Manage ABR video traffic with granular policy control and signature detection powered by machine learning.
Video traffic will account for 76% of all network traffic by 2025. Manage it intelligently to minimize network CapEx and OpEx, identify new services, increase average revenue per user (ARPU). With ABR video streaming, the client device selects the appropriate bit rate based on current network conditions and device capabilities.
ABR video detection and control can save up to 71% of bandwidth.
F5 has tools to manage ABR video, making it easy, for example, to rate limit an uncontrolled 4K on-demand video (a 40 Mbps stream) to 1080p HD (a 7 Mbps stream).
Detect and classify a wide range of applications with a traffic classification engine that can identify over 4,000 application signatures. Accurately identify video stream resolution and source—along with many other criteria—using a purpose-built machine learning algorithm.
A mobile phone user doesn’t need 1080p HD resolution (a 5 Mbps stream), so rate pace mobile phone traffic, for example, to a maximum of 720p HD. Rate pace OTT HD video flows to reduce bandwidth based on policies you create.
Create policies based on subscriber information—such as IP address and user name—to correlate them with application flows and apply granular policies per subscriber, application, or other criteria.
Tailor services based on subscriber profiles to monetize services based on time, resolution, peak usage, or other options. For instance, rate limit video streams to manage peak congestion, or restrict HD streams for heavy video users after certain download thresholds.
Apply bandwidth control policies based on traffic classifications to both rate limit TCP video traffic and rate pace fast-growing Quick UDP Internet Connections (QUIC) video traffic such as YouTube and Hulu. BIG-IP Policy Enforcement Manager (PEM) makes it easy.
Export detailed statistics and generate granular reports about subscriber and application flows. For instance, obtain data about activity at per-session and per-application levels, as well as event-driven, flow-based records.