Open AccessPractical algorithms for allocating the roadside units in VANET
Author(s) -
Wu Jingli,
Wu Yong
Publication year - 2020
Publication title -
iet intelligent transport systems
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.579
H-Index - 45
eISSN - 1751-9578
pISSN - 1751-956X
DOI - 10.1049/iet-its.2019.0617
Subject(s) - vehicular ad hoc network , metric (unit) , computer science , heuristic , wireless ad hoc network , software deployment , greedy algorithm , algorithm , trace (psycholinguistics) , constraint (computer aided design) , quality of service , performance metric , computer network , engineering , artificial intelligence , wireless , telecommunications , mechanical engineering , linguistics , operations management , philosophy , operating system , management , economics
The vehicle‐to‐infrastructure communication is a kind of communication mode of the vehicular ad‐hoc network (VANET). It is paramount to allocate roadside units (RSUs) appropriately to assure the quality of communication service. Sarubbi et al. have proposed a greedy heuristic allocation algorithm Delta‐r. It examines the Δρ 2ρ 1‐deployment constraint, a metric for evaluating the performance of VANET, with relative contact time. Unfortunately, the relative contact time of vehicles at urban cells may present ‘false high’, which would play an opposite effect on making a decision. In order to solve the problem, a new allocation algorithm Delta‐uc is presented, which examines the Δρ 2ρ 1‐deployment constraint with a useful contribution. It gets rid of the negative effect of ‘extra’ contact time by means of retaining merely the contributing relative duration of a vehicle at an urban cell. Furthermore, an effective recombination operator is introduced basing on algorithm Delta‐uc, and a parthenogenetic algorithm UCPGA is presented to solve the allocation problem. Experimental comparisons were performed on the real mobility trace of Cologne, Germany. The experimental results demonstrate that in many Δρ 2ρ 1‐deployments, the number of RSUs produced by the Delta‐uc and UCPGA algorithms are, respectively, fewer than those produced by algorithms Delta‐r and Delta‐GA.