In this paper, we propose algorithms for the placement phase and distribution phase of caching in wireless networks using Random Linear Network Coding (RLNC). The goal is to minimize, the load on the link from BS to the devices and the latency as a Quality of Experience (QoE) of the users. In the first algorithm, the devices place the files in their cache in a distributed fashion based on the popularity of files, so that the multicasting opportunities in the delivery phase are maximized. In the second algorithm, the devices share the content of their cache among each other and the remaining coded packets of the files are multicasted from the base station (BS). We exploit the sharing opportunities for different file requests made possible by the intelligent cache placement and the spatial diversity of the Device to Device (D2D) networks for obtaining global caching gain. An implementation of the algorithms in Virtual Machines (VMs) is carried out using an RLNC library and the average number of packet transmissions from the server VM to the client device VMs is measured. Unlike previous works, we consider a realistic setting where there are packet losses in the wireless links and utilize the forward error correction capabilities of RLNC.