Mobile users in urban HetNet environments contend with reduced Quality of Experience (QoE) while streaming video due to fluctuations in available bandwidth. Adaptive Bit Rate technologies have been developed to deal with changes in wireless link conditions but they do not address the problem of streaming HD video over bandwidth constrained links. Service providers impose data-caps on pay-as-you-go accounts which is problemat-ic since much of the downloaded data consists of video files and breaching the data-cap can result in degraded performance and high excess usage charges. User devices are typically dual-homed having both a cellular interface and a Wi-Fi inter-face although some devices may be equipped with additional interfaces such as Bluetooth Low Energy. Urban HetNets provide users with the opportunity to switch to a network that best suits their needs. Users seek to protect their data allowances by downloading as little data as possible over cellular networks by connecting to Wi-Fi whenever possible. However, under certain conditions it can be detrimental to the user’s QoE to attempt to connect to Wi-Fi APs as the connection periods might be so short that little or no data can be downloaded or the connection attempt might fail. In addition, mobile devices are con-strained by their battery capacity and it is important that energy consumption be reduced where possible. This thesis presents a solution for managing video streaming in urban HetNets for mobile users, protecting their data-caps, optimising QoE and reducing energy consumption. The proposed solution consists of the following three novel components: a) the Scan-Or-Not-to-Scan (SONS) framework containing the SONS utility function b) MPEG-DASH-based Framework (MDF) which manages SD video streams c) Adaptive Interface Selection (AIS) which manages HD video streams over bandwidth constrained wireless links The proposed components were evaluated using modelling, simulation and hybrid test solutions consisting of Linux Virtual Machines and Network Simulator 3 networks. The results showed that the proposed solution can reduce the amount of data transferred over cellular networks, maintain QoE of SD video streams by reducing the number of stalling events and enable streaming of HD video over bandwidth constrained links.