This paper provides a comprehensive survey on the security and privacy aspects of Decentralized Federated Learning (DFL), a paradigm that combines Federated Learning (FL) with blockchain technology to enhance security and privacy. The authors review the state-of-the-art DFL methods, focusing on their security robustness and employed technologies. They discuss various threats to DFL, including attacks on performance and privacy, and propose potential defense mechanisms to mitigate these threats. The paper also explores the verifiability of DFL and highlights future research directions. Key contributions include: 1. **State-of-the-Art DFL Methods**: The survey reviews recent advancements in DFL, emphasizing the integration of blockchain and the use of technologies like Smart Contracts (SC) and Differential Privacy (DP) to enhance security and privacy. 2. **Threats to DFL**: The paper identifies and analyzes various threats to DFL, such as data poisoning, model poisoning, routing attacks, and consensus attacks. It also discusses privacy risks like model inversion, membership inference, and private key hijacking. 3. **Defense Mechanisms**: The authors propose defense mechanisms to protect against these threats, including homomorphic encryption (HE), secure multiparty computation (SMC), and differential privacy (DP). Each mechanism is evaluated based on its ease of implementation, effectiveness, and defensibility. 4. **Verifiability of DFL**: The paper discusses the importance of verifiability in DFL, ensuring that the system can be trusted and that participants are verified. 5. **Future Research Directions**: The authors suggest areas for future research, including the development of hybrid approaches to balance privacy and utility, and the exploration of new attack surfaces and defense mechanisms. The survey aims to provide a thorough understanding of the security and privacy challenges in DFL and to guide researchers and practitioners in designing more secure and robust systems.This paper provides a comprehensive survey on the security and privacy aspects of Decentralized Federated Learning (DFL), a paradigm that combines Federated Learning (FL) with blockchain technology to enhance security and privacy. The authors review the state-of-the-art DFL methods, focusing on their security robustness and employed technologies. They discuss various threats to DFL, including attacks on performance and privacy, and propose potential defense mechanisms to mitigate these threats. The paper also explores the verifiability of DFL and highlights future research directions. Key contributions include: 1. **State-of-the-Art DFL Methods**: The survey reviews recent advancements in DFL, emphasizing the integration of blockchain and the use of technologies like Smart Contracts (SC) and Differential Privacy (DP) to enhance security and privacy. 2. **Threats to DFL**: The paper identifies and analyzes various threats to DFL, such as data poisoning, model poisoning, routing attacks, and consensus attacks. It also discusses privacy risks like model inversion, membership inference, and private key hijacking. 3. **Defense Mechanisms**: The authors propose defense mechanisms to protect against these threats, including homomorphic encryption (HE), secure multiparty computation (SMC), and differential privacy (DP). Each mechanism is evaluated based on its ease of implementation, effectiveness, and defensibility. 4. **Verifiability of DFL**: The paper discusses the importance of verifiability in DFL, ensuring that the system can be trusted and that participants are verified. 5. **Future Research Directions**: The authors suggest areas for future research, including the development of hybrid approaches to balance privacy and utility, and the exploration of new attack surfaces and defense mechanisms. The survey aims to provide a thorough understanding of the security and privacy challenges in DFL and to guide researchers and practitioners in designing more secure and robust systems.