Establishment and optimization of wireless sensor network for micro-grid demand response system and body area networks
Date of Issue2015
School of Mechanical and Aerospace Engineering
Sensor network has attracted worldwide attention over two decades, particularly with the proliferation in Micro-Electro-Mechanical Systems (MEMS) technology which has facilitated the development of smart sensors. Unlike traditional networks, wireless sensor network has its own resource constraints which include a limited amount of energy, short communication range, low bandwidth, and limited processing and storage in each node. This dissertation presents the development of the cross layer optimization to solve the compensation for overload, latency or other mismatch of requirements and resources without considering the hierarchical structure of Open System Interconnection (OSI) model with strict boundaries between layers. To implement the cross layer optimization of link layer and MAC layer, in this thesis, interference mitigation without feedback channel in the wireless body area network (WBAN) and feedback overhead reduction in the smart electric vehicle charging system are investigated as applications. Firstly, the link layer optimization in the wireless body area network is presented to maximize link capacity between transmitter and receiver by considering interference mitigation. Pseudo orthogonal resource allocation patterns are proposed to mitigate the interference in the wireless body area network. To design pseudo orthogonal resource allocation patterns, it is necessary to analyze the properties of a pseudo orthogonal code (POC). By considering the properties of the distance matrix of a pseudo orthogonal code, the low bound of the length of the pseudo orthogonal code is theoretically calculated. However, to optimize data throughput of the resource allocation, the trade-off relationship between the orthogonal properties of a POC and the quantity of resource allocation of WBAN is considered. To mitigate the collisions, the two dimensional resource allocation patterns are proposed, where random assignment of codewords to time domain and that of frequency sub-channel offsets are considered. In addition, based on the properties of correlation and cyclic shift on the POC, the transmission efficiency of WBAN is calculated. Based on the low bound of a POC and the transmission efficiency, the optimal quantity of resource allocation of each WBAN is theoretically analyzed, which is exactly the ratio of the length of a POC to the number of WBANs in the system. The simulation results illuminate that the theoretical low bound of the proposed scheme is more close to the real structure of POC compared to the conventional schemes. The other application of WSN is the electric vehicle (EV) charging system in the MAC layer, which optimizes the scheduling of data transmission. The penetration of EV charging system imposes a heavy burden on the conventional power distribution system, so that the coordination of EV charging demands is required via a bi-directional real time monitoring system. However, it may cause the performance degradation in the whole EV charging system due to the significant control signal overhead. To mitigate the control signal overhead from electric vehicles to a central controller, the threshold based charging scheme in electric vehicle charging system is proposed, which only EVs with low level of battery state of charge below the predetermined charging threshold can transmit their charging request to controller. In the simulation results, the proposed threshold based charging scheme can achieve the reduction of the control signaling overhead and the satisfactory charging performance.
DRNTU::Engineering::Electrical and electronic engineering