In this paper, we propose an efficient handoff decision algorithm to mitigate the number of unnecessary handoffs in femtocell networks, called OHDP. The rationale behind our proposal is to optimize the list of FAPs target based on the prediction of the direction of the User Equipment (UE) using linear regression. OHDP integrates further a Call Admission mechanism to carefully select a FAP target with enough resources. We show through extensive simulations that OHDP outperforms two prominent state-of-the-art schemes in terms of number of handoffs, and dwell time while ensuring a low handoff decision time.