Federated Learning (FL) has emerged as a promising technique for edge devices to collaboratively learn a shared prediction model, while keeping their training data on the device, thereby decoupling the ability to do machine learning from the need to store the data in the cloud. However, FL is difficult to implement and deploy in practice, considering the heterogeneity in mobile devices, e.g., different programming languages, frameworks, and hardware accelerators. Although there are a few frameworks available to simulate FL algorithms (e.g., TensorFlow Federated), they do not support implementing FL workloads on mobile devices. Furthermore, these frameworks are designed to simulate FL in a server environment and hence do not allow experimentation in distributed mobile settings for a large number of clients. In this paper, we present Flower 1 , a FL framework which is both agnostic towards heterogeneous client environments and also scales to a large number of clients, including mobile and embedded devices. Flower's abstractions let developers port existing mobile workloads with little overhead, regardless of the programming language or ML framework used, while also allowing researchers flexibility to experiment with novel approaches to advance the state-of-the-art. We describe the design goals and implementation considerations of Flower and show our experiences in evaluating the performance of FL across clients with heterogeneous computational and communication capabilities. 1 https://flower.dev/ Preprint. Under review.

No abstract

We adjust the Proof of Work (PoW) consensus mechanism used in Bitcoin and Ethereum so that we can build on its strength while also addressing, in part, some of its perceived weaknesses. Notably, our work is motivated by the high energy consumption for mining PoW, and we want to restrict the use of PoW to a configurable, expected size of nodes, as a function of the local blockchain state. The approach we develop for this rests on three pillars: (i) Proof of Kernel Work (PoKW), a means of dynamically reducing the set of nodes that can participate in the solving of PoW puzzles such that an adversary cannot increase his attack surface because of such a reduction; (ii) Practical Adaptation of Existing Technology, a realization of this PoW reduction through an adaptation of existing blockchain and enterprise technology stacks; and (iii) Machine Learning for Adaptive System Resiliency, the use of techniques from artificial intelligence to make our approach adaptive to system, network and attack dynamics. We develop here, in detail, the first pillar and illustrate the second pillar through a real use case, a pilot project done with Porsche on controlling permissions to vehicle and data log accesses. We also discuss pertinent attack vectors for PoKW consensus and their mitigation. Moreover, we sketch how our approach may lead to more democratic PoKW-based blockchain systems for public networks that may inherit the resilience of blockchains based on PoW.

Training Automatic Speech Recognition (ASR) models under federated learning (FL) settings has recently attracted considerable attention. However, the FL scenarios often presented in the literature are artificial and fail to capture the complexity of real FL systems. In this paper, we construct a challenging and realistic ASR federated experimental setup consisting of clients with heterogeneous data distributions using the French Common Voice dataset, a large heterogeneous dataset containing over 10k speakers. We present the first empirical study on attention-based sequence-to-sequence E2E ASR model with three aggregation weighting strategies -standard FedAvg, lossbased aggregation and a novel word error rate (WER)-based aggregation, are conducted in two realistic FL scenarios: crosssilo with 10-clients and cross-device with 2k-clients. In particular, the WER-based weighting method is proposed to better adapt FL to the context of ASR by integrating the error rate metric with the aggregation process. Our analysis on E2E ASR from heterogeneous and realistic federated acoustic models provides the foundations for future research and development of realistic FL-based ASR applications.

No abstract