Virtual Laboratory: Companding in Digital Communication

Theory of Companding

Companding (compressing + expanding) is a non-linear technique used in digital communication systems to improve the signal-to-noise ratio (SNR) for signals with wide dynamic range. It is particularly important in pulse code modulation (PCM) systems.

Why Companding is Needed

In digital communication systems, analog signals are converted to digital form through sampling and quantization. Uniform quantization allocates equal step sizes across the entire amplitude range, which is inefficient for signals like speech that have a non-uniform amplitude distribution.

Without companding, weak signals suffer from poor quantization SNR while strong signals have excess SNR. Companding solves this by:

Compressing the signal at the transmitter using a non-linear amplifier
Applying uniform quantization
Expanding the signal at the receiver with an inverse non-linear characteristic

Companding Process in PCM Systems

Input Signal → Compressor → Uniform Quantizer → Encoder → Channel → Decoder → Expander → Output Signal

Types of Companding Laws

Feature	μ-law Companding	A-law Companding
Standard	North America & Japan	Europe & International
Compression Characteristic	y = (ln(1+μ\|x\|)) / (ln(1+μ)) × sign(x)	y = (A\|x\|)/(1+ln(A)) for 0≤\|x\|≤1/A y = (1+ln(A\|x\|))/(1+ln(A)) for 1/A≤\|x\|≤1
Typical Parameter Value	μ = 255	A = 87.6
Implementation	15-segment piecewise linear approximation	13-segment piecewise linear approximation
SNR Performance	Better for low-level signals	Better for high-level signals

Mathematical Representation

μ-law Compression Characteristic

y = sign(x) × [ln(1 + μ|x|) / ln(1 + μ)]

where: -1 ≤ x ≤ 1, μ is compression parameter (typically 255)

A-law Compression Characteristic

For 0 ≤ |x| ≤ 1/A: y = sign(x) × [A|x| / (1 + ln(A))]

For 1/A ≤ |x| ≤ 1: y = sign(x) × [(1 + ln(A|x|)) / (1 + ln(A))]

where: A = 87.6 for standard implementation

Advantages of Companding

Improves SNR for weak signals
Reduces the required number of bits for a given SNR
Provides robust performance for signals with wide dynamic range
Reduces quantization noise for low-amplitude signals

Frequently Asked Questions

What is the purpose of companding in PCM systems?

Companding is used to improve the signal-to-quantization-noise ratio (SQNR) for low-amplitude signals in PCM systems. Without companding, uniform quantization would result in poor SQNR for weak signals. Companding compresses the signal before quantization and expands it after reconstruction, effectively making the quantization steps non-uniform.

What is the difference between μ-law and A-law companding?

μ-law (used in North America and Japan) provides better performance for low-level signals, while A-law (used in Europe and internationally) offers better performance for high-level signals. μ-law uses a 15-segment piecewise linear approximation, while A-law uses a 13-segment approximation. The mathematical compression characteristics also differ between the two standards.

How does companding affect bandwidth requirements?

Companding itself doesn't change the bandwidth of the transmitted signal. However, by improving the SQNR for low-amplitude signals, it allows for fewer quantization bits to achieve the same overall SQNR performance, which reduces the bit rate and thus the bandwidth required for transmission.

Companding Simulation

This interactive simulation allows you to explore the effects of companding on signal quantization. Adjust the parameters to see how different compression laws affect the signal-to-noise ratio.

Input Signal and Quantization

Sinusoidal input signal with amplitude: 1.0 V

Companded Signal and Quantization

Companded output signal with μ-law (μ = 255)

Simulation Controls

Companding Type:

Compression Parameter (μ or A): 255

1 127 255

Input Signal Amplitude: 1.0 V

0.1 V 1.0 V 2.0 V

Quantization Bits: 8

4 bits 8 bits 12 bits

Simulation Results

Signal-to-Quantization-Noise Ratio (SQNR): 38.2 dB

Compression Ratio: 2.5:1

Quantization Levels: 256

Observation: With μ-law companding, low-amplitude signals show improved SQNR compared to uniform quantization.

Laboratory Procedure

Follow these steps to conduct the companding experiment and understand its effects on digital communication systems.

Understand the Theory

Review the theory of companding, uniform quantization, and pulse code modulation (PCM). Understand why companding is necessary for signals with wide dynamic range like speech.

Study the mathematical expressions for μ-law and A-law companding
Understand the difference between compression and expansion
Review the block diagram of a PCM system with companding

Set Up the Simulation

Use the simulation tool to set up different test scenarios:

Select the companding type (μ-law, A-law, or no companding)
Adjust the compression parameter (μ or A)
Set the input signal amplitude and frequency
Choose the number of quantization bits

Run Experiments

Conduct the following experiments using the simulation:

Experiment 1: Compare SQNR for different input amplitudes with and without companding
Experiment 2: Study the effect of changing the compression parameter (μ or A) on SQNR
Experiment 3: Compare performance of μ-law vs A-law companding for different signal levels
Experiment 4: Analyze the effect of changing the number of quantization bits

Record Observations

For each experiment, record the following:

Input parameters (amplitude, frequency, companding type, μ/A value, quantization bits)
Output SQNR values
Visual observations from the signal waveforms
Comparison between companded and non-companded signals

Analyze Results

Based on your observations, answer the following questions:

How does companding affect the SQNR for low-amplitude signals?
What is the optimal value of μ for speech signals?
How does the number of quantization bits affect the performance of companding?
Which companding law (μ or A) provides better performance for very low amplitude signals?
What are the practical implementation advantages of piecewise linear approximation?

Safety and Precautions

This is a virtual laboratory, so no physical safety precautions are needed. However, ensure you:

Save your simulation data regularly
Document each experiment thoroughly
Compare your results with theoretical expectations
Verify calculations for SQNR and compression ratios

Laboratory Report Guidelines

A well-structured lab report is essential for documenting your experiment. Follow these guidelines to create a comprehensive report on companding.

Report Structure

Title Page: Experiment title, your name, course details, date, and instructor's name
Abstract/Summary: Brief overview of the experiment (100-150 words)
Introduction: Purpose of the experiment and background theory
Objectives: Clear statement of what the experiment aims to achieve
Theory: Explanation of companding concepts and mathematical formulas
Experimental Setup: Description of the simulation parameters and procedure
Results and Observations: Data tables, graphs, and analysis
Discussion: Interpretation of results and comparison with theory
Conclusion: Summary of key findings and learning outcomes
References: Citations for any external sources used

Detailed Requirements for Each Section

Introduction (Approx. 200 words)

Explain the importance of companding in digital communication systems, particularly in PCM. Mention the problems with uniform quantization for signals with wide dynamic range and how companding addresses these issues.

Theory Section (Approx. 300 words)

Include:

Mathematical expressions for μ-law and A-law companding
Explanation of compression and expansion processes
Discussion of SQNR and how it's affected by companding
Block diagram of PCM system with companding

Results and Observations

Present your data in clear tables and graphs:

Table comparing SQNR for different amplitudes with and without companding
Graph showing SQNR vs. input amplitude for different companding types
Table showing effect of μ/A parameter on compression performance
Waveform diagrams showing input, compressed, quantized, and reconstructed signals

Sample Data Tables

Table 1: SQNR Comparison for Different Input Amplitudes (8-bit quantization)
Input Amplitude (V)	No Companding (dB)	μ-law (μ=255) (dB)	A-law (A=87.6) (dB)
0.1	18.5	34.2	32.8
0.5	30.1	36.8	37.2
1.0	36.1	38.2	38.5
1.5	38.6	39.1	39.4

Table 2: Effect of μ Parameter on SQNR (Input amplitude = 0.2V, 8-bit quantization)
μ Value	SQNR (dB)	Compression Ratio	Observation
10	26.4	1.2:1	Minimal compression
100	32.8	2.1:1	Good improvement for weak signals
255	34.2	2.5:1	Optimal for speech signals

Discussion Points

In your discussion section, address the following:

How do your experimental results compare with theoretical expectations?
Why is μ=255 considered optimal for speech signals in μ-law companding?
What are the practical implementation challenges of companding?
How does companding affect the bandwidth requirements of a PCM system?
What are the advantages and disadvantages of μ-law vs A-law companding?

Evaluation Criteria

Clarity and organization (20%): Logical flow, proper sectioning, readability
Theoretical understanding (25%): Accurate explanation of concepts, correct formulas
Experimental methodology (20%): Clear description of procedure, appropriate parameters
Results and analysis (25%): Proper data presentation, insightful interpretation
Conclusion and references (10%): Concise summary, proper citations

                Submission Guidelines
                Report should be typed, not handwritten
Include all graphs, tables, and diagrams with proper captions
Use proper units and significant figures in measurements
Cite at least 3 relevant references (textbooks, research papers, standards)
Submit as a single PDF file by the deadline

            