“One or more sensors obtain signals, and the system sends and processes the data obtained from the sensors. Sensor nodes can exchange data with each other through wired or wireless networks or gather them to one point.
One or more sensors obtain signals, and the system sends and processes the data obtained from the sensors. Sensor nodes can exchange data with each other through wired or wireless networks or gather them to one point.
Therefore, sensor networks can be classified as a data collection network and a data distribution network. A typical data acquisition system consists of sensors and lines that process actual information. The data distribution network involves communication protocols, network topology, and data transmission and processing methods. The basic network topology used is star, ring, bus and mesh structure, as shown in Figure 1.
Figure 1: Network topology The choice of sensor network topology depends on the application, data processing type and requirements. The need to improve the connectivity between PCs and the real world is gaining momentum. Many sensors and stimuli are used, and it is becoming inevitable to integrate available data to interconnect them.
The number of sensor network nodes continues to increase, and wired connections may not be available because some sensors have to be placed in remote areas. The cost of each node is also falling, and the range of sensor nodes can reach is wider. There are many advancements in low-power wireless technology, which can be used to design more efficient systems.
Compared with wired networks, wireless networks can also provide better scalability, and it is easier to add a new node to the wireless network. Sensor networks need to balance the performance and lifetime of sensor nodes. The wireless node can be dynamically configured to weigh, and can be automatically operated for local control and power management. Many wireless protocols can be used as sensor networks, such as Zigbee, Bluetooth technology, GSM, Wi-Fi, etc. The choice of wireless protocol depends on the needs of sensor network applications.
Wireless sensor network nodes with low power consumption require very little maintenance, and can work for several days, sometimes several months, using the same battery. Therefore, low-power design is very demanding in real-world wireless sensor networks, and there is a basic requirement that sensor nodes have very low power consumption when processing and transmitting sensor data.
The sensor in the sensor node usually measures the slowly changing analog quantity, and the node only needs to be activated for a short period of time to transmit data, and then go to sleep. This means the standby current that the sensor node must have. And, most of the data transmission is from sensor nodes to base stations.
The network architecture and communication protocol must use this asymmetric sensor node to base station sensor communication. Low-power sensor design is crucial. Micro-electromechanical systems (MEMS) with low-power sensors are also critical. Sensor nodes may work in different node environments with densely distributed sources. Sensor nodes also need to use very low power to transmit in noisy environments.
Data collection of sensor nodes Data from the sensor network must be collected and processed in a centralized location. Data processing in sensor networks can be divided into data transmission and data collection.
Data transmission is processed by information routed in the sensor network. This information may also be data obtained from sensors or data requests from other sensors. There are many algorithms for data transmission in sensor networks.
The number of data collection algorithms is particularly large in communication, and the communication is carried out from the beginning to the end of the sensor node. In this case, the trade-off is delay and power consumption. In the case of direct transmission, each node directly sends the collected data to the central network (if the node has GSM capabilities). An operating system will run on the wireless sensor network node. This makes it easier to expand and add more wireless sensors.
The operating system of the sensor network is similar to the embedded operating system, because they are specially developed for a certain application and are not universal. Moreover, due to the low power and low cost requirements of the system, a general-purpose operating system cannot be used. Considering that most sensor networks do not require real-time capabilities, a smaller operating system can be selected, such as TinyOS designed specifically for sensor nodes.
Figure 2 shows a typical sensor network implemented using a GSM (Mobile Communication System) modem. All the sensors here send their data to a centralized server. The server controls each sensor node; however, there is no communication between sensor nodes. The server must interfere with the communication between any two sensor nodes.
Figure 2: Typical implementation of sensor network GSM modem GSM (Mobile Communication System) is one of the existing standards for mobile phones. Although initially only used for voice communication, it has achieved data communication capabilities through GPRS (General Packet Radio Service Technology) and EDGE (Enhanced Data Rate GSM Evolution Technology). GSM modem is a type of modem that accepts SIM (Subscriber Identity Module) and works like a mobile phone. The GSM modem can be used in low power consumption mode, or it can be turned off when not in use.
The cost of data transmission via the GSM network is also rapidly decreasing. And, the GSM modem can be easily connected to the microcontroller using standard communication protocols. There are more and more cases in which mobile phones process voice and data. Most GSM modems have a TCP/IP stack, which can be used to transmit data in a secure channel. This also reduces the complexity of application development, allowing the microcontroller to simply connect to the GSM modem.
The GSM modem can connect to any IP (Internet Protocol) address and transmit data. Multiple modems can send data to a single IP address, and all data can be centralized and processed anywhere in the world. The user can remotely dynamically configure each modem based on the data sent over the network. In some networks, a single node has GSM capabilities. Other nodes will send the data to this special node, and then to the central server. This can reduce the cost of the overall system, but if the nodes are not clustered together, this cannot be achieved.
The GSM modem can also provide instant alerts, and can transmit data to a different high-priority IP address via SMS (Short Message Service) or depending on the situation. These features can be used for fault tolerance and redundancy checking.
Let us look at an example of a GSM modem (SIM300), which communicates with a microcontroller via a serial port. The modem has a standard command set, AT command. These instructions control the operation between the modem and the microcontroller. The microcontroller sends these commands at a specific baud rate through the UART (Universal Asynchronous Receiver Transmitter) interface.
The modem can be configured using specific instructions, and then the data sent through the serial port can be transmitted to the central server. Therefore, the interface with the GSM modem simplifies the data collection and processing in the sensor network.
The complete system implementation Figure 3: The block diagram of the sensor node using PSoC Figure 4: The internal design of the sensor node using PSoC Cypress’ PSoC mixed-signal programmable controller can provide analog and digital systems, which can be configured to each The functions required by the node. This eliminates the need for special hardware for different types of sensors. This system-on-chip (SoC) microcontroller can also handle all the tasks required by the sensor node with a single chip, including ADC, DAC, PGA, comparator, amplifier, digital filter function, DMA, LCD controller ,etc.
The system design using SoC can be quickly designed and frequently changed to match the customization of sensor nodes. PSoC sleep mode has very low power consumption and is very suitable for this type of application.