An End-to-End Overload Control Method in SIP Networks

SIP protocol is considered as signaling protocol for IP multimedia subsystem (IMS) and it is introduced by 3GGP as signaling foundation in the next generation networks (NGN). SIP protocol despite having features such as: text based, IP based, independent of the data transmission, support for mobility and end-to-end has not the suitable mechanism for dealing with Overload. This challenge will cause that a wide range of NGN users face with a sharp dropin the quality of service. Multi agent systems are suitable method for solving the complex systems. Since a distributed SIP network is a complex system, which is composed of subsystems that interacting with each other we propose multi agent system for overload control in SIP networks with each scale. An approach to reduce the complexity of such systemsis using organization based multi agent system. In this paper, we use an organization called holonic multi agent system (HMAS) to model a large SIP network. Each holon covers a geographical area of the SIP network in the first level. Holons in the second level control the first level holons and third level holons controls the second level holons hence the hierarchical structure is formed for holons. The overload control is achieved by communication and knowledge exchange between the holons. Experimental results show that the holonic multi agent overload control prevent overload in SIP network while it causes increase overall throughput and reduce delay

Keywords:

SIP Network , Overload Control , Holonic Multi-Agent