TLQ-Geo: a two-level q-learning-based geographic routing protocol for flying ad hoc networks

A Novel RL-Driven Zone-Based Leader-Aware Energy-Efficient Routing Protocol for Dynamic MANET Environments

The paper introduces RML-ZEREM to solve existing limitations, which functions as a Reinforcement Learning (RL) based Zone-Based Leader-Aware Energy-Efficient Routing Protocol for MANETs, which serves next-generation MANET applications.

A Q-learning Weighted Cluster Based Routing Protocol for VANETs

Simulation results obtained demonstrate that Q-WeCBR outperforms CBR, DSDV, and GPSR in terms of packet delivery ratio and throughput, confirming the effectiveness of clustering combined with learning-based routing for dynamic vehicular networks.

QL-6GRP: A Lightweight Q-Learning-Based Routing Protocol for Dynamic MANETs in 6G-Oriented Environments

Mobile Ad Hoc Networks (MANETs) are expected to support highly dynamic and decentralized communication scenarios in future 6G-oriented wireless systems. However, routing remains challenging because of mobility, topology variability, and resource constraints. Reinforcement learning (RL) offers a promising alternative by enabling adaptive routing decisions based on observed network conditions. This paper presents QL-6GRP (Q-Learning for 6G Routing Protocol), a lightweight Q-learning-based routing protocol designed for fully distributed MANET environments. The protocol enables each node to learn next-hop forwarding decisions using local observations, including link quality, residual energy, hop progress, and neighborhood density. A complete implementation of QL-6GRP was developed within the NS-3 simulator, supporting online learning through hop-level feedback signaling and bounded-memory operation. The protocol was evaluated under multiple parameter settings and network sizes using Random Waypoint mobility and UDP constant-bit-rate traffic to examine both routing performance and computational behavior. The experimental results demonstrate the feasibility of adaptive routing with moderate signaling overhead under carefully tuned moderate-scale scenarios while revealing key trade-offs between feedback frequency, routing quality, and computational scalability. Moderate periodic feedback provides the most favorable balance, whereas excessive feedback increases overhead without improving performance. In addition, reinforcement-learning operations incur substantial computational costs, with the wall-clock runtime increasing by approximately 13 times when the network size increases from 50 to 100 nodes. These findings reveal the operating limits and practical design trade-offs of lightweight tabular RL-based MANET routing and provide useful guidelines for future scalable learning-driven protocols in dynamic wireless environments.

An Optimization of Multiple Route Selection in Flying Ad Hoc Network

Simulation results indicate that HOA-MEPFL-CLCT-RP outperforms existing models in terms of Packet Delivery Ratio (PDR), energy efficiency, End-to-End Delay (E2D), and routing overhead.

Reliable Session-Oriented Multi-Path Routing for LEO Satellite Networks: A Multi-Agent Learning Approach

Reliability-Aware Local-Grid-Based Multipath Routing with Q-Learning Adaptation for Wireless Sensor Networks with a Mobile Sink

Multipath routing in wireless sensor networks (WSNs) improves reliability by providing alternative forwarding paths when a route fails. However, mobile sinks make path maintenance difficult because sink movement can invalidate previously constructed source-to-sink routes. Existing protocols typically depend on either global path reconstruction, which increases control overhead, or footprint-chaining, which accumulates detours through previous sink positions and may weaken path independence. To address this problem, this paper proposes QL-LGMPRP, a reliability-aware local-grid-based multipath routing protocol that combines a sink-centered local grid, two-path delivery, link-quality-aware forwarding, and lightweight tabular Q-learning for waypoint adaptation. Mobility-related route changes are confined to the sink-centered grid, whereas a compact tabular Q-learning policy adjusts the primary-path direction using grid, link-quality, and energy-related state variables. The sink constructs a local grid around its current position, with cells sized to keep in-grid forwarding locally bounded. When an event occurs, the source computes an entry point on the grid perimeter and constructs two greedy paths: a primary path through a Q-learning-selected waypoint near the grid boundary and a backup path toward the current sink position. The Q-learning agent uses a compact tabular state representation that includes the boundary-cell index, residual-energy level, sink-grid position, and local link-quality information, and learns waypoint offsets using a reward that combines delivery success, transmission energy, and delay. This design confines routing adaptation to the sink-centered grid while allowing the waypoint policy to respond to heterogeneous link conditions. Simulation results under different sink speeds and interference conditions show that QL-LGMPRP maintains high delivery reliability while reducing detour-related forwarding costs relative to footprint-chaining and showing lower weak-link exposure than the geometric-forwarding comparison schemes.