[Paper Review] Hardware and Interference Limited Cooperative CR-NOMA Networks under Imperfect SIC and CSI

Sultangali Arzykulov, Galymzhan Nauryzbayev+2 more|arXiv (Cornell University)|Apr 22, 2020

Advanced Wireless Communication TechnologiesEngineering47 references29 citations

TL;DR

This paper analyzes the outage performance of a hardware and interference-limited underlay cognitive radio non-orthogonal multiple access (CR-NOMA) network with imperfect successive interference cancellation (SIC) and channel state information (CSI). It derives closed-form expressions for end-to-end outage probability (OP) under practical impairments, showing that CR-NOMA outperforms orthogonal multiple access (OMA) but performance degrades significantly under high hardware impairments, SIC errors, and CSI inaccuracy.

ABSTRACT

The conflation of cognitive radio (CR) and nonorthogonal multiple access (NOMA) concepts is a promising approach to fulfil the massive connectivity goals of future networks given the spectrum scarcity. Accordingly, this letter investigates the outage performance of imperfect cooperative CR-NOMA networks under hardware impairments and interference. Our analysis is involved with the derivation of the end-to-end outage probability (OP) for secondary NOMA users by accounting for imperfect channel state information (CSI), as well as the residual interference caused by successive interference cancellation (SIC) errors and coexisting primary/secondary users. The numerical results validated by Monte Carlo simulations show that CR-NOMA network provides a superior outage performance over orthogonal multiple access. As imperfections become more significant, CR-NOMA is observed to deliver relatively poor outage performance.

Motivation & Objective

• To analyze the end-to-end outage performance of a cooperative CR-NOMA network under realistic hardware and channel impairments.
• To model and quantify the impact of imperfect successive interference cancellation (SIC), channel state information (CSI) errors, and hardware impairments on secondary user (SU) reliability.
• To derive closed-form expressions for the outage probability (OP) of NOMA users under interference-limited conditions and interference temperature constraints (ITC).
• To compare the performance of CR-NOMA against orthogonal multiple access (OMA) under various practical impairments.
• To validate analytical results through Monte Carlo simulations under varying levels of hardware distortion, SIC error, and CSI inaccuracy.

Proposed method

• Proposes a two-timeslot half-duplex decode-and-forward (DF) relaying model for a downlink underlay CR-NOMA network with one source (S), one relay (R), and B secondary users (U1 to UB).
• Models channel coefficients using minimum mean square error (MMSE) estimation, with estimation error variance ζi ∝ ρ−κ, where ρ is the transmit SNR.
• Incorporates hardware impairments via aggregate distortion noise φ(·) = √(φ²t + φ²r), modeling transmitter and receiver impairments.
• Derives instantaneous signal-to-interference-distortion-noise ratio (SIDNR) expressions for both relay decoding (γR,j) and user detection (γb,j) stages, accounting for residual interference from SIC errors.
• Uses the CDF of the minimum of two SIDNRs (γR,j and γb,j) to compute the outage probability (OP) via FγR,j(ψj) + Fγj(ψj) − FγR,j(ψj)Fγj(ψj), with ψj = 2²ᴿᵗʰʲ − 1.
• Derives closed-form CDF expressions for γR,j and γb,j using exponential integral functions (Ei[·]) and Laplace transforms, validated via Monte Carlo simulations.

Experimental results

Research questions

• RQ1 How does imperfect SIC affect the outage performance of secondary NOMA users in a CR-NOMA network?
• RQ2 What is the impact of hardware impairments (HIs) on the end-to-end outage probability of NOMA users under interference constraints?
• RQ3 How does CSI inaccuracy (modeled via ζi ∝ ρ−κ) influence the system's reliability and saturation behavior at high SNR?
• RQ4 How does the presence of coexisting primary users and interference temperature constraints (ITC) affect NOMA user outage?
• RQ5 Does CR-NOMA still outperform OMA under severe hardware and SIC impairments?

Key findings

• At 20 dB transmit SNR, U1’s outage probability increases from 0.09 (no primary interference) to 0.12 (PT = 10 dB) and 0.62 (PT = 25 dB), showing significant performance loss under primary interference.
• With ϵ = 0.03 (high SIC imperfection), U2’s outage probability reaches 0.45 at 30 dB SNR, compared to 0.09 under perfect SIC, indicating severe performance degradation.
• At φ = 0.15 (high hardware impairment), U1’s OP degrades by 0.025 and U2’s by 0.3 at 30 dB SNR, showing U2 is more sensitive to distortion.
• When θ = 0.1, both NOMA and OMA users experience full outage at all SNR levels, indicating intolerable CSI uncertainty.
• For κ = 1.5 and θ = 10, the system approaches perfect CSI performance at high SNR, showing that SNR-dependent CSI error improves outage performance.
• Even under high hardware impairments (φ = 0.15), NOMA still outperforms OMA, though the OMA user suffers full outage, highlighting NOMA’s robustness under practical conditions.