article page
| 1
| 2
| 3
|
hardware based Deep Packet Inspection (DPI) to inspect every packet and extract the user plane KPIs.
Device Monitoring OSS utilizing DPI to ascertain the service quality of every session allows mobile operators to monitor new smartphone devices and services as they are being deployed. The components of a Device Monitoring OSS include:
- DPI deployed within the mobile network that extracts metadata about the quality of each session,
- Service Detailed Records (SDR) measurements inserted into a database,
- Per session KPIs and KQIs aggregated by service delivery component,
- Filtering based on preferences, dashboards and geographical map displays and
- KPI/KQI interface to the mobile operator's data warehouse.
Legacy KPIs have traditionally focused on voice dropped calls as the primary network/device quality indicator. Mobile broadband sessions are much more complex to monitor because a session drop may not affect customer experience if the user wasn't transmitting data, while other factors, like throughput, latency, packet loss, and retransmissions, often times have a much greater effect. As a result, we need the measurements within the SDRs. These detail measurements are critical to understanding customer experience of smartphone device users and can fuel a successful device launch.
User plane monitoring focuses on measuring the customer experience of the actual user content service quality, not just on control plane signaling messages that detect network drops and server access problems. Mobile broadband customers have come to realize you can have a data session that doesn't necessarily drop but still have a bad customer experience because of lousy throughput, latency, and packet loss.
USER PLANE INDICATORS
Packet Loss
Packet loss is a basic indicator for broadband data service quality with new patent pending algorithms that improve applicability to mobile broadband networks. Simply stated, if you drop too many packets it will cause serious quality problems resulting in white screens and unusable content. Packet loss is also extremely difficult to isolate since it occurs at any point in the network from any of hundreds of servers or thousands of cell sites in the Radio Access Network (RAN) or the smartphone device itself.
|
|
We need the measurements within the SDRs. These detail measurements are critical to understanding customer experience of smartphone device users and can fuel a successful device launch. |
|
Mobile service providers that proactively resolve statistically relevant packet loss problems increase user satisfaction. Furthermore, resolving packet loss problems on smartphone devices associated with revenue generating services eliminates revenue leakage.
Real-Time Throughput
Monitoring uplink and downlink throughput in near real-time (e.g. 5 min.) by subscriber and statistically aggregating by smartphone devices is a basic requirement for verifying service levels. Users view throughput as a critical performance indicator and expect promised throughput levels to be maintained for the session (e.g. 300 Kbps). Verifying the throughput for a certain user at a particular point in time is a basic requirement to perform up to the user expectations of valued smartphone device early adopter subscribers.
Real-time end-to-end throughput monitoring is also required in monitoring streaming and other advanced smartphone device data services to verify service levels. For example, monitoring real-time throughput levels of a particular server or service and correlating them with statistically aggregated smartphone device indicators assists in problem resolution.
By providing the ability to isolate, prioritize and proactively resolve service affecting problems, mobile operator engineers can significantly enhance user satisfaction and resolve early adopter subscriber issues. Furthermore, proactively resolving problems associated with smartphone device early adopter subscribers is a proven method to increase user satisfaction of early adopters and increases adoption take rates and revenue ramp up rates.
Latency
Considered the key to providing user satisfaction in fixed-line broadband networks, latency is crucial to understanding user satisfaction in a mobile broadband network. While some mobile operators manually ping a smartphone device to perform basic latency reports, this method is flawed, error prone, and may not provide the level of granularity needed as data volume and service complexity increase.
Isolating the root cause component within the mobile network with latency over threshold provides actionable results to
article page
| 1
| 2
| 3
|
|
|
|
