Feasibility Study of Using Predictive LTE Connection Selection from Multi-Operator for Teleoperated Vehicles

Abstract

Service depending on good connection is growing and so its sensitivity, like Advanced Driver-Assistance System (ADAS). ADAS is the most common technological feature in the modern car, and the hope to reach a dependable anonymous car is the ultimate target. We (From end user and manufacture perspectives) are evaluating Teleoperated Driving as the most promising achievable feature to support emerging needs for traffic headache avoidance and health & safety cautions, with human to human sense & interaction proven to be better than Human to Machine in handling (Human driving vs. Machine driving). Since this whole service is depending on sensors (Already covered by different car manufactures) and connectivity (Varying in the sense of coverage and capacity). In this paper, we study the applicability of predicting the most preferable market operator within a certain area (Satisfying a previous studied criteria) to use as a primary data connection before getting into a new measurement delay. For this purpose, a long measurement period was preformed with a connection prediction reaching from 87% to 93% using variant models.