For internet routing, theory and reality might be very distant. RED in particular has a terrible record of making things worse on live systems. But that was a f

For internet routing, theory and reality might be very distant. RED in particular has a terrible record of making things worse on live systems. But that was a few years ago and perhaps newer variants are better. Also there's a huge problem on routing algorithms due to hundreds of patents covering very trivial uses. For example, there is an awful patent covering the very important case of not dropping tiny ACK packets. For congestion control, it's more useful to drop big TCP data packets to make the endpoints adjust. And the key networking gear players are full on supporting those crazy software patents. So my prediction is things will stay the same and nothing meaningful will happen for 10 years until a few key patents expire. Nobody will risk exposure to a multimillion IP lawsuit. And anyway, we the users will just pay the bills either way so telcos don't care. You can thank Big Telcos, Cisco and many others. Of course big players with a huge patent war-chest (i.e. Google/Amazon/Microsoft) will advance their WANs.