Efficient and Secure Authentication Key Establishment Protocol Using Chaotic Map and PUF in Smart Environments

With the rapid growth and popularization of the Internet of Things (IoT), it has been applied to numerous fields such as smart industry, smart agriculture and smart home. Designing practical and robust authentication key agreement (AKA) schemes for smart environments has become a pressing problem to be solved. Due to differences in security requirements and resource allocation in smart environments, which we call security-asymmetry and resource-asymmetry, it is necessary to design specific AKA schemes for this environment. Since the design of remote AKA protocols does not fully consider security-asymmetry and resource-asymmetry, many existing schemes are not practical in smart environments. With regard to security-asymmetry, compared with traditional public-key techniques applied in AKA schemes, chaotic map is more effective than modular exponentiation and scalar multiplication, and it supplies many feasible attributes such as unpredictability, unrepeatability, uncertainty, which can be used to achieve communication security between users and gateways, while security operations based on hash function are sufficient to secure communications between gateways and smart devices. In view of resource-asymmetry, the complex operations in the authentication process can be completed by the gateway, so as to make full use of the rich resources on gateway side and reduce the use of resources on user side and device side. Based on such considerations, an efficient and secure authentication key agreement scheme based on chaotic map and physical unclonable function (PUF) for smart environments is proposed. We present a rigorous informal analysis of the proposed scheme. Moreover, the formal security verification is accomplished using the Automated Validation of Internet Security Protocols and Applications (AVISPA) tool. Finally, performance evaluations indicate the proposed protocol consumes less communication cost and computation cost while achieving more security functions compared to other four state-of-the-art related schemes.