Source-linked AI summary
Wideband Spectrum Sensing in Cognitive Radio Networks
Zhi Quan, Shuguang Cui, Ali H. Sayed, H. Vincent Poor
TL;DR
Cognitive radios need wideband sensing that finds spectrum holes while avoiding harmful interference, but prior work did not jointly decide occupancy across multiple bands. The paper introduces multiband joint detection, optimizes multiple narrowband detectors through hidden-convexity reformulations, and numerically shows higher opportunistic rates and lower interference than uniform-threshold sensing.
Problem
Wideband sensing methods had not addressed joint occupancy decisions across multiple frequency bands, despite the need to improve spectrum use while protecting primary users.
Method
The paper jointly optimizes a bank of narrowband detectors by formulating sensing as throughput-maximization problems and exploiting hidden convexity to obtain convex programs under practical conditions.
Results
Numerical evaluations show that optimized multiband thresholds achieve higher opportunistic rates and lower induced interference than uniform thresholds.
Takeaways & Limitations
Multiband joint detection improves spectrum efficiency by balancing spectrum utilization against interference to primary users.
Abstract
from arXiv · showhide
Spectrum sensing is an essential enabling functionality for cognitive radio networks to detect spectrum holes and opportunistically use the under-utilized frequency bands without causing harmful interference to legacy networks. This paper introduces a novel wideband spectrum sensing technique, called multiband joint detection, which jointly detects the signal energy levels over multiple frequency bands rather than consider one band at a time. The proposed strategy is efficient in improving the dynamic spectrum utilization and reducing interference to the primary users. The spectrum sensing problem is formulated as a class of optimization problems in interference limited cognitive radio networks. By exploiting the hidden convexity in the seemingly non-convex problem formulations, optimal solutions for multiband joint detection are obtained under practical conditions. Simulation results show that the proposed spectrum sensing schemes can considerably improve the system performance. This paper establishes important principles for the design of wideband spectrum sensing algorithms in cognitive radio networks.
I. INTRODUCTION
Wideband spectrum sensing has been studied mainly for single-band detection, leaving joint decisions across multiple bands unaddressed. The paper introduces multiband joint detection, formulating threshold selection as optimization problems that improve opportunistic throughput while limiting primary-user interference.
- Motivation: Spectrum sensing enables cognitive radios to detect weak, possibly unknown primary signals and identify opportunities for spectrum access.The paper adopts non-coherent energy detection as a simple building block for wideband sensing.
- Research gap: Prior cooperative spectrum-sensing studies mainly detect signals on a single frequency band, while wideband methods estimate PSD without making joint multiband decisions.The paper identifies joint decisions across multiple frequency bands as an unresolved requirement for wideband sensing.
- Proposed approach: The proposed framework jointly optimizes multiple narrowband detectors to improve opportunistic throughput and reduce interference to primary communication systems.It applies to individual cognitive radios and jointly considers detection across a bank of narrowband channels.
- Proposed approach: The sensing problem is formulated as optimization, then reformulated as convex programs by exploiting hidden convexity under practical conditions.This yields optimal solutions for the multiband joint-detection problems described in the paper.
- Evaluation: The paper evaluates the proposed algorithms numerically and concludes that multiband joint detection improves use of unused frequency bands while limiting interference.The paper is organized around system modeling, algorithm development, numerical evaluation, and conclusions.
A. Wideband Spectrum Sensing
The wideband channel is divided into K non-overlapping narrowband subchannels, some of which may be unused by primary users. The paper jointly detects occupancy across these bands to identify spectral holes for opportunistic access.
- Channel model: A wideband primary communication channel is modeled as K non-overlapping narrowband subchannels, with some potentially unused in a given region and time.Multiuser OFDM is identified as suitable because it enables flexible subband manipulation.
- Occupancy model: Each subchannel k is represented by binary hypotheses H0,k for primary-signal absence and H1,k for primary-signal presence.The resulting hypothesis vector indicates which subchannels are allowed or prohibited for opportunistic access.
- Detection objective: The sensing task is to identify spectral holes across K subchannels while assuming cognitive radios remain silent during the detection interval.Under this assumption, received spectral power comes from primary users and background noise; the proposed detector jointly considers multiple bands.
B. Received Signal
The received signal model captures multipath fading, additive Gaussian noise, and frequency-selective channel responses across the wideband subchannels. Frequency-domain observations are obtained using the DFT while assuming slowly varying channel responses and independent signal, channel, and noise terms.
- Time-domain model: In a multipath fading environment, h(l) denotes the discrete-time channel impulse response and L is the number of resolvable paths.The received baseband signal is modeled as a convolution of the transmitted signal with the channel response plus additive noise.
- Time-domain model: The additive noise v(n) is complex white Gaussian noise with zero mean and variance σ2.This noise is added to the primary transmitted signal after channel convolution.
- Frequency-domain model: Multipath fading produces frequency-selective channel behavior, represented through the discrete frequency response Hk over the K subchannels.The channel is assumed slowly varying, so its frequency responses remain constant during a detection interval.
- Frequency-domain model: The receiver estimates each subchannel's frequency-domain signal by computing a DFT of the received samples.The resulting representation is used to obtain frequency-domain observations for subsequent detection.
- Frequency-domain model: The frequency-domain received signal combines the transmitted subchannel signal, channel gain, and noise as HkSk + Vk.Sk denotes the primary transmitted signal at subchannel k, while Vk is the frequency-domain noise term.
- Assumptions: The transmitted signal Sk, channel gain Hk, and additive noise Vk are assumed mutually independent.The model also uses L ≤ N and a channel response that remains constant during each detection interval.
C. Signal Detection in Individual Bands
Individual-band sensing forms the building block for multiband detection by testing signal absence or presence from received energy. Its threshold controls a throughput–interference tradeoff through false alarms and misses.
- Binary detection: Single-band detection tests H0,k, representing primary-signal absence, against H1,k, representing primary-signal presence.This binary test is the per-subchannel building block for multiband joint detection.
- Energy detection: The detector summarizes received energy over M samples and compares the resulting statistic with a subchannel-specific threshold γk.The threshold determines the occupancy decision for subchannel k.
- Statistic model: The model assumes unit transmitted signal power and uses an asymptotic normal distribution for the energy statistic under the central limit theorem.The statistic is represented as Yk ∼ N(E(Yk), Var(Yk)).
- Threshold tradeoff: A higher threshold lowers the false-alarm probability but increases the miss probability, while a lower threshold has the opposite effect.This threshold tradeoff directly affects spectrum access and primary-user protection.
- Network implications: Low false-alarm probability preserves access to unused spectrum, whereas low miss probability limits interference from cognitive radios to primary users.False alarms reduce opportunistic throughput, while misses measure interference risk to primary users.
III. MULTIBAND JOINT DETECTION
Multiband joint detection jointly selects detection thresholds across subchannels to balance opportunistic throughput against interference constraints. The resulting optimization can be transformed into a tractable convex program under practically relevant low-error conditions.
- Framework: The framework optimizes a threshold vector across multiple narrowband detectors to use unoccupied spectral segments while limiting primary-user interference.The threshold vector is γ = [γ0, γ1, . . . , γK−1]T.
- Framework: Aggregate opportunistic throughput weights each subchannel’s achievable rate by its spectrum utilization, which depends on the threshold vector.Higher false-alarm probabilities reduce opportunistic utilization, while higher miss probabilities increase interference risk.
- Optimization formulations: The optimization maximizes aggregate opportunistic throughput subject to aggregate interference constraints for primary users and individual subband constraints.A second formulation instead minimizes interference subject to a required minimum aggregate throughput.
- Threshold allocation: Threshold allocation reflects subchannel priorities: higher opportunistic rates favor higher thresholds, whereas higher-priority primary users favor lower thresholds.These choices trade off lower false alarms against lower miss probabilities across subchannels.
- Convex reformulation: Although the original objective and constraints are generally nonconvex, hidden convexity yields convex formulations in the low-probability false-alarm and miss regime.Under these conditions, efficient methods such as interior-point algorithms can obtain globally optimal solutions.
IV. SIMULATION RESULTS
Simulations evaluate multiband joint detection in an eight-subchannel single-user OFDM system against optimized uniform-threshold sensing. The proposed method achieves higher throughput under interference constraints and lower interference at fixed throughput, improving spectrum efficiency.
- Simulation setup: The simulation uses a wideband channel divided into 8 subchannels and compares optimized multiband thresholds with a uniform threshold selected for maximum aggregate throughput.The study randomly generates channel conditions, subchannel throughput rates, and interference cost coefficients.
- Throughput-constrained sensing: Multiband joint detection achieves a much higher opportunistic rate than uniform-threshold sensing as the aggregate interference constraint varies.Aggregate opportunistic rate increases when the interference constraint ε is relaxed.
- Interference-constrained sensing: For any given opportunistic throughput, multiband joint detection induces less interference to primary users than uniform thresholds.The comparison corresponds to the alternative formulation that minimizes aggregate interference subject to throughput constraints.
- Overall outcome: The numerical results show that multiband joint detection considerably improves spectrum efficiency by making more efficient use of spectral diversity.Optimized thresholds and associated miss and false-alarm probabilities are illustrated for both optimization formulations.
V. CONCLUSION
The paper proposes multiband joint detection for wideband spectrum sensing, jointly considering primary users across narrowband subchannels. It formulates this as optimization problems and obtains optimal solutions under practical conditions.
- Multiband joint detection considers primary-user detection across a bank of narrowband subchannels jointly rather than one band at a time.
- The joint detection problem is formulated as a class of optimization problems targeting improved spectral efficiency and reduced interference.
- Hidden convexity in the seemingly nonconvex formulations enables optimal solutions under practical conditions.
- Numerical examination shows that the proposed spectrum sensing algorithms perform well.