The application of IoT has given rise to the adoption of mobile edge computing network architecture by many IT-based organizations. However, the problem of limited energy supply, processing capacity, storage of mobile devices, and the inherent problem of bandwidth and transmission delay incurred in offloading tasks to the remote cloud server have been the challenges. In this research, we propose Energy Efficient Canonical Polyadic Decomposition based Offloading Algorithm. This algorithm utilizes canonical Polyadic decomposition based on rank_1 to decompose the input task data to produce the intermediate output with reduced task data size. The reduced data from the mobile layer is offloaded to the remote processing device where the processing capacity, energy, and storage are almost unlimited. The reduction method reduces the tasks deployed at the mobile layer, thereby reducing the mobile devices busy time and energy consumption. The data reduction ratio is used to simulate the computation offloading in mobile edge computing. The result from the simulation shows that the proposed algorithm improved the mobile devices energy consumption by 83.38% than the existing algorithms. The better result is achieved through the reduced data reduction ratio and reduced number of deployed task on mobile devices which are achieved through canonical Polyadic decomposition.

Index Terms— Computation Offloading, Mobile Computing, Energy consumption, Task Scheduling, Resource Allocation