Abstract: When enjoying location-based services, users need to submit location information and query requests frequently, which may lead to privacy disclosure. Therefore, a trusted prediction cache architecture based on blockchain(TPCB) for privacy protection is proposed by using the ability of cognitive engine to perceive user needs. First, the query request is predicted based on the LSTM model, and the requester relies on the service provider to broadcast the cached data to obtain the service. Secondly, the trust mechanism is used to evaluate the trust value to solve the problem of untrusted interaction when communicating with different neighbors. Finally, based on the characteristics of blockchain, a large number of trust data and transaction data generated in the transaction process are stored in the block. The experimental results show that TPCB can effectively improve the cache hit rate, and has good performance in restraining malicious providers, encouraging providers who refuse to participate and privacy protection.

Key words: cognitive internet of vehicles, cache, privacy protection, blockchain, trust mechanism