Many of today’s Internet services and data applications encrypt traffic to guarantee user security and prevent carriers from identifying and controlling the data flow. For example, P2P and VoIP services such as BitTorrent and Skype obfuscate their protocol structure to avoid detection, traffic shaping, or metering. Likewise, many OTT applications use SSL, or other techniques, to encrypt communications and hide user information.
While encryption is reassuring for privacy purposes, it compromises the service provider’s visibility to manage traffic and provide comprehensive security.
Allot’s Encrypted Traffic Classification solution empowers you to overcome these challenges. It analyzes and classifies all traffic, irrespective of encryption, so you can manage it. As a result, you can improve your network’s efficiency, make it more secure, and ensure your users enjoy the Quality of Experience (QoE) they expect.
Traffic you can see is traffic you can control
Part of the Allot SmartTraffic QoE suite, our Encrypted Traffic Classification solution is based on Allot’s Dynamic Actionable Recognition Technology (DART). It employs multiple inspection methods to identify traffic according to Layer-7 application, user, device, access, video and contextual attributes.
By adopting Allot’s combination of inspection methods, you can achieve highly granular and accurate recognition, even at maximum speeds and peak loads.
You’ll be able to:
- Quantify and understand network usage
- Optimize network efficiency
- Assure high-quality service delivery and experience
- Create value-add services packages and charge for their use.
More about DART
DART proactively learns and adapts to evolving encryption tactics using many methods, including:
- Pattern and numerical property analysis of packet contents
- Heuristic analysis of behavior statistics from inspected transactions
- Learning peer system behavior to identify P2P seeders (multiple transmission of files) and popular peers.
- Classification of Port, IP address, or range of IP addresses
- Machine-learning based on statistical distribution patterns for thousands of traffic attributes.
- Automatic IP discovery through analysis of service hosts and subscriber records.
Allot’s patented Predictive DPI (PDPI) technology enables the classification engine to learn from the patterns and behavior of traffic that it recognizes with 98% accuracy and compares them with encrypted flows to improve identification.