Now a days, emerging trends in the field of wireless sensor networks (WSNs) tend to work on more complex scenarios and flexible network models as the conventional WSN systems that are based on a classical arrangement of sensors. Generally, these networks have different limitations such as control node election, data aggregation, load balancing during data collection etc. The load balancing depends on the effective routing techniques which provide an optimum path to transmit the data such that the minimum amount of energy should be consumed. The control nodes are responsible for assigning the task and data transmission in the cluster-based routing techniques and the selection of the control node is an NP-hard problem. To resolve this problem, an adaptive particle swarm optimization (PSO) ensemble with genetic mutation-based routing is proposed to select control nodes for IoT based software-defined WSN. The proposed algorithm plays a significant role in selecting the control nodes by considering energy and distance parameters. The proposed work is implemented for the heterogeneous networks having different computing power accompanied by single and multiple sinks. The experiment was carried out on the scale of the performance matrix such as fitness value, stability period, average residual energy, etc. The simulation result of the proposed algorithm outperforms over other algorithms under the different arrangements of the network.

现在，无线传感器网络 (WSN) 领域的新兴趋势倾向于在更复杂的场景和灵活的网络模型上工作，就像基于经典传感器布置的传统 WSN 系统一样。通常，这些网络具有不同的限制，例如控制节点选举、数据聚合、数据收集期间的负载均衡等。负载均衡取决于有效的路由技术，这些路由技术提供了传输数据的最佳路径，从而使能量应该最小化。消耗。在基于集群的路由技术中，控制节点负责分配任务和数据传输，控制节点的选择是一个NP-hard问题。为了解决这个问题，提出了一种具有基于基因突变的路由的自适应粒子群优化 (PSO) 集成，用于为基于物联网的软件定义 WSN 选择控制节点。所提出的算法通过考虑能量和距离参数在选择控制节点方面起着重要作用。所提出的工作是针对具有不同计算能力的异构网络实现的，这些网络伴随着单个和多个接收器。实验在适应度值、稳定期、平均剩余能量等性能矩阵的尺度上进行了实验，在网络的不同安排下，所提算法的仿真结果优于其他算法。所提出的算法通过考虑能量和距离参数在选择控制节点方面起着重要作用。所提出的工作是针对具有不同计算能力的异构网络实现的，这些网络伴随着单个和多个汇。实验在适应度值、稳定期、平均剩余能量等性能矩阵的尺度上进行了实验，在网络的不同安排下，所提算法的仿真结果优于其他算法。所提出的算法通过考虑能量和距离参数在选择控制节点方面起着重要作用。所提出的工作是针对具有不同计算能力的异构网络实现的，这些网络伴随着单个和多个接收器。实验在适应度值、稳定期、平均剩余能量等性能矩阵的尺度上进行了实验，在网络的不同安排下，所提算法的仿真结果优于其他算法。