Pocket switched networks (PSNs) take advantage of human mobility to distribute data. Investigations on real-world trace data indicate that human mobility follows a simple reproducible pattern: a human being usually visits a few places at high frequencies. These most frequently visited places form the home of a node, which is exploited in this paper to design two HomE based ROuting (HERO) algorithms. In the basic HERO, the first encountered relay whose home contains the place where the destination resides is selected to deliver the data. The enhanced HERO, on the other hand, continuously selects a better relay that visits the destination place at a higher frequency. In both algorithms, each node only needs to maintain and exchange its relatively stable home information and/or the corresponding visiting frequencies; therefore no global networking information and no frequent information update are needed, resulting in a low burden on the network due to its low communication and storage overheads. Moreover, HERO involves only simple arithmetic operations, thus causing little computation overhead at the mobile nodes. The simulation results indicate that both HERO algorithms outperform the state-of-the art.