English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

In this paper, we generate a transmit power pool (PP) for Internet of things(IoT) networks with semi-grant-free non-orthogonal multiple access (SGF-NOMA)via multi-agent deep reinforcement learning (MA-DRL) to enable open loop powercontrol (PC). The PP is mapped with each resource block (RB) to achievedistributed power control (DPC). We first formulate the resource allocationproblem as stochastic Markov game, and then solve it using two MA-DRLalgorithms, namely double deep Q network (DDQN) and Dueling DDQN. Each GF useras an agent tries to find out the optimal transmit power level and RB to formthe desired PP. With the aid of dueling processes, the learning process can beenhanced by evaluating the valuable state without considering the effect ofeach action at each state. Therefore, DDQN is designed for communicationscenarios with a small-size action-state space, while Dueling DDQN is for alarge-size case. Moreover, to decrease the training time, we reduce the actionspace by eliminating invalid actions. To control the interference and guaranteethe quality-of-service requirements of grant-based users, we determine theoptimal number of GF users for each sub-channel. We show that the PC approachhas a strong impact on data rates of both grant-based and GF users. Wedemonstrate that the proposed algorithm is computationally scalable tolarge-scale IoT networks and produce minimal signalling overhead. Our resultsshow that the proposed MA-Dueling DDQN based SGF-NOMA with DPC outperforms theexisting SGF-NOMA system and networks with pure GF protocols with 17.5\% and22.2\% gain in terms of the system throughput, respectively. Finally, we showthat our proposed algorithm outperforms the conventional open loop PCmechanism.

A Power-Pool-Based Power Control in Semi-Grant-Free NOMA Transmission