This thesis investigates the downlink resource allocation problem in Orthogonal Frequency Division Multiple Access (OFDMA) Heterogeneous Networks (HetNets) consisting of macro-cells and small-cells sharing the same frequency band. Dense deployment of small-cells overlaid by a macro layer is considered to be one of the most promising solutions for providing hotspot coverage in future 5G networks. The focus is to devise optimised policies for small-cells’ access to the shared spectrum, in terms of their transmissions, in order to maximise small-cell served users sum data rate while ensuring that certain level of quality of service (QoS) for the macro-cell users in the vicinity of small-cells is provided. We obtain the optimal solution to the resource allocation problem by employing the well-known Dual Lagrangian method. The formulated resource allocation problem is decomposed into N sub-problem at each Resource Block (RB). The optimal transmit power and RB allocation for each small-cell is obtained by updating the dual variables based on sub-gradient method. Furthermore, a low complexity heuristic solution based on binary integer linear programming is proposed for practical systems, and its performance is analysed in comparison with Reuse-1 and orthogonal frequency reuse cases. Alongside considering the data channel protection for macro-cell served users in the vicinity of small-cells, we also cater for control channel constraints. Since the control channel holds key information to decode data channel information, and if control channel is lost the data channel performance will be severely affected. We formulate the joint control and data channel resource allocation problem in HetNets. The solution to the complex problem is addressed by two low complexity heuristic solutions. The proposed interference aware heuristic solution is based on a progressive iterative approach which has significantly lower complexity compared to the optimal case. Whilst the aim is to maximise the net data-rate performance of the small-cells, we also address the energy efficiency issues in dense small-cell networks. We formulate the optimisation problem to minimise the small-cells’ energy consumption by making use of sleep-mode capabilities. The solution to the problem is provided by another heuristic algorithm which addresses the interference problem as well as the energy consumption minimisation. The interference minimisation phase is updated on a shorter intervals, whereas sleep mode phase takes place on larger time scales, considering the wake-up delays associate with practical networks. The potential energy saving gains are extensively examined by hypothetically making use of call data records from City of Milan. This real data helps highlight the potential of energy savings by making use of small-cells’ sleep-mode capabilities along with our proposed algorithm.