Dynamic Coded Convolution with Privacy Awareness for Mobile Ad Hoc Computing

The Internet of Things (IoT) devices generate massive data that contain a lot of valuable information. To enable efficient on-device data analysis, the mobile ad hoc computing (MAHC) is a promising solution. It allows IoT or other edge devices to share resources among each other via device-to-device communication links and carry out computation-intensive tasks in a collaborative manner. This paper investigates the vector convolution problem using computing nodes in a MAHC system. A novel dynamic coded convolution strategy with privacy awareness is developed to address the unique features of MAHC systems, including node heterogeneity, frequently changing network typologies, time-varying communication and computation resources. Simulation results show its high efficiency and resilience to uncertain stragglers.