The most prevalent Smart Grid application today is connecting the consumer premises to utilities for Automatic Meter Reading (AMR), which requires a very limited amount of bandwidth. Emerging Smart Grid applications employ periodic readings and active control in an attempt to manage the load on the grid (also called Advanced Metering Infrastructure or AMI). Other rapidly emerging applications include Street Light Control (SLC), Vending Machines, Solar Panels, Electrical Vehicle Charging, Smart Appliances and in general any application that involves an electrically connected device requiring monitoring and control. The bandwidth demand of such applications is higher and typically requires between 15Kbps and 30Kbps of reliable data.
As applications develop, the communication techniques employed over the power line have evolved. Initially deployed schemes often included variations of basic single carrier Frequency Shift Keying (FSK) and Phase Shift Keying (PSK) techniques. Such techniques provide limited bandwidth and are limited in their ability to cope with the harsh power line environment reliably.
illustrates FSK and Binary PSK modulations. FSK is highly affected by impulse noise that can spread over a number of bits. PSK is more resistant to noise, but is affected by phase distortion and impedance variation. Combining the two schemes concurrently provides additional level of robustness. Some devices (such as the SM6401 from Semitech Semiconductor) provide the flexibility of the combined approach.
Figure 3 - FSK and BPSK modulations
Other techniques deployed in early systems to avoid impulse and tonal noise involve “spreading” of the communication signal over a wide band or Spread Spectrum. Spread Spectrum technology has its roots in the military. It intentionally uses broad, randomized (noise like) signals that are much wider band than the information they are carrying to make them more noise-like. Spread Spectrum signals use fast codes that run many times the information bandwidth or data rate. These special "spreading" codes are called "pseudo random" or "pseudo noise" codes. Spread Spectrum reception is then performed by correlating the received spread spectrum signal with a replica of the expected waveform. Spread Spectrum communication techniques perform well in the presence of Gaussian noise. However, they tend to struggle with propagation delays and tonal interference that are common in power line environments.
Just like in other domains, the new wave of N-PLC implementations adopts advanced modulation approaches like Orthogonal Frequency Division Multiplexing (OFDM) to better address the increasing data bandwidth and reliability needs. Multiple emerging N-PLC standards, such as ITU G.hnem and IEEE 1901.2, are using OFDM as their underlying technique.
OFDM is a technique for transmitting large amounts of digital data over a noisy channel. OFDM gained considerable success in wireless and other noisy communication environments, as it combines many slow data rate carriers to form an overall higher data rate. The technology works by splitting the signal into multiple smaller sub-signals that are then transmitted simultaneously at different (orthogonal) frequencies. Each smaller data stream is then mapped to an individual data sub-carrier and modulated using PSK or QAM (Quadrature Amplitude Modulation). The primary advantages of OFDM over single carrier schemes are its ability to cope with severe channel conditions and higher data rates. If parts of the spectrum are blocked by noise, with error correction, the data can still be received without errors. Figure 4 illustrates the spectrum of a typical OFDM modem with a single and five sub-carriers. Using orthogonal sub-carriers assures that there is no crosstalk between the sub-carriers. Compared to single carrier modems, OFDM implementations take advantage of more advanced digital signal processing techniques, such as Fast Fourier Transform (FFT).
Figure 5 - Single OFDM frame structure in time and frequency domains
Figure 4 - Examples of OFDM spectrum
OFDM is a well-established and well researched technology that no doubt takes N-PLC to the next level of performance. However, as we have seen, the power line noise environment has unique enough characteristics that may make conventional OFDM insufficient. While OFDM is an inherently adaptive technique, it relies on successful communication over sufficient number of carriers and in particular successful transmission of the frame header and preamble (the exact number depends on the error correction techniques employed and the structure of the frame). The harsh noise conditions of the power line and the fact that many frequencies are temporarily or permanently blocked to communication still present a challenge in that regard. The emerging OFDM based IEEE 1901.2 standard recognized the issue of scattered usable frequencies and has implemented a sub-banding mechanism to filter out noisy portions of the available spectrum (Figure 5
illustrates the structure of an OFDM frame without (a) and with (b) sub-banding). This is a step in the right direction; however, it still does not resolve the vulnerability of the frame header. There is room for even more flexible schemes that can adjust to the power line noise that is time and frequency dependent by adapting modulations, frequencies and the power spread to achieve better communications performance. As an example, the SM2200 device from Semitech Semiconductor implements an OFDMA-like scheme that improves on OFDM by allowing complete independence between the communication channels and enables dynamic channel selection that adapts continuously to changes in the channel characteristics. It is being successfully deployed in China as part of one of its first Advanced Metering Infrastructure deployments.