SDIO Multiplexing Header for IP Packet Buffering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing SDIO interface in mobile computing devices incurs high control overhead and CPU interrupts due to the need for multiple transactions per IP packet, limiting data transfer rates.
Innovation Solution
Implementing a method to buffer multiple IP packets and attach a multiplexing header, allowing for a single SDIO transaction to transfer multiple packets, thereby reducing protocol overhead and CPU interrupts by amortizing these costs across multiple packets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple SDIO transactions are used per IP packet, then protocol control information can be properly transmitted, but data transfer rate is limited due to high overhead
Solution Approach 1:
The patent combines multiple IP packets into a single SDIO transaction by buffering packets and attaching multiplexing headers. This merging approach allows multiple data units to be transferred together, reducing the number of control overhead instances and improving overall data transfer rate while maintaining reliable protocol control information transmission.
Solution Approach 2:
The patent performs preliminary actions by buffering multiple IP packets before transmission and pre-attaching multiplexing headers that contain routing information. This preparation allows the system to initiate fewer SDIO transactions with larger data payloads, reducing overhead and improving transfer efficiency while ensuring reliable delivery.
2Productivity
If single SDIO transaction transfers multiple IP packets, then data transfer rate increases, but packet routing complexity increases
Solution Approach 1:
The patent introduces multiplexing headers as intermediary elements that carry routing information for individual packets within an aggregated transfer. These headers act as mediators that enable the receiving end to correctly route and separate packets without requiring complex routing logic during transmission, thus improving data transfer rate while managing routing complexity through structured header information.
3Ease of operation
If DMA is used for data transfer, then CPU overhead is reduced, but interrupt frequency increases limiting transfer rate
Solution Approach 1:
The patent merges multiple IP packet transfers into a single SDIO transaction initiated by DMA. This consolidation reduces the number of interrupts required, as one interrupt can now complete multiple packet transfers instead of one interrupt per packet. This approach maintains the CPU efficiency benefits of DMA while reducing interrupt frequency to levels that no longer limit transfer rates.
Data Source
AI summary
One embodiment of the invention comprises a non-transitory, tangible computer readable storage medium encoded with processor readable instructions to perform a method of transferring SDIO data. One method comprises buffering multiple IP packets to transfer from one of a SDCC host and a SDIO client to the other of the SDCC host and the SDIO client. A multiplexing header is attached to each of the multiple IP packets and one of at least one SDIO read command and at least one SDIO write command issued. The multiple IP packets are then transferred in a single SDIO transfer between the one of a SDCC host and a SDIO client to the other of the SDCC host and the SDIO client.