Task Offloading Strategy Algorithm for Realistic Internet of Vehicles Scenarios
Liang Dong, Di Zhang, Yian Yu, Yan Huo · 2023
In the Internet of Vehicles (IoV), which is characterized by vehicle mobility and distributed positioning of edge servers, balancing the load among distributed servers and reducing system costs become challenges. The paper addresses the establishment of a comprehensive system cost model to enhance flexibility and usability. In the joint optimization of server load, pricing, and energy consumption, the Analytic Hierarchy Process (AHP) is used to allocate weights to these three indicators. The paper proposes a Chebyshev and Reverse Learning Improved sparrow search algorithm (COSSA) to solve this model. Firstly, the algorithm's initialization utilizes Chebyshev hybrid mapping improvement, enhancing the population's distribution for later improvement of solution quality. Then, a reverse learning perturbation is applied to the optimal individuals in the population, preserving diversity and enhancing the algorithm's exploration capability. COSSA algorithm is compared with three other algorithms, validating the practicality and effectiveness of the model and the algorithm. First, experimental results comparing different indicators and system utilities demonstrate the algorithm's effectiveness. A comparison between equal weight allocation and AHP weight allocation illustrates the importance of differentiated weight allocation in real-world application scenarios.