StreamLens/debug_specific_outlier.py

#!/usr/bin/env python3
"""Debug specific outlier around frame 1001"""

import sys
sys.path.append('.')

from analyzer.analysis import EthernetAnalyzer
from analyzer.utils import PCAPLoader

def debug_specific_outlier(pcap_file="1 PTPGM.pcapng", src_ip="192.168.4.89"):
    """Debug specific outlier around frame 1001"""
    
    print("=== Debugging Specific Outlier Around Frame 1001 ===")
    
    # Initialize analyzer
    analyzer = EthernetAnalyzer(enable_realtime=False, outlier_threshold_sigma=3.0)
    
    # Load and process packets
    loader = PCAPLoader(pcap_file)
    packets = loader.load_all()
    
    print(f"Loaded {len(packets)} packets")
    
    # Process packets
    for i, packet in enumerate(packets, 1):
        analyzer._process_single_packet(packet, i)
    
    # Calculate statistics
    analyzer.calculate_statistics()
    
    # Find the test flow
    test_flow = None
    for flow_key, flow in analyzer.flows.items():
        if flow.src_ip == src_ip:
            test_flow = flow
            break
    
    if not test_flow:
        print(f"❌ No flow found from {src_ip}")
        return
    
    print(f"\n✅ Found flow: {test_flow.src_ip}:{test_flow.src_port} → {test_flow.dst_ip}:{test_flow.dst_port}")
    
    # Check all frame types for outliers around frame 1001
    target_frame = 1001
    print(f"\n=== Searching for outliers around frame {target_frame} ===")
    
    for frame_type, ft_stats in test_flow.frame_types.items():
        if hasattr(ft_stats, 'enhanced_outlier_details') and ft_stats.enhanced_outlier_details:
            for frame_num, prev_frame_num, delta_t in ft_stats.enhanced_outlier_details:
                if abs(frame_num - target_frame) <= 5:  # Within 5 frames of target
                    deviation = (delta_t - ft_stats.avg_inter_arrival) / ft_stats.std_inter_arrival if ft_stats.std_inter_arrival > 0 else 0
                    print(f"  {frame_type}: Frame {frame_num} (from {prev_frame_num}): {delta_t * 1000:.3f} ms ({deviation:.1f}σ)")
    
    # Also check the raw outlier data for any issues
    print(f"\n=== All CH10-Data Outliers ===")
    ch10_data_stats = test_flow.frame_types.get('CH10-Data')
    if ch10_data_stats and hasattr(ch10_data_stats, 'enhanced_outlier_details'):
        print(f"Total CH10-Data outliers: {len(ch10_data_stats.enhanced_outlier_details)}")
        for i, (frame_num, prev_frame_num, delta_t) in enumerate(ch10_data_stats.enhanced_outlier_details):
            deviation = (delta_t - ch10_data_stats.avg_inter_arrival) / ch10_data_stats.std_inter_arrival if ch10_data_stats.std_inter_arrival > 0 else 0
            print(f"  {i+1}. Frame {frame_num} (from {prev_frame_num}): {delta_t * 1000:.3f} ms ({deviation:.1f}σ)")
    
    # Let's also check if there might be confusion between different data sources
    # Check if there are any outlier frames with frame# around 1001 and prev_frame# around 49
    print(f"\n=== Searching for any outlier with prev_frame_num around 49 ===")
    found_suspicious = False
    for frame_type, ft_stats in test_flow.frame_types.items():
        if hasattr(ft_stats, 'enhanced_outlier_details') and ft_stats.enhanced_outlier_details:
            for frame_num, prev_frame_num, delta_t in ft_stats.enhanced_outlier_details:
                if prev_frame_num >= 45 and prev_frame_num <= 55:  # Around 49
                    deviation = (delta_t - ft_stats.avg_inter_arrival) / ft_stats.std_inter_arrival if ft_stats.std_inter_arrival > 0 else 0
                    print(f"  {frame_type}: Frame {frame_num} (from {prev_frame_num}): {delta_t * 1000:.3f} ms ({deviation:.1f}σ)")
                    found_suspicious = True
    
    if not found_suspicious:
        print("  No outliers found with prev_frame_num around 49")
    
    # Check the frame sequence around 1001 to understand the context
    print(f"\n=== Frame sequence context around {target_frame} ===")
    ch10_data_stats = test_flow.frame_types.get('CH10-Data')
    if ch10_data_stats:
        if target_frame in ch10_data_stats.frame_numbers:
            frame_index = ch10_data_stats.frame_numbers.index(target_frame)
            start_idx = max(0, frame_index - 2)
            end_idx = min(len(ch10_data_stats.frame_numbers), frame_index + 3)
            
            print(f"CH10-Data frames around index {frame_index}:")
            for i in range(start_idx, end_idx):
                marker = " -> " if i == frame_index else "    "
                ts = ch10_data_stats.timestamps[i] if i < len(ch10_data_stats.timestamps) else "N/A"
                print(f"{marker}[{i}] Frame {ch10_data_stats.frame_numbers[i]}: {ts}")

if __name__ == "__main__":
    if len(sys.argv) > 1:
        debug_specific_outlier(sys.argv[1])
    else:
        debug_specific_outlier()