Why the Internet of Things is developing so slowly

The term "Internet of things" comes from the definition of the European Union, which is the things, the most well known English expression in the United States, which is generally called cyber-physical Bae, or CPS, which is more academic. The research direction of the Internet of Things is a system direction.

Why is it developing slowly? Mainly because sensor receptacle basic research is still not up to now.

This direction has always been my track and attention, sensor development and data processing should be regarded as my old line. To know what the internet of Things is, and why it is so slow, it is necessary to understand its technological evolution and development process.

The internet of things is not a very new technology, it is the essence of the last century academia began to rise sensor networks, self-organization and Multi-hop network (wireless sensor receptacle, hoc receptacle, wireless multi-hop). The application of RFID in intelligent Logistics is only the most basic application scenario, the current research is far more complex than this.

With the rapid development of microelectronic technology, especially MEMS devices, wireless RF modules have been highly integrated, miniaturized and more stable, and the use of sensor networks to achieve distributed data acquisition has a realistic demand, especially in some environments with high cost of artificial collection (such as Forest, fire scene, High-voltage transmission lines , Battlefield, space, and more difficult to collect, such as tracking wildlife, air detection, etc. It is the main direction of the development of IoT technology that the application in the scene of the high cost of manual detection.

However, back to the beginning, to achieve a distributed sensor collection network that sensor receptacle, looks simple, but in fact it is not so easy. Traditional wireless networks, such as honeycomb, WLAN, etc., optimize routing and Bandwidth allocation strategy to achieve better communication quality in high speed mobile. But there are problems with the corresponding solution on the sensor network:

1, Wireless sensor network sensor nodes are often not directly to provide energy to the scene, that is, frequently can only use batteries or solar power supply, therefore, the focus of sensor network design is to ensure the full function of the maximum energy saving and increase the network life cycle;

2, wireless module launch distance and transmit power is positively correlated, when the communication two nodes between the distance is greater than the maximum launch distance what to do?

3, when the wireless node without electricity how to solve?

It should be said that the energy supply problem is the most important problem in the research of wireless sensor networks, and the most classical theory in this field is to solve these three basic problems. Although now the academic community began to play big data, get a few sensor, set a network mining point data, to a large data analysis looks very tall, but if these three core issues do not solve, then the internet of Things is far from commercialization.

Back to these three questions

First routing strategy issues academia has studied a lot, such as flood (flooding), SPIN, directional diffusion and so on.

Second, when the communication distance between two wireless nodes is too far, or this communication distance will consume a lot of energy how to do? To solve this problem, there are multi-hop networks (multi-hop) and self-organizing networks (hoc), that is, to minimize the wireless emission distance of a node and thereby reduce energy consumption, the node sends the packet to the nearest node before sending it to another Communication in the form of this multi-node relay and baton, thus minimizing the energy consumption of communications, which is multi-hop; Since the organization is more complex, in the hoc, each node is both a network node and a router, each node can be as a gateway and extranet communication, but also can forward packets, Thus, wireless sensor networks have more flexibility and optimization of energy consumption in data transmission. But networking and multi-hop strategies are a challenge, and how to choose the optimal route and hop approach is a problem.

Third question, how does the node supply? Solar power is the first thought, but solar panels are limited in size. So then there is a temperature difference for energy, and so the wonderful way. The easiest way to do this is to change the battery. The year before last, U Washington's Gollakota group realized backscatter should be the general direction of the future, this group realizes a kind of passive sensor network, namely does not need the power supply, through the reverse scatter link modulation space electromagnetic signal carries on the energy and the communication, thus thoroughly solves the power supply question. This should be a breakthrough, but there is still work to be done in business.

No matter how the concept of the current Internet of things hype how fire, how fancy, its essence is the laboratory of these things, RFID this thing in the Intelligent Logistics application is only the most basic scene, and academic research has experienced three stages, the first phase is based on MEMS technology, the miniaturization of nodes, The most typical representative is the smart dust intelligent dust of the Berkeley David Culler Group, which communicates with scattering light between nodes, and the sensor nodes are only 1mm³ size; the second stage is to solve the problem of networking, that is, the self-organizing and multi-hop networks of hoc and multi-hop modes. To the third stage, academia began to look for apps, start with big data, and play something that looked "fun." In fact, the large data concept was first proposed, but also because of the rise of the Internet of things, sensor access to the network, greatly increased the amount of data can be mined, the data on the network includes not only the social network of this data from users, but also from the physical world of data.

Many of Daniel in this direction have turned to mobile computing or into industry. As the direction of the sensor network begins to fade, the domestic internet industry is starting to cool down, the root cause is:

1, supply the bottleneck is too large, used smart phone should be deeply aware of the current lithium battery power supply limited;

2, sensor miniaturization encountered bottlenecks, although the MEMS to the sensor miniaturization provides an effective technical thinking, but this also relies on the process of upgrading, digital devices can use a very small process, but analog devices, especially CMOS, miniaturization still have difficulties;

3, Wireless sensor network, as well as the emergence of self-organizing and multi-hop network of typical systems, such as smart dust,m³ is actually more suitable for special applications, such as military and fire rescue, space and so on, commercial applications need to dig more application scenarios, a few days ago Baidu Investment Indoor Atlas uses the geomagnetic monitoring method to do indoor positioning, which is a good scene, but the current mainstream indoor positioning technology does not actually have a great development, has been a technical bottleneck.

In short, the Internet of Things or CPS, this direction in the future undoubtedly has broad prospects for development, but the current basic research and related technology has yet to be developed, so it seems that slow development, in fact, is stagnant, are waiting for a subversive application can let sensor receptacle Zha.