Abstract
Networking protocols are designed through long-standing and hard-working human efforts. Machine Learning (ML)-based solutions for communication protocol design have been developed to avoid manual effort to adjust individual protocol parameters. While other proposed ML-based methods focus mainly on tuning individual protocol parameters (e.g. contention window adjustment), our main contribution is to propose a new Deep Reinforcement Learning (DRL) framework to systematically design and evaluate networking protocols. We decouple the protocol into a set of parametric modules, each representing the main protocol functionality that is used as a DRL input to better understand and systematically analyze the optimization of generated protocols. As a case study, we introduce and evaluate DeepMAC a framework in which the MAC protocol is decoupled into a set of blocks across popular 802.11 WLANs (e.g. 802.11 a/b/g/n/ac). We are interested to see which blocks are selected by DeepMAC across different networking scenarios and whether DeepMAC is capable of adapting to network dynamics.
Highlights
The proliferation of existing Internet and mobile communications networked devices, systems and applications has contributed to increasingly large, heterogeneous, dynamic and systematically complex networks
Ii) In order to show the feasibility of our framework, we propose DeepMAC, a novel deep reinforcement learning-based framework that targets the design of 802.11 Medium Access Control (MAC) protocols based on the given networking scenario
Reinforcement learning is a machine learning technique where the agent interacts with a time-variant environment that can be modeled as a Markov Decision Process (MDP), a Partially Observable MDP (POMDP), a game, etc
Summary
The proliferation of existing Internet and mobile communications networked devices, systems and applications has contributed to increasingly large, heterogeneous, dynamic and systematically complex networks. A protocol is decoupled into a set of parametric modules as DRL inputs, each representing the main protocol functionality referred to as Building Blocks(BBs) This modularization technique helps to better understand the protocols generated, optimize the protocol design and analyze them systematically. Our framework could be utilized as a multivariate optimization tool that helps in alleviating the current protocol design process Using this framework, domain experts provide the required specifications (objective) for a specific scenario as DRL input and could identify/capture the role that each protocol component (block) plays in varying scenarios for different objectives. I) We motivate a novel Reinforcement Learning (RL)-based protocol design approach that tunes the protocol parameters and optimizes the protocol design across all network stack layers by leveraging the concept of demodulating a protocol into its set of main functionalities referred to as building blocks. Iii) We present the future trends and opportunities that such a novel RL-based framework can bring to protocol design approaches that are more robust and adaptive to varying network conditions, application requirements, and heterogeneous device characteristics
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
