KPI Packet loss and Jitter

In both Telco and IP network domains, maintaining consistent application performance depends heavily on understanding packet delivery reliability and latency behavior across dynamic network segments. Whether in a 5G Core, Data Center (DC), or FTTH aggregation network, Aviz Service Node delivers precise KPI measurements for TCP Packet Loss and TCP–UDP Jitter — enabling end-to-end insight into real user experience. These metrics are critical for engineers troubleshooting throughput degradation, latency spikes, or congestion-related impairments that traditional flow counters often fail to expose.

Aviz Service Node computes TCP Packet Loss by passively analyzing mirrored traffic flows at the packet level. It tracks TCP sequence numbers, retransmissions, and acknowledgment gaps to accurately determine the volume and rate of lost packets. The analysis is performed per flow and correlated with user session metadata such as IP addresses, VLANs, QoS tags, and tunnel identifiers (e.g., GTP in Telco or VXLAN/GRE in IP networks). This cross-domain correlation allows engineers to pinpoint where in the path — between RAN and Core, or between leaf-spine nodes in a DC fabric — the loss originates. Similarly, TCP–UDP Jitter is computed by evaluating inter-packet timestamp variations, providing real-time insight into delay fluctuations for both connection-oriented (TCP) and connectionless (UDP) traffic streams.