Local SIP overload control: Controller design and optimization by Extremum Seeking

The Session Initiation Protocol (SIP) is a signaling framework allowing two or more parties to establish, alter, and terminate various types of media sessions. An open issue is the proper handling of overload situations that in SIP servers occur when the incoming flow of requests overcomes the processing resources. Due to overload, call establishment times increase and retransmissions are triggered causing an uncontrolled increase of the total input rate. In this paper we present a local overload control system, made of two PI controllers, aiming at regulating both the queue length and the CPU load of the SIP proxy to provide high goodput and low call establishment delays. The proposed control system has been implemented in Kamailio (OpenSER) and the controllers parameters have been tuned by employing the Extremum Seeking algorithm to minimize a cost function. A performance evaluation and comparison with Ohta and Occupancy local overload controllers has shown the following main results: 1) the proposed controller is able to counteract overload situations providing a goodput which is close to the optimal while maintaining low call establishment delays and retransmission ratios; 2) the proposed control system significantly outperforms Ohta and OCC both in terms of goodput, call establishment delays and retransmission ratios.