Throughput and Delay Scaling in Supportive Two-Tier Networks

Long Gao, Rui Zhang, Changchuan Yin, Shuguang Cui

Introduction

In traditional cognitive radio networks, secondary (unlicensed) users are typically viewed as opportunistic guests that must avoid interfering with primary (licensed) users. This "non-supportive" paradigm ensures the primary network's performance is not degraded, but it offers no benefit to the legacy infrastructure. This paper explores a paradigm shift: supportive two-tier networks.

In this model, a dense secondary network actively assists the primary network by relaying its traffic. The authors investigate whether this cooperation can break the fundamental throughput and delay scaling limits of the primary network without compromising the secondary network's own performance.

The Supportive Protocol and Interference Management

The researchers model a network where primary nodes have density $n$ and secondary nodes have a significantly higher density $m = n^\beta$ (with $\beta \geq 2$). To facilitate cooperation, they propose a time-slotted multi-hop protocol where secondary nodes act as relays for primary packets.

When a primary source transmits, the packet is divided into segments and routed in parallel through the dense secondary mesh. To ensure this doesn't lead to catastrophic interference, the protocol employs "preservation regions." These are specific spatial areas around active primary transmitters where secondary nodes are required to remain silent, effectively creating a protective buffer for primary communications.

Spatial layout of preservation and collection regions used to manage interference between the two tiers.

Scaling Analysis and Mobility

The study employs mathematical scaling analysis to determine how throughput and delay change as the number of nodes $n$ grows toward infinity. They consider three main scenarios for node movement:

  1. Static Case: All nodes are fixed in position.
  2. i.i.d. Mobility: Secondary nodes are reshuffled to new random positions in every time slot.
  3. Random Walk (RW): Secondary nodes move according to a standard random walk model.

The core insight is that the extreme density of the secondary tier allows for massive parallelism in relaying, which fundamentally alters the scaling laws of the primary tier.

Key Results: Breaking the Scaling Barriers

The primary network in a stand-alone static configuration typically suffers from a throughput scaling of $\Theta(1/\sqrt{n \log n})$. By utilizing the supportive secondary tier, this paper proves the primary throughput can be improved to $\Theta(1/\log n)$.

Furthermore, mobility acts as a significant catalyst for performance. In the i.i.d. mobility scenario, the secondary relays can achieve constant delay $\Theta(1)$ for the primary tier. This is a "win-win" scenario: the primary tier receives a massive performance boost, while the secondary tier still achieves its own optimal throughput scaling.

Throughput-delay scaling trade-offs for the primary network, illustrating the improvements achieved through supportive relaying.
Scenario Primary Throughput ($\lambda_p$) Primary Delay ($D_p$)
Stand-alone (Static) $\Theta(1/\sqrt{n \log n})$ $\Theta(\sqrt{n / \log n})$
Supportive (Static) $\Theta(1/\log n)$ $\Theta(\sqrt{n^\beta \log n} \lambda_p(n))$
Supportive (i.i.d. Mobile) $\Theta(1/\log n)$ $\Theta(1)$
Supportive (Random Walk) $\Theta(1/\log n)$ $\Theta(1/S)$

Conclusion and Future Outlook

This research demonstrates that dense secondary networks can serve as a powerful infrastructure resource for legacy primary networks. By transitioning from interference avoidance to active cooperation, both tiers can achieve superior scaling performance. Future work remains to explore regimes where the secondary network is less dense ($1 < \beta < 2$) and to address the practical overheads and security implications of using secondary nodes as honest relays.

Summary updated

BibTeX
@article{gao2012throughput,
  author = {Long Gao and Rui Zhang and Changchuan Yin and Shuguang Cui},
  title = {{Throughput and Delay Scaling in Supportive Two-Tier Networks}},
  journal = {IEEE J. Sel. Areas Commun.},
  year = {2012},
  url = {https://openreview.net/forum?id=CfQlQPFSHnd},
}
Figure