Implementasi modifikasi metode Frequency Hopping Spread Spectrum pada Audio Watermarking

Abstract

Today, through the development of technology, digital data is easier to make, manipulate, and even misused for harmful purposes, such as the falsification of property or copyright infringement. The misuses can be overcome by using digital watermarking technique. In this paper, wave audio files (.wav) used as digital data was given watermark using modified frequency hopping spread spectrum (FHSS) method. The audio files were given watermark “Indonesia” with variant multiplier (scaling factor), spread to the audio files by hopping. The quality of watermarked audio was measured using signal to noise ratio (SNR) and questionnaire from 30 respondents. Analysis is concerned on the process of the running time of watermark insertion and detection, changes in quality watermarked audio related to the use of scaling factor and the bit error watermark detection results. To test the resilience of the watermark, watermarked audio were attacked using several operations. Based on the analysis results, increasing the value of scaling factor causes a decrease in the quality of watermarked audio, marked by the decline in value of SNR. The highest SNR value is 70.5147 dB for rock music. Based on the results of the questionnaire, 86.00 % of respondents said that watermarked audio volume the equal to the original audio files and 94.67 % of respondents said that noise were not heard from the watermarked audio were. For the bit error watermark, increasing the value of scaling factor causes a decrease in percentage of bit error watermark detection. The percentage of the highest bit error watermark occurs on instrument music (5.010 %). The results of resilience of the watermark test showed modified FHSS method was resistant against resampling operation, addition of noise with small amplitude and multiple watermarks, but this method was not resistant to cropping operations.