China 
Africa 
Explore chapters and articles related to this topic
Radio Wave Propagation
Published in Jerry C. Whitaker, The RF Transmission Systems Handbook, 2017
Other HF propagation anomalies are difficult to predict and occur without warning. Many of these anomalies are the result of solar flares. A shortwave fade (SWF) is a sudden and complete absorption of HF radio waves in the D region of the ionosphere. It occurs on the sunlit side of the Earth and is caused by ultraviolet and X-ray emissions from a solar flare. The fade occurs approximately 8 minutes after the solar event and can last from a few minutes to a few hours. Protons entering the ionosphere near the magnetic poles can cause a complete loss of HF propagation in these regions. This effect may occur several hours after the flare. Ionospheric storms are another potential effect that can drastically alter the expected MUF and may occur one to two days after the solar event. Sporadic ionization in the E layer, known as sporadic E, may occur at any time and is not necessarily solar related. This effect can isolate the F layer, altering the path characteristics such that communication is disrupted or perhaps enhanced. Even VHF frequencies may be affected by sporadic E conditions causing a significant increase in effective range.
Design and Implementation of a Truly-Wireless Real-Time Sensor/Actuator Interface for Discrete Manufacturing Automation
Published in Richard Zurawski, Networked Embedded Systems, 2017
Guntram Scheible, Dacfey Dzung, Jan Endresen, Jan-Erik Frey
The ISM Band also has a good balance between possible disturbances (more likely at lower frequencies, see Figure 28.9) and a good range. Effective range is lower at higher frequencies with a given transmit power. Electromagnetic interference is a concern, however measurements in different environments (see Section 28.4.3) have shown that typical industrial electromagnetic noise is significantly reduced above 1 GHz, with only arc welding providing broad band noise up to 1.5 GHz.
Spatial Statistical Downscaling for Constructing High-Resolution Nature Runs in Global Observing System Simulation Experiments
Published in Technometrics, 2019
Pulong Ma, Emily L. Kang, Amy J. Braverman, Hai M. Nguyen
The basic idea is as follows. First, we define a finite set of locations spread out across the entire domain of interest. This set is referred to as a set of candidate centers, denoted by . Suppose that we pre-specify a set of basis functions with centers and bandwidths as the initial sets of centers and corresponding bandwidths, where the superscript stands for the iteration of the algorithm. A possible choice is a small set of equally spaced basis functions over the domain. The centers and bandwidths of new basis functions are added automatically at each iteration of the forward algorithm. At the beginning of the ith iteration (), the current set of basis functions is used to fit data with the FRK model, which gives the estimated trend and the estimated random effects, . We use the pseudo-residuals defined as , where is the observation at location and denotes the matrix resulting from the current set of basis functions, to assess where basis functions should be added to improve the fit. As in classical geostatistics (Cressie 1993), these pseudo-residuals can be used to carry out the local semivariogram analysis for each observation location. The empirical-local-mean-squared error (ELMSE) is defined for each point in the set of candidate centers : , where denotes a local neighborhood surrounding the location and is chosen based on the effective range obtained from the semivariogram analysis. The effective range is defined as the distance at which the semivariogram value achieves 95% of the sill. New basis functions are placed where the ELMSE is large, and the bandwidths of these basis functions are chosen corresponding to the effective range. We also impose a separation criterion to avoid substantial overlap among the supports of the newly added basis functions at each iteration. We repeat these steps until the upper bound of the number of basis functions, , is reached, or the ELMSE does not change substantially. In practice, we recommend using both together as a stopping criterion, and can be chosen as large as computational constraints allow. When computational limits are less constrained, users can set larger, or even let the stopping criterion depend solely on the change in ELMSE.