Method for predicting one or more events e.g. network congestion, involves feeding trained model with received performance data, selected features and configured hyperparameters, and predicting events utilizing trained model by processors
2023-10-11
专利权人JIO PLATFORMS LTD (RELN-C)
申请日期2023-10-11
专利号IN202321068459-A
成果简介NOVELTY - The method involves receiving a request from multiple sources by multiple processors. Multiple features are selected based on the received request by the processors. A trained model is fetched from a storage unit based on a received request. The trained model is fed with the received performance data, selected features and configured hyperparameters. Multiple events are predicted utilizing the trained model. The request includes performance data and details of a trained model to be utilized, where the performance data is a network function performance data including key performance indicators (KPIs) and counters. USE - Method for predicting one or more events such as network congestion, network equipment failures. ADVANTAGE - The accuracy of predictions is improved and allows better resource allocation. The model can adapt to varying conditions and performance metrics, potentially leading to more relevant and timely predictions. The method easily integrates into existing network management frameworks, facilitating improved data-driven decision-making, and contributes to better resource utilization and network optimization, ultimately leading to cost savings and improved service quality. DETAILED DESCRIPTION - An INDEPENDENT CLAIM is included for a system for predicting one or more events. DESCRIPTION OF DRAWING(S) - The drawing shows a block diagram of an environment for predicting one or more events. 100Environment 102User equipment 104Server 106Network 108Computer system
IPC 分类号G06F-011/34 ; G06F-009/54 ; G06N-020/00 ; G08G-005/00 ; H04L-041/147
国家印度
专业领域信息技术
语种英语
成果类型专利
文献类型科技成果
条目标识符http://119.78.100.226:8889/handle/3KE4DYBR/20056
专题中国科学院新疆生态与地理研究所
作者单位
JIO PLATFORMS LTD (RELN-C)
推荐引用方式
GB/T 7714
DANGE G,LOBO R,TELGOTE K,et al. Method for predicting one or more events e.g. network congestion, involves feeding trained model with received performance data, selected features and configured hyperparameters, and predicting events utilizing trained model by processors. IN202321068459-A[P]. 2023.
条目包含的文件
条目无相关文件。
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。