Menu:

Mobile Access Point Deployment in Workflow-based Mobile Sensor Networks

People

Timeline

Project Description

With the advances in mobile networks, we see the emergence of mobile sensor networks that enable flexible and wide monitoring and collection of sensory data in diverse indoor and outdoor environments. Furthermore, we observe a wide spread of mobile sensor networks, where sensing devices such as smartphones, tablets or other types of sensors are carried by robots or people (MUs) that accomplish mission-critical tasks, hence move according to predefined workflows. Note that workflows define mobility patterns, i.e., sequences of mission locations that MUs are scheduled to visit.

The efficient and effective gathering of sensory data in the aforementioned scenarios is of great importance. One possible solution is to utilize existing commercial cellular networks to collect these data. However, in many of the aforementioned scenarios such as infrastructure monitoring and military surveillance, cellular network coverage is typically not available. Even in scenarios where cellular coverage is guaranteed, it is highly costly to purchase long-term unlimited data plans for MUs because of the large amount of data that need to be gathered. In this project, we consider a potentially much more cost efficient alternative that deploys wireless access points working as sink nodes to gather data from MUs. 

The problems of deploying stationary and mobile wireless APs for data collection in sensor networks have been widely studied. The stationary AP deployment problem has been extensively studied in several previous works to provide coverage to both mobile and stationary sensor nodes in a cost and energy efficient manner. However, stationary AP deployment, supporting mobile users, leads to AP-resource overprovisioning, hence great inefficiencies since when mobile devices move out of APs’ coverage regions, those stationary APs become heavily underutilized. Therefore, the idea of placing APs on vehicles or mobile robots to make APs mobile so as to alleviate such underutilization. However, different from the existing mobile AP deployment schemes that minimize data collection latency, maximize throughput of the sensor network, and minimize energy consumption, our work takes a very different perspective on the mobile AP deployment problem. We take into consideration MUs’ workflows and explore the cost-minimizing mobile AP deployment problem in workflow-based mobile sensor networks.

Publication

Funding Agencies

This project is supported by Boeing.