Link for ModelSim

Decoding with Inphase and Quadrature Convolution

Decoding a received Manchester signal can occur in several ways, but the approach taken in the model for this tutorial is to consider Manchester Encoding as a digital phase modulation with two symbols: +180 and –180 degrees. By convolving the incoming signal with a reference inphase (I) and quadrature (Q) waveform at the modulation frequency, it is possible to extract the data and retrieve information about any phase errors in the received waveform. After one data cycle, the receiver computes two values (referred to as isum and qsum in the VHDL code), which are measurements of the inphase and quadrature convolution values. The receiver then decodes the values to predict

• The original transmitted data value for the cycle

• An estimate of the phase error between the incoming signal and the receiver's data period

A critical aspect of this design is the interpretation of the I/Q convolution measurements. At the end of a data receive cycle, the decoder translates the I/Q values into an estimate of the transmitted data and phase error. One way to visualize the receiver's condition is to plot I/Q measurements. This tutorial presents the I/Q maps of a receiver design.

Data is considered invalid if isum and qsum are completely ambiguous about the data value of the received waveform.

In a similar way, you can generate an I/Q mapping of the phase adjustment value in plot format. Such a plot gives a visual representation of the decoding block. In practice, the details of this mapping have strong impact on the stability and performance of the Manchester receiver. In the ideal case where the receiver is perfectly locked to the incoming waveform, the receive cycle is 16 cycles long and the measured I/Q convolution values are easy to interpret. However, data cycles that are 15 or 17 cycles long create some bias in the measurement of the I/Q convolution. It is possible to customize the I/Q measurement during these cycles, but that would increase the size and complexity of the receiver. Instead, the data acquisition cycle is extended or shortened with no change in decoding the resulting values. However, this decoder bias can create problems with dithering or reduced noise immunity. This tutorial examines these issues.

The Receiver

Setting Up Tutorial Files