Reliable Federated Learning Systems Based on Intelligent Resource Sharing Scheme for Big Data Internet of Things

Sa Math,Seokhoon Kim,Prohim Tam

doi:10.1109/access.2021.3101871

Abstract

Federated learning (FL) is the up-to-date approach for privacy constraints Internet of Things (IoT) applications in next-generation mobile network (NGMN), 5 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sup> generation (5G), and 6 <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">th</sup> generation (6G), respectively. Due to 5G/6G is based on new radio (NR) technology, the multiple-input and multiple-output (MIMO) of radio services for heterogeneous IoT devices have been performed. The autonomous resource allocation and the intelligent quality of service class identity (IQCI) in mobile networks based on FL systems are obligated to meet the requirements of privacy constraints of IoT applications. In massive FL communications, the heterogeneous local devices propagate their local models and parameters over 5G/6G networks to the aggregation servers in edge cloud areas. Therefore, the assurance of network reliability is compulsory to facilitate end-to-end (E2E) reliability of FL communications and provide the satisfaction of model decisions. This paper proposed an intelligent lightweight scheme based on the reference software-defined networking (SDN) architecture to handle the massive FL communications between clients and aggregators to meet the mentioned perspectives. The handling method adjusts the model parameters and batches size of the individual client to reflect the apparent network conditions classified by the k-nearest neighbor (KNN) algorithm. The proposed system showed notable experimented metrics, including the E2E FL communication latency, throughput, system reliability, and model accuracy.

Highlights

INTRODUCTIONmobile edge computing (MEC) is the crucial enabler of technologies for network slicing (NS) and local computing resource for mobile user services
This paper proposed an intelligent resource allocation based on a lightweight machine learning (ML) algorithm, namely k-nearest neighbor (KNN)
CONTRIBUTIONS In this paper, we propose a lightweight approach for effective resource allocation to enhance the reliability of the Federated learning (FL) model in big data Internet of Things (IoT) communication networks

Summary

INTRODUCTION

MEC is the crucial enabler of technologies for network slicing (NS) and local computing resource for mobile user services. S. Math et al.: Reliable FL Systems Based on Intelligent Resource Sharing Scheme critical for next-generation self-organizing networking (SON) and efficient resource configuration for heterogeneous user services. In the HetIoT environments, big data will be in local areas since the client has computing resource limitations to train the model. B. CONTRIBUTIONS In this paper, we propose a lightweight approach for effective resource allocation to enhance the reliability of the FL model in big data IoT communication networks. While the virtual computing will attach with the NR stations for radio service operations, the failures of model transferring can occur whenever radio gateway and attached MEC server resources are overloaded operations.

SYSTEM ARCHITECTURE

RESOURCE ADJUSTMENT

10 Update caching metrics

Findings

CONCLUSION