Advanced Technologies Development
The advanced technology development center not only looks ahead at future technologies but also decides themes to be either shared between all business sections in the Anritsu Group or to be implemented horizontally.
The center handles antenna-related OTA measurements [*1] which are a key technology for deploying fifth-generation (5G) mobile communications system, as well as X-ray analysis technologies using machine learning [*2] to improve inspection accuracy.
[※1] OTA Measurement
This is the abbreviation for Over The Air, meaning measurement of the radio send and receive performance over the air via the antenna. Generally, measurement is performed by connecting a cable to the radio antenna connector to be measured.
[※2] Machine Learning
This is one R&D subject in artificial intelligence; it is a technology/method for implementing human learning ability in a machine using a computer.
Research into fifth-generation (5G) mobile communications system is advancing with focus on using massive MIMO and beam forming technologies, etc., in the microwave and mm-Wave bands. Additionally, RF ports may disappear with mm-Wave deployment, increasing the importance of OTA measurements for evaluating antenna directivity, power, etc.
The previous Far-Field Measurement (FFM) method requires long-term measurement in a large-scale measurement environment. Additionally, at mm-Wave band measurements expected to be adopted by fifth-generation (5G) mobile communications system due to the large transmission capacity, FFM suffers from decreased sensitivity due to attenuation of the signal strength as the signal propagates over longer distances.
The advance technology development center is working to develop a low-cost Near-Field Measurement (NFM) method that solves this issue.
Development of technologies for measuring mm-Wave bands up to frequencies of 300 GHz used by fifth-generation (5G) mobile communications system is in progress.
In 2014, we successfully developed a pre-selector (variable filter) for frequencies above 100 GHz, which had proved difficult to implement previously, and Anritsu was the first company in the world to develop a spectrum analyzer incorporating a variable filter. A 300-GHz band spectrum analyzer is currently being developed to support future measurement needs using the expanded mm-Wave band.
In the image recognition field, the focus is on machine learning as typified by deep learning. Trained neural network using Big Data have excellent problem-solving abilities and high applicability and are spreading into many fields.
Since pre-trained network models have powerful feature-extraction abilities (non-linear functions), they are a candidate signal-analysis technology and are being developed for various applications.