|a英文題名:Investigation of Parallel Multiple Pattern Matching Algorithm based on GPU for Network Packet Payload Scanning
320
|a參考書目
328
|a碩士論文--靜宜大學資訊管理學系
330
|a隨著網路科技的蓬勃發展,許多新型態的網路服務被開發用以提供各種應用程式以及計算功能,這些服務皆仰賴網際網路進行連結,因此,透過靈活的封包分類系統來進行封包分析是網路安全議題當中相當重要的一環,封包分析是需要大量的計算時間的密集型應用程序。因此,我們提出一個可以比對不同長度模式的比對方法,使用GPGPU設備來進行封包內容特徵比對之加速,並於不同的高性能運算平台上進行性能比較,實驗結果證明,我們的方法比起透過CPU平台,可以達到超過十倍的性能差異。|uWith the rapid growth of network technologies, many new web services have been developed to provide various applications and computing functions. These services rely deeply on the internet. Therefore, packet analysis is an important issue of network security that typically adopts a flexible packet filtering system to classify each processed packet. Packet analysis is a computation-intensive application that requires hung computing time to process large amount of internet packets. Hence, we propose an efficient method to classify packets with pattern of various lengths simultaneously by using GPGPU device. We also evaluate the performance of implement on various memory architectures of GPGPU device. The experiment results demonstrate that the proposed method can achieve 10X speed up over the sequential packet classification software on single CPU.
517
1
|aInvestigation of Parallel Multiple Pattern Matching Algorithm based on GPU for Network Packet Payload Scanning