LAUSR

Abstract In this paper, we target the problem of image transmission in Communicating Things Networks (CTNs) where the ultimate objective is to preserve both the visual quality of the received data at the destination node and the network resources, energy in particular. Nevertheless, visual quality is known to be proportional to the image volume (i.e., better visual quality leads to voluminous data and vice-versa) and hence inversely proportional to the energy cost. Furthermore, as CTNs are open infrastructures, they are prone to transmission errors which are almost managed through costly data re-transmissions. We propose a novel approach, composed of two components: (a) a low layer communication protocol coupled with (b) an image errors recovery filtering mechanisms; which avoids re-transmission due to communication errors. The key idea behind our proposal is to manage errors in transmitted data at the application layer, rather than at the network layer. In other words, corrupted packets are transferred from the lower layers to the upper layers, and a filtering algorithm is in charge to reconstruct the corrupted image. The results of the extensive experiments we conducted have demonstrated the powerful strength of the proposed solution to resist against various types of errors and to enhance the visual quality of images whilst ensuring both data confidentiality and network resources preservation.