Spatial Analysis of Opportunistic Downlink Relaying in a Two-Hop Cellular System

We consider a two-hop cellular system in which the mobile nodes help the base station by relaying information to the dead spots. While two-hop cellular schemes have been analyzed previously, the distribution of the node locations has not been explicitly taken into account. In this paper, we model the node locations of the base stations and the mobile stations as a point process on the plane and then analyze the performance of two different two-hop schemes in the downlink. In one scheme the node nearest to the destination that has decoded information from the base station in the first hop is used as the relay. In the second scheme the node with the best channel to the relay that received information in the first hop acts as a relay. In both these schemes we obtain the success probability of the two hop scheme, accounting for the interference from all other cells. We use tools from stochastic geometry and point process theory to analyze the two hop schemes. Besides the results obtained a main contribution of the paper is to introduce a mathematical framework that can be used to analyze arbitrary relaying schemes. Some of the main contributions of this paper are the analytical techniques introduced for the inclusion of the spatial locations of the nodes into the mathematical analysis.

💡 Research Summary

The paper addresses the problem of dead‑spots in cellular downlink coverage by exploiting opportunistic two‑hop relaying, where mobile users assist the base station (BS) in delivering data to destinations that experience poor direct reception. While prior works have examined two‑hop cellular architectures, they typically assume deterministic or regular node placements and ignore the spatial randomness inherent in real networks. To fill this gap, the authors model both BSs and mobile stations (MSs) as independent Poisson point processes (PPPs) on the Euclidean plane, thereby capturing the stochastic geometry of a large‑scale cellular system.

Two distinct relay selection policies are investigated:

1. Nearest‑Neighbour Relay (NNR) – After the first hop, all MSs that successfully decode the BS transmission form a thinned PPP. The relay is the MS among this set that is geometrically closest to the final destination. This policy leverages spatial proximity and allows the authors to use the well‑known nearest‑neighbour distance distribution of a PPP to characterize the relay‑to‑destination link.

2. Best‑Channel Relay (BCR) – From the same set of successfully decoded MSs, the relay is the one with the highest instantaneous channel gain (including path‑loss and Rayleigh fading) toward the destination. This selection is based on channel quality rather than distance, requiring order‑statistics to model the distribution of the maximum channel gain.

For each policy the end‑to‑end success probability is defined as the probability that both hops satisfy a target SINR threshold θ. The analysis proceeds by first deriving the conditional success probability of the first hop, which depends on the distance from the BS to the decoding MS and on the aggregate interference generated by all other BSs. Using the Laplace functional of the PPP, the authors obtain a closed‑form expression for the Laplace transform of the interference, L_I(s). The set of potential relays is then treated as a new PPP (thinned by the first‑hop success event), and the distribution of the relay‑to‑destination distance (NNR) or the maximum channel gain (BCR) is derived analytically.

In the second hop, interference originates from two sources: (i) transmissions from BSs in other cells and (ii) transmissions from relays in other cells that are simultaneously forwarding data. Both sources are modeled as independent PPPs, and their combined interference Laplace transform is again expressed in closed form. By integrating over the random relay locations (for NNR) or over the distribution of the maximal channel gain (for BCR), the authors obtain the overall success probability P_success = P₁·P₂ for each scheme.

Numerical evaluation and Monte‑Carlo simulations are performed to validate the analytical results. The authors vary key parameters such as BS density (λ_B), user density (λ_M), path‑loss exponent (α), transmit power (P_t), and SINR threshold (θ). The findings can be summarized as follows:

• Performance Gap – BCR consistently outperforms NNR in dense networks or at high transmit powers, achieving up to a 10 % higher success probability because it exploits instantaneous channel fluctuations. However, BCR requires real‑time channel state information (CSI) exchange and additional signaling overhead, which may be impractical in some deployments.
• Complexity vs. Gain – NNR offers a low‑complexity, location‑based relay selection that performs nearly as well as BCR in sparse networks where interference is limited. Its reliance on geographic proximity eliminates the need for CSI feedback.
• Interference Sensitivity – Both schemes are highly sensitive to inter‑cell interference. As the density of BSs or relays increases, the Laplace transform of the interference decays rapidly, leading to a steep drop in success probability. This underscores the importance of interference management techniques (e.g., power control, frequency reuse) in two‑hop cellular designs.
• Analytical Accuracy – The derived expressions match simulation results across a wide range of parameters, confirming that the stochastic‑geometry framework accurately captures the impact of spatial randomness.

Beyond the specific results, the paper’s primary contribution is methodological: it establishes a general analytical framework that can accommodate arbitrary relay selection rules within a PPP‑based cellular model. By expressing success probability in terms of Laplace transforms of interference and by treating the set of eligible relays as a thinned PPP, the authors enable systematic performance evaluation of more sophisticated schemes such as cooperative MIMO relaying, dynamic spectrum sharing, or user‑centric clustering.

The authors conclude by suggesting several avenues for future work, including: (i) extending the model to non‑homogeneous point processes (e.g., clustered or determinantal processes) to capture hotspot traffic; (ii) incorporating user mobility and time‑varying relay sets; (iii) jointly optimizing relay selection and power allocation for energy‑efficient operation; and (iv) evaluating the impact of advanced interference mitigation (e.g., coordinated multipoint, beamforming) within the same stochastic‑geometry framework.