Abstract
We analyze the asymptotic performance of a generalized multiuser diversity scheme for an interference-limited secondary multiuser network of underlay cognitive radio systems. Assuming a large number of secondary users and that the noise at each secondary user's receiver is negligible compared to the interference from the primary transmitter, the secondary transmitter transmits information to the $k$-th best secondary user, namely, the one with the $k$-th highest signal-to-interference ratio (SIR). We use extreme value theory to show that the $k$-th highest SIR converges uniformly in distribution to an inverse gamma random variable for a fixed $k$ and large number of secondary users. We use this result to derive asymptotic expressions for the average throughput, effective throughput, average bit error rate and outage probability of the $k$-th best secondary user under continuous power adaptation at the secondary transmitter, which ensures satisfaction of the instantaneous interference constraint at the primary receiver caused by the secondary transmitter. Numerical simulations show that our derived asymptotic expressions are accurate for different values of system parameters.
Introduction: The Challenge of Multiuser Diversity in CR Networks
In underlay Cognitive Radio (CR) systems, a secondary network shares the radio spectrum with a primary network, provided the secondary transmitter (ST) strictly limits the interference it causes to the primary receiver (PR). To maximize performance, the ST typically employs "multiuser diversity" by transmitting to the secondary user (SU) with the best channel conditions.
However, in practical scenarios, the absolute best user is not always available due to scheduling delays, handoffs, or other service requirements. Systems often need to select the $k$-th best user. Analyzing the exact performance of this selection scheme for an arbitrary number of users ($N$) is mathematically complex, especially when considering the continuous power adaptation required to protect the primary network. This paper addresses this gap by deriving tractable, closed-form asymptotic expressions for large networks.
System Model and Methodology
The study considers an interference-limited regime where thermal noise at the SU receiver is negligible compared to the interference from the primary transmitter (PT).
Instead of traditional exact probability distributions, the authors apply Extreme Value Theory (EVT) to find the limiting distribution of the $k$-th highest Signal-to-Interference Ratio (SIR) as $N \to \infty$. They prove that in an interference-limited CR network, the $k$-th highest SIR converges to an inverse gamma distribution, a significant departure from the log-gamma distribution seen in noise-limited networks.
Using this limiting distribution, the authors derived closed-form asymptotic expressions for key performance metrics, including average throughput, effective throughput (under delay constraints), average Bit Error Rate (BER), and outage probability.
Key Insights: From Theory to Practice
A major finding is that "asymptotic" does not imply "unrealistic." While EVT relies on the limit as $N \to \infty$, the derived expressions are remarkably accurate for practical network sizes. The approximations tightly match exact simulations with as few as 30 to 40 users.
Furthermore, the research highlights an interference bottleneck. When the ST has a high maximum transmit power limit, the effective throughput eventually hits a ceiling dictated by the maximum tolerable interference level at the primary receiver, rather than scaling indefinitely with power.
Performance Validation
Monte Carlo simulations confirmed the accuracy of the EVT approach. As shown in the results, the theoretical asymptotic curves closely follow the simulation data points across various parameters.
Key results include:
- Closed-Form Success: Expressions for performance metrics were successfully derived using standard special functions like the Meijer G-function, making them easily computable for system design.
- Power vs. Interference Limits: For a limited ST power (e.g., $P_S = -20$ dB), throughput saturates once the primary receiver's interference tolerance threshold ($T$) is sufficiently high, as the ST reaches its physical power limit.
- Impact of Imperfect CSI: The framework remains robust even with outdated Channel State Information (CSI), requiring only minor adjustments to fading parameters to maintain predictive accuracy.
Limitations and Future Work
The accuracy of these asymptotic expressions degrades if the chosen rank $k$ is too close to the total number of users $N$. The EVT approximations require $N$ to be relatively large compared to $k$. Additionally, the model assumes an interference-limited environment; if the primary transmitter is silent, the system reverts to a noise-limited regime. Future work could extend this framework to multi-antenna (MIMO) secondary networks or scenarios with multiple primary transmitters.
Summary updated