By enkatesh Mutalik, Marcel Schemmann, Craig Malone, Zhijian Sun, Long Zou, Mary Margaret Ong, Alexei Chkourko, Philips Broadband Networks
Analog-to-digital conversion techniques are finding increasing application in HFC transmission networks as a means of transporting analog signals reliably. Among the advantages of these techniques are a range of multiplexing options, mass-produced devices, and the ability to perform additional signal processing in the digital domain.
The role of Digital Signal Processing (DSP) in an analog-to-digital converted CATV system may be divided into two broad sections. In the first, the processing is independent of the specific nature of payload (e.g., the data could be of QPSK, QAM16, QAM64 or QAM256 types). Here, we would primarily be designing general compression schemes as well as automating various self-test and EMS routines, activities which are roughly independent of the digitized information content. In the second, the signal processing can take advantage of known characteristics of the payload data. For example, here, we would be interested in designing appropriate filters, demodulators, and other devices more appropriate for distributed CMTS-type applications.
The authors have implemented a hybrid combination of Digital Signal Processing (DSP) and traditional RF technology to improve HFC return path performance. This approach uses an A/D optical return path transmitter and receiver which, in addition to normal "brute force" coding, adds DSP algorithms that enhance the achievable dynamic range and conserve bit rate. Test results on prototypes show dramatic improvement in Noise Power Ratio (NPR), and simulations closely match the measured performance.
This paper will present some of these results and address digital signal processing of the payload agnostic type. It will also describe some of the technical details in relation to