⚡ electrical engineering

Error Rate Analysis and Low-Complexity Receiver Design for Zero-Padded AFDM

This paper proposes and analyzes low-complexity MMSE and MRC-TD detectors for zero-padded AFDM systems that leverage the channel's lower triangular structure to achieve performance comparable to conventional matrix inversion methods while significantly reducing computational complexity.

Original authors: Qin Yi, Zeping Sui, Zilong Liu

Published 2026-04-29

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Qin Yi, Zeping Sui, Zilong Liu

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Picture: Sending Messages on a Bumpy Road

Imagine you are trying to send a long string of letters (data) to a friend using a walkie-talkie. The problem is that you are both moving very fast (like on a high-speed train), and the signal bounces off buildings and mountains. This causes two main problems:

The "Echo" Problem (Inter-Symbol Interference): The letters start to blur into each other because the signal takes different amounts of time to arrive.
The "Doppler" Problem: Because you are moving, the pitch of the voice changes (like a passing siren), which scrambles the message.

Current technology (called OFDM) tries to fix this by adding a "guard space" between letters, but it's not perfect when things are moving super fast. A newer technology called AFDM (Affine Frequency Division Multiplexing) is better at handling this motion, but it's usually complicated and slow to decode.

What This Paper Does

The authors of this paper propose a smarter way to use AFDM. They introduce a method called Zero-Padding (ZP).

The Analogy: Imagine you are sending a package.
- The Old Way (Cyclic Prefix): You wrap the package in a special bubble wrap that repeats the last few inches of the package. It helps, but it wastes space and energy.
- The New Way (Zero Padding): Instead of wrapping the end, you just leave a few inches of empty space (zeroes) at the end of the package.
- The Benefit: This empty space creates a very neat, organized structure inside the receiver's computer. It's like organizing a messy desk by putting everything in a specific, predictable order.

The Two New "Detectors" (The Decoders)

The main goal of the paper is to build a "decoder" that can read these messages quickly without getting tired (low complexity). They designed two new tools:

The "Cholesky" Decoder (Low-Complexity MMSE):
- The Metaphor: Imagine you have a giant, complicated math puzzle (a matrix) that usually takes a supercomputer hours to solve. Because of the "Zero Padding" trick, the puzzle turns out to be shaped like a triangle with lots of empty space.
- The Trick: Instead of solving the whole giant puzzle at once, the authors found a way to break it down into smaller, easy steps (like peeling an onion layer by layer). They call this "Cholesky decomposition."
- Result: It solves the puzzle just as accurately as the supercomputer method, but it's much faster and uses less energy.
The "Iterative Refiner" (MRC-TD Detector):
- The Metaphor: Imagine you are trying to hear a whisper in a noisy room.
- The Process:
  1. You guess what the first word is.
  2. You subtract that guess from the noise to see what's left.
  3. You use the clearer signal to guess the next word.
  4. You repeat this process a few times, getting cleaner and cleaner with each pass.
- Result: This method also gets the message right, and you can choose to stop after 2 passes (very fast) or 12 passes (very accurate), giving you flexibility.

What They Found (The Results)

The authors ran computer simulations to test their ideas. Here is what they discovered:

Accuracy: Their new, fast decoders work exactly as well as the slow, heavy-duty decoders used today. They didn't lose any accuracy to gain speed.
Speed: The new methods are significantly less complex. In computer terms, they reduced the work required from a massive "cubic" amount to a much smaller "linear" amount.
Better than the Old Stuff: Their "Zero-Padded" system performed better than the standard "Cyclic Prefix" system.
- Why? Think of it like a runner. The old system wastes energy carrying extra weight (the prefix). The new system puts all its energy into the actual message (the data), so it runs faster and further.
The Math Checks Out: They wrote down complex mathematical formulas to predict how often errors would happen. When they compared these formulas to their computer simulations, the lines matched up perfectly.

Summary

This paper is about making high-speed wireless communication (like for trains or satellites) more efficient. By leaving a little bit of "empty space" at the end of data packets, the authors created a system that is easier to decode. They built two new "decoders" that are fast, energy-efficient, and just as accurate as the current heavy-duty methods. It's like upgrading from a heavy, slow truck to a sleek, fast sports car that gets the same job done without breaking a sweat.

1. Problem Statement

The next generation of wireless communication systems (e.g., high-speed trains, V2X, LEO satellites) requires reliable data transmission in high-mobility scenarios. In these environments, conventional Orthogonal Frequency Division Multiplexing (OFDM) suffers from severe performance degradation due to Doppler-induced Inter-Carrier Interference (ICI) in doubly-selective (time- and frequency-selective) channels.

While Affine Frequency Division Multiplexing (AFDM) has emerged as a robust, Doppler-resilient waveform, existing research primarily focuses on AFDM with a Cyclic Periodic Prefix (CPP). CPP-based systems suffer from power inefficiency (energy wasted on the prefix) and do not fully exploit the structural properties of the channel matrix. Conversely, Zero Padding (ZP) is known in OFDM to eliminate Inter-Symbol Interference (ISI) while preserving a sparse, lower-triangular time-domain (TD) channel structure, offering higher power efficiency and better performance in deep fades.

The Core Challenge: There is a lack of low-complexity receiver designs specifically for Zero-Padded AFDM (ZP-AFDM). Conventional detection methods (like Matrix Inversion MMSE) have high computational complexity ( $O(N^3)$ ), making them impractical for real-time high-mobility applications. The paper aims to design low-complexity detectors for ZP-AFDM and rigorously analyze their error rate performance.

2. Methodology

The authors propose a comprehensive framework for ZP-AFDM, covering system modeling, detector design, and theoretical performance analysis.

A. System Model

Transceiver Architecture: Information symbols are modulated via the Inverse Discrete Affine Fourier Transform (IDAFT) onto chirp subcarriers. A Zero Padding (ZP) sequence is appended to the signal to prevent ISI.
Channel Characterization: The system operates over a doubly-selective channel with $P$ paths.
Structural Insight: Unlike CPP-AFDM (which results in a circulant channel matrix), the ZP-AFDM system yields a lower-triangular banded channel matrix ( $H$ ) in the time domain. This structure is key to reducing complexity.

B. Proposed Detection Algorithms

The paper introduces two novel low-complexity detectors that exploit the lower-triangular structure of $H$ :

Low-Complexity MMSE Detector:
- Instead of inverting the full effective channel matrix in the affine frequency domain (complexity $O(N^3)$ ), the detector operates in the time domain.
- It utilizes the Cholesky decomposition of the matrix $\Psi = H^H H + \frac{1}{\gamma_s}I_N$ .
- Since $\Psi$ is a banded Hermitian matrix (bandwidth $2Q+1$ ), its Cholesky factor $L$ is also banded.
- Detection is performed via forward and backward substitution ($Lz=b$ and $L^H\hat{s}=z$ ), reducing complexity significantly.
MRC-TD Detector (Maximum Ratio Combining in Time Domain):
- An iterative detector that directly exploits the sparse lower-triangular structure of $H$ .
- It sequentially updates symbol estimates by coherently combining non-zero channel coefficients.
- It performs Iterative Interference Cancellation (IC): after estimating a symbol, its contribution is subtracted from the residual signal to refine subsequent estimates.
- The process iterates until convergence or a maximum iteration count $K$ is reached.

C. Theoretical Performance Analysis

ML Detection: The authors derive a closed-form Bit Error Rate (BER) upper bound for the Maximum Likelihood (ML) detector using the Union Bound technique and Pairwise Error Probability (PEP) analysis.
MMSE Detection: The BER is approximated by analyzing the Signal-to-Interference-plus-Noise Ratio (SINR) of each chirp subcarrier, leading to a closed-form expression involving the complementary error function.

3. Key Contributions

Novel Transceiver Design: Established the generic architecture for ZP-AFDM, highlighting the unique lower-triangular structure of the TD channel matrix.
Low-Complexity Detectors:
- Proposed a Cholesky-based MMSE detector that avoids matrix inversion, reducing complexity from $O(N^3)$ to $O(NQ^2 + N \log_2 N)$ .
- Proposed an iterative MRC-TD detector with complexity $O(KNQ + N \log_2 N)$ , offering a flexible trade-off between performance and complexity.
Closed-Form BER Analysis: Derived theoretical BER expressions for both ML and MMSE detectors, providing tight upper bounds that match simulation results.
Performance Validation: Demonstrated that ZP-AFDM outperforms CPP-AFDM due to better power allocation (no energy wasted on prefixes) and superior diversity.

4. Simulation Results

The authors validated their theories through extensive Monte Carlo simulations:

Accuracy of Theory: The derived theoretical BER curves for both ML and MMSE detectors perfectly overlap with simulation results, confirming the accuracy of the analytical models.
ZP vs. CPP: ZP-AFDM consistently outperforms CPP-AFDM across various modulation schemes (BPSK, QPSK) and path counts. At a BER of $10^{-3}$ , ZP-AFDM provides significant $E_b/N_0$ gains (approx. 3.8 dB) over OFDM counterparts.
Detector Performance:
- The proposed Low-Complexity MMSE achieves identical BER performance to the conventional (high-complexity) MMSE detector.
- The MRC-TD detector converges rapidly (within 2–12 iterations depending on SNR) to achieve BER performance nearly identical to the conventional MMSE.
Complexity Reduction: The proposed detectors offer substantial computational savings. The MRC-TD detector, in particular, allows system designers to tune the number of iterations to balance latency/complexity against error performance.

5. Significance

This work is significant for the advancement of 6G and high-mobility communication systems for several reasons:

Enabling High-Mobility Comms: It provides a practical, low-complexity solution for AFDM, a leading candidate waveform for high-Doppler environments, making it feasible for real-time implementation in hardware.
Efficiency: By adopting Zero Padding, the system maximizes power efficiency and ensures reliable recovery even under deep channel fades (spectral nulls), addressing a key limitation of CP-based systems.
Theoretical Rigor: The paper bridges the gap between theoretical performance bounds and practical receiver design, offering closed-form BER expressions that can guide system optimization without relying solely on time-consuming simulations.
Scalability: The reduction in complexity from cubic to linear/quadratic order makes the technology scalable for large bandwidths and high-order modulations required in future networks.