- Machine learning
- Wireless communication
The present invention is in the field of machine learning and artificial intelligence,applied to a critical problem in wireless communication.
As the world is set to embrace the 5G network, multi-point connectivity is foreseen as an important technology for improving the perceived quality of experience for mobile users. Radio signals can help to map the location of the user and their destination. Thereby identifying the hot spots with network traffic congestions or poor services and optimizing the location for placing an antenna. However, the current technique is flawed, specially in dense or urban environments, as it relies on delay signal strength and their reliance on multipath propagation channels.In the current invention we have developed a mapping technique to convert radio signals from one or many mobile equipment (ME) into low dimensional representation, potentially 3D or 2D spatial geometry, and vice versa. This process is called “Channel Charting”. In a radio communication device, an input in the form of multiple received signals transmitted from multiple MEs are collected at one or multiple locations. Thereafter it is processed to extract a low-dimensional representation of transmitter characteristics as well as a forward and inverse channel mapping function. The forward channel mapping function helps in localizing and targeting services and understanding communication contexts of users. The inverse channel mapping function enables one to extract information in the high-dimensional radio geometry from the low-dimensional representation, which can be used for channel estimation, rate adaption, BS hand over, as well as for optimizing antenna placement or identify hot spots, etc. The mapping can adapt dynamically and autonomously to changes in the environment both indoors (LoS) and rural areas (non-LoS).
- No use of more costly Global Navigation Satellite System (GNSS)
- It can be used in a rich scattering environment such as indoors, densities, street canyons, and for various weather and traffic conditions.
- Implementation in mobile equipment like GPS, smart phones, vehicular communication, telematics, for traffic safety, and for sensor nodes on the Internet of Things.
- Used for channel estimation, rate adaption, BS hand over, hot-spot identification and optimizing antenna placement.
Name: Ryan Luebke