Intellegent Dynamic Timeout For Efficient Flow Table Managemnet In Software Defined Satellite Network

Satellite communication play a vital role in information transmission and it has been widely used for military purposes,live television,weather broadcast and internet of things.However,traditional satellite network led the resources underutilization in terms of advance satellite technologies and it does not guarantee the requirements of future information transmission.Software Defined Network(SDN)is a new network paradigm which change the architecture of current distributed network into centralized control architecture and this concept has used to change the architecture of traditional satellite network into Software Defined Satellite Network(SDSN).This research focuses on deployment of Open Flow protocol and presents a novel threeplanes architecture of SDSN.The proposed SDSN model of this work is a combination of satellites and ground stations and it is divided into three planes: Data plane,Control plane and Management plane.The satellite network is a part of the data and control planes while the ground stations is a part of the management plane.Low Earth Orbit(LEO)satellites reside in data plane for data forwarding function and Geostationary Earth Orbit(GEO)satellites reside in control plane for controlling the whole network.We consider two serious issues faced by SDSN.The first issue is the limited resources of Ternary Content Addressable Memory(TCAM)in SDSN due to which connection of more users to network is difficult.The second issue is the frequent handover which occurs in LEO satellite.When satellite handover occurs the flow rule entries from previous cell may reside in flow table and thus it increases the size of flow table which rises the chance of flow drop.To solve these issues proper management of flow table is necessary.This work proposes a heuristic algorithm of Intelligent Dynamic Timeout(IDT)for efficient use of satellite resources and proper flow table management.The algorithm considers the key points of limited TCAM space,traffic classification according to different priorities of users and satellite link handover to predict a dynamic idle timeout based on the occupied space of flow table in TCAM of the switch.This dynamic idle timeout is used to evict the unused flow rule entries from flow table and keep TCAM space free to connect more users to the network.Simulation results and performance analysis show the best performance achieved in terms of reduction in number of flow table entries to ensure the connection of more users to the network,reduction in number of packets drop and reduction in number of Table Miss Packets.The average size of flow table and the average drop flow rate are reduced by 39.55% and 11.2 % respectively when comparing the performance of IDT with static idle timeout 50 Seconds.Similarly,average decrease in number of Table Miss Packets is 10.18% when comparing the performance of static idle timeout 15 Seconds with IDT.