LoRa is a leading Low-Power Wide-Area Network technology, designed for Internet of Things applications that require communication over long distances at low power. While a lot of research has been done on the performance, scalability and security analysis of LoRa networks, the important problem of planning and deploying LoRa networks has not received much attention in the research community yet. In the first part of this thesis, we address the problem of planning LoRa networks, which consists of gateway placement and end device configuration. We formulate the problem as a mixed-integer non-linear optimization problem, which is shown to be NP-Hard. By theoretically analyzing the properties of the optimal solution in simplified and regularly-structured network topologies, we develop an approximate algorithm for planning large-scale LoRa networks efficiently, which is shown to outperform a commonly-used configuration algorithm in LoRaWAN in terms of the overall throughput and energy efficiency of the network. In the second part of this thesis, we propose a cooperative packet detection scheme which can be implemented in LoRa networks in order to improve network throughput. The proposed method aligns with LoRa requirements in that it does not add any complexity to end devices which have limited power constraints. We analyze the effects of cooperation mathematically and through simulations in regularly-structured and arbitrary network topologies. Evaluations show that cooperation can lead to significant improvements in throughput in real-world inspired LoRa networks.