Spectrum handoff strategy for cognitive radio-based Mac in industrial wirless sensor and actuator networks
In this thesis, a Cognitive Radio(CR)-based MAC for Industrial Wireless Sensor and Actuator Network (IWSAN) applications is proposed. IWSANs are typically used for closed-loop control applications, and they demand strict requirements in terms of time and robustness. Low latency and low error rates a...
| Idioma: | English |
|---|---|
| Publicación: |
Mondragon Unibertsitatea. Goi Eskola Politeknikoa
2019
|
| Materia: | |
| Acceso electrónico SOLREBILTEGIA: |
https://hdl.handle.net/20.500.11984/1197 https://doi.org/10.48764/jzmf-qz71 https://katalogoa.mondragon.edu/janium-bin/janium_login_opac.pl?find&ficha_no=126067 |
| Resumen: |
In this thesis, a Cognitive Radio(CR)-based MAC for Industrial Wireless Sensor and Actuator Network (IWSAN) applications is proposed. IWSANs are typically used for closed-loop control applications, and they demand strict requirements in terms of time and robustness. Low latency and low error rates are required in order not to endanger persons or machinery. Moreover, these applications operate in industrial environments such as factories or transport scenarios (as aeronautics or railway) where multipath fading and shadowing are present due to metal surfaces. Furthermore, interference from other communication systems or industrial machinery is also common in these environments. The proposed MAC, based on the CR paradigm, is capable of ensuring time and robustness requirements in industrial channels.
In the process of designing the CR-based MAC for IWSAN applications, a comparison between several non-CR-based MACs and CR-based MACs has been carried out. This comparison, which allows stating the benefits of CR for these applications, is presented in this thesis. The performance of every MAC is determined theoretically using Network Calculus, and it is validated through OPNET simulations. CR solutions, due to their adaptability characteristics, are capable of avoiding interference and ensuring robustness in industrial environments. However, none of the selected MACs are capable of ensuring robustness without comprising time requirements, since interference is avoided but not in a bounded time. On the other hand, the MAC proposed in this thesis is capable of avoiding interference ensuring time and robustness requirements at the same time. This MAC is therefore suitable for IWSAN applications.
To ensure a deterministic behavior against interference, a novel handoff algorithm, which detects interference and hops to another channel, has been proposed. This algorithm has been designed to be used jointly with one of the evaluated MACs. The detection of the interference and the hop to another channel is done in a bounded time, because the proposed algorithm detects interference while the system is transmitting. The performance of this proposal is evaluated using Network Calculus and OPNET simulations, and the results are compared with the system without the proposed handoff algorithm. The comparison of the results shows how the evaluated MAC is only capable of ensuring both time and robustness requirements when the proposed handoff strategy is used.
Moreover, the spectrum sensing algorithm used to obtain information about the environment is delved and its performance is measured through MATLAB simulations. An energy detector has been chosen due to its simplicity. Also, a cyclostationary Modulation Classifier is presented and a simplification has been carried out allowing its implementation on real hardware. The Modulation Classifier is capable of distinguishing between OFDM, QPSK and GFSK signals. The performance of the algorithm is presented in this thesis for different signals and for different receiver impairments such as frequency offset, DC offset and I/Q imbalance.
Finally, a cognitive platform to validate the spectrum sensing algorithms is presented. This platform has been designed using a Xilinx Virtex 6 FPGA by a working group composed of researchers from IK4-Ikerlan and Mondragon Unibertsitatea. The platform, which uses both spectrum sensing algorithms, is an Ethernet-to-RF bridge. It has been designed to replace an Ethernet wired link by a wireless one for IWSAN applications. The proposed platform ensures a reliable communication link against interference. In the proposed implementation, the energy detector is used by the transmitter in order to find a free channel to transmit data, whereas the modulation classifier is used by the receiver in order to distinguish between the signal transmitted by the RF-Ethernet bridge and other signals. In this way the receiver can find the channel where the transmitter is carrying out the communication. |
|---|