# US 5,706,404 A · Neural network using inhomogeneities in a medium as neurons and transmitting input signals in an unchannelled wave pattern through the medium

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 5,706,404 A
- Document ID: US5706404A
- Status: Granted
- Kind: A
- Application number: 08/688,462
- Filed: 1996-07-30
- Granted: 1998-01-06
- Earliest claimed: 1993-03-03
- Assignees: U.S. Philips Corporation
- Inventors: Sel B. Colak
- Web page: https://patentgrep.com/patent/US5706404A

## CPC

- G06N 99/007: {Molecular computers, i.e. using inorganic molecules (using biomolecules G06N3/002)}
- B82Y 10/00: Nanotechnology for information processing, storage or transmission, e.g. quantum computing or single electron logic
- G06N 3/063: using electronic means
- G06N 3/065: Analogue means
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- B82Y: SPECIFIC USES OR APPLICATIONS OF NANOSTRUCTURES; MEASUREMENT OR ANALYSIS OF NANOSTRUCTURES; MANUFACTURE OR TREATMENT OF NANOSTRUCTURES

## Abstract

Neural net with spatially distributed functionalities. An information processing system comprises a neural net with fully distributed neuron and synapse functionalities in a spatially inhomogeneous medium to propagate a response field from an input to an output. The response field is a reaction of the medium to a plurality of input signals and depends non-linearly on the input signals. The response field is also determined by the inhomogeneities. The value of the field at one or more particular locations is indicative of one or more output signals of the neural net.

## Claims

### Claim 1

A neural network processor comprising:
input means for receiving a plurality of input signals;
output means for providing at least one output signal;
neural network means coupled to the input means and the output means, the neural network means comprising a physical body, the physical body having physical properties allowing for propagation of an electrical WAVE in an unchanneled fashion therethrough from the input means to the output means, the body comprises:
at least one region, which is suitable for affecting electrical conduction through the physical body;
training means for altering a state of the region during a learning mode of the neural network means; and wherein:
the physical body comprises a semiconductor body; and
the semiconductor body comprises:
a semiconductor substrate;
an insulating layer on a major surface of the substrate;
at least one charge storage region within the insulating layer for inducing the at least one region in the substrate.

### Claim 2 (depends on 1)

The processor of claim 1 wherein the training means comprises at least one conductor situated on the opposite side of the insulating material from the semiconductor substrate.

### Claim 3 (depends on 2)

The processor of claim 2 further comprising an additional doped region within the semiconductor substrate, which doped region is in electrical contact with the conductor, via an aperture in the insulating material.

### Claim 4 (depends on 3)

The processor of claim 3 wherein the semiconductor substrate has a doping of opposite conducting type to that of additional doped region.

### Claim 5 (depends on 2)

The processor of claim 2 wherein the conductor is situated directly over the charge storage region and interacts capacitively or galvanically therewith.

### Claim 6 (depends on 2)

The processor of claim 2 wherein the conductor is situated over a thin portion of the insulating material and interacts capacitively or galvanically with the semiconductor substrate.

### Claim 7 (depends on 1)

The processor of claim 1 wherein the training means comprises
a piezo-electric layer on the insulating layer for receiving mechanical stress waves which alter the electrical properties of the piezo-electric layer;
a second insulating layer on the piezo-electric layer; and
a conductive layer on the second insulating layer, so that the piezo-electric layer alters an effect of the conductive layer on the at least one region.

### Claim 8 (depends on 1)

The system of claim 1, wherein the training means comprises an upper surface containing photoelectric elements for receiving photons for altering electrical properties of the upper surface.

### Claim 9 (depends on 1)

The system of claim 1 wherein the training means comprises a ferroelectric layer on the insulating layer.