ConPA: A Contention-free Mechanism with Power Adaptation for Beyond Listen-Before-Talk
Francesc Wilhelmi, Paolo Baracca, Gianluca Fontanesi, Lorenzo Galati-Giordano
TL;DR
This work tackles the challenge of achieving low latency and high reliability in unlicensed spectrum by moving beyond traditional LBT-based access. It introduces Contention-free with Power Adaptation (ConPA), a mechanism that bypasses contention and adapts transmit power based on measured interference to enhance airtime and throughput. An analytical framework based on Continuous-Time Markov Chains (CTMC) is developed to derive metrics such as throughput, airtime, and SINR, and ConPA is evaluated against DCF and 802.11ax OBSS/PD-based Spatial Reuse. Results show up to 76% throughput gains and substantial airtime improvements through spatial reuse, with some SINR tradeoffs at short range, underscoring ConPA’s potential for latency-critical deployments in the unlicensed bands.
Abstract
In view of the need to find novel means to utilize the unlicensed spectrum to meet the rising latency and reliability requirements of new applications, we propose a novel mechanism that allows devices to transmit anytime that a packet has to be delivered. The proposed mechanism, Contention-free with Power Adaptation (ConPA), aims to bypass the contention periods of current Listen-Before-Talk (LBT) approaches, which are the main source of unreliability in unlicensed technologies like Wi-Fi. To assess the feasibility of ConPA, we provide an analytical method based on Markov chains, which allows deriving relevant performance metrics, including throughput, airtime, and quality of transmissions. Using such a model, we study the performance of ConPA in various scenarios, and compare it to baseline channel access approaches like the Distributed Coordination Function (DCF) and the IEEE 802.11ax Overlapping Basic Service Set (OBSS) Packet Detect (PD)-based Spatial Reuse (SR). Our results prove the effectiveness of ConPA in reusing the space to offer substantial throughput gains with respect to the baselines (up to 76% improvement).
