Spectrum Spreading Gain in Multiuser Systems

Updated 22 December 2025

Spectrum spreading gain is defined as the ratio of spread to data bandwidth, enhancing interference suppression and robustness.
It enables improved SNR and multiuser orthogonality through techniques such as DS-SS, CDMA, OFDM, and Galois-domain methods.
Increasing the gain offers enhanced resilience and diversity, but introduces trade-offs with reduced spectral efficiency and data throughput.

Spectrum spreading gain, often termed "processing gain," is a fundamental metric in spread-spectrum and multiuser communication systems. It quantifies the improvement in interference rejection, noise averaging, and multi-path resilience that results when a narrowband signal is intentionally dispersed over a much wider spectral or time-chip space using spreading codes, signatures, or time modulation. The gain is commonly defined as the ratio of spread bandwidth to the data bandwidth, and is closely tied to orthogonality, redundancy, and coding structure. Spectrum spreading gain impacts not only error-rate performance, but also system capacity, stealth, and signal robustness across diverse domains such as OFDM, CDMA, random access, LDPC-coded systems, and even radar/microwave camouflaging.

1. Mathematical Definition and Quantification

Spectrum spreading gain $G$ is rigorously defined as the ratio of the total spread-signal bandwidth ( $B_\text{spread}$ ) or spreading sequence length ( $N$ ) to the original data bandwidth ( $B_\text{data}$ ) or symbol rate. In standard direct-sequence spread-spectrum (DS-SS) systems, if each data symbol is mapped to an $N$ -chip pseudorandom or orthogonal spreading code, then

$G = \frac{B_\text{spread}}{B_\text{data}} = N$

This definition extends to OFDM and multilevel Galois-field systems, where $G$ may be modified by spectral compression or orthogonalization, e.g.,

$G_\mathrm{proc} = \frac{N}{v}$

for Galois-field Division Multiplex (GDM), where only $v$ cyclotomic representatives are transmitted out of $N$ (Oliveira et al., 2015). In unsourced random access (URA) using random spreading signatures of length $n_p$ , the processing gain is simply $G_p = n_p$ (Hu et al., 19 Dec 2025). In LDPC-coded DSSS systems, the linear processing gain is $N_s = R_c / R_d$ , and in dB notation it becomes $G_s(\text{dB}) = 10\log_{10} N_s$ (Jooshin et al., 2021). For metasurface-based radar camouflaging, $G_s = B_\text{spread} / B_\text{signal} \approx N$ (Wang et al., 2019).

2. Interference Suppression and SNR Improvement

Spectrum spreading gain underpins robust rejection of narrowband interference and noise. In DS-SS overlay scenarios, for an interferer with original power $P_i$ , the output post-despreading variance is reduced to $P_i / G$ , i.e., the interference is “thinned out” by a factor $G = W_s / R_b$ (Vouyioukas, 2012). The same effect arises with multi-user interference (MAI): mean and variance of the MAI in shared systems are scaled inversely with $G$ , directly impacting bit-error-rate (BER). For AWGN, despreading reduces the noise variance per bit to $N_0/(2N)$ (Jooshin et al., 2021). In Galois-domain CDMA, output symbol SNR is multiplied by $G_\mathrm{proc}$ over input SNR due to averaging over $N$ orthogonal finite-field carriers (Oliveira et al., 2015). For radar and camouflaging, the scattered power spectral density is suppressed by $1/N$, guaranteeing unobservability to adversaries and SNR gain of $10\log_{10}N$ dB for keyed receivers (Wang et al., 2019). In multiuser random access with orthogonal or nearly orthogonal codes, the cross-correlation power scales as $1/n_p$ , yielding large anti-interference capabilities (Hu et al., 19 Dec 2025).

3. Impact on Diversity, Multiuser Orthogonality, and Spectral Efficiency

Spreading codes allow exploitation of frequency, time, and index diversity. In SS-SIM-OFDM systems, spreading over $K$ active subcarriers leads to a diversity order $G_d = K$ , enabled by MRC-like subcarrier combining. The diversity order is determined by the minimum Hamming distance between spreading vectors, and careful subcarrier index mapping (OSI method) can maintain $G_d = K$ universally (2207.14454). Orthogonality in multilevel spreading designs eliminates multiuser interference, as the Welch/Massey–Mittelholzer bound guarantees exactly zero cross-correlation between users’ carriers (Oliveira et al., 2015), which can be exploited for massive CDMA and Galois-based multiplexing. Spectral efficiency is critically modulated by spreading gain: increasing $G$ reduces peak data rate for fixed bandwidth, but multilevel compression (e.g., cyclotomic cosets in GDM) allows spectral efficiency $\eta_\text{GDM} = \frac{N}{v}\log_2 p$ that can exceed classical binary CDMA or TDMA by several times (Oliveira et al., 2015).

4. Trade-Offs in Coding, Spreading, and Spectral Utilization

Spreading gain introduces a fundamental trade-off between interference resilience and spectral efficiency. Increasing $G$ improves interference rejection but reduces effective data rate and increases occupied bandwidth. In coding-spreading systems, the optimal suppression of narrow-band interference is achieved not by spreading per se but by minimizing code rate $R_c$ via powerful low-rate coding (e.g., turbo-Hadamard codes), as shown theoretically and in simulation by Wu & Yang (Wu et al., 2015). Spreading only provides linear improvement, whereas coding leverages large minimum distance. In OFDM-based ISAC ranging, adding a P-DPSS orthogonal spreading layer on top of OFDM can reduce inter-band leakage energy (and integrated sidelobe levels) by up to $10$ dB, but the spectral utilization drops by $\eta$ (e.g., $\eta = 0.9$ ) (Said et al., 4 May 2025). Thus, system design must balance spectral compactness, interference suppression, diversity, and coding gain.

5. Algorithmic Realizations and Orthogonal Spreading Architectures

Practical spreading gain depends on the construction, orthogonality, and redundancy of spreading sequences:

In SS-SIM-OFDM, the precoding vector $p$ is constructed by zeroing all inactive subcarriers and assigning orthogonal chips to active ones. The transmit block is $x = p \cdot s$ , ensuring orthogonality and diversity gain (2207.14454).
DS-SS overlays utilize long maximal-length m-sequences, with the correlator integrating against the spreading code to average out interference (Vouyioukas, 2012).
LDPC-DSSS systems leverage parity-check matrix columns for chip sequences, requiring only lightweight multiply-accumulate operations at the receiver for despreading, followed by standard SPA decoding (Jooshin et al., 2021).
Random spreading in URA builds a codebook of Gaussian unit-norm signatures; iterative Gaussian-approximation decoders perform MMSE filtering and LLR exchange, with the cross-correlation matrix yielding fast convergence (<5 iterations) to near-Gaussian MAC optimality (Hu et al., 19 Dec 2025).
In Galois-domain CDMA, spreading is performed via finite-field transforms (GFT/FFHT), with only $v$ cyclotomic coset leaders transmitted, offering redundancy reduction, and O( $N\log N$ ) implementation (Oliveira et al., 2015).
In metasurface camouflaging, chip-state switching according to a periodic PRN sequence spreads radar echoes, suppressing PSD and enabling robust detection by matched-filter correlation (Wang et al., 2019).

6. Performance Metrics and Empirical Results

Spectrum spreading gain manifests empirically as SNR improvement, BER reduction, and multiuser robustness. In SS-SIM-OFDM (with OSI), processing gain yields $3-15$ dB improvement at BER $\sim 10^{-4}$ relative to conventional OFDM-IM and related benchmarks (2207.14454). LDPC-DSSS simulations show $2.3$ dB Eb/N0 savings and up to $36$-dB BER improvements for moderate spreading factors ( $N_s=4$ –$10$) (Jooshin et al., 2021). In random spreading URA, theoretical and simulation curves confirm that with $n_p=114$ , the iterative decoder achieves near-Gaussian MAC performance (within $0.1$–$0.5$ dB of the bound) for $25$–$175$ active users (Hu et al., 19 Dec 2025). For Galois-domain CDMA, spectral efficiency may rise by $2.6\times$ to $4.3\times$ compared to Walsh or TDMA for suitably chosen parameters (Oliveira et al., 2015). In camouflaged radar, $N=127$ chips achieve SNR boosts of $18$–$27$ dB (simulation and experimental), and adversary detection is reduced below the noise floor (Wang et al., 2019).

7. Limitations, System-Level Design, and Outlook

While increasing spectrum spreading gain improves robustness and SNR, it is constrained by bandwidth availability, synchronism requirements, and diminishing returns in specific coding regimes:

Excessive spreading with insufficient coding offers no asymptotic advantage for narrow-band interference suppression when powerful low-rate codes are feasible (Wu et al., 2015).
Larger $G$ decreases data throughput for a fixed spectral band, forcing trade-offs in overlay and multi-service settings (Vouyioukas, 2012, Said et al., 4 May 2025).
Orthogonality is critical; imperfect symbol or chip synchrony can impair interference rejection in Galois or binary systems.
In joint sensing-communication frameworks (OFDMA-ISAC, camouflaging), spreading gain must be balanced against spectral utilization and application-specific constraints.

Current trends emphasize hybrid architectures combining optimized spreading (e.g., orthogonal DPSS, random Gaussian, Galois transform-based) with high-minimum-distance coding, dynamic access, and multi-domain diversity exploitation. The parameter $G$ remains central for quantifying gains and guiding system design across radio, optical, and non-traditional electromagnetic contexts.