A new neural network AI from MIT CSAIL has been making headlines recently for using machine learning to solve some issues associated with buffering. We’ve all experienced buffering before, either in the form of pixilation, long wait times while loading a video or audio file, and everyone’s favorite nemesis: the spinning pinwheel of death. Buffering occurs because it’s impossible for your computer (or TV) to receive data all in one lump for immediate playback. Therefore, data is broken up into smaller packets and sent to its destination, sequentially. So while you’re enjoying your favorite Spotify playlist played to wireless speakers throughout your home, or streaming the latest episode of Game of Thrones in your living room, your entertainment media will always be sent bit by bit. If all goes well, you’ll never notice this is even happening. But more often than not, you’ll experience some sort of indication of file buffering. If there isn’t enough bandwidth, you’ll either experience pixilation, longer buffering times, or drop outs because your network can’t transmit data fast enough to maintain a sufficient “buffer.”
Essentially, what MIT CSAIL’s AI, dubbed the “Pensieve” neural network, does is use machine learning to switch between pixilation and buffering so your videos aren’t over buffering when they don’t need to, or pixelating when they don’t need to. According to MIT, the neural network will tune itself over time based on a system of rewards and penalties, allowing streaming services to customise this for their content—with priorities for buffering or resolution. If the streaming service is able to predict that a user watching a video on a handheld device is about to walk into a poor connectivity area, the system will be able to reduce the streaming resolution sufficiently, creating enough of a buffer for (potentially) stutter-free streaming (livemint). This is all fine and good, but it’s essentially like putting a Band-Aid on a festering wound: it may cover up the problem, but it by no way solves the underlying issue.
The real problem with buffering lies with your WiFi network. Conventional WiFi runs on TCP (Transmission Control Protocol), designed in the 1960’s for transferring files down wired Ethernet lines – certainly not for streaming real-time video and wireless audio throughout the Smart Home. (For more information on the shortcomings of TCP, check out this blog post). As long as your WiFi runs on this outdated protocol, it doesn’t matter what techniques are being innovated to combat the annoyance of buffering – the cause of the issue still needs to be addressed, not the symptoms.
Blackfire Research understands this. That is why we developed Real-Time Packet Management (RPM), the Blackfire Research solution to buffering. For whole home, wireless audio, RPM uses a special multipoint, real-time feedback signal from each speaker to monitor the effects of noise on the audio data stream, allowing for a much shorter queue and much less buffering. RPM is part of the Blackfire Realtime Entertainment Distribution (RED) framework, a revolutionary new protocol designed to stream both HD 5.1 audio and 4K video, simultaneously, across multiple devices around your home- all over the standard WiFi – with precise synchronization, low latency for lip sync, and overall reliability.
RPM can be found in any Blackfire powered device. Partnering with Blackfire Research means you’re ahead of the pack, and most of all, one step closer to defeating your nemesis: that darn spinning pinwheel of death.
