Viavi Solutions

Systems, methods, and devices are provided for implementing artificial intelligence (AI) and machine learning (ML)-based model switching techniques for neural network (NN)-based receivers. In one aspect, a system may deploy a plurality of neural network receivers associated with a receiver and trained to operate in different environments, receive sensing data associated with the receiver, and predict, based on the received sensing data, an environment for the receiver from a predetermined set of different environment types. The system may also select a neural network receiver of the plurality of neural network receivers based on a probability of a successful decoding and switch to the selected neural network receiver.

In some implementations, a device may obtain, as part of an inference phase, a software log associated with a network test. The device may generate a constellation image based on the software log, wherein the constellation image indicates a channel constellation associated with the software log. The device may provide the constellation image to an artificial intelligence or machine learning (AI/ML) system. The device may identify, via the AI/ML system, one or more features associated with the constellation image. The device may identify, via the AI/ML system, a channel impairment associated with the constellation image, wherein the channel impairment is based on the one or more features. The device may identify, via the AI/ML system, a remedy to resolve the channel impairment. The device may provide an output of the remedy.

A device may include a filter array disposed on a substrate. The filter array may include a first mirror disposed on the substrate. The filter array may include a plurality of spacers disposed on the first mirror. A first spacer, of the plurality of spacers, may be associated with a first thickness. A second spacer, of the plurality of spacers, may be associated with a second thickness that is different from the first thickness. A first channel corresponding to the first spacer and a second channel corresponding to the second spacer may be associated with a separation width of less than approximately 10 micrometers (µm). The filter array may include a second mirror disposed on the plurality of spacers.

A management system may receive a wireless signal transmitted to a receiver. The management system may determine, based on the wireless signal or a set of decoded bits associated with the wireless signal that were determined using an artificial intelligence (AI) receiver module of the receiver, whether a cyclic redundancy check (CRC) associated with the wireless signal is passed. The management system may perform, based on determining whether the CRC associated with the wireless signal is passed, at least one of: a data storage operation associated with the wireless signal, a fallback operation associated with the wireless signal, or an online label recovery operation associated with the wireless signal. The management system may perform, based on determining whether the CRC associated with the wireless signal is passed, a model modification operation to update a neural network (NN) of the AI receiver module of the receiver.

An optical article printed on a substrate may include an organic binder; and a plurality of reflective magnetic platelets provided in the organic binder, wherein the plurality of reflective magnetic platelets are substantially aligned in accordance with at least part of a surface of revolution, and wherein the plurality of reflective magnetic platelets are aligned to cause a first reflective effect of the optical article when the substrate is rotated around a first axis and to cause a second reflective effect of the optical article when the substrate is rotated around a second axis, wherein the first reflective effect is different from the second reflective effect.

An optical element comprising a body having a surface, wherein the surface has a plurality of regions periodically arranged in a tessellation, and wherein each region of the plurality of regions has a random spatial distribution of microstructures is disclosed. An optical system comprises a light source; and the optical element is also disclosed. Methods of making and using the optical element and the optical system are also disclosed.

In some implementations, a metamaterial absorber (MMA) is configured to absorb a particular range of electromagnetic radiation. The MMA includes a first metal or semiconductor material; a dielectric material disposed on the first metal or semiconductor material; and a second metal material disposed on the dielectric material. A length dimension associated with the MMA is less than or equal to 200 micrometers (µm), a width dimension associated with the MMA is less than or equal to 200 µm, and a thickness dimension associated with the MMA is less than or equal to 8 µm.

Transfer learning (TL)-based systems, methods, and devices are provided for neural network and/or neuromorphic network transmitters/receivers with a set of desired modulation orders. In one aspect, a source system trains a full neural net modulation/demodulation model, from which one or more upper/output layers are removed, and the remaining base layers are transferred into a target system. A set of one or more upper/output layers are generated for the set of desired modulation orders, then transferred into, and trained in, the target system. The target system may store the transferred base layers and the trained set of one or more upper/output layers for the set of desired modulation orders, and use them to modulate/demodulate any transmission having one of the set of desired modulation orders.

A device may include a sensor window. The sensor window may include a substrate. The sensor window may include a set of layers disposed onto the substrate. The set of layers may include a first subset of layers of a first refractive index and a second set of layers of a second refractive index different from the first refractive index. The set of layers may be associated with a threshold transmissivity in a sensing spectral range, and may be configured to a particular color in a visible spectral range and associated with a threshold opacity in the visible spectral range. The device may include a spectral sensor device aligned to the sensor window and including at least one sensor element to receive light in the sensing spectral range and provide a plurality of sensing functionalities based on at least one measurement of the light in the sensing spectral range.

In some implementations, a base station testing system obtains model information associated with a neural network (NN) of an artificial intelligence (AI) receiver module of a receiver, training information associated with the NN; and the NN. The base station testing system determines one or more test cases for testing the NN and determines respective testing conditions associated with the one or more test cases. The base station testing system generates, based on the respective testing conditions, respective test case data associated with the one or more test cases. The base station testing system determines, based on testing the NN using the respective test case data associated with the one or more test cases, performance information associated with the NN, and generates, based on determining the performance information, a performance report associated with the NN.