Linear Precoding Design for OTFS Systems in Time/Frequency Selective Fading Channels

Basic Information

• Paper ID: 2501.00272
• Title: Linear Precoding Design for OTFS Systems in Time/Frequency Selective Fading Channels
• Authors: Yao Ge, Lingsheng Meng, David González G., Miaowen Wen, Yong Liang Guan, Pingzhi Fan
• Classification: eess.SP (Signal Processing)
• Submission Date: December 31, 2024 to arXiv
• Paper Link: https://arxiv.org/abs/2501.00272

Abstract

Although Orthogonal Time-Frequency Space (OTFS) modulation has been demonstrated as a promising modulation scheme for high-mobility doubly selective fading channels, its capability to achieve full diversity order in time or frequency selective fading channels remains unclear. Through pairwise error probability (PEP) analysis, the authors observe that the original OTFS system cannot always guarantee full exploitation of the embedded diversity in time or frequency selective fading channels. To address this issue and further improve system performance, this paper proposes linear precoding solutions for OTFS systems in time and frequency selective fading channels, respectively, based on algebraic number theory. The proposed linear precoding OTFS system guarantees maximum diversity and potential coding gain in time/frequency selective fading channels without transmission rate loss and without requiring channel state information (CSI) at the transmitter.

Research Background and Motivation

Problem Background

1. Channel Characteristics Challenges: High data rates and multipath propagation lead to frequency selectivity in wireless channels, while carrier frequency offset (CFO) and Doppler effects caused by mobility induce time selectivity
2. OFDM Limitations: Although traditional OFDM can convert frequency selective fading channels into parallel flat fading subchannels, it cannot guarantee subcarrier orthogonality in Doppler time-selective fading channels, resulting in significant performance degradation
3. OTFS Advantages and Shortcomings: OTFS modulation, as an alternative, requires only one cyclic prefix for the entire OTFS frame, achieving high spectral efficiency; however, its full diversity capability in time/frequency selective fading channels lacks theoretical proof

Research Motivation

Through PEP analysis, it is discovered that the original OTFS system cannot always guarantee full exploitation of embedded diversity in time or frequency selective fading channels. Therefore, effective methods that guarantee performance and high spectral efficiency need to be developed.

Core Contributions

1. Theoretical Analysis: First clarifies the diversity performance limitations of OTFS systems in time/frequency selective fading channels through PEP analysis
2. Precoding Design: Proposes linear precoding schemes for time and frequency selective fading channels based on algebraic number theory
3. Performance Guarantee: The proposed scheme guarantees maximum diversity order and potential coding gain without transmission rate loss
4. Practicality: Does not require transmitter CSI and is applicable to arbitrary system dimensions
5. Comprehensive Verification: Simulation results validate the superiority of the proposed scheme compared to original unprecoded and existing phase rotation OTFS systems

Methodology Details

Problem Formulation

Design a linear precoding matrix V such that the precoded OTFS system achieves maximum diversity gain and coding gain in time/frequency selective fading channels while maintaining constant transmission rate.

System Model

Transmitter

Information stream x ∈ A^(MN×1) undergoes linear precoding to obtain:

x̄ = Vx

where V ∈ C^(MN×MN) is the precoding matrix, and M and N are the numbers of delay and Doppler dimension resources in the OTFS grid, respectively.

Channel Model

Frequency Selective Fading Channel:

h = [h[0], h[1], ..., h[L-1]]^T ∈ C^(L×1)

Time Selective Fading Channel: Employs basis expansion model (BEM):

h[c] = Σ(q=0 to Q) c_q e^(jω_q c)

Receiver

Frequency selective channel:

r[c] = Σ(p=0 to L-1) h[p]s[[c-p]_MN] + n[c]

Time selective channel:

r[c] = h[c]s[c] + n[c]

Precoding Design

Frequency Selective Fading Channel

Through PEP analysis, the end-to-end input-output relationship is:

y = HVx = Φ(x)h

To achieve maximum diversity gain L, the matrix C = (Φ(x)-Φ(x̂))^H(Φ(x)-Φ(x̂)) must be full rank.

Vandermonde Precoding Matrix:

Θ = (1/ξ)[1 α₁ ... α₁^(MN-1)
           1 α₂ ... α₂^(MN-1)
           ⋮  ⋮  ⋱    ⋮
           1 αₘₙ ... αₘₙ^(MN-1)]

Parameter Selection:

• If MN = 2^d (d ≥ 1): αₖ = e^(j(4k-3)π/(2MN))
• If MN = 3×2^d (d ≥ 0): αₖ = e^(j(6k-1)π/(3MN))
• If MN = 2^d×3^t (d,t ≥ 1): αₖ = e^(j(6k-5)π/(3MN))

Precoding Matrix:

V = (F_N ⊗ I_M)F_MN^H Θ

Time Selective Fading Channel

Similarly, to achieve maximum diversity gain Q+1, the precoding matrix is:

V = (F_N ⊗ I_M)Θ

Technical Innovations

1. Theoretical Breakthrough: First theoretically analyzes the diversity performance of OTFS in single-selective channels
2. Algebraic Construction: Constructs Vandermonde matrices based on algebraic number theory to ensure full rank conditions
3. Unified Framework: Provides a unified precoding design framework for time and frequency selective channels
4. CSI-Free Design: Precoding design is independent of transmitter channel information

Experimental Setup

Simulation Parameters

• Carrier Frequency: 4 GHz
• Subcarrier Spacing: Δf = 15 kHz
• Modulation Scheme: QPSK
• Perfect channel knowledge assumed at receiver

Comparison Methods

1. Original unprecoded OTFS system
2. Existing phase rotation OTFS system
3. Proposed precoded OTFS system

Detectors

1. Maximum Likelihood (ML) detector (small-dimension systems)
2. Memory Approximate Message Passing (Memory AMP) detector (large-dimension systems)

Experimental Results

Frequency Selective Channel Results

• Small-Dimension System (M=4, N=2): As the number of resolvable paths L increases, the proposed scheme achieves corresponding diversity gain slopes (-2, -3, -4)
• Large-Dimension System (M=128, N=16): Using Memory AMP detector, the proposed scheme maintains performance advantages

Time Selective Channel Results

• Different Mobility Speeds (60/300/600 km/h): Higher speeds yield greater Doppler diversity gains, with the proposed scheme consistently outperforming comparison methods
• Large-Dimension Verification: Performance advantages are maintained under practical detectors

Key Findings

1. The proposed precoding scheme achieves the maximum diversity gain predicted by theoretical analysis
2. Compared to unprecoded and phase rotation schemes, significant performance improvements are observed under various channel conditions
3. Performance advantages remain evident under practical low-complexity detectors

Related Work

Traditional Solutions

1. Linear Constellation Precoding for OFDM 3: Addresses frequency selective channels but incurs spectral efficiency loss due to CP overhead
2. Space-Time Doppler Coding 4: Guarantees maximum space-Doppler diversity but also introduces overhead
3. Block Precoding Transmission 5: Targets doubly selective channels but requires spreading techniques

OTFS-Related Research

1. Diversity Performance Analysis 7,8: Primarily focuses on uncoded and coded OTFS systems in doubly selective channels
2. Existing Precoding: Simple schemes such as phase rotation that fail to guarantee maximum diversity

Advantages of This Work

Compared to existing work, this paper is the first to theoretically analyze and address the diversity problem of OTFS in single-selective channels, providing a solution without rate loss.

Conclusions and Discussion

Main Conclusions

1. The original OTFS system cannot always achieve full diversity in time/frequency selective channels
2. Linear precoding based on algebraic number theory can guarantee maximum diversity gain and coding gain
3. The proposed scheme has no transmission rate loss and does not require transmitter CSI
4. Performance advantages are demonstrated under various channel conditions and detectors

Limitations

1. Assumes perfect channel knowledge at receiver; practical implementation requires channel estimation
2. Simulations primarily focus on QPSK modulation; performance with higher-order modulation requires further verification
3. Complexity analysis lacks detail; computational burden of practical implementation needs further assessment

Future Directions

1. Robust precoding design incorporating channel estimation
2. Extension to higher-order modulation and MIMO systems
3. Algorithm design for further reducing detection complexity

In-Depth Evaluation

Strengths

1. Significant Theoretical Contribution: First theoretically analyzes diversity performance of OTFS in single-selective channels, filling an important theoretical gap
2. Strong Method Innovation: Precoding design based on algebraic number theory demonstrates mathematical rigor and practicality
3. Comprehensive Experimental Verification: Covers diverse channel conditions, system dimensions, and detectors
4. High Practical Value: CSI-free requirement and rate-loss-free characteristics offer promising practical prospects

Weaknesses

1. Idealized Channel Assumptions: Perfect CSI assumption is difficult to satisfy in practice
2. Missing Complexity Analysis: Computational complexity of precoding and detection lacks detailed analysis
3. Limited Applicability: Primarily addresses single-selective channels; performance in doubly selective channels requires further investigation

Impact

This work provides important contributions to theoretical analysis and practical implementation of OTFS systems, expected to advance OTFS technology applications in 5G/6G communication systems.

Applicable Scenarios

1. High-speed mobile communication scenarios (e.g., high-speed rail, vehicular networks)
2. Multipath-rich indoor/urban environments
3. Communication systems with high spectral efficiency requirements

References

The paper cites 13 important references covering classical works in wireless communication theory, OFDM precoding, OTFS modulation, and related fields, providing a solid theoretical foundation for this research.