JumboNet Elephant Tracking Using Delay-Tolerant Routing with Multiple Sinks

Abstract

Wildlife monitoring is vital in areas where humans and animals share the same living area. For instance, in Sri Lanka there are a high number of deaths among elephants and people due to the human-elephant conflict. A possible solution to reduce the number of deaths is to monitor the elephants' locations in real-time, and to promptly intervene when elephants approach human populated areas. Tagging elephants with GPS collars represents a viable solution, but obtaining location information can quickly drain tag batteries. As a result, systems that utilize energy harvesting have been proposed as a promising alternative to ensure very long lifetimes of the tags. As the amount of energy that can be harvested in a given time period is limited, communication protocols must be designed to be energy efficient, while still ensuring a high packet delivery ratio and low delay. In this paper, we analyze the performance of different delay-tolerant network protocols in a real scenario using actual elephant movement data from JumboNet and examine the effects of increasing the number of sinks in the network on the packet delivery ratio, average packet delay, and average energy consumption. We find that epidemic routing outperforms other protocols in terms of packet delivery ratio when using low transmission powers. On the other hand, if enough energy is available and the tracking system can support high transmission powers, direct delivery is superior compared to a multi-hop routing approach.