Development and evaluation of an autonomous wirelesssensor actuator network in logistic systems

Abstract

In this work, definition, development and evaluation of Autonomous Wireless Sensor Actuator Network (AWSAN) is discussed. Firstly the central and distributed systems are described. Then characteristic for autonomous system are considered. Based on the feedback model it is explained how the autonomous concept is applied in sensor actuator network. The necessity of developing the intelligent container to establish autonomous logistic system and reinforcing the system performance is explained.In order to develop an autonomous network a target-oriented routing algorithm is invented. By this routing algorithm the message is routed over the minimum length tree of the network graph and directed by sequential coordinates. This routing algorithm is called SCAR. Moreover to implement the AWSAN in automation process application it was required to find an optimal sample number for the sensor. A method is introduced by which it can be answered how often the sensors should take samples from the environment. A typical sensor node for implementation is also introduced.To evaluate the sensor network, it is simulated by a Probabilistic Wireless Network Simulator (Prowler). The network energy consumption and its distribution over the nodes in network with central and autonomous structure are studied. It is shown that the autonomous network consumes less energy than the central network. In the autonomous network the network energy consumption is distributed over the nodes more even compare to central network. This claim implies that the autonomous network is more sustainable. In another simulation an apple garden humidity is taken as case study to compare the robustness of the autonomous and central networks. The result is shown that the autonomous network is more robust. According to the optimal sample number method, it is also shown that with the autonomous network better control quality is achievable and the network is more scalable. During this work the AWSAN is seen as a micro scale of large autonomous systems and with some resemblances the result are generalized to autonomous system as well. Finally it is discussed that if decision by entities becomes dependent on conditions, the above results will face challenges and shows that the merits of autonomous network are not absolute.