Multilink Wireless Communication Low Latency Data Handling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The 802.11be standard's multi-link operation (MLO) feature is not optimized for bandwidth-demanding and low-latency communication services, as it delays the transmission of low-latency data due to shared acknowledgment and sequence numbering policies across multiple links, impacting services like gaming and streaming.
Innovation Solution
Implementing a communication method where data units on specific links are managed independently of the conventional acknowledgment and sequence ordering policies, allowing for immediate removal from buffers and delivery to upper layers without acknowledgments, while maintaining conventional management for other links to ensure reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional acknowledgment and sequence numbering policies are applied across all multiple links, then reliability of data transmission is improved, but latency for low-latency services deteriorates
Solution Approach 1:
The patent segments the multiple communication links into two distinct sets: a first set of links for reliable data transmission and a second set of links for low-latency data transmission. This segmentation allows different buffer management policies to be applied to different link sets, resolving the contradiction between reliability and latency by enabling simultaneous optimization for both requirements on different links.
Solution Approach 2:
The patent applies different buffer management qualities locally to different link sets. For the first set of links, conventional acknowledgment-based buffer management is applied to ensure reliability. For the second set of links, immediate buffer removal without acknowledgment is applied to minimize latency. This local differentiation allows each link set to be optimized for its specific performance requirement.
2Reliability
If data units wait for acknowledgment before removal from transmit buffer, then transmission reliability is improved, but buffer space availability for new data deteriorates
Solution Approach 1:
The patent divides the transmit buffer into two separate buffers: a first transmit buffer for data units transmitted over the first set of links (reliability-oriented) and a second transmit buffer for data units transmitted over the second set of links (latency-oriented). This segmentation enables independent buffer management for each link set, allowing the second buffer to immediately remove data units without waiting for acknowledgment, thereby improving buffer space utilization efficiency for low-latency services.
Solution Approach 2:
The patent implements dynamic buffer management policies tailored to different link characteristics. The second transmit buffer dynamically removes data units as soon as they are transmitted over the low-latency links, without waiting for acknowledgment, adapting the buffer management behavior to the specific requirements of low-latency services and improving overall buffer space utilization efficiency.
3Reliability
If sequence numbering is enforced for all data units, then data ordering reliability is improved, but delivery speed for low-latency data deteriorates
Solution Approach 1:
The patent segments the data transmission system into two independent paths: one for ordered data delivery and one for speed-optimized delivery. The second set of links operates without sequence numbering enforcement, allowing data units to be delivered immediately to upper layers without waiting for preceding data, thereby significantly improving data delivery speed for low-latency services while maintaining ordering reliability through alternative means.
Data Source
AI summary
In a multilink system, block acknowledgment is performed at TID level. It is commonly accepted that acknowledgment policy refrains low latency delivery. However, it is often that traffic streams combine low-latency data and data requiring a high level of reliability. A need exists to facilitate the transmission of low-latency data units of a TID when a multilink operation, such as in 802.11be, is implemented with an acknowledgment policy. One specific link from amongst the multiple links may be dedicated to low latency data for which the originator and the recipient no longer take into account the acknowledgment. It turns that the originator can remove the low latency data from its transmit buffer immediately after transmission, while the recipient can deliver these data to the upper layer without waiting for the good reception of all preceding data, according to a sequence numbering. Low-latency services are consequently improved.