Network Probes
Deliver high quality user-experience by ensuring high performance
A probe is a device or software that converts packet-based network communications into an analyzable metadata format. Different techniques for accomplishing this are used depending on the traffic and application. Deep Packet Inspection (DPI) and Machine Learning are used for user traffic as well as large volume traffic such as video content. For signaling traffic, the packets are parsed and the relevant data extracted.
The purpose for using a probe is to reduce data volume to allow processing of the data. On average, the data reduction ranges from a factor of 1000 to 10,000, depending on the specific method. Cubro’s time-window based solution is specially designed for producing high-quality data with a low volume
Cubro’s time-window aggregation metadata is built around Cubro’s proprietary deep packet inspection (DPI) engine that detects L7 applications regardless of encryption. This provides detailed information describing the communication including the network endpoint, data transmitted and received, duration of the session, and the specific application or service being used.
Scalable agnostic solution
The metadata output is platform-agnostic and utilizes industry-standard data formats. It provides valuable data points for numerous use cases, including security applications, performance management, network planning, general usage monitoring, customer experience analysis, and more. The Cubro solution offers standardized and agnostic interfaces to any kind of 3rd party solution.
The volume of traffic that must be handled by a probe can reach dozens of Terabytes per second, depending on the size of the customer. Even mid-size companies can easily generate multiple tens of Gbit/s of traffic that must be analyzed. Cubro uses high-performance hardware appliances to address these throughput needs while also minimizing footprint, cost, and power consumption. We offer solutions designed for everything from Gbit to multiple Terabits of performance.
Hardware Platforms
Cubro probing /meta data extraction software is running on different hardware appliances including Omnia Series and Omnia NIC.
Cubro probes are typically used by Datacenters/Enterprises and Service Providers.
Each use case differs in some major ways and requires a different approach.
Datacenter/Enterprises
Datacenter and Enterprise customers tend to have less traffic than service providers, however they often have more sites and greater geographic distribution.
Within enterprises, probes have to provide data to multiple 3rd party analytic platforms (such as security and performance management). Hence, there is a need for more standardization of the output format.
Cubro offers IPFIX, IPFIX with DPI, and the Custos time-window based format. Custos 3D-style user interface provided insightful, immediately actionable network information, stored network data dramatically more efficiently than NetFlow/IPFIX, and implemented a human-oriented data structure that could be easily integrated into 3rd-party systems.
Service Providers
Service providers are constantly increasing traffic throughput to and beyond the Tbit/sec range. There are two major categories of traffic, user plane (user data) and control plane (signaling traffic). The two types of traffic must be correlated (enhancing the user plane data by incorporating the control plane information) to make the best use of the user plane data. As a result, Cubro offers probes that process signaling traffic, in addition to user plane traffic, for service providers.
To be cost effective and power efficient in a service provider environment, probes need to have a very small footprint and minimize power consumption.
Furthermore, the platform also offers industry standard/platform agnostic interfaces to 3rd party systems. For example, by leveraging Kafka, any database system can receive the metadata output.
Our probes improve network quality, leads to increased customer satisfaction, and can augment information security systems. The Cubro Probe platform generates comprehensive data records for applications such as business analytics, service assurance, business assurance, security assurance. The Probe analyses and processes user and signalling traffic in real-time.
Technology Partnership with SilverEngine
To leverage the value of the network data extracted by Cubro Probe, Cubro is collaborating with SilverEngine. Silverengine provides consultancy services related to optimizing or fixing the network, understanding network and user behavior and on helping how to segment subscribers.
As an end result the operators are able to get actionable insights with the data generated by Cubro Probe. The insights include for example how to improve subscribers experience and mobile network performance using Analytics and Machine Learning techniques.
Cubro together with Silverengine offer service packages with actionable insights including:
- Identify underperforming cells, information for Network Optimisations
- Core NE elements issues
- Poor performing User Equipment and Devices
- Subscribers with poor experience, ensure VIP, high ARPU and In-bound roaming subscribers getting the service they expect
- Suggested marketing campaigns for churn preventions and upsells
Specific service packs:
VoLTE
- Example KPIs: IMS Registration Success rate, Speech quality (MOS), mute calls, dropped call rate, repeat calls
- Use Machine Learning to find out Radio KPIs “associated” with mute calls and poor speech quality
Mobile Broadband (MBB)
- Example KPIs: Attach issues, APN failures, Streaming data throughput, DNS resolution issues
- Speed test performance of subscribers e.g. Ookla
Customer Experience Management (CEM)
- Clustering of subscriber’s experience and forecasting
- QoE index [0-100] for each subscriber consisting of relevant C-plane U-plane KPIs
- Optimized QoE index model based on Machine Learning
- Subscriber segmentation
Applications
- Application profiling, e.g. access patterns and protocols used
- Application QoE
- Improve relevant KPIs for each popular application
- Video performance e.g. YouTube and Netflix
- Rationale of local video or gaming cache
5G
- Insights on time spent on 5G vs LTE in 5G NSA deployment per subscriber
- Performance advantage in 5G compared to LTE
- Machine Learning used to find out relevant Radio and Backhaul KPIs associated with long RTT and low throughput
IoT
- Understand and learn communication patterns of different classes of IoT devices
- Detect anomalies and outages
Use case for Mobility
With the insights, mobility data sales packages can be created for advertisement, commercials, real estate and traffic planning. The insights help deciding whether to open a new branch store in a specific spot or as an example where to focus infrastructure improvements based on biggest traffic jams on roads and streets
- Rationale for open new store branch in certain spot.
- Identify main traffic jam roads and streets – for infrastructure improvements
Customer Benefits include better network performance, lower costs for network infrastructure, reduced subscriber churn due to better network quality and improved return on investment (ROI).
Looking for product support?
Get in touch with our technical team now
Our newsletter provides thought leadership content about the industry. It is concise and has interesting content to keep you updated with what’s new at Cubro and in the industry. You can unsubscribe anytime with a single click.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.