Congestion Control for Radio Networks

Abstract

Nowadays, we have witnessed an exponential growth of the traffic generated by users in mobile communications networks. This need for a greater capacity in mobile networks raises problems for operators, because the demand for network capacity is not always accompanied by the evolution of the available technologies. This represents problems in access networks that result in failures in the customer service and losses to the operator. It is quite important that the service is the best possible for both client and the operator: preventing and resolving network congestion is a common interest. The quality of service is also an important point of great concern, that requires to evaluate the service provided in order to achieve the best satisfaction from the user. However, many challenges still need to be surpassed on current mobile networks.

This paper identifies and resolves cases of congestion in mobile networks, as well as it infers the quality of services provided to the customer. It proposes and implements an architecture that is able to identify problems in access networks, through network monitoring, using metrics gathered by a network management system that gives Key Performance Indicators (KPIs) evaluating the network, and through clients’ information for a specific cell, obtained through probing. Then, a Call Admission Control (CAC) module in the 3GPP Policy and Charging Rules Function (PCRF) is in charge of applying a rule set designed to resolve the case of the detected congestion. The policies chosen directly affect the users in the congested cell, causing the levels of occupancy to drop, and consequently, the cell resources to become stable. Several congestion use cases are identified and demonstrated, which show the effectiveness of the rules applied to the network, services and users.