Method and apparatus for low power chip-to-chip communications with constrained ISI ratio

• US 10,348,436 B2
• Filed: 02/09/2016
• Issued: 07/09/2019
• Est. Priority Date: 02/02/2014
• Status: Active Grant

• Current Assignee: Kandou Labs, SA
• Original Assignee: Kandou Labs, SA
• Inventors: Shokrollahi, Amin, Hormati, Ali, Ulrich, Roger
• Primary Examiner(s): Ohri, Romani
• Application Number:
  US15/019,868
• Publication Number:
  US 20160173219A1
• Time in Patent Office: 1,246 Days
• CPC Class Codes: H04J 13/004 Orthogonal
• CPC Class Codes: H04L 25/0272 Arrangements for coupling to multiple lines, e.g. for differential transmission
• CPC Class Codes: H04L 25/0276 Arrangements for coupling common mode signals
• CPC Class Codes: H04L 25/03343 Arrangements at the transmitter end
• CPC Class Codes: H04L 25/14 Channel dividing arrangements , i.e. in which a single bit stream is divided between several baseband channels and reassembled at the receiver
• CPC Class Codes: H04L 25/4919 using balanced multilevel codes H04L25/4927 takes precedence

Abstract

An efficient communications apparatus is described for a vector signaling code to transport data and optionally a clocking signal between integrated circuit devices. Methods of designing such apparatus and their associated codes based on a new metric herein called the “ISI Ratio” are described which permit higher communications speed, lower system power consumption, and reduced implementation complexity.

Claims

1. A method comprising:

receiving a set of input signals;

generating symbols of a normalized-gain codeword based on a weighted sum of a set of subchannels, each subchannel weighted with a corresponding input signal, wherein the set of subchannels correspond to rows of a set of rows selected from a normalized orthogonal matrix, each row of the set of rows (i) mutually orthogonal and (ii) orthogonal to an all-one row, wherein a first number of symbols of the normalized-gain codeword affected by a first subchannel is different from a second number of symbols of the normalized-gain codeword affected by a second subchannel, the first and second subchannels having equal subchannel gain; and

transmitting the symbols of the normalized-gain codeword on a plurality of wires of a multi-wire bus, each symbol of the normalized-gain codeword transmitted via a respective wire of the plurality of wires of the multi-wire bus.

2. The method of claim 1, wherein the normalized orthogonal matrix is represented by the matrix:

3. The method of claim 1, wherein the normalized orthogonal matrix is generated from an unnormalized matrix, and wherein generating the normalized matrix comprises:

generating a normalized set of signals by adjusting signal levels of the received input signals, and weighting subchannels of the unnormalized matrix with corresponding signals of the normalized set of signals.

4. The method of claim 3, wherein the unnormalized matrix is represented by the matrix:

5. The method of claim 3, wherein adjusting the signal levels of input signals comprises adjusting signal levels of lower-magnitude signals to a magnitude of a highest-magnitude signal of the set of input signals.

6. The method of claim 3, wherein adjusting the signal levels of input signals comprises adjusting signal levels of higher-magnitude signals to a magnitude of a lower-magnitude signal of the set of input signals.

7. The method of claim 1, further comprising:

receiving the symbols of the normalized-gain codeword via the multi-wire bus;

generating a set of comparator outputs, using a set of multi-input comparators (MICs), each comparator output generated based on a respective combination of a respective subset of symbols of the normalized-gain codeword according to a respective set of input-weighting coefficients, and wherein each comparator output has an equal magnitude; and

generating a set of output bits by slicing the set of comparator outputs.

8. The method of claim 7, wherein the respective set of input-weighting coefficients corresponds to a respective row of the set of rows of a code receiver matrix.

9. The method of claim 4, wherein the normalized set of input signals are selected from a basis vector [0, ±

⅜

, ±

¼

, ±

⅜

, ±

¼

, ±

⅜

].

10. An apparatus comprising:

an encoder configured to receive a set of input signals, and to responsively generate symbols of a normalized-gain codeword based on a weighted sum of a set of subchannels, each subchannel weighted with a corresponding input signal, wherein the set of subchannels correspond to rows of a set of rows selected from a normalized orthogonal matrix, each row of the set of rows (i) mutually orthogonal and (ii) orthogonal to an all-one row, wherein a first number of symbols of the normalized-gain codeword affected by a first subchannel is different from a second number of symbols of the normalized-gain codeword affected by a second subchannel, the first and second subchannels having equal subchannel gain; and

a plurality of line drivers configured to transmit the symbols of the normalized-gain codeword on a plurality of wires of a multi-wire bus, each symbol of the normalized-gain codeword transmitted via a respective wire of the plurality of wires of the multi-wire bus multiwire bus.

11. The apparatus of claim 10, wherein the normalized orthogonal matrix is represented by the matrix:

12. The apparatus of claim 10, wherein the normalized orthogonal matrix is generated from an unnormalized matrix, and wherein the apparatus further comprises:

a scaling unit configured to generate a normalized set of input signals by adjusting signal levels of the received input signals, and to weight subchannels of the unnormalized matrix with corresponding signals of the normalized set of input signals.

13. The apparatus of claim 12, wherein the unnormalized matrix is represented by the matrix:

14. The apparatus of claim 13, wherein the normalized set of input signals are selected from a basis vector [0, ±

⅜

, ±

¼

, ±

⅜

, ±

¼

, ±

⅜

].

15. The apparatus of claim 12, wherein the scaling unit is configured to adjust signal levels of lower-magnitude signals to a magnitude of a highest-magnitude signal of the set of input signals.

16. The apparatus of claim 12, wherein the scaling unit is configured to adjust signal levels of higher-magnitude signals to a magnitude of a lower-magnitude signal of the set of input signals.

17. The apparatus of claim 10, further comprising:

a plurality of multi-input comparators (MICs) configured to receive the symbols of the normalized-gain codeword via the multi-wire bus, and to generate a set of comparator outputs, each comparator output generated based on a respective combination of a respective subset of symbols of the normalized-gain codeword according to a respective set of input-weighting coefficients, and wherein each comparator output has an equal magnitude; and

a plurality of slicers configured to generate a set of output bits by slicing the set of comparator outputs.

18. The apparatus of claim 17, wherein the respective set of input-weighting coefficients corresponds to a respective row of the set of rows of a code receiver matrix.