適用於窄頻電力線通訊 A類干擾 下之適應 性能量分配

Abstract

The ozone layer has been destructed for a long time and eco-tech gets more discussions and attention. Effective resource management in the world become more important. Power management is one important issue of green technology. Through the rapid demand of smart power use and management, issues of power line gradually gets more concern. Smart Grid, which rely on mass power line infrastructure, has rapid development in recent years. An energy-efficiency and real-time power management is critical now and ever. Hence, power line communication (PLC) becomes widely discussed these years. The use of power line distribution grid for data communication has gained lot of interest over the past several years. By establishing communication on the same power line infrastructure that delivers electricity, there is no need to create new communication paths through obstacles such as buildings, hills. Thus, installation and maintenance costs are lower than Ethernet or wireless communication. Nevertheless, PLC suffer many challenges such as multi-path fading, and time-varying impulse noise, which is modeled as class-A noise in this thesis. The proposed receiver solve most of the problem. Unfortunately, impulse noise is sometimes the most system degradation factor that cause transmission fail. Many works shows adding certain processor to deal with it for some system enhancement. For instance, impulse noise suppressor or Viterbi preprocessor is added is their designed receiver. However, lots of hardware overhead are simultaneously produced. Resource allocation is a useful and effective way to improve the system without much hardware overhead. A proposed power allocation/loading scheme added in my transmitter is for reducing the signal distortion by fading and impulse noise. A well-known method to improve data rate about bis/power allocation/loading is waterfilling algorithm. By contrary, my work prefer to ii provide power-saving or bit error rate (BER) improvement Both standards, PRIME and G3- PLC are tested with my algorithms ad others. In severe class-A noise channel, my solution provides 5dB gain over the best power algorithms in class-A scheme and 0.2 dB in well class- A noise channel for PRIME system. In G3-PLC, there’re about 2.8 dB and 0.1dB respectively. This reveals my algorithms is strongly applied in severe class-A case.