Research Of Cost Minimization For Code Offloading With Cellular Traffic Aggregation

With the rapid development of mobile internet, the performance of intelligent terminal has been significantly improved. For the explosive growth of mobile applications, the computation and battery capacity are still hardly to meet requirement. Code offloading has become a hot topic to solve the problem by sending heavy computation tasks to resourceful cloud, instead of executing all tasks on local mobile devices. Unfortunately, current code offloading techniques are not efficient enough because of high communication cost, resulted by so-called tail energy in cellular network. Motivated by the fact that significantly less energy is consumed by mobile devices during good connectivity, and the low-power device-to-device communication techniques like Wi-Fi or Bluetooth, we propose a novel code offloading strategy with cellular traffic aggregation, COTA, to minimize the total cost, which has important theoretical and practical significance.Based on the energy consumption of data transmission characteristics of cellular network, Wi-Fi and Bluetooth etc., COTA aggregates offloaded codes to several mobile devices with low-power device-to-device communication techniques like Wi-Fi or Bluetooth before sending them to cloud, and the device transfers data to cloud in combination, which can significantly reduce tail energy effect. Firstly, the feasibility of the scheme is discussed, theoretical analysis shows that COTA can realize energy saving. Then, we propose three problems to be solved, that is the cooperation of multi terminal, the decision of offloading strategy and the choice of hot polymerization. Secondly, we propose an optimization framework by jointly considering code partition and traffic aggregation. Due to the hardness of this problem, we design an efficient heuristic algorithm solve the cost minimization problem. The basic idea is to iteratively improve code partition and traffic aggregation to achieve reduced total cost, such that we only need to deal with a simplified linear programming problem in each iteration.Extensive simulations are conducted to evaluate the performance of our algorithm. The results present that COTA algorithm can save energy by 30% in most of the situations which outperforms existing offloading strategies. Experiments demonstrate that the proposed algorithm COTA is more effective when the computation load is larger, the communication capacity is better or when Wi-Fi is used as the device-to-device interface.