US Patent Application 18004867. FEDERATED LEARNING OF AUTOENCODER PAIRS FOR WIRELESS COMMUNICATION simplified abstract
Contents
FEDERATED LEARNING OF AUTOENCODER PAIRS FOR WIRELESS COMMUNICATION
Organization Name
Inventor(s)
June Namgoong of San Diego CA (US)
Taesang Yoo of San Diego CA (US)
Naga Bhushan of San Diego CA (US)
Pavan Kumar Vitthaladevuni of San Diego CA (US)
Jay Kumar Sundararajan of San Diego CA (US)
Alexandros Manolakos of Escondido CA (US)
Krishna Kiran Mukkavilli of San Diego CA (US)
Hwan Joon Kwon of San Diego CA (US)
Tingfang Ji of San Diego CA (US)
Wanshi Chen of San Diego CA (US)
FEDERATED LEARNING OF AUTOENCODER PAIRS FOR WIRELESS COMMUNICATION - A simplified explanation of the abstract
- This abstract for appeared for US patent application number 18004867 Titled 'FEDERATED LEARNING OF AUTOENCODER PAIRS FOR WIRELESS COMMUNICATION'
Simplified Explanation
The abstract describes a wireless communication system where a client device uses two autoencoders to determine and transmit information about its environment. The first autoencoder calculates a feature vector based on the client's surroundings, while the second autoencoder generates a latent vector based on the feature vector. Both the feature vector and the latent vector are then transmitted by the client device.
Original Abstract Submitted
Various aspects of the present disclosure generally relate to wireless communication. In some aspects, a client may determine, using a first client autoencoder, a feature vector associated with one or more features associated with an environment of the client. The client may determine a latent vector using a second client autoencoder and based at least in part on the feature vector. The client may transmit the feature vector and the latent vector. Numerous other aspects are provided.
