Some top up topics here since there was very interesting conversation in the comments. Just want to bring that topic to the front. What is data assessment & prioritization? How does traditional vs new mobile n/w handle this ever rising need to prioritize the data?
https://twitter.com/Finstor85/status/1478021479530909699
Real-time analysis & prediction of certain kinds of data is getting more importance in communication between Machine to Machine. Let's refer that as M2M from now on. The M2M data is giving more imp. in certain application context than P2P (person to person)
Esp. geospatial data related to self-driving, driving assistance, navigation, traffic & accident info., license plate detail etc. has more imp. than any kind of data transferred over the network. It isn't the network that is capable to assess & prioritize at protocol level.
Real-time cognitive analysis is just going to explode as we get into smart/always connected world. Vector data is far more important than Scalar data.
SDN can help deploy a solution that can do traffic prediction. You need AI/ML algo do that for you just before you send data over the network. You assign a score to certain kinds of data segments & then control which data gets priority. Your protocol can't differentiate this.
There is something called as "Feature Selection Method" which is used to do data assessment, optimization, and finally prioritization. Sensor/Mobile device controller can prioritize data!
I can go into detail but then it would be too technical. The point is - your network/connectivity/communications provider can enable such solutions as a part of their network services! It isn't part of network itself, but part of solution/offering as a whole.
Prioritization is offered as a Class of Service. You can tag a certain types of data packets as drop eligible. It is very commonly done when network is congested. Low priority packets are dropped. Network capacity assessment is a key here.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
