Siva The IPFS Search Engine

Abstract

In the world of increasing data, it becomes necessary to upgrade the present structure of the network and go for the decentralized network. For this, IPFS (InterPlanetary File System)
made it possible to make a pure Peer-to-Peer (P2P), censor resistant and permanent Web in which links will never die. Despite good points of IPFS, finding data without knowing it's unique name is impossible since there is no search engine to index the data and respond to queries.
Our aim is to provide a decentralized search engine to IPFS in which the nodes can search for any content at any time and anywhere.
we propose using DHT to handle the indexed data and strengthen the search with adding result cache layer to the search engine alongside with a hash filter to make cache lookup process fast and efficient.
We used simulation to test the caching effect on the search engine using a sampled dataset from Yahoo Webscope. The simulation results show the effect of caching on reducing
query latency and effect of different cache sizes on the cache hit-rate as the hit-rate decreases with decreasing cache size.