Coordinate hiding in solving the generalized dangerous proximity problem

Abstract

In connection with the rapid development of autopilot transportation systems (driverless cars, unmanned aerial vehicles – drones) all the more urgent becomes the problem of collision avoidance (the problem of dangerous proximity), especially in conditions of a large number of members of the traffic and its high intensity. If there is an attacker who is able to control one or more participants in the traffic, which strategy is to create an emergency on the road or collisions of unmanned vehicles there is a need to protect against such malicious behavior. One of the options of such protection is the use of the concept of multilateral confidential calculations to solve this particular problem. The future autopilot systems assume in some local traffic control zone the possibility of interaction of participants according to the "everyone-with-everyone" scheme, which is a necessary condition for multilateral confidential computing systems in conditions of various unauthorized actions. In contrast to the previously published results using this concept in the model with the so-called semi-honest enemy, this article for the first time proposes some solutions to protect the traffic participants from the malicious actor, who, having revealed the coordinates of moving objects, tries to drive them to collision. It is also considered the expansion of this problem to three dimensions. At the same time, it is shown that the implementation of these systems in practice, most likely, does not help to avoid the problematic situations that are not solved only by mathematical methods. There is a large set of problems that must be solved by various organizational, technical and managerial methods. Nevertheless, the prospects of using the methods of provably safe coordinate hiding look, in the author's opinion, quite attractive for experts in the field of information security, communications, transport, and distributed systems.

Keywords

• autopilot transport systems, geographic coordinates hiding, secure computation, secure function evaluation, two-party and multi-party protocols of secure computation.