Exploring Relay Cooperation for Secure and Reliable Transmission in Two-Hop Wireless Networks

This work considers the problem of secure and reliable information transmission via relay cooperation in two-hop relay wireless networks without the information of both eavesdropper channels and locations. While previous work on this problem mainly studied infinite networks and their asymptotic behavior and scaling law results, this papers focuses on a more practical network with finite number of system nodes and explores the corresponding exact result on the number of eavesdroppers one network can tolerant to ensure desired secrecy and reliability. We first study the scenario where path-loss is equal between all pairs of nodes and consider two transmission protocols there, one adopts an optimal but complex relay selection process with less load balance capacity while the other adopts a random but simple relay selection process with good load balance capacity. Theoretical analysis is then provided to determine the maximum number of eavesdroppers one network can tolerate to ensure a desired performance in terms of the secrecy outage probability and transmission outage probability. We further extend our study to the more general scenario where path-loss between each pair of nodes also depends the distance between them, for which a new transmission protocol with both preferable relay selection and good load balance as well as the corresponding theoretical analysis are presented.

💡 Research Summary

The paper addresses secure and reliable information delivery in two‑hop relay wireless networks when the legitimate nodes have no knowledge of the eavesdroppers’ channel state information (CSI) or locations. While prior work on this topic has largely focused on asymptotic scaling laws in infinite networks, this study concentrates on a practical finite‑node setting and derives exact expressions for the maximum number of eavesdroppers that can be tolerated while meeting prescribed secrecy outage probability (SOP) and transmission outage probability (TOP) constraints.

The authors first consider the simplified scenario where all inter‑node links experience identical path‑loss, i.e., the average signal‑to‑noise ratio (SNR) is the same for any pair of nodes. Within this framework two transmission protocols are examined:

Optimal Relay Selection (ORS) – The relay that maximizes the instantaneous SNR at the destination is chosen. This yields the smallest possible SOP and TOP but requires global CSI, incurs high computational complexity, and tends to concentrate traffic on a few relays, leading to poor load balance.
Random Relay Selection (RRS) – A relay is drawn uniformly at random from the pool of available relays. The selection overhead is O(1) and traffic is naturally balanced, yet the chosen relay may have a sub‑optimal channel, which can increase outage probabilities.

For each protocol the authors formulate SOP and TOP as functions of the transmit power (P), noise variance (\sigma^2), the number of relays (N), and the number of eavesdroppers (M). By imposing SOP (\le \epsilon_s) and TOP (\le \epsilon_t) (with (\epsilon_s,\epsilon_t) being design‑level thresholds), they solve for the largest integer (M_{\max}) that satisfies both constraints. The resulting expressions reveal that (M_{\max}) grows roughly linearly with (N) and logarithmically with the SNR gap (P/\sigma^2), but the proportionality constants differ markedly between ORS and RRS. ORS achieves the highest (M_{\max}) at the cost of heavy CSI exchange, whereas RRS sacrifices a modest amount of security/reliability for dramatically reduced signaling and perfect load balancing.

The analysis is then extended to a more realistic distance‑dependent path‑loss model, where the average SNR of a link decays as (d^{-\alpha}) (with (\alpha) the path‑loss exponent). In this setting the spatial locations of relays become critical. To reconcile the need for good channel quality with load balancing, the authors propose a Preferred Relay Selection with Load Balancing (PRSLB) protocol. The procedure consists of two steps: (i) compute a distance‑based weight for each relay, forming a candidate set of “good” relays whose weighted SNR exceeds a predefined threshold; (ii) select uniformly at random from this candidate set. This hybrid approach retains the low‑complexity and balancing advantages of RRS while approaching the SOP/TOP performance of ORS.

The theoretical derivations rely on Markov’s inequality, order‑statistics of exponential channel gains, and bounding techniques to obtain closed‑form upper bounds for SOP and TOP under the three protocols. These bounds are then inverted to yield explicit formulas for (M_{\max}) as a function of system parameters. The authors validate the analytical results through extensive Monte‑Carlo simulations. In the equal‑loss scenario, with (N=20) relays, transmit SNR of 20 dB, and target (\epsilon_s=\epsilon_t=0.01), ORS tolerates up to 12 eavesdroppers, PRSLB up to 11, and RRS up to 9. In the distance‑dependent case (path‑loss exponent (\alpha=3)), PRSLB’s performance remains within 5 % of ORS while preserving the O(1) selection cost. Load‑balance metrics (standard deviation of per‑relay traffic) confirm that PRSLB and RRS achieve near‑uniform relay utilization, whereas ORS exhibits a variance roughly three times larger.

Key insights emerging from the work are:

Finite‑node analysis matters – Exact (M_{\max}) expressions enable network planners to dimension the number of relays and transmit power to meet concrete security/reliability targets, rather than relying on asymptotic scaling trends.
Load balancing versus optimality – In many practical deployments, the modest loss in tolerable eavesdroppers when using random or hybrid selection is outweighed by the benefits of reduced CSI exchange, lower latency, and even energy consumption across relays.
Spatial awareness is essential – When path‑loss depends on distance, blindly applying ORS can waste resources on far‑away relays; a distance‑aware candidate set dramatically improves the security‑reliability trade‑off.

The paper concludes by highlighting possible extensions: multi‑antenna relays, mobile eavesdroppers, cooperative jamming, and cross‑layer scheduling that jointly optimize relay selection, power allocation, and network coding. Overall, the study bridges the gap between theoretical scaling laws and actionable design guidelines for secure two‑hop relay networks in realistic, finite‑size environments.