Network Packet Prioritization for Downlink Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current cellular communication systems lack effective methods for prioritizing downlink data transmissions based on various attributes, leading to suboptimal Quality of Service (QoS) and user experience, especially in multi-user and multi-technology environments like 4G and 5G networks.
Innovation Solution
Implementing network-based and sender-based data packet prioritization techniques using machine learned models that tag data packets with priority levels based on attributes such as application type, user preferences, network conditions, and device attributes, allowing for optimized transmission through the radio protocol stack.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If data packets are transmitted without prioritization in current cellular systems, then the system operation is simple, but the Quality of Service and user experience deteriorate
Solution Approach 1:
The patent applies preliminary action by tagging data packets with priority levels before transmission. The network device or sending device assigns priority tags to packets based on application type, user preferences, and network conditions before the packets enter the transmission queue. This pre-prioritization ensures that when packets are transmitted, they are already organized by importance, improving QoS without adding complexity during the transmission phase.
Solution Approach 2:
The patent implements local quality by applying different priority levels to different data packets based on their specific characteristics. Each packet is individually tagged with a priority level that reflects its specific requirements (e.g., real-time video vs. background synchronization). This allows the system to provide customized transmission treatment to each packet type, enhancing overall QoS while maintaining manageable system complexity through standardized tagging mechanisms.
2Ease of operation
If data packets are prioritized based on multiple attributes, then the user experience improves, but the processing complexity increases
Solution Approach 1:
The system performs preliminary action by determining priority levels for data packets before transmission based on multiple attributes including application type, user preferences, and network conditions. The network device or sending device analyzes these attributes in advance and assigns appropriate priority tags, so that the actual transmission process can proceed efficiently without real-time complex decision-making.
Solution Approach 2:
The patent introduces an intermediary mechanism in the form of priority tags or labels that mediate between the complex attribute analysis and the simple transmission execution. The tagging system acts as an intermediary layer that captures the results of complex multi-attribute analysis in a simplified format that can be easily processed during transmission, thus improving user experience while controlling processing complexity.
3Productivity
If machine learned models are used for packet prioritization, then the transmission optimization improves, but the computational requirements increase
Solution Approach 1:
The patent uses machine learned models as an intermediary that processes complex prioritization decisions. The model is trained offline to recognize patterns in application types, user behaviors, and network conditions, then applies these learned patterns to assign priority levels. This intermediary approach allows the system to achieve high transmission efficiency through intelligent prioritization while keeping online computational energy requirements manageable by leveraging pre-trained model knowledge.
Solution Approach 2:
The machine learned model performs preliminary action by pre-processing and analyzing packet attributes to determine priority levels before transmission. The model captures complex relationships between multiple attributes and priority outcomes during offline training, then applies this knowledge rapidly during actual transmission. This preliminary intelligent analysis improves transmission efficiency while the model's cached knowledge reduces real-time computational energy consumption.
Data Source
AI summary
Techniques for network-based and sender-based data packet prioritization for downlink transmissions are discussed herein. Packets can be tagged or associated with information indicative of a priority of the packet for downlink transmission to a user equipment (UE). A priority level can be determined based on an application identifier or type associated with a data request, a level of user interaction with the UE, network conditions, and other factors. Packets can be received by a PDCP layer of a base station for sending based on the priority. Packets may be associated with a primary priority level associated with a QCI level and a secondary priority level based on UE and/or network factors discussed herein. Packets associated with a same QCI may be prioritized to optimize transmission of downlink data associated with a single UE or between transmission of downlink data associated with a plurality of UEs.