Compression method using Vector Quantization in cable TV HFC network

Abstract

Due to the convergence of broadcasting and communication, the cable broadcasting network is changing from the existing Hybrid Fiber Coaxial (HFC) to the Fiber To the Home (FTTH) due to the emergence of new types of services. RF-signal over IP (RoIP) technology has also emerged in order not to change the existing equipment and transmission method. However, when using RoIP technology, there is a problem that the data is greatly increased in the process of digitizing the upstream RF signal to the optical IP network. Centralized Radio Access Network (C-RAN), which is an Long Term Evolution (LTE) environment, requires high bandwidth in the process of digitizing IQ data according to the Common Public Radio Interface (CPRI) standard. To solve this problem, IQ data compression conditions and compression methods such as Up/Down Sampling and Non-linear Quantization are proposed in Open Radio Interface (ORI). In addition, several methods such as block scaling and Huffman coding are emerging. Thesis compresses the uplink RF signal based on DOCSIS 3.0 in HFC environment using up/down sampling and vector quantization. The conventional nonlinear quantization is scalar quantization, and I and Q are quantized separately. However, the vector quantization can quantize I and Q at the same time by vector quantization, thereby improving the compression rate. From these results, it is confirmed that the ORI compression condition is satisfied. Compared with the conventional nonlinear quantization, a good EVM was obtained for the same compression rate. Also, the existing vector quantization increases the code book search complexity as the code book size K increases. Therefore, instead of calculating the squared Euclidean distance for all existing clusters, we first find clusters close to the input data and calculate the squared Euclidean distance only for the corresponding clusters. As a result, code book search complexity can be reduced.