Vehicle Control Module RDMA Synchronization Using Mutex Locks

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing vehicle control systems face challenges in efficiently coordinating memory access across different nodes without relying on RDMA network interface card hardware enhancements, leading to potential race conditions and increased processor load and communication latency.

Innovation Solution

Implementing a software-based process synchronization mechanism using mutex locks and atomic operations to manage memory access between vehicle control module nodes over Ethernet links, allowing synchronized access without requiring RDMA network interface card enhancements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If software-based process synchronization mechanism using mutex locks is implemented, then processor load and communication latency are reduced, but device complexity increases due to additional synchronization software components

Engineering Contradiction:
Improvedata transmission efficiencyVSAvoidsynchronization mechanism complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent introduces a mutex server as an intermediary component that manages synchronization between multiple vehicle control module nodes. The mutex server receives lock requests from mutex clients, manages the mutex state (locked/unlocked), and coordinates access to shared memory regions. This intermediary approach centralizes synchronization logic, reducing the burden on individual nodes and improving overall system efficiency while maintaining manageable complexity through clear separation of concerns.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements self-service through automatic mutex lock management where the mutex server autonomously handles lock acquisition, holding, and release operations based on atomic operation results. The system also performs self-synchronization by automatically detecting race conditions and coordinating access without external intervention, allowing the synchronization mechanism to manage itself and reducing the need for complex external control logic.

Inventive Principle:
Principle #25Self-service

2Adaptability or versatility

If mutex lock control is implemented without RDMA network interface card hardware enhancements, then adaptability to hybrid network configurations is improved, but communication latency increases due to software-based synchronization overhead

Engineering Contradiction:
Improvehybrid network configuration compatibilityVSAvoidcommunication latency
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The mutex server and client architecture provides universal synchronization capability that works across different network interface card types (RDMA-capable and non-RDMA-capable). The software-based approach allows the same synchronization mechanism to function in homogeneous RDMA environments, heterogeneous hybrid environments, and traditional Ethernet environments, making the system universally adaptable without requiring hardware-specific modifications while maintaining acceptable performance through optimized software implementation.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The system performs preliminary synchronization setup by establishing mutex locks and registering memory regions before actual data transmission occurs. The mutex server pre-configures synchronization parameters and atomic operation handlers in advance, so that during actual communication, the synchronization overhead is minimized. This preliminary action ensures that when data transmission is needed, the synchronization infrastructure is already in place and optimized, reducing real-time latency.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If atomic operation instructions are used for mutex lock management, then reliability of memory access coordination is improved, but processor load increases due to execution of atomic instructions

Engineering Contradiction:
Improvememory access coordinationVSAvoidprocessor load
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent extracts the computationally intensive atomic operation execution from the critical path of data transmission. The mutex server handles atomic operations separately from the main data transfer flow, using them only for synchronization state management (lock acquisition and release). This extraction allows atomic instructions to provide reliable coordination without continuously burdening the processor, as they are executed only when synchronization state changes are needed, not during every data transmission operation.

Inventive Principle:
Principle #2Taking out (Extraction)

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Reduces processor load and communication latency, enabling efficient data transmission for critical vehicle functions such as automated emergency braking and active damping, while supporting hybrid network configurations with both RoCE and SoftRDMA NICs.

Implementation Method 1

the mutex server of the second vehicle control module node is configured to access the mutex lock by executing an atomic operation instruction

Methodology Applied
Scientific EffectAtomic operation:

Data Source

PatentUS12488831B2Remote direct memory access with process synchronization mechanisms for vehicle control modules
Publication Date: 2025.12.02 GM GLOBAL TECHNOLOGY OPERATIONS LLC
  • US12488831B2 patent drawing
  • US12488831B2 patent drawing
  • US12488831B2 patent drawing

AI summary

A vehicle electrical architecture includes a first vehicle control module node including a first remote direct memory access (RDMA) capable network interface card and a mutex client, and a second vehicle control module node including a second RDMA capable network interface card, a mutex server, and a memory, with an Ethernet link in communication between the first and second RDMA capable network interface cards. The first vehicle control module node is configured to receive a write request for a specified memory region of the second vehicle control module node, communicate with the mutex server of the second vehicle control module node, via the mutex client, to request a mutex lock associated with the specified memory region, and in response to a notification of the mutex lock being obtained, execute an RDMA write operation to the specified memory region.