Loss Differentiation Algorithms (LDA) are currently used to determine the cause of packet losses with an aim of improving TCP performance over wireless networks. In this work, we propose a cross-layer solution using two complementary LDA schemes in order to classify the loss origin on an 802.11 link and then to react consequently. The first LDA scheme, acting at the MAC layer, allows differentiating losses due to signal failure caused by displacement or by noise from other loss types. Moreover, in the case of a signal failure, this scheme adapts the behavior of the MAC layer in order to avoid a costly end-to-end TCP resolution. The objective of the second LDA scheme, which acts at the TCP layer, is to distinguish between losses due to interferences and those due to congestions, then accordingly adapt the TCP behavior. We finally demonstrate, through simulation, the efficiency of each LDA scheme as well as the whole cross-layer solution.