China
Africa
Explore chapters and articles related to this topic
Why digital?
Published in John Watkinson, The Art of Digital Audio, 2013
The capacity of memories and storage media is measured in bytes, but to avoid large numbers, kilobytes, megabytes and gigabytes are often used. As memory addresses are themselves binary numbers, the wordlength limits the address range. The range is found by raising two to the power of the wordlength. Thus a four-bit word has sixteen combinations, and could address a memory having sixteen locations. A ten-bit word has 1024 combinations, which is close to one thousand. In digital terminology, 1K = 1024, so a kilobyte of memory contains 1024 bytes. A megabyte (1 MB) contains 1024 kilobytes and a gigabyte contains 1024 megabytes.
Communications, Signal Processing, and Data Handling
Published in Douglas O. J. deSá, Instrumentation Fundamentals for Process Control, 2019
MAP uses AM/PSK at 10 Mbit/s, two-channel operation on a directional bus with remodulator, 75-Ω trunk and drop cables. Note: A bit (binary digit) is a single character that could be either 0 or 1, and M = mega or 1 million; hence, Mbit is 1 million bits, also abbreviated to Mb. A byte is a collection of bits, most commonly 6 or 8, but the number depends on the transmission packaging, etc. Bytes are usually measured in millions (mega) and abbreviated to MB, or Mbyte. Transmission is usually timed over a one-second period; therefore, abbreviations are Mbit/s, Mbps, or Mbyte/s.
Functions and performance of sensors for slope monitoring in opencast coal mines – laboratory experimentation
Published in Petroleum Science and Technology, 2023
Sathish Kumar Mittapally, Ram Chandar Karra
ESP8266 is a microchip developed by “Espressif” that is used as a Wi-Fi module in IoT projects. However, this ESP8266 chip can also be programmed to work as a microcontroller. It uses System on Chip (SOC) that does not require additional microcontrollers like Arduino boards to draw inputs (Bhardwaj, 2021). It is proficient in hosting an application or offloading all Wi-Fi networking functions to another application processor and is already pre-programmed with a set of AT commands (Sruthy et al. 2020). It is easy to run and program through a USB port with Arduino IDE (Integrated Development Environment) software. The NodeMCU ESP8266 (Figure 6) on the board can power up the node, or the 3.3 V pin can also be used; GND (ground) pins are provided for grounding and Vin pins for the external power supply. The EN and RST buttons are used to reset and control the microcontroller. AO analog pin is used to measure the range of 0 to 3.3 V Analog signals, and GPIO (General-purpose Input Output) pins are used for the input-output purpose. SPI (Serial Peripheral Interface) pins are used for SPI communications or short-distance communications. It has a 4 MB flash memory, 128 KB RAM, and 80 MHz clock speed.
Evaluation of nine machine learning methods for estimating daily land surface radiation budget from MODIS satellite data
Published in International Journal of Digital Earth, 2022
Shaopeng Li, Bo Jiang, Shunlin Liang, Jianghai Peng, Hui Liang, Jiakun Han, Xiuwan Yin, Yunjun Yao, Xiaotong Zhang, Jie Cheng, Xiang Zhao, Qiang Liu, Kun Jia
Among the nine ML models, the implementation cost for the ResNet model was the highest, with the most training time (∼24.83 h, 89,389 s) spent and the largest memory (∼5 GB, 5,280 MB) needed. For the other eight models, the time needed for model training differed greatly, with the top three being the SVM (72,723 s, ∼ 20.20 h), the DBN (67,725 s, ∼18.81 h), and the RBF (24,042 s, ∼ 6.67 h), and the least time needed was the XGBoost (28 s). The computer memory needed was similar for these models, ranging from approximately 800–1,200 MB except for the DBN, which had the least memory requirements of 160 MB. Relatively speaking, the implementation costs for the three tree methods (RF, AdaBoost and XGBoost) were the smallest, which is one of the most important reasons why the tree methods are so popular. Moreover, SVM is not suitable for large samples (Mountrakis, Im, and Ogole 2011) because it took more than 3 h for training when only 50% of the total training samples were used, which is due to the complex mapping process in SVM of the high dimensional space (Mountrakis, Im, and Ogole 2011). Therefore, the results demonstrate that the implementation costs for the DL methods are not always expensive except for ResNet, and a trade-off must be made between the running costs and model performance when selecting ML methods.
Arrangement and Accomplishment of Interconnected Networks with Virtual Reality
Published in IETE Journal of Research, 2022
A University of Southwestern Louisiana's Mechatronics and Automated Laboratories (RAL) uses an experimental multi-user connected VE to teleoperate a mobility robot. Computer work has taken place in the Virtually Realities and Media Centre. These helmet or operational platforms are used by testbed elements: I a portable robotics platform having a common storage multi-processor management platform, such as the Nomad 200. I Silicon Graphics workstation O2, running IRIX 6.3; (ii) numerous Gateway 2000 PCs (Pentium II 333 MHz, 384 MB RAM), running Metropolitan Window 95/NT/98; (iii) multiple Gateway 2000 PCs (Pentium II 333 MHz, 384 MB RAM), running Microsoft Windows 95/NT/98. Numerous Silicone Graphic desktops and Pentium-based PCs attached to a 10 Mb/s Gigabit LAN are also employed in a multi-user setup. To connect without additional machines on your LAN, the robots employ a wirelessly Internet network via RangeLAN2/Access Points. VRML, DIS, and Java were used to create the VE.