1. Boundary Cutting for Packet Classification Author: Hyesook Lim, Nara Lee, Geumdan Jin, Jungwon Lee, Youngju Choi, Changhoon Yim Publisher: Networking, IEEE/ACM Transactions Presenter: Chih-Hsun Wang Date: 2014/3/4 Department of Computer Science and Information Engineering National Cheng Kung University, Taiwan R.O.C.

2. Introduction National Cheng Kung University CSIE Computer & Internet Architecture Lab 2 The computation of HiCuts and HyperCuts required for the preprocessing consumes much memory and construction time. Moreover, the cutting is based on a fixed interval, which does not consider the actual space that each rule covers; hence it is ineffective. In this paper, we propose a new efficient packet classification algorithm based on boundary cutting.

3. Proposed Algorithm (BC) National Cheng Kung University CSIE Computer & Internet Architecture Lab 3 Building a BC Decision Tree When the cutting of a prefix plane according to rule boundaries is performed, both the starting and the ending boundaries of each rule can be used for cutting.

4. Proposed Algorithm (BC) National Cheng Kung University CSIE Computer & Internet Architecture Lab 4 Fig. 1. Rules in a prefix plane

5. Proposed Algorithm (BC) National Cheng Kung University CSIE Computer & Internet Architecture Lab 5 binth = 3 R 1 (011100, 011000) R 2 (010000, 011000) R 3 (000000, 111000) R 4 (000000, 000000) R 5 (100000, 100000) R 6 (011100, 000000) R 7 (001000, 000000) R 8 (000000, 011000) R 9 (011100, 110000) R 10 (000100, 100000) R 11 (110000, 000000) By padding zeros

6. National Cheng Kung University CSIE Computer & Internet Architecture Lab 6 Proposed Algorithm (BC) Searching in the BC Algorithm Binary search is required to determine the proper edge to follow for a given input. During the binary search, the pointer to the child node is remembered when the input matches the entry value or when the input is larger than the entry value.

7. National Cheng Kung University CSIE Computer & Internet Architecture Lab 7 Proposed Algorithm (BC) Input packet with headers (000110, 111100, 19, 23, TCP) Remembered:

8. National Cheng Kung University CSIE Computer & Internet Architecture Lab 8 Proposed Algorithm (BC) Input packet with headers (000110, 111100, 19, 23, TCP) Remembered: 000100

9. National Cheng Kung University CSIE Computer & Internet Architecture Lab 9 Proposed Algorithm (BC) Input packet with headers (000110, 111100, 19, 23, TCP) Remembered: 100000 111000

10. Proposed Algorithm (Selective BC) In the BC algorithm, if a subspace becomes an internal node, every starting boundary of the rules included in the subspace is used for cutting. We propose a refined structure using the binth to select or unselect the boundary of a rule at an internal node. National Cheng Kung University CSIE Computer & Internet Architecture Lab 10

11. Proposed Algorithm (Selective BC) National Cheng Kung University CSIE Computer & Internet Architecture Lab 11 binth=3

12. Data Structure National Cheng Kung University CSIE Computer & Internet Architecture Lab 12

13. Simulation Results Determine the proper binth value. We first constructed the BC algorithm setting binth as an arbitrary small number and then identified jumbo leaves, those here that have more rules than binth. Among the number of rules included in the jumbo leaves, the largest number is identified, and we name the number lower bound binth, meaning that binth should not be smaller than this number to avoid unnecessary cutting. National Cheng Kung University CSIE Computer & Internet Architecture Lab 13

14. Simulation Results Compared Method BC SBC HiCuts with node merging HyperCuts_1 With node merging, space compaction, pushing upward during the building of the decision tree HyperCuts_2 With node merging, space compaction, pushing upward after the building of the decision tree National Cheng Kung University CSIE Computer & Internet Architecture Lab 14

15. Simulation Results National Cheng Kung University CSIE Computer & Internet Architecture Lab 15

16. Simulation Results National Cheng Kung University CSIE Computer & Internet Architecture Lab 16

17. Simulation Results National Cheng Kung University CSIE Computer & Internet Architecture Lab 17

18. Simulation Results National Cheng Kung University CSIE Computer & Internet Architecture Lab 18 Search performance -- ACL Worst caseAverage case

19. Simulation Results National Cheng Kung University CSIE Computer & Internet Architecture Lab 19 Search performance -- IPC Worst caseAverage case

20. Simulation Results National Cheng Kung University CSIE Computer & Internet Architecture Lab 20 Search performance -- FW Worst caseAverage case