Abstract: Described is a method of reducing uplink inter-cell interference in a cellular communications network. The method comprises identifying in a cell any user equipments (UEs) which are susceptible to interference (ISUs) from another UE, particularly another UE in a neighboring cell. The method includes identifying in said cell any UEs causing interference (ICUs) to at least one neighboring cell and, more particularly, causing interference to a UE in a neighboring cell. The identified ISUs and ICUs are combined into a priority list for frequency resource reservation or allocation in said cell. The method may include limiting the identification of ISUs and ICUs to a selected number of UEs within the cell.

Abstract: Described is a method and apparatus for processing an uplink (UL) signal at a Physical Uplink Control Channel (PUCCH) in a wireless communication system to determine a discontinuous transmission (DTX) state. The method comprises receiving a UL channel signal at a PUCCH receiver apparatus and, after resource element (RE) demapping of said received UL channel signal in said PUCCH receiver apparatus, normalizing a signal power of at least one signal element or resource. The normalized power is compared to a selected, calculated or predetermined threshold and, based on said comparison, a determination is made on whether or not a DTX state has occurred.

Abstract: A deep neural network models semiconductor devices. Measurements of test transistors are gathered into training data including gate and drain voltages and transistor width and length, and target data such as the drain current measured under the input conditions. The training data is converted by an input pre-processor that can apply logarithms of the inputs or perform a Principal Component Analysis (PCA). Rather than use measured drain current as the target when training the deep neural network, a target transformer transforms the drain current into a transformed drain current, such as a derivative of the drain current with respect to gate or drain voltages, or a logarithm of the derivative. Weights in the deep neural network are adjusted during training by comparing the deep neural network's output to the transformed drain current and generating a loss function that is minimized over the training data.

Abstract: A switch-less transceiver has parallel LC resonances that act as OFF switches and series LC resonances act as ON switches when resonating at the transmit (TX) or receive (RX) frequency. When the transmitter is disabled, no current flows through series LC filters. Instead, series impedances to ground provide an RF ground to the transmitter output node. A TX inductor between the transmitter output node and the antenna is in parallel with a TX blocking capacitor to ground, which together form a parallel resonance to ground that has a high impedance when resonating at the receiver frequency. This high impedance acts as an OFF switch to block antenna signals from entering the transmitter. The two paths are in parallel, presenting a high impedance to the antenna and forming an OFF switch when the receiver is disabled.

Abstract: A method of generating an AR user manual in an electronic 3D viewing environment, comprising: recording a moving trajectory of the 3D viewing environment's optical sensor; receiving a landmark location information; executing an iterative target object pose estimation comprising: estimating an estimated target object pose from each of the optical sensor poses in the recorded moving trajectory and the landmark location; calculating an estimation error from a 3D model being arranged in the estimated target object pose and projected onto the target object in the real-world scene; calculating a mean of the estimation errors; and reiterating the iterative target object pose estimation to optimize the estimated target object pose for a minimum mean estimation error; if the minimum mean estimation error is within a predefined estimation error threshold, then rendering the AR user manual onto the target object according to the optimized estimated target object pose.

Abstract: A method for applying image encoding recognition in the execution of natural language processing (NLP) tasks, comprising the processing steps as follows. A sentence from a textual source is extracted by an NLP-based feature extractor. A word vector is generated in response to the sentence by the NLP-based feature extractor. The word vector is converted into a feature vector {right arrow over (b)} by the NLP-based feature extractor, in which the feature vector {right arrow over (b)} satisfies {right arrow over (b)}?m and the parameter m is a positive integer. The feature vector is transformed into an image set having a plurality of two-dimensional images by a transformer. The image set is fed to a neural network to execute image recognition by a processor, so as to analyze the sentence.

Abstract: Provided herein is a method for measuring a concentration of a metal ion in an electrodeposition solution. The method of the present disclosure can substantially reduce the interference of organic additives and different electrode conditions on voltammetric metal ion concentration measurements.

Abstract: An assembly for power converting includes a circuit board, a power switching circuit mounted on the circuit board, an inductor coil that includes a winding and two ends, a magnetic core that is surrounded by the winding of the inductor coil, and a magnetic mixture that encapsulates the circuit board, the power switching circuit, the inductor coil and the magnetic core. The winding of the inductor coil is stacked above the power switching circuit and is sufficiently large to fill up a size of the assembly.

Abstract: A bit-depth optimization engine reduces the hardware cost of a neural network. When training data is applied to a neural network during training routines, accuracy cost and hardware costs are generated. A hardware complexity cost generator generates costs for weights near bit-depth steps where the number of binary bits required to represent a weight decreases, such as from 2N to 2N?1, where one less binary bit is required. Gradients are generated from costs for each weight, and weights near bit-depth steps are easily selected since they have a large gradient, while weights far away from a bit-depth step have near-zero gradients. The selected weights are reduced during optimization. Over many cycles of optimization, a low-bit-depth neural network is generated that uses fewer binary bits per weight, resulting in lower hardware costs when the low-bit-depth neural network is manufactured on an Application-Specific Integrated Circuit (ASIC).

Abstract: A method and apparatus for transporting packets over a wide area network (WAN) in a communications network are provided where the WAN comprises a plurality of interconnected nodes including at least a first communication node, a second communication node and a WAN controller node. The method comprises establishing virtual private network (VPN) tunnel connections on communication links between some or all of the communication nodes comprising the WAN, using a non-stream-oriented transport layer protocol to establish a non-stream-oriented association for each VPN tunnel connection, and, on receiving a packet connection from a source device at said first communication node, encapsulating packets from said packet connection into one or more non-stream-oriented associations between the first communication node and the second communication node to thereby transport said packets from the source device to the second communication node.

Abstract: An Electrical Protective Device for Low-Voltage DC networks (EPDL) has a controller that on start-up first closes a mechanical relay in a ground line, waits a first delay, then closes a bipolar transistor to allow current to flow through a positive supply line, then turns on a Metal-Oxide-Semiconductor Field-Effect Transistor (MOSFET) connected in parallel with the bipolar transistor once an output voltage reaches a minimum target. The controller reverses the start-up order to open the MOSFET, bipolar transistor, and finally the relay when an over-voltage, input-power, or integrated-current fault is detected. The output current is integrated over time and compared to a threshold, allowing for capacitor charging or other temporary over-loads. When the output voltage dips below the minimum target, the MOSFET is turned off until the voltage recovers. If the voltage does not recover within a time period, the bipolar transistor and then the relay are turned off.

Abstract: A method of tracking a point of interest (POI) in an electronic three-dimensional (3D) viewing environment, comprising: capturing via an optical sensor and recording motions of an onsite user, wherein each motion comprises an image surrounding and a pose of the onsite user; displaying a snapshot to a remote user, wherein the snapshot is one of the recorded motions; receiving a POI indicator in the snapshot from the second user; estimating a 3D position of the POI in the electronic 3D viewing environment using the POI indicator data, a moving trajectory from each of the recorded motions to the snapshot, and an estimation of distance between the optical sensor to the POI center; and rendering and superimposing the POI indicator in the electronic 3D viewing environment to be displayed to the onsite user using the estimated 3D position of the POI, the moving trajectory, and the recorded motions.

Abstract: Described is a controller for an AC to DC or a DC to AC multi-phase power converter of a type having N power converter phases, where N is greater or equal to 2. The controller comprises a control module configured to change or vary a phase shift angle of the input current or output current for each of the N power converter phases such that an average phase shift value for each of said N power converter phases over a control module AC line cycle is about, near or substantially the same value. In an embodiment of an AC/DC or a DC/AC multi-phase power converter of a type having N power converter phases arranged in parallel, an advantage of arranging the average phase shift value for each power converter phase to be substantially equal or about equal over an AC line cycle is that it reduces or eliminates any imbalances in the input currents or output currents of the N power converter phases.

Abstract: A method and system for tracking a position and orientation of a target object that moves along a two-dimensional plane, a camera to generate video frames of the target object, and a processor configured to correct perspective scaling of a current one of the video frames; estimate a current location and angle of the target object using a motion model or its previous location/angle; cut out a region of interest from the current video frame around the estimated current location; resize the region of interest to shrink its size by a predetermined ratio M (width=width/M, height=height/M) or equivalently reducing the image resolution of the region of interest by the ratio M (resolution=resolution/M); track a new location of the target object by template matching the current region of interest with a previously stored template; and either conclude the tracked target object location and angle are accurate and updating the template if a stopping criterion is reached with indication of successful tracking, or proceed

Abstract: A rotational geometric phase hologram has geometric phase optical elements (GPOEs) serially cascaded along a common optical axis to form a GPOE cascade used for receiving a linearly-polarized light beam and generating output light beams at an exit surface of the last GPOE. Interference occurred in the output light beams creates a polarization interference pattern on the exit surface. A photoalignment substrate, when positioned in close proximity to the exit surface, records the pattern. Advantageously, each GPOE is rotatable about the common optical axis. Respective rotation angles of the GPOEs are determined according to a spatially-varying linear polarization orientation distribution selected to be generated for the polarization interference pattern.

Abstract: An image volume formed by plural anatomical images each having plural image slices of different imaging modalities is segmented by a 2D convolutional neural network (CNN). An individual anatomical image is preprocessed to form a mixed-context image by incorporating selected image slices from two adjacent anatomical images without any estimated image slice. The 2D CNN utilizes side information on multi-modal context and 3D spatial context to enhance segmentation accuracy while avoiding segmentation performance degradation due to artifacts in the estimated image slice. The 2D CNN is realized by a BASKET-NET model having plural levels from a highest level to a lowest level. The number of channels in most multi-channel feature maps of a level decreases monotonically from the highest level to the lowest level, allowing the highest level to be rich in low-level feature details for assisting finer segmentation of the individual anatomical image.

Abstract: A DC-DC power converter has an auxiliary tank cascaded to share an efficiency tank's inductor, capacitor, and transformer. Switching transistors pump the auxiliary tank at startup to provide a boost current. The switching frequency is reduced in steps and the voltage gain and power of the converter sensed until the voltage gain matches a voltage gain calculated for the efficiency tank. Then tank switchover occurs and transistors to the efficiency tank are pumped with the last switching frequency used by the auxiliary tank, and the auxiliary tank is not pumped. Since the voltage gains before and after tank switchover are equal, no output voltage deviation or current spike occurs. A voltage sag or failure switches back to the auxiliary tank at a switching frequency determined by a dynamic contour line where the voltage gains of the two tanks are equal for the current power state.

Abstract: A wireless power transmitter that supplies power to a wireless power receiver. The wireless power transmitter includes a control unit, a resonator and an amplifier. The control unit generates an identification information of the wireless power transmitter. The resonator transmits power to the wireless power receiver. The amplifier drives the resonator and is controlled by the control unit such that the power transmitted by the resonator includes a first signal that carries the identification information of the wireless power transmitter. The wireless power transmitter further includes a wireless communication unit that sends a second signal having the identification information of the wireless power transmitter.

Abstract: A method for annotating a batch of original samples is provided. A first subset of original samples, selected from the batch and determined by minimizing an entropy-mean difference between the first subset and the batch, is used for human annotation to yield human-annotated samples. The human-annotated samples are used as training data to configure an annotation process for annotating an input sample to yield an annotated output sample, and a check process for verifying annotation accuracy of the annotated output sample. Remaining original samples in the hatch are processed by the annotation process to yield machine-annotated samples, whose accuracy is verified by the check process. In one embodiment, part of the original samples corresponding to erroneous machine-annotated samples are selected for human annotation. Resultant additional human-annotated samples are used to update the two processes. The remaining original samples not yet annotated are then processed by the two processes.

Abstract: A two-point modulation Phase-Locked Loop (PLL) has a dual-input Voltage-Controlled Oscillator (VCO). A digital data modulation signal is combined with a carrier and input to a feedback divider. The data modulation signal is also input to an offset Digital-to-Analog Converter (DAC) to generate an analog voltage to a second input of the VCO. The loop path through the VCO has a higher gain than the DAC path through the VCO, which has better linearity. A calibration unit divides the VCO output and counts pulses. The offset DAC has a data input and a gain input. During calibration, the data input of the DAC is set to minimum and then maximum values and VCO output pulses counted, and repeated for two values of the gain input to the DAC. From the four counts a K(DAC) calculator calculates the calibrated gain to apply to the gain input of the offset DAC.