Energy constrained optimization for spreading factor allocation in lorawan

Abstract

This paper discusses a spreading factor allocation for Long Range Wide Area Network (LoRaWAN). Because Long Range (LoRa) is based on chirp spread spectrum that each spreading factor is approximately orthogonal to each other, the performance of LoRaWAN can be enhanced by allocating the spreading factor appropriately to end devices (EDs). Several spreading factor allocation techniques have been reported. Techniques shown in existing studies can improve some characteristics (e.g. throughput or packet reception probability (PRP)); however, there are a few studies that have focused on the energy consumption of the EDs. The LoRa communication offers a low power communication and this enables the improvement of the performance in exchange for the energy consumption. This paper presents a performance improvement technique via spreading factor allocations for LoRaWAN. We define the optimization problem for the spreading factor allocation to maximize the PRP under a constraint for the average energy consumption of all the EDs. It enables for the performance improvement under the constraint of the average energy consumption of all the EDs by solving the problem. This study further develops a method to solve the defined problem based on a distributed genetic algorithm, which is metaheuristics method. Although the techniques shown in the existing studies give the average energy consumption as a result of the performance improvement by the spreading factor allocation, the presented technique can enhance the LoRaWAN performance by allocating the spreading factor to EDs under the constraint for the average energy consumption of all the EDs. Numerical examples validate the effectiveness of the presented technique. The PRP performance of the presented technique is superior to that of the techniques shown in the existing studies despite that the average energy consumption of all the EDs of the presented technique is less than that of the techniques shown in the existing studies.