Nowadays all existing public key cryptosystems are classified into three categories relied on different mathematical foundations. The first one is based on the difficulty of factoring the product of two big prime numbers. The representatives are the RSA and the Rabin cryptosystems. The second one such as the ElGamal cryptosystem is based on the discrete logarithm problem. The last one is based on the NP-completeness of the knapsack problem. The first two categories survived crypto attacks, whereas the last one was broken and there has been no attempt to use such a cryptosystem. In order to save the last category, Kiriyama proposed a new public key cryptosystem based on the non-linear knapsack problem, which is an NP-complete problem. Due to the non-linear property of the non-linear knapsack problem, this system resists all known attacks to the linear knapsack problem. Based on his work, we extend our research in several ways. Firstly, we propose an encrypted secret sharing scheme. We improve the security of shares by our method over other existing secret sharing schemes. Simply speaking, in our scheme, it would be hard for outsiders to recover a secret even if somehow they could collect all shares, because each share is already encrypted when it is generated. Moreover, our scheme is efficient. Then we propose a multiple identities authentication scheme, developed on the basis of the non-linear knapsack scheme. It verifies the ownership of an entity's several identities in only one execution of our scheme. More importantly, it protects the privacy of the entities from outsiders. Furthermore, it can be used in resource-constrained devices due to low computational complexity. We implement the above schemes in the C language under the Linux system. The experimental results show the high efficiency of our schemes, due to low computational complexity of the non-linear knapsack problem, which works as the mathematical foundation of our research.