Mobile Network Latency Assessment via Passive Traffic Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for assessing mobile communications network latency are incomplete, as they fail to accurately measure the latency experienced by each subscriber for each communication session and often skew results by including low-speed channel delays, network buffering delays, and misconfigured routes, making it difficult to identify issues in the underlying network infrastructure.
Innovation Solution
A network monitoring apparatus that filters out low-speed data path traffic and focuses on high-speed data paths to measure round-trip time (RTT) metrics, providing more accurate latency metrics by aggregating data on a service component basis, thereby identifying specific service components contributing to latency without using active pinging or probes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If active ICMP probes or average TCP RTT probes are used to measure network latency, then network traffic delay can be measured, but the results are skewed by low-speed channel delays and network buffering delays, obscuring the underlying network infrastructure latency
Solution Approach 1:
The patent segments the network path into distinct components (access network, core network, service network) and measures latency for each segment separately using distributed monitoring agents deployed at different network points. This allows isolation of infrastructure latency from end-to-end latency that includes application and user equipment delays.
Solution Approach 2:
The patent introduces passive monitoring agents as intermediaries that observe and measure traffic without actively injecting probes. These agents capture latency measurements from actual user traffic flows, eliminating the distortion caused by active probing while still providing measurable latency data for network infrastructure assessment.
2Productivity
If average TCP RTT measurements are used to assess network performance, then overall network delay can be obtained, but the measurements include both low-speed and high-speed channel delays, skewing the results
Solution Approach 1:
The patent implements local quality monitoring by deploying monitoring agents at specific network locations (access network, core network, service network) to measure latency characteristics local to each segment. This provides differentiated latency measurements for each network portion rather than a single averaged value, enabling precise identification of infrastructure latency issues.
Solution Approach 2:
The patent replaces active mechanical probing (ICMP pings, TCP RTT probes) with passive observation of actual user traffic flows. By substituting active injection of test packets with monitoring of real traffic, the system eliminates the artificial delays introduced by probing mechanisms while maintaining measurement capability.
3Reliability
If comprehensive end-to-end latency monitoring is implemented, then overall network performance can be assessed, but it becomes difficult to identify specific service components contributing to latency
Solution Approach 1:
The patent divides the end-to-end network path into segmentable portions (access network, core network, service network) with monitoring agents deployed at each segment. Each agent measures latency for its local segment and reports to a central system, enabling both comprehensive end-to-end assessment and detailed identification of specific latency-contributing components.
Solution Approach 2:
The patent implements feedback mechanisms where monitoring agents continuously report latency measurements for different network segments to a central management system. This feedback loop enables real-time identification of which specific service components are contributing to latency, allowing targeted optimization while maintaining overall network performance visibility.
Data Source
AI summary
A network monitoring apparatus in a mobile communications network monitors, for a predetermined time, TCP packets in the network that tend to comprise end-user traffic using the high-speed data path but not the low-speed data path to provide monitored packet information. The network monitoring apparatus uses this monitored packet information to identify the fastest end user-experienced communication exchanges within the mobile communications network. By one approach, the network monitoring apparatus effectively filters out most or all of the end-user traffic that uses the low-speed data path when providing that monitored packet information. This can comprise, for example, tending to utilize data pertaining to TCP packets that correspond to higher traffic volume (and hence that are likely being conveyed via a high-speed data path) while also tending to discard data that pertains to TCP packets to that correspond to low traffic volume (and hence that are likely being conveyed via a low-speed data path).