Mobile terminals (MTs) within the coverage of both the WiFi and cellular network can offload computation tasks via the WiFi access point to efficiently release the cellular network congestion and reduce the load on base stations. In this paper, a novel scheme of joint WiFi and cellular offloading is proposed to optimally reduce the latency and energy consumption of MTs in task processing. Based on the statistical characteristics of MTs’ task generation, our scheme serves as strategic guidance for computation offloading without frequent execution of the optimization algorithm. MTs with the different channel access priorities are also considered in our scheme. A balance factor is introduced to flexibly adjust the minimization between the energy consumption of MT and the processing latency of its tasks. We design an iterative algorithm combining the linear programming with the Alternating Optimization technique to efficiently solve the non-convex problem for computation tasks offloading decision. The numerical results demonstrate that our scheme can largely improve the system performance, and the superiority of our scheme is shown in all scenarios.

WiFi和蜂窝网络覆盖范围内的移动终端（MT）可以通过WiFi接入点来减轻计算任务的负担，从而有效地释放蜂窝网络的拥塞并减轻基站的负载。本文提出了一种新的WiFi和蜂窝卸载联合方案，以最佳地减少任务处理中MT的等待时间和能耗。基于MT任务生成的统计特征，我们的方案可作为计算卸载的战略指导，而无需频繁执行优化算法。我们的方案中还考虑了具有不同信道访问优先级的MT。引入了一个平衡因子来灵活地调整MT的能量消耗与其任务的处理延迟之间的最小化。我们设计了一种将线性规划与交替优化技术相结合的迭代算法，以有效地解决计算任务分流决策的非凸问题。数值结果表明，该方案可以极大地提高系统性能，并且在所有情况下均显示出该方案的优越性。