[Paper Review] Noncoherent Short-Packet Communication via Modulation on Conjugated Zeros

Philipp Walk, Peter Jung+1 more|arXiv (Cornell University)|May 21, 2018

Blind Source Separation Techniques31 references3 citations

TL;DR

This paper proposes Modulation on Conjugate-Reciprocal Zeros (MOCZ), a noncoherent short-packet communication scheme that encodes data in the conjugate-reciprocal zero pairs of a signal's z-transform. The method achieves robustness against unknown finite impulse response (FIR) channels and additive noise by exploiting zero structure invariance, with a low-complexity decoder enabling linear scaling in bit length and superior bit-error rate performance over noncoherent OFDM-IM and pilot-based M-QAM schemes.

ABSTRACT

We introduce a novel blind (noncoherent) communication scheme, called modulation on conjugate-reciprocal zeros (MOCZ), to reliably transmit short binary packets over unknown finite impulse response systems as used, for example, to model underspread wireless multipath channels. In MOCZ, the information is modulated onto the zeros of the transmitted signals $z-$transform. In the absence of additive noise, the zero structure of the signal is perfectly preserved at the receiver, no matter what the channel impulse response (CIR) is. Furthermore, by a proper selection of the zeros, we show that MOCZ is not only invariant to the CIR, but also robust against additive noise. Starting with the maximum-likelihood estimator, we define a low complexity and reliable decoder and compare it to various state-of-the art noncoherent schemes.

Motivation & Objective

• Address the challenge of reliable short-packet communication in massive IoT and tactile internet scenarios where channel estimation overhead is prohibitive.
• Overcome the limitations of conventional coherent schemes that require pilot overhead and are sensitive to time-varying channels.
• Develop a noncoherent modulation scheme invariant to unknown FIR channel impulse responses (CIRs), enabling blind transmission without channel state information.
• Design a low-complexity decoder for practical implementation while maintaining high reliability, especially for short-length data.
• Demonstrate performance gains over existing noncoherent schemes like noncoherent OFDM-IM and pilot-based M-QAM in terms of bit-error rate (BER).

Proposed method

• Encode binary data by placing conjugate-reciprocal zero pairs in the z-domain, forming a polynomial whose coefficients represent the baseband signal.
• Leverage the fact that zero structure is preserved at the receiver regardless of the CIR, enabling CIR-invariant transmission.
• Design a maximum-likelihood (ML) decoder based on the channel's power delay profile and noise variance, achieving optimal performance.
• Propose a low-complexity decoder that decodes each zero independently by evaluating the received signal against a codebook of zero patterns, reducing complexity from exponential to linear in the number of bits.
• Use root neighborhood analysis and stability bounds to derive noise robustness conditions, ensuring reliable detection under additive noise.
• Apply Huffman sequence-based codebooks to optimize zero spacing and improve minimum distance, enhancing noise resilience.

Experimental results

Research questions

• RQ1 Can a noncoherent communication scheme be designed that is invariant to unknown FIR channel impulse responses and robust to additive noise?
• RQ2 How can the zero structure of a signal be exploited to enable blind transmission without channel estimation?
• RQ3 What is the achievable bit-error rate performance of such a scheme compared to state-of-the-art noncoherent methods like OFDM-IM and pilot-based M-QAM?
• RQ4 Can a low-complexity decoder be constructed that scales linearly with the number of bits while preserving high reliability?
• RQ5 What are the theoretical noise robustness bounds for the proposed scheme, and how do they depend on zero spacing and signal radius?

Key findings

• The MOCZ scheme achieves CIR-invariant transmission because the zero structure of the signal is preserved at the receiver regardless of the channel impulse response.
• The proposed low-complexity decoder achieves linear complexity in the number of bits, enabling practical implementation for short-packet communication.
• Numerical results show that MOCZ outperforms noncoherent OFDM-IM and pilot-based M-QAM in terms of bit-error rate across various signal-to-noise ratios.
• Noise power bounds derived via root neighborhood analysis indicate that robustness decreases with increasing signal length N, especially when zeros are closely spaced.
• For the worst-case Huffman sequence (all zeros inside the unit circle except one), the noise power bound scales with R, N, and the minimum distance between zero neighborhoods.
• Analytical bounds suggest that a minimum separation of δmax ≈ 0.455 (for N=4), 0.288 (for N=8), and 0.165 (for N=16) is required to maintain disjoint zero neighborhoods and ensure stability under noise.