The objective of this application is to provide funding to compensate Dr. Shahram ShahbazPanahi, who is a Full Professor at the Department of Electrical and Computer Engineering, University of Ontario Institute of Technology, Oshawa, Ontario, Canada. Dr. ShahbazPanahi will be on research leave from his institution starting July 1, 2018 until July 1, 2019, and intends to spend part of this period at the Interdisciplinary Centre for Security Reliability and Trust (SnT), University of Luxembourg. During his stay, Dr. ShahbazPanahi focuses on distributed optimization and dynamic programing for wireless caching in wireless networks. Recently there has been a surge for streaming video – a content that is produced in advance but requires significant bandwidth so it can be delivered with low latency. It is expected that as a major source of content delivery traffic, video-on-demand constitute about 70 percent of mobile network traffic by 2019. Another characteristic of content delivery traffic is that at times some contents are requested more widely due to their popularity. The high demand for the most popular contents calls for strategic caching of those contents to avoid significant delays in their delivery. The recent literature on wireless caching rarely focus on one component, namely the users’ Quality of Experience (QoE). Any resource allocation and transmission strategy have to take into account what kind of experience the users are experiencing. Given the time-varying and broadcast features of wireless channels, the users’ QoE in the network can change drastically during the content delivery period. As such caching policies are required to adapt to the fluctuations of the wireless channel in a dynamic manner based on the users’ QoE. Hence, the network has to be equipped with software and hardware needed to collect the users’ feedbacks on the quality they are experiencing. Once such feedbacks are collected, one has to deal with the challenging research task of how to incorporate these feedbacks in the design of wireless caching policies. It is worth mentioning that while all users’ will have the opportunity to provide their feedback, not all of them will do so. As a result that state of the network is only partially known to the caching policy maker. Given these characteristics of content delivery networks, a dynamic programming approach with hidden or partially observable states appear to be a promising approach to design wireless caching schemes and transmission strategies. And this is exactly what this proposal is concerned with: a partially observable Markov decision process (POMDP) approach to QoE-based design of wireless caching policies.