Opportunistic data gathering in IoT networks using an energy-efficient data aggregation mechanism

Abstract

Internet of Things (IoT) applications rely on data collection and centralized processing to assist decision-making. Nevertheless, in multi-hop Low Power and Lossy Networks (LLN) scenarios, data forwarding can be troublesome as it imposes multiple retransmissions, consuming more energy. This paper revisits the concept of mobile agents to collect data from sensors more efficiently. Upon receiving a data request, the IoT gateway performs a cache lookup and promptly dispatches a mobile agent to collect data if this is not available. Data collection then uses closed-loop itineraries computed using a Traveling Salesman Problem (TSP) heuristic starting at the network gateway. The itinerary goes through nodes producing solicited and unsolicited data. We assume that the unsolicited data will be requested soon, and opportunistically collecting it avoids future agent transmissions. We limit the collection capacity of each agent using a knapsack problem approach. Simulation results show that our proposal reduces the network traffic and energy consumption compared with a traditional mobile agent without opportunistic data collection. In addition, we show that data aggregation can further improve the performance of our proposal.