I need a research assistant to do research on packet sniffer. Need to proposed / design efficient packet sniffer to detect and classify normal packet and packet contain anomaly (can focus on current threat ie online banking threat like phishing or 4 major attacks ie dos, probe, U2R, R2L). Preferable use hybrid method SOM + MCDM (multi criteria decision making). For scoring method/ prioritization, need to specify in details the criteria & alternatives..Suggestion of other method/technique can also be discussed.. Need justification for every method used. Accuracy is a very important issue as we need to compare result with existing IDS. Need to create scenario/matrix in order to test the proposed design.