Stochastic-Network-Calculus-Theory-Based Road Side Unit Location Optimization at an Intersection

This paper deals with the problem of stochastic network calculus-based road side unit (RSU) location optimization at an intersection. Considering the stochasticity of data arrival rate at RSUs and RSU service capacity, the upper bound of vehicle-to-RSU (V2R) communication delay is derived using stochastic network calculus. Further, to hedge against the impact of the randomness of the traffic density on RSU location optimization, the problem is formulated as a two-stage nonlinear mixed-integer stochastic program. The objective function of this program is to minimize the investment cost of RSU location and the expectation of the penalty cost of the V2R communication delay upper bound. In the first stage of the program, the number and location of RSUs are optimized when the traffic density at the intersection is uncertain. In the second stage, when the traffic density is realized, the subareas of the intersection are assigned to the located RSUs to minimize the penalty cost. To find a global optimal solution to the problem, a Benders decomposition algorithm is proposed. The experiment results show that the proposed model is able to achieve 19.4 ms/per cost lower V2R communication delay, compared with the average-communication-delay-based model.