MIMO-OFDM Neuromorphic Inter-Device Communication

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Neuromorphic systems face limitations in data rate, capacity, and reliability due to point-to-point connectivity issues, such as deadlock, livelock, and memory constraints in address event representation (AER) protocols, which affect performance and accuracy in spike-timing dependent plasticity (STDP) and lead to system failures.

Innovation Solution

Implementing a Multiple Input Multiple Output (MIMO) Orthogonal Frequency Division Multiplexing (OFDM) system that separates data into parallel channels based on destination addresses and frequency bands, using a central router to combine and transmit data efficiently, reducing the need for physical hardware connections and improving signal stability through space-time signal processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If point-to-point connectivity is used in AER protocols, then data can be delivered sequentially through routers, but data rate and capacity are limited and deadlock/livelock may occur

Engineering Contradiction:
Improvedata rateVSAvoiddeadlock and livelock occurrence
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent segments the communication system into multiple independent frequency channels instead of using sequential point-to-point routing. Each frequency channel operates independently, allowing parallel data transmission and eliminating the sequential bottlenecks that cause deadlock and livelock in traditional AER protocols.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a frequency dimension to the communication system by implementing OFDM modulation. This transforms the single-dimensional sequential transmission into multi-dimensional parallel transmission across multiple frequency channels, dramatically increasing data rate while preventing deadlock through independent channel operation.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Ease of operation

If look-up tables are implemented in each node for packet delivery, then routing can be performed, but substantial memory is consumed

Engineering Contradiction:
Improverouting capabilityVSAvoidmemory consumption
Core Design Contradiction:
Ease of operationVSQuantity of substance

Solution Approach 1:

The patent extracts the complex routing logic from individual nodes and replaces it with simple frequency-based channel assignment. Instead of implementing full look-up tables in each node, the system uses centralized frequency allocation where packets are routed simply by assigning them to appropriate frequency channels, dramatically reducing memory requirements while maintaining routing capability.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the routing parameter from complex destination-based look-up table queries to simple frequency channel assignments. This parameter transformation reduces the information storage requirement at each node from substantial memory to minimal frequency identification, while maintaining effective routing through the network.

Inventive Principle:
Principle #35Parameter changes

3Productivity

If multiple neuromorphic devices are connected to increase neuron capacity, then processing power increases, but communication issues such as traffic jams and system failures increase

Engineering Contradiction:
Improveneuron processing capacityVSAvoidsystem failure rate
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent segments the network traffic into multiple independent frequency channels, allowing parallel communication between multiple neuromorphic devices. This segmentation prevents traffic jams by distributing load across channels and eliminates system failures caused by communication bottlenecks, while maintaining high neuron processing capacity through scalable device interconnection.

Inventive Principle:
Principle #1Segmentation

4Device complexity

If time-multiplexing is used to encode spiking data, then data can be transmitted over a single bus, but data rate is limited

Engineering Contradiction:
Improvecommunication bus structureVSAvoiddata rate
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The patent introduces frequency as an additional dimension for data transmission by implementing OFDM modulation. Instead of using time-multiplexing on a single bus, the system transmits multiple data streams simultaneously across multiple frequency channels, dramatically increasing data rate while maintaining relatively simple bus structure through parallel frequency-based communication.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP2962416B1A MIMO-OFDM system for robust and efficient neuromorphic inter-device communication
Publication Date: 2018.06.27 HRL LAB
  • EP2962416B1 patent drawingFigure 1
  • EP2962416B1 patent drawingFigure 2
  • EP2962416B1 patent drawingFigure 3

AI summary

A Multiple Input Multiple Output (MIMO) Orthogonal Frequency Division Multiplexing (OFDM) system for inter-device communication is described. Information data from each neuromorphic chip is coded and modulated, on the basis of destination, into different channels. The parallel signals in different channels are sent serially using TDM to a central router. After signal grouping by a central switching controller, each group of signals may be delivered to corresponding transmitter in the central router for transmission to a corresponding receiver in the neuromorphic chip using TDM.