A parallel processing packet inspection by centralized multiple robust distribution system

Many emerging network application have been using pattern matching for detecting malicious intrusion in the computer networks model multiple platform which becoming unbalanced size distribution and create workload size distribution and create workload hard in the CPU which resulting low performance scalability. However uneven network flow size are being increased for retaining the packet order in each flow and the existing parallel processing model is basic unit for partition the workload which cannot be quite advantage of multicore platform, therefore it creates low CPU utilization and increasing number of CPU in the network. For overcome this problem we present in this paper we proposed a novel parallel processing inspection model which is called CMRDS (Centralized multiple robust distribution system). This implies balanced parallel detection by adding multiple dimensions on pattern matching workload division from various different types of nodes. CMRDS allows multiple pattern matching algorithm and creates simultaneously different pattern set and handle variable pattern length efficiently for packet inspection is completed in O(N log M) and N is the size of the packet M is the longest packet length inspect packet for optimizing flow in the different network. According to this algorithm subset divisions are descended in the packet for maximum use of core utilization and avoiding unnecessary overhead. The simulation experiment result is achieving for increasing distinct performance against traditional balancing.