application of IOT. Wireless transmission technology can be divided into two categories according to transmission distance: one is wireless LAN technology represented by ZigBee, WiFi and Bluetooth. Subject to technical limitations, a single communication mode has limitations in different degrees and directions, and is mostly applied to indoor scenes such as intelligent homes and intelligent buildings with combined schemes; The other is mobile Internet of things technology, that is, Wan communication technology. In May 2020, the Ministry of industry and information technology issued the notice on further promoting the all-round development of mobile Internet of things, aiming to establish a mobile Internet of things integrated ecosystem with the coordinated development of Nb IOT, 4G and 5g, meet the needs of most low-speed scenarios with Nb IOT, meet the needs of medium and constant speed Internet of things and voice with lte-cat1, meet the needs of high-speed and low delay networking with 5g technology, and continue to promote the construction of 5g network infrastructure. picture Internet of things platform: data value incubator Linkage between perception layer and application layer to provide general PAAS services to simplify application development The Internet of things platform can provide general PAAS services across different devices and data sources, play an intermediary role in the whole Internet of things architecture, link all interactions between the perception layer and the application layer – connect and manage the Internet of things device side downward, complete the collection and storage of perception data, and provide application developers and system integrators with unified data interface and common module tools for application development. Based on the realization of “IOT”, the data generated during the frequent interaction between the perception layer and the application layer has the characteristics of large volume, many types and dynamic rolling. As the core hub in the industrial chain, the Internet of things platform is the soil for application integration and data value incubation. In addition to providing data exchange between infrastructure services and supporting equipment, through the processing, analysis and visualization of platform data, Greatly advance the data enabling process, give full play to the scale effect, realize data production and processing, facilitate the rapid application and landing of data, simplify the complexity of Internet of things solutions and reduce the cost of solutions, act as an “acceleration layer” and promote the landing speed and process of each layer in the application scenario. picture The four types of platforms are processed level by level to realize the cumulative promotion of data value from bottom to top The Internet of things platform plays a key role in the architecture of the Internet of things. According to its functions, the Internet of things platform can be divided into four parts: connection management platform, device management platform, application enabling platform and business analysis platform. Among them, the equipment management is basically dominated by communication modules and communication equipment providers, and the network management platform is dominated by telecom equipment manufacturers and operators. Leading enterprises have built open Internet of things platforms, opened important components and continued to improve openness, so as to better aggregate industrial partners and developer resources and empower all walks of life. Horizontal general platform to deepen the application of key vertical fields through partner ecology. Through strategic cooperation, vertical industry giants and Internet enterprises strengthen platform interconnection, improve platform service functions, share industry resources and enhance industry competitiveness. chart
0
