Space-Air-Ground Integrated Multi-Domain Network Resource Orchestration Based on Virtual Network Architecture A DRL Method
Abstract : Traditional ground wireless communication networks cannot provide high-quality services for artificial intelligence (AI) applications such as intelligent transportation systems (ITS) due to deployment, coverage and capacity issues. The space-air-ground integrated network (SAGIN) has become a research focus in the industry. Compared with traditional wireless communication networks, SAGIN is more flexible and reliable, and it has wider coverage and higher quality of seamless connection. However, due to its inherent heterogeneity, time-varying and self-organizing characteristics, the deployment and use of SAGIN still faces huge challenges, among which the orchestration of heterogeneous resources is a key issue. Based on virtual network architecture and deep reinforcement learning (DRL), we model SAGIN's heterogeneous resource orchestration as a multi-domain virtual network embedding (VNE) problem, and propose a SAGIN cross-domain VNE algorithm. We model the different network segments of SAGIN, and set the network attributes according to the actual situation of SAGIN and user needs. In DRL, the agent is acted by a five-layer policy network. We build a feature matrix based on network attributes extracted from SAGIN and use it as the agent training environment. Through training, the probability of each underlying node being embedded can be derived. In test phase, we complete the embedding process of virtual nodes and links in turn based on this probability. Finally, we verify the effectiveness of the algorithm from both training and testing.
? Most existing works focus on evaluating either only one single network segment in space, air, or ground, or the integration of space-ground network or air-ground network.
? There exist extensive works evaluating the performance of terrestrial communications by using MATLAB/Simulink, Network Simulator NS.3, OPNET network simulation platform, and so forth.
? A SAGIN simulation platform should not only simulate the communication and networking protocols, but also support various existing and potential SAG applications and services.
? It is easy to modify the existing modules or add new modules to extend the functionality of the simulation platform.
? AI-based techniques are then investigated to address intractable research problems in the SAGVN to enhance resource management efficiency.
? Different AI methods are suitable for different SAGVN resource management problems because of the diversified approaches, input data requirements, and outputs.
? The high dimension of the input information makes the problem intricate and can hamper the convergence performance when adopting AI-based methods.
? We propose an AI-based two-layered optimization scheme to solve the joint optimization problem with different optimization granularities.
• In general, the proposed integrated network architecture may provide a good platform for future network standardization, which is expected to be reliable, real-time, efficient and safety.
• The optimized bandwidth solution was proposed to enable heterogeneous deployment of UAV based floating relay cells inside the macro cell and achieved dynamic and adaptive coverage.
• A genetic algorithm was proposed to optimize the positions of UAV-BSs with the goal of maximizing the network throughput.
• An intelligent solution utilizing the priority-wise dominance and the entropy approaches was proposed for the accurate and efficient placement of the UAVs.
? In urban/suburban scenarios, on the other hand, the resource management is sophisticated and SDN controllers are required to enhance resource utilization efficiency and optimize the network performance.
? In particular, the RSCs can act as actors and independently make decisions based on local information, and LC functions as a global critic to provide performance feedback to RSCs within its coverage range.
? Predicting the individual vehicle mobility is of vital importance for high resource scheduling performance.
? Therefore, in this article, we focus on the efficient resource management in the SAGVN with an SDN-based control architecture to facilitate programmable, scalable, and adaptable network optimization.
|