Consider a predictive coding system that predicts current data f(n) by f(n-1). Suppose that the encoder...

     

Transcribed Image Text:

Consider a predictive coding system that predicts current data f(n) by f(n-1). Suppose that the encoder encoder quantizes e(n) = f(n)-f(n-1) and generates quantized error e(n)= e(n) + d(n) where d(n) is quantization noise. Determine the output of the decoder when n = 10. You may assume f(-1) = 0 which is known to both encoder and decoder. Discuss the problem of this system and design an improved system by changing the predictor of the encoder. What is the output of the decoder when n = 10 for the improved system? Consider a predictive coding system that predicts current data f(n) by f(n-1). Suppose that the encoder encoder quantizes e(n) = f(n)-f(n-1) and generates quantized error e(n)= e(n) + d(n) where d(n) is quantization noise. Determine the output of the decoder when n = 10. You may assume f(-1) = 0 which is known to both encoder and decoder. Discuss the problem of this system and design an improved system by changing the predictor of the encoder. What is the output of the decoder when n = 10 for the improved system?

Answer rating: 100% (QA)

In the given predictive coding system the output of the decoder at time n denoted as fn is obtained ... View the full answer