Non-Centralised and Non-GPS Navigation Mechanism using IoT sensors: challenges and trade-offs

Abstract

In this paper, we propose a non-centralised navigation setup framework using the ubiquitous IoT sensor nodes as landmarks for non-GPS navigation where GPS or cellular coverage is poor or non-existent. Navigation will not require central control like legacy cellular networks but the IoT nodes can communicate among themselves. Such scenarios may arise in dense urban landscape and forest regions where device to device communication can enable navigation. These IoT sensors for navigation purposes will be referred to as navigation anchor points (NAP). When a navigation request arrives, these NAPs send distance information to the hand-held device of the requester. This hand-held device runs a computationally inexpensive shortest path algorithm to find out the NAPs between the source and the destination. The NAPs estimate the distances between themselves by using the received signal strength when they exchange beacon packets. The distance estimation between these NAPs can be plagued by measurement noise due to the multi-path effect. Measurement noise can be reduced by increasing the number of beacon packets used to measure the signal strength or by changing the transmission power. In this paper, we conducted simulations with a variable number of beacon packets and power and include all types of wireless channel conditions. We argued that radio clutter can change without any notice due to objects, humans and change in vegetation. Therefore it is necessary to run navigation setup every time a request arrives and not rely on a fixed set of NAPs for a source-destination pair. These NAPs are mostly battery operated and therefore, energy conservation for an extended period of operation is necessary. Hence we can see a trade-off between the accuracy of the number of NAPs used in navigation to the energy used in setting up the navigation path. This paper has explored the possibility of using the non-GPS non-centralised navigation scheme in scenarios where there is inadequate satellite coverage, and the NAPs are power-constrained, and the navigable device cannot support energy-expensive GPS based navigation. The simulation results with different network densities and uncertainties in distance measurement show that there can be lower bound limit to the transmission power irrespective of the number of beacon packets that can be used to reach the optimal number of NAPs. This framework is an excellent example of use-case for 5G device-to-device (D2D) communication.