Optimized Quantization for Convolutional Deep Neural Networks in Federated Learning | IEEE Conference Publication | IEEE Xplore