In cloud storage systems, the index system, which maps a queried item to its storage location, is a key component. Due to the limited space in the storage nodes, it requires to compact the index for fast response. Moreover, since the index may be published on many nodes of the system, the privacy of the nodes' ownership information needs to be protected. In this paper we introduce an indexing modelthe Lightweight Bloom filter Array (LBA) to improve the index performance while preserves item information on the nodes of the cloud storage systems. We conduct theoretical analysis to find the performance of the index system. We also make theoretical comparisons with other index method. Simulations show that the theoretical deduction is in good accordance with the expected performance.