Approach

Technique

Main Idea

Zhou and Tang [62]

SVGR Algorithm

Develop a novel differentially private distributed algorithm based on the stochastic variance reduced gradient technique.

Van Dijk et al. [64]

Asynchronous federated learning algorithm

Develop a novel algorithm that eliminates waiting times and reduces overall network communication.

Girgis et al. [65]

CLDP-SGD

Develop a distributed communication-efficient and locally differentially private stochastic gradient descent algorithm along with a detailed analysis of its communication, privacy, and convergence tradeoffs.

Zhang et al. [66]

Mechanism Design

Develop a federated learning scheme based on differential privacy and mechanism design under which high-quality clients are selected to improve the accuracy of the model.

Denisov et al. [67]

Optimal Private Linear Operators

Develop improved differential privacy for SGD that achieves significant improvements in a notable problem in federated learning with differential privacy at the user level.

Lian et al. [68]

COFEL

Develop a novel federated learning system that reduces communication time through layer-based parameter selection and enhances privacy protection through local differences in privacy.

Amiri et al. [69]

Universal Vector Quantization

Develop a novel algorithm for achieving differential privacy and reduced communication overhead through compression of client-server communication by quantization.

Liu et al. [70]

FL-PFA

Develop a novel framework, named FL-PFA, that achieves communication cost minimization.

Zhang et al. [32]

Clipping

Develop a novel federated learning framework with clipping technique

Truex et al. [71]

Security Multi-party Computation

Develop a novel approach that combines differential privacy and SMC, thus enabling users to reduce the growth of noise injection.