In 5G Ultra-Dense Cells, a distributed wireless backhaul is an attractive solution for forwarding traffic to the core. In this thesis, we investigate Gateway Location Problem and show that finding near optimal gateway locations improves the Backhaul Network Capacity(BNC). Toward this end, the p-median problem has been formulated as Integer Linear Program. Subsequently, we use Genetic Algorithm in combination with K-means algorithm to find gateway locations that maximize the BNC. We evaluate the performance of our new heuristic, K-GA, in comparison with six different approaches in terms of Average Number of Hops(ANH) and BNC at different node densities through extensive Monte Carlo simulations. All approaches are tested under different small cell distribution scenarios. K-GA achieves ANH and BNC within 3% of optimal and saves on-average 95% of time. We also analyze the effect of the number of gateways on ANH and BNC. The results show that more gateways are beneficial.