SYSTEMS AND METHODS FACILITATING USER LEVEL TRAFFIC TYPE CLASSIFICATION IN PACKET DATA NETWORKS FOR NETWORK RESOURCE OPTIMIZATION IN OPEN RADIO ACCESS NETWORKS

Smart overview of the Invention

The patent application details a system designed to enhance network resource optimization within open radio access networks (ORAN) by classifying user-level traffic types in packet data networks. This system involves acquiring data traffic from an open radio unit network (ORAN) associated with user equipment (UE) and analyzing it through a traffic classification application, known as a rApp. The rApp classifies the data traffic into various service types, which then informs the allocation of RAN slices to optimize network resources for these service types. The allocation decisions are communicated to a near-real-time radio access network intelligent controller (near-RT RIC) to implement the necessary adjustments in the network.

Background

Network providers face increasing challenges as the demand for network services grows, leading to potential slowdowns, data loss, and performance issues. The rise of encrypted network traffic adds complexity, making it difficult for providers to understand and manage traffic effectively. Without proper traffic classification, optimal utilization of network resources is limited, impacting the quality of service experienced by users. This system aims to address these issues by enabling better traffic classification and management.

System Description

The system employs a processing system with a processor and memory to execute operations that include acquiring and analyzing data traffic from ORANs. A traffic classification application, or rApp, classifies this data into service types, which guides the allocation of RAN slices optimized for these services. The system then transmits these allocations to a near-RT RIC, which implements them across open distributed units (O-DUs) and open radio units (O-RUs) in the ORAN. This approach allows for efficient network resource allocation, improved traffic management, and enhanced user experience.

Technical Implementation

The patent describes a non-transitory machine-readable medium containing executable instructions for a processing system. These instructions facilitate operations such as receiving data traffic through an O1 interface, classifying it into service requirements, and transmitting resource slice allocations via an A1 interface to a near-RT RIC. The system processes real-time data traffic using a machine learning model to determine optimized network resource slices and generates policies to guide their implementation.

Benefits and Challenges

By classifying user-level traffic types, network providers can gain insights into usage patterns and better manage network traffic, even when encrypted. This capability allows providers to offer tailored services, enhancing the user experience. However, the increasing volume and encryption of network traffic present ongoing challenges in maintaining effective traffic management. The proposed system seeks to address these challenges by optimizing network resources and improving the understanding and management of network traffic.