Towards Network Automation: A Multi-Agent Based Intelligent Networking System

This P.h.D thesis has three parts. The first part is the mathematical modeling of softwarized network. We studied network softwarzation through Virtual Network Function (VNF) placement considering 5G and B5G’s stringent requirements of latency, reliability, and support for heterogeneous devices. Since the existing wireless network architecture is limited to fulfill these constraints, a cloud radio access network (C-RAN), along with network function virtualization, is suggested to provide flexibility and network agility. C-RAN decouples network functions, such as firewall and packet gateway, from hardware to software deployed in the cloud. Thus, a comprehensive end-to-end formulation of this architecture is required for VNF placement. Most of the existing works focus on virtual function placement with different objectives, addressing different service requirements separately. Six 5G constraints are considered simultaneously to find optimal VNF placement with service differentiation. The selected six parameters reflect services' requirements, network constraints, and computing constraints. We first model the overall cloud radio access network as a multi-layer loopless-random hypergraph and we provide the overall formulation of the system. Then, we reformulate such a model considering backup virtual functions and CPU over-provisioning techniques to improve both virtual function's reliability and processing latency. Finally, we propose service differentiation to reduce CPU utilization and energy consumption, while using the above techniques. The results suggest that the application of service differentiation can significantly improve the assignment of computing resources and energy efficiency. We have also demonstrated a VNF design for IoT interoperability developing a translator as a VNF. Furthermore, we have tested the emulation of LTE, LTE-A, and 5G over a lightweight open platform considering delay analysis. The second part of the work focus on network automation. The advent of network softwarization is enabling multiple innovative solutions through software-defined networking (SDN) and network function virtualization (NFV). Specifically, network softwarization paves the way for autonomic and intelligent networking, which has gained popularity in the research community. Along with the arrival of 5G and beyond, which interconnects billions of devices, the complexity of network management is significantly increasing both investments and operational costs. Autonomic networking is the creation of self-organizing, self-managing, and self-protecting networks, to manage complex and heterogeneous networks. To achieve full network automation, various aspects of networking need to be addressed. So, we proposed a novel architecture called multi-agent-based network automation of the network management system (MANA-NMS). The architecture rely on network function extit{atomization}, which defines extit{atomic} decision-making units. Such units could represent VNFs. These extit{atomic} units are autonomous and adaptive. In this part, first, we present a theoretical discussion of the challenges arisen by automating the decision-making process. Next, the proposed multi-agent system is presented along with its mathematical modeling. And then MANA-NMS architecture is mathematically evaluated from functionality, reliability, latency, and resource consumption performance perspectives. As an example for extit{atomic} agent design, we have developed an autonomous network traffic classifier agent (NTCA). We design and implement an NTCA using a machine learning algorithm as a cognitive component of the agent. To compare, we used K-Nearest Neighbors (K-NN), Decision Tree, Support Vector Machine (SVM), and Naive Bayes in the agent design. We perform an evaluation using classification accuracy, training latency, and classification latency. We also tested the performance of the NTCA by implementing it in the MANA-NMS conceptual framework. The third part of this P.h.D. work is to use MANA-NMS principles to decomposition SDN controllers and other monolithic systems and incorporate intelligence in the subfunctions, creating loosely coupled units in the service-oriented architecture. The existing controllers are monolithic, resulting in code inefficiency for distributed deployment. extit{microONOS} controller has been proposed, showing a decomposed controller architecture into logical subfunctions. These functions are implemented as microservices and deployed as VNF, enabling flexible deployment. However, the microONOS controller is in early-stage development and the full controller decomposition is not availed. Moreover, the communication interface between the decomposed components of the controller is based on gRPC. Our proposed architecture implements Ryu controller decomposition. In the decomposition, we used REST API as a communication interface between the decomposed functions. Moreover, we compared the performance using gRPC and WebSocket. We also further proposed a multi-agent architecture for the next-generation network. In this regard, recently, the 3GPP standard defines the service-based architecture (SBA) framework, where the architecture elements are defined in terms of Network Functions (NFs). This approach provides flexibility in terms of dynamic scaling and backup deployment of functions. However, to fully utilize the flexibility and dynamicity that the architecture provides, intelligence should be introduced in the decomposed functions. Here, we propose to unify well-defined standards for the 5G architecture such as Software-defined Networking (SDN), ETSI network function virtualization (NFV), ETSI generic autonomic networking architecture (GANA), and ETSI multi-access edge computing (MEC) in a unified intelligent architecture. Moreover, we define network functions and applications as {atomic units}, as in the case of MANA-NMS. Using these agents as building blocks, we provide an intelligent pool of networking resources and applications that can collaborate to form next-generation architectures for future 6G networks.