I would doubt that any of the traditional models, such as the Hata model, are applicable to femtocells. Reason being, the distances are way too short, and antenna height has completely different meaning, in these cases. You have short distances, implying the free space propagation loss model would be best, but with obstacles and scatter.

This paper covers the subject and proposes an empirical model.

http://www.ijcsi.org/papers/IJCSI-11-3-1-128-133.pdf

Adding to Albert, the thesis in the given link cover all aspects of channel models related to femtocells: http://projekter.aau.dk/projekter/files/52683437/Master_Thesis.pdf

Best wishes

I have used a path loss model with considering shadowing effect in my paper. It may be useful for you.

Article QoS-Aware Downlink Radio Resource Management in OFDMA-Based ...