الفهرس | Only 14 pages are availabe for public view |
Abstract During the last few years, Software-Defined Networking (SDN) gets the focus of network vendors. Nowadays, many researchers turned to do a lot of researches on this paradigm. SDN decouples the control plan of data plane. So, forwarding devices become simple packet filtering and forward devices in which hold flow tables with flow rules. SDN supports networks with programmability through executing applications or services with packets. It promises to solve today network issues like vendor dependency and high cost. The intelligence in SDN is located on external server. So, switches became very simple in packet forwarding and high-speed in packet filtering and classification. This is also due to the Ternary Content Addressable Memory (TCAM) chips that are located within each device in which use parallel search technology. TCAMs is the standard chips for packet filtering and classification.In fact, TCAMs have some limitations such as power consumption, heat generation, delay, limited size and high cost. All these factors affect SDN scalability. Thus, many researchers proposed solutions for TCAMs OpenFlow rules minimization either through software or hardware solutions.In this work, a redundancy removal algorithm is introduced. The model consists of some match cases in that depends on rules overlapping. The proposed algorithm aims at reducing the total number of TCAM OpenFlow table entries through merging the overlapped rules. Real-life packet filtering and classification system is used from medium enterprise and the compression ratio is calculated to test the efficiency of proposed algorithm. As a result of the experiment, our proposed solution achieves total compression ratio 96%. |