[Paper Review] OTFS -- Predictability in the Delay-Doppler Domain and its Value to Communication and Radar Sensing

Saif Khan Mohammed, Ronny Hadani+2 more|arXiv (Cornell University)|Feb 17, 2023

PAPR reduction in OFDM19 references4 citations

TL;DR

This paper establishes that OTFS modulation in the delay-Doppler (DD) domain achieves predictable, non-fading communication and radar sensing when the crystallization condition—where pulsone periods exceed channel spreads—is satisfied. It demonstrates that Zak-OTFS enables model-free operation by directly learning the input-output relation from pilot symbols, outperforming conventional MC-OTFS, especially under high Doppler spread.

ABSTRACT

In our first paper [2] we explained why the Zak-OTFS input-output (I/O) relation is predictable and non-fading when the delay and Doppler periods are greater than the effective channel delay and Doppler spreads, a condition which we refer to as the crystallization condition. We argued that a communication system should operate within the crystalline regime. It is well known that it is possible to identify a linear time varying (LTV) channel if and only if it is under-spread. The crystallization condition is more restrictive than the under-spread condition, so identification is always possible. In the crystalline regime, we show that Zak-OTFS pilot sequences minimize the complexity of identifying the effective DD domain channel filter. We demonstrate that the filter taps can simply be read off from the response to a single Zak-OTFS pilot. In general, we provide an explicit formula for reconstructing the Zak-OTFS I/O relation from a finite number of received pilot symbols in the delay-Doppler (DD) domain. This reconstruction formula makes it possible to study predictability of the Zak-OTFS I/O relation for a sampled system that operates under finite duration and bandwidth constraints. We analyze reconstruction accuracy for different choices of the delay and Doppler periods, and of the pulse shaping filter. Reconstruction accuracy is high when the crystallization condition is satisfied, implying that it is possible to learn directly the I/O relation without needing to estimate the underlying channel. This opens up the possibility of a model-free mode of operation, which is especially useful when a traditional model-dependent mode of operation (reliant on estimation of the underlying physical channel) is out of reach (for example, when the channel comprises of unresolvable paths, or exhibits a continuous delay-Doppler profile such as in presence of acceleration). Our study clarifies the

Motivation & Objective

• To establish the theoretical foundation for predictability in OTFS systems operating in the delay-Doppler domain.
• To identify the crystallization condition as the key enabler for non-fading, predictable system behavior in doubly selective channels.
• To demonstrate that the input-output relation of Zak-OTFS can be directly learned from pilot signals without estimating the underlying physical channel.
• To compare Zak-OTFS with its multicarrier approximation (MC-OTFS) and show superior predictability and performance under high Doppler spread.
• To clarify the role of aliasing in the DD domain as the root cause of non-predictability and to introduce the crystalline decomposition of the channel.

Proposed method

• Derives an explicit reconstruction formula for the Zak-OTFS input-output relation using a finite number of pilot symbols in the DD domain.
• Introduces the concept of the crystalline decomposition, separating the effective DD channel into predictable (crystalline) and non-predictable components.
• Uses the Zak transform to map time-domain signals to quasi-periodic representations in the DD domain, enabling analysis of pulse train waveforms (pulsones).
• Analyzes reconstruction accuracy as a function of delay and Doppler periods and pulse shaping filter design, showing high accuracy under the crystallization condition.
• Compares Zak-OTFS and MC-OTFS by analyzing their respective I/O relation predictability and BER performance under imperfect channel knowledge.
• Employs a canonical decomposition of the channel to isolate the impact of aliasing due to mismatched pulsone and channel spreads.

Experimental results

Research questions

• RQ1 Under what conditions is the input-output relation of a sampled OTFS system predictable in the delay-Doppler domain?
• RQ2 How does aliasing in the DD domain lead to non-predictability, and what is its fundamental origin?
• RQ3 To what extent can the input-output relation be learned directly from pilot symbols without estimating the underlying physical channel?
• RQ4 How does the predictability of Zak-OTFS compare to that of MC-OTFS, especially under high Doppler spread?
• RQ5 What is the impact of the crystallization condition on bit error rate (BER) performance in both perfect and imperfect knowledge scenarios?

Key findings

• The crystallization condition—where the delay and Doppler periods exceed the effective channel delay and Doppler spreads—ensures that the Zak-OTFS input-output relation is predictable and non-fading.
• When the crystallization condition holds, the effective DD channel filter taps can be directly read from the response to a single Zak-OTFS pilot, enabling direct learning of the I/O relation.
• Reconstruction of the I/O relation from pilot symbols achieves high accuracy when the crystallization condition is satisfied, enabling model-free operation.
• Performance under model-free operation is only slightly worse than with perfect I/O knowledge when the crystallization condition holds, demonstrating robustness.
• Zak-OTFS exhibits superior predictability and BER performance compared to MC-OTFS as Doppler spread increases, due to reduced aliasing and better channel resolution.
• The non-predictable component of the channel vanishes exactly when the crystallization condition is met, confirming that aliasing is the root cause of non-predictability.