Article Details

Abstract

Aiming at the challenges of strong network dynamics, multi-objective conflicts and poor Quality of Service (QoS) guarantee faced by real-time audio-video transmission in emergency communication scenarios, a hybrid intelligent routing mechanism integrating Deep Reinforcement Learning (DRL) and multi-population hierarchical Ant Colony Optimization (ACO) is proposed. Firstly, this paper establishes a stochastic time-varying model of emergency networks, uses Markov chains to characterize link availability evolution, and describes performance parameter fluctuation characteristics through first-order autoregressive processes. Secondly, it constructs a two-layer collaborative architecture of Deep Dueling Double Q-Network (D3QN) and multi-population ant colony algorithm, and realizes deep fusion of value assessment and path search through Q-value driven pheromone modulation mechanism. Furthermore, a dual-gate stability control mechanism is designed to effectively suppress frequent path switching while maintaining rapid response. The experimental results show that the proposed method reduces the average end-to-end delay, packet loss rate, and path switching frequency by 43.2%, 65.4%, and 67.5% respectively compared to the shortest path algorithm, while increasing bandwidth utilization by 78.9%, demonstrating excellent adaptability and robustness in burst congestion and link failure scenarios.

Keywords

emergency communication; Deep Reinforcement Learning; Ant Colony Optimization; routing algorithm

References

1. [1] YANG A P，HUAN J，WANG Q，et al. ST-D3QN: Advancing UAV Path Planning with an Enhanced Deep Reinforcement Learning Framework in Ultra-low Altitudes [J].IEEE Access，2025，13：65285-65300.
2. [2] HE Y H， XIAO G Y，ZHU J，et al. Reinforcement Learning-based SDN Routing Scheme Empowered by Causality Detection and GNN [J/OL].Frontiers in Computational Neuroscience，2024，18：1393025（2024-04-29）[2025-06-18].https://doi.org/10.3389/fncom.2024.1393025.
3. [3] ZHANG J，XU H，LIU D，et al. Advancing Dynamic Emergency Route Optimization with a Composite Network Deep Reinforcement Learning Model [J/OL].Systems，2025，13（2）：127（2025-02-17）[2025-06-18].https://doi.org/10.3390/systems13020127.
4. [4] CASAS-VELASCO D M，RENDON O M C，DA FONSECA N L S. DRSIR: A Deep Reinforcement Learning Approach for Routing in Software-defined Networking [J].IEEE Transactions on Network and Service Management，2022，19（4）：4807-4820.
5. [5] CUI J G，WU L，HUANG X D，et al. Multi-strategy Adaptable Ant Colony Optimization Algorithm and its Application in Robot Path Planning [J/OL].Knowledge-Based Systems，2024，288：111459（2024-03-15）[2025-06-18].https://doi.org/10.1016/j.knosys.2024.111459.
6. [6] HAN H Z，TANG J，JING Z J. Wireless Sensor Network Routing Optimization Based on Improved Ant Colony Algorithm in the Internet of Things [J/OL].Heliyon，2024，10（1）：e23577（2024-01-15）[2025-06-18].https://doi.org/10.1016/j.heliyon.2023.e23577.
7. [7] NIE W Z，CHEN Y，WANG Y H，et al. Routing Networking Technology Based on Improved Ant Colony Algorithm in Space-air-ground Integrated Network [J/OL].EURASIP Journal on Advances in Signal Processing，2024，2024：34（2024-03-13）[2025-06-18].https://doi.org/10.1186/s13634-024-01131-5.
8. [8] TAWFEEK M A，ALRASHDI I，ALRUWAILI M，et al. Improving Energy Efficiency and Routing Reliability In Wireless Sensor Networks Using Modified Ant Colony Optimization [J/OL].EURASIP Journal on Wireless Communications and Networking，2025，2025：22（2025-04-08）[2025-06-18].https://doi.org/10.1186/s13638-025-02449-w.
9. [9] GÖK M. Dynamic Path Planning via Dueling Double Deep Q-network (D3QN) with Prioritized Experience Replay [J/OL].Applied Soft Computing，2024，158：111503[2025-06-18].https://doi.org/10.1016/j.asoc.2024.111503.
10. [10] LIAO X，LI L，HUANG C X，et al. Noisy Dueling Double Deep Q-network Algorithm for Autonomous Underwater Vehicle Path Planning [J/OL].Frontiers in Neurorobotics，2024，18：1466571（2025-10-14）[2025-06-18].https://doi.org/10.3389/fnbot.2024.1466571.
11. [11] MAO K. Multi-agent Reinforcement Learning Driven Resource Game Optimization for Network Slicing in MEC-enabled HetNets [J/OL].Scientific Reports，2025，16：3231（2025-12-23）[2025-06-18].https://www.nature.com/articles/s41598-025-33190-5.
12. [12] MA C D，LI A，DU Y L，et al. Efficient and Scalable Reinforcement Learning for Large-scale Network Control [J].Nature Machine Intelligence，2024，6：1006-1020.
13. [13] HYDHER H，JAYAKODY D N K，HEMACHANDRA K T，et al. UAV Deployment in WSN System for Emergency/Remote Area Applications [J/OL].Computer Networks，2025，257：110977[2025-06-18].https://doi.org/10.1016/j.comnet.2024.110977.
14. [14] KHAN M A，ULLAH F，ALSIRHANI A，et al. Communication Architecture and Operations for SDR-enabled UAVs Network in Disaster-stressed Areas [J/OL].Ad Hoc Networks，2024，160：103506（2024-07-01）[2025-06-18].https://doi.org/10.1016/j.adhoc.2024.103506.
15. [15] JIANG W W，ZHANG Y，HAN H，et al. Network Simulation Tools for Unmanned Aerial Vehicle Communications: A Survey [J/OL].International Journal of Communication Systems，2024，37（15）：e5878（2024-06-17）[2025-06-18].https://doi.org/10.1002/dac.5878.
16. [16] CHANDRAN I，VIPIN K. Multi-UAV Networks for Disaster Monitoring: Challenges and Opportunities from a Network Perspective [J/OL].Springer Nature Computer Science，2024，5：519（2024-05-02）[2025-06-18].https://link.springer.com/article/10.1007/s42979-024-02788-3.
17. [17] WANG S，SONG R，ZHENG X，et al. A3C-R: A QoS-oriented Energy-saving Routing Algorithm for Software-defined Networks [J/OL].Future Internet，2025，17（4）：158（2025-04-03）[2025-06-18].https://doi.org/10.3390/fi170401
18. 58.

About the authors

Chen Kun (1981-), male, Han nationality, from Suining, Sichuan, holds an intermediate professional title and a bachelor's degree. His research direction is emergency communication.