TRA Trust Relations Architecture 5G
In-lab we run a new vPolice model. It consists of 3 elements to improve the now.
#1 PreCrime SafetyMonitoring (of critical city fragments)
#2 PreAccident TrafficActors (collecting BigData of Inter-Actor physical spaces)
#3 PreDrone CitySkyMonitoring (catch starting locations of eFlight objects)
#1 is a non-privacy-violationg system of CVPR and AI complexity reduction. No faces only but fast 'there is something > GoFast'. We call this WeNo-SysKnows. We only got this cleared by law that the system collects but does not reveal in general, only postcrime by warrant.
#2 For years we observe that the individual physical space of traffic actors shrinks. We early want to be alerted of RRA Rising Risk of Accidents. This RADAR system runs fully vertical and no ALPR, no video at all, just spaces.
#3 A small slice horizontal (like in a tank) video system detects drones and birds. The hard part is to sort out the birds. A district got an advanced PreDrone system, the private citicens pay monthly for fast police showup. A Wifi directional small horizontal layer jammer was dropped as to much RF power required.
This the important intro to come to the point. All this runs with multiple partners outside of LEO and designed for 5G network slice for LEO. Related to security we started with 'there is no security'. So we now are interested in a public forum of what you think how to design and structure new types of trust relations. Whom be trusted, in which respect and to what extent.
Please give us feedback of the trust architecture based on 'there is no security'.
Thank you in advance.
Nobody working on PreCrime, PreAccident and PreDrone? Cannot be.
I don’t think we have the funds, manpower or time to do much pro-active work here in the UK. All of our work is reactive.
Fully understand, we too were reactive for decades. But since getting a new boss things changed from hunting to trapping (I reported, my boss No more hunting, lets build traps).
To design the Trust Relations Architecture 5G is the crux. Sensitive citizen data have to be protected in general until crime occurs and the attorney opens the law. The following players are involved
a) MNO 5G CH
b) Video hardware supplier DE
c) Video Analytics software supplier IL
d) Electrical Infrastructure provider locally CH-ZH (second abbr is state (Kanton))
e) Data Protection Representatives locally, national CH-ZH, CH-BE
f) RADAR supplier DE
Dropping the setup and electrical running of this combined 3-elements vPolice system lets only focus on the communication layer of RADAR data and Video data. The network management layer is in our own internal ICT dept. hands.
The risk of criminal hacking-in we set up from the beginning to 100% possibility. Immediately you see the risk of destroying our reputation and promisses we legally made. But its unavoidable to lose. So we prepare for 'early morning news Police system hacked, TBs stolen from UR. What would we say? To defend we want to present a top quality and highly professional TRA documentation.
The only shield to defend would be We did everything possible, the remaining risk of hacking was considered as much as possible. Reminder we started with 'there is no security'.
But back to a)-g). How to design the level of trust, the limitations of trust and the layers of translucent data insight? There is always a minimum and a maximum. So trust should be limited to the minimum. To legally bind all players by NDAs is obvious, but how to operate?
We also designed a communications roadmap. This included to pre-event commit We were hacked before it already occured. To announce the bad as soon as possible was consulted from our legal, the pre-event commitment our lab proposal (yes, very! controversal).
Up from May 25th, 2018 the new GDPR regulation forces us to adjust the TRA documentation.
Who runs a Trust Relations Architecture TRA?