This paper first analyzes the existing optimization models and methods based on low-latency and high-concurrency 5G wireless sensor networks and focuses on some research results based on the physical layer, data link layer, and network layer. After that, the energy-constrained cross-layer optimization algorithm of the joint physical layer and network layer and the cross-layer optimization algorithm based on reliable transmission are analyzed and compared. According to the ideological and political education cycle and network connectivity of wireless sensor networks, the existing definition of the ideological and political education cycle of wireless sensor networks was questioned and its irrationality was demonstrated. Firstly, we introduce the system model, energy capture model, and energy consumption model of this scheme; then, calculate the energy consumption of the node sending and receiving data and analyze the maximum throughput that the node can provide for data transmission to perform mathematical modeling. In a balanced ideological and political education cycle, the optimal solution is given. Finally, it introduces distributed high-throughput routing for EH-WSNs based on low latency and high concurrency and gives a performance comparison with the shortest path routing scheme. The energy consumption of each part of the node is analyzed in a comprehensive and in-depth manner, and the direction of the energy optimization strategy of the network node and its sensor network is pointed out. On this basis, this article proposes a new definition of the network ideological and political education cycle: the ideological and political education cycle of the wireless sensor network is equal to the ideological and political education cycle of the first joint in the network that fails due to energy exhaustion. Based on this definition, the wireless sensor network is analyzed, and the cross-layer (physical layer, data link layer, and network layer) optimization mathematical model of the wireless sensor network ideological and political education cycle is established. The two data collection schemes proposed in this paper are simulated under the conditions of no-delay ideological and political education cycle and different time-delayed ideological and political education cycles. The simulation results show that the model is better than both in terms of network life and data integrity.