This handbook aims to serve as a one-stop, reliable resource, including curated surveys and expository contributions on federated learning. First introduced in 2016, federated learning allows devices to collaboratively learn a shared model while keeping raw data localized, thus promising to protect data privacy.