A Secure and Privacy Preserved Parking Recommender System Using Elliptic Curve Cryptography and Local Differential Privacy

Abstract

The privacy preservation has received considerable attention from organizations as the growing population is apprehensive regarding personal data being preserved. Smart Parking is a parking strategy that combines technology and human innovation in an effort to use as few resources as possible(such as time and space) to achieve faster and easier parking spots of vehicles. Smart parking systems utilize third-party parking recommender systems to offer customized parking space recommendations to its users based on their past parking experience. However, indiscriminately sharing a user's data with a third party recommendation system may expose their personal information. As their activity and node mobility can be deduced from previous paring experience. There are several privacy and security issues in existing systems, such as identity and location disclosure, availability and authenticity issues. Another problem with existing solutions is that most distributed systems need a third party to anonymize user data for privacy preservation. Therefore, this article fills the described research gaps by introducing parking recomender systems using Local Differential Privacy (LDP) and Elliptic Curve Cryptography (ECC). Based on ECC we proposed the mutual authentication mechanism using Hash-based message authentication code (HMAC) to provide anonymity and integrity during communication. Moreover, given the risks to security and privacy posed by untrustworthy third parties. We used LDP which uses the Laplace distribution technique to add noise randomly and eliminates any necessity for a third party for data perturbation. In addition to LDP , we utilized the IOTA distributed ledger technology (DLT) to provide a new level of security that ensures immutability, scalability, and quantum secrecy and decentralized the system. Our experiments demonstrate that, in addition to preserving the driver's privacy and security, our proposed model has low storage overheads, computation, and communication costs.

Federated Learning-Based Personalized Recommendation Systems: An Overview on Security and Privacy Challenges

“…Unauthorized access to the system's resources is one of the most common adversarial practices performed by the bad elements. Despite such unauthorized access, the extensive operational process of PRS may have consequences that may disclose the confidentiality, authenticity, and even integrity of consumers' data in some scenarios [41]. Data breaching is another possible malicious approach that may be performed by attackers to obtain valuable information from consumers.…”

Internet of Things: Security and Solutions Survey

“…This is accompanied by the prediction that the majority of current asymmetric cryptography will be broken by the advent of quantum technology employing Shor’s Algorithm [ 200 ]. Winternitz One Time Signature (WOTS) Scheme [ 201 ], Supersingular Isogeny Diffie–Hellman Key Exchange (SIDH) [ 200 ], etc., are post-quantum resistance. It is required to standardize a model to preserve integrity.…”

“…For smart parking, Khaliq et al [ 201 ] proposed an authentication method using ECC and location differential privacy based on Laplace distribution. The method needs to use third-party app and also hide the sensitive data in local settings before sharing with service provider.…”

ML-DDoSnet: IoT Intrusion Detection Based on Denial-of-Service Attacks Using Machine Learning Methods and NSL-KDD

“…An algorithm combining interactive machine learning and active learning for HBR prediction was proposed by Wu et al [29]. Using local differential privacy (LDP) and elliptic curve cryptography (ECC), Khaliq et al [30] describe parking recommender systems with research gaps. Recent SBR prediction models have performed poorly due to mislabeled instances in five publicly available datasets, according to Wu et al [31].…”