Abstract : Long range (LoRa), being non operator-based, has been one of the most innovative LPWANs technology offering long range and low power capability. It has proved to be excellent in remote and rural area deployments. With LoRa's low data rate property, small payload sensor nodes are mainly deployed with LoRa. However, the integration of large data transmission such as images along with small payload size data, can contribute to realizing the goals of improved productivity and reduced management cost for various IoT applications. LoRa wide area network (LoRaWAN) does a random channel selection of usable channels, to transmit data frames to gateways, as recommended by the adaptive data rate MAC command. This introduces multiple scenarios where nodes make channel selections that usually ends up affecting the data delivery times of other transmissions, especially of large data nodes, and even data frame drops. Also, LoRa network operates in the sub-GHz bands where duty-cycle regulation is imposed on the frequency channels and time offs from frequency-use need to be adhere to. As we experience multiple frame transmissions for large data, the airtime to transmit data increases. Coupled with the time-off periods and transmission time limitation due to the duty-cycle, throughput is limited and further delay is introduced. In this thesis, we tackle these two problems by scheduling channels on time slots for data transmission of nodes that minimizes the data delivery delay specifically for large data (images) sensor nodes that transmits delay-constrained periodic images in a LoRa network. Considering that large data traffic model is predictable, we propose a priority-group based and channel scheduling methodology using a combined knowledge of the network for informed channel selection instead of a random channel selection. The proposed channel scheduling algorithm is priority-group based where large data transmissions have the highest priority for channel selection. Our simulation results on a LoRa Network Simulator (LoRaSim) indicates an increase in successful data frame reception rates of large data delivered, at data delivery times that satisfies LoRa application-specific deadlines while adhering to the duty-cycle limits imposed