FFT and Slicing Block for an OFDM System

Overview

This project implements a 128-point FFT block and a slicing unit for an OFDM (Orthogonal Frequency-Division Multiplexing) communication system. It decodes complex waveforms into frequency bins and extracts binary data from specific tones.

The design meets the following specifications:

• 128-point FFT for complex signals.
• 24 data tones, each coded as 2 bits.
• Synchronization tones in bins 55 and 57 for full-scale reference.

---

Features

1. 128-Point Complex FFT:

   • Processes signed 16-bit inputs in 1.15 fixed-point format.
   • Converts time-domain signals into frequency bins.

2. Slicing Block:

   • Maps frequency bin magnitudes to 2-bit binary values.
   • Decodes energy levels into 00, 01, 10, or 11.

3. Data Format:

   • Input: 16-bit real and imaginary parts.
   • Output: 48-bit binary data for 24 tones.

---

Specifications

1. Input Data:

   • Real and imaginary parts in 1.15 fixed-point format.
   • Range: ±1 (normalized).

2. Output Data:

   • 48 bits per FFT block (2 bits per tone for 24 tones).
   • Energy levels:
     • 00: 0% (<25% of full scale)
     • 01: 33% (≥25% and <50%)
     • 10: 66% (≥50% and <75%)
     • 11: 100% (≥75%)

3. Synchronization Tones:

   • Fixed tones in bins 55 and 57 for full-scale reference.

4. Module Ports:

   Port Name	Direction	Bits	Description
   Clk	Input	1	Positive edge system clock.
   Reset	Input	1	Active-high asynchronous reset.
   PushIn	Input	1	Data present on input.
   FirstData	Input	1	Indicates the first FFT block.
   DinR	Input	16	Real part of input (1.15 fixed-point).
   DinI	Input	16	Imaginary part of input (1.15 fixed-point).
   PushOut	Output	1	Data present on output.
   DataOut	Output	48	Decoded 48-bit output.

---

Design Flow

1. Input Processing:

   • Accepts complex signals (real and imaginary parts) in fixed-point format.
   • Applies a 128-point FFT to convert time-domain data to frequency bins.

2. Slicing:

   • Evaluates magnitudes of frequency bins (4–52).
   • Categorizes energy levels into binary values based on full-scale reference.

3. Output Data:

   • Produces 48-bit binary data for every FFT block.
   • Data corresponds to 24 tones with 2 bits per tone.

---

Challenges and Solutions

1. Precision:

   • Managed fixed-point arithmetic for accurate FFT results.

2. Synchronization:

   • Utilized bins 55 and 57 to establish full-scale reference levels.

3. Energy Mapping:

   • Implemented robust slicing logic for consistent decoding.

---

How to Use

1. Setup:

   • Integrate the module into your OFDM communication system.
   • Provide 128-point complex waveform input via DinR and DinI.

2. Run:

   • Connect Clk and Reset signals.
   • Use PushIn and FirstData to manage input flow.

3. Extract Output:

   • Read 48-bit binary data from DataOut for each FFT block.

---

Future Enhancements

• Support for larger FFT sizes (e.g., 256 or 512 points).
• Dynamic tone mapping for enhanced system flexibility.
• Optimization of fixed-point arithmetic for reduced latency.

---

License

This project is licensed under the MIT License. Contributions and feedback are welcome!