Multi-query Computationally-Private Information Retrieval with Constant Communication Rate

A fundamental privacy problem in the client-server setting is the retrieval of a record from a database maintained by a server so that the computationally bounded server remains oblivious to the index of the record retrieved while the overall communication between the two parties is smaller than the database size. This problem has been extensively studied and is known as computationally private information retrieval (CPIR). In this work we consider a natural extension of this problem: a multi-query CPIR protocol allows a client to extract m records of a database containing n `-bit records. We give an information-theoretic lower bound on the communication of any multi-query information retrieval protocol. We then design an efficient non-trivial multi-query CPIR protocol that matches this lower bound. This means we settle the multiquery CPIR problem optimally up to a constant factor.