6G Traceable Spatial Messaging in Resident Domains—A Cell-Free MIMO UDNs for Hybrid BilSTM & GRU RNN Enabled Architectural Reference Model (ARM)

Autor: Azeemi, Naeem Z., Azeemi, Naveed A., Umaira Abdullah
Rok vydání: 2023
Předmět:
DOI: 10.5281/zenodo.7874999
Popis: In the era of 6G, a game-changing approach will redefine the communications and networks depending on the required services. A large volume of geo-tagged data can be fundamental to providing applications of location-based services (LBSs). One of the important LBS applications is to provide continuous spatial keyword queries. A continuous spatial keyword query monitors a designated region with a set of keywords. In the designated region, if mobile objects contain all the keywords of the query, they are the answer set for the query. The query continuously monitors the spatial region and reports its up-to-date query result. In order to support new requirements and services, mature technologies are needed to embed such as, Artificial intelligence (AI) and Machine Learning (ML). There are several algorithms for text classification, ranging from ML to Deep Learning (DL). Since the advent of the high-end computational facility (HiPC), numerical crunching has become much easier with lesser computational time. This has paved way for evolution of Complicated Network Architecture (CAN) which can be trained to achieve Higher Accuracy, Precision and Recall (HiAP&R). A cumulative performance of HiAP&R proportionately affects the F1 score based on which the performance of the Neural Network (NN) model can be assessed. The present work attempts to explore the proposed neural network, Hybrid RNN model with two BiLSTM layers and two GRU layer and compare the performance with other hybrid models. We exposed our results for Unsupervised Learning (UL) includes Hierarchical Clustering (HC), Partitioned Clustering (PC), Association Rule Mining (ARM), and Dimensionality Reduction (DR). Supervised Learning (SL) is comparatively explored to cover decision tree, K-nearest neighbouring, and Support Vector Machine (SVM). 6G use cases, requirements, and key enabling techniques are discussed in Reinforcement Learning (RL), both Model-Based Approaches (MBA) and Model-Free Approaches (MFA) are investigated. Deep learning is improvised to tailor 6G communication Social Media Data Volume (SMDV) from a perceptron to neural networks, convolutional neural networks, and recurrent neural networks. The GloVE dataset is employed to train the models, and their performance is evaluated by accuracy, precision, recall and F1-score. The performance of the proposed models is compared with other models by using F1 score. We expect that our results are useful for researchers and technicians seeking an optimal, sub-optimal, or trade-off solution for each B5G communications and 6G networks problem using AI techniques. E.g., 6G use case proposed here AI with GRU and RCNN trade-offs makes communications and networks design and management smarter and safer. Key question is not about whether but when and how to implement AI in 6G communication systems.
6G mobile network communication provides innovative internet ecosystems going beyond the smart phone that must be created, and new multiple application sectors including potential new players and service providers need to collaborate in order to take advantage together of the technological progress. This work explicitly address the scenarios where a large volume of geo-tagged data can be fundamental to providing applications of location-based services (LBSs). We discuss how the important LBS applications are to provide continuous spatial keyword queries. It is shown here a continuous spatial keyword query monitors a designated region with a set of keywords. In the designated region, if mobile objects contain all the keywords of the query, they are the answer set for the query. The query continuously monitors the spatial region and reports its up-to-date query result. The current comprehensive research sought to address four text categorization methodologies by tweaking the models and assessing their performance in relation to accuracy, recall, and F1 score. In this work we concerned with latent variable models for discrete data, embedded such as bit vectors, sequences of categorical variables, count vectors, graph structures, relational data, etc. These models can be used to analyze voting records, text and document collections, low-intensity images, movie ratings, etc. However, we will mostly focus on text analysis, and this will be reflected in our terminology.
Databáze: OpenAIRE
Externí odkaz: