Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Fault Tolerance for Access Point Failures in Smart Greenhouse Networked Control Systems
The need for improving and increasing crop production, especially in harsh environmental conditions, has promoted the use of technological solutions, tools and automated methodologies inside a greenhouse to provide the best growing conditions for crops. This thesis presents an innovative fault-toler...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
AUC Knowledge Fountain
2022
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| Summary: | The need for improving and increasing crop production, especially in harsh environmental conditions, has promoted the use of technological solutions, tools and automated methodologies inside a greenhouse to provide the best growing conditions for crops. This thesis presents an innovative fault-tolerant model on the access point level for a Networked Control Systems (NCS) greenhouse. NCS provides the ability to monitor and control the internal environmental conditions of a greenhouse such as temperature, relative humidity, and soil moisture. However, there is a risk of failures occurring at the access point level, that must be taken into consideration. Therefore, this thesis provides a comprehensive study on all access point failure cases including single, double, triple and quadruple access point failures. Riverbed Modeler is used to evaluate system performance in terms of Packet Loss Rate (PLR). For single and double access point failure scenarios, the proposed model succeeded in meeting system requirements with lower PLR values below the threshold value with 95% confidence. For triple and quadruple access point failure scenarios, Riverbed simulations showed that the system can tolerate up to three and four access point failures if the remaining active access points are placed in their ideal position within the greenhouse to provide an optimal coverage and lower PLR values. Additionally, a technique was proposed to help system designers balance between system cost (regarding access points) and the cost of system downtime. Finally, a use case is presented to find the point of diminishing returns, in which investing in access points does not only depend on their cost but also on system availability. Markov models are used to measure system downtime by calculating steady state availability (AVss). System availability was calculated by modeling several scenarios using SHARPE. As expected, it was shown that system availability increases with the increase of the number access points inside the greenhouse. While some cases showed that having three or two access points can achieve the same system availability with lower system cost but with a slight increase in system downtime. |
|---|