The
multiplication of streaming video services has led to ferocious competition on
the commercial and technological front. While streaming services visibly compete
on their pricing and content attractiveness, a more insidious technological
battle has also taken place. The best way to describe it is to compare video to
a gas. Video will take up as much capacity in the network as is available.
When you
start a streaming app on your phone, it will assess the available bandwidth and
try to deliver the highest definition video available. Smartphone vendors and
streaming providers try to provide the best experience to their users, which in
most cases means getting the highest bitrate available. When several users in
the same cell try to stream video, they are all competing for the available
bandwidth, which leads in many cases to a suboptimal experience, as some users
monopolize most of the capacity while others are left with crumbs.
In recent
years, technologies have emerged to mitigate this issue. Network slicing, for
instance, when fully implemented could see dedicated slices for video
streaming, which would theoretically guarantee that video streaming does not
adversely impact other traffic (video conferencing, web browsing, etc…). However,
it will not resolve the competition between streaming services in the same
cell.
Open RAN offers another tool for efficiently resolving these issues. The RIC (RAN Intelligent Controller) provides for the first time the capability to visualize
in near real time a cell’s congestion and to apply optimization techniques with
a great level of granularity. Until Open RAN, the means of visualizing network
congestion were limited in a multi-vendor environment and the means to
alleviate them were broad and coarse. The RIC allows to create policies at the
cell level, on a per connection basis. Algorithms allow traffic type inference
and policies can be enacted to adapt the allocated bandwidth based on a variety
of parameters such as signal strength, traffic type, congestion level, power
consumption targets…
For
instance, an operator or a private network for stadiums or entertainment venues
could easily program their network to not allow upstream videos during a show,
to protect broadcasting or intellectual property rights. This can be easily
achieved by limiting the video uplink traffic while preserving voice, security
and emergency traffic.
Another
example would see a network actively dedicating deterministic capacity per
connection during rush hour or based on threshold in a downtown core to
guarantee that all users have access to video services with equally shared
bandwidth and quality.
A last
example could see first responder and emergency services get guaranteed
high-quality access to video calls and broadcasts.
When
properly integrated into a policy and service management framework for traffic
slicing, Open RAN can be an efficient tool for adding fine grained traffic
optimization rules, allowing a fairer apportioning of resource for all users,
while preserving overall quality of experience.
