# US 6,731,706 B1 · Square root raised cosine symmetric filter for mobile telecommunications

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

- Number: US 6,731,706 B1
- Document ID: US6731706B1
- Status: Granted
- Kind: B1
- Application number: 09/429,058
- Filed: 1999-10-29
- Granted: 2004-05-04
- Earliest claimed: 1999-10-29
- Assignees: Intel Corporation
- Inventors: Tinku Acharya; George J. Miao
- Web page: https://patentgrep.com/patent/US6731706B1

## CPC

- H03H 17/06: Non-recursive filters
- H03H 17/0283: {Filters characterised by the filter structure (H03H17/0202, H03H17/0219 - H03H17/0248 take precedence)}
- H03H: IMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS (waveguides, resonators, lines or other devices of the waveguide type H01P)

## Abstract

A symmetric discrete-time square-root-raised-cosine filter may be utilized to process a serial input signal and to produce an output square in each clock cycle. The filter may be adapted to decrease the computational complexity of the resulting calculations, making the filter amenable to use in power sensitive applications and in those situations where the available processor's performance is relatively limited.

## Claims

### Claim 1

A discrete-time square-root-raised-cosine filter comprising a symmetric finite impulse response filter, wherein coefficients of the filter are not equal to zero for even integer sample numbers and wherein coefficients of the self-convolution of the filter are approximately equal to zero for even integer sample numbers.

### Claim 2 (depends on 1)

The filter of claim 1 having odd symmetric coefficients.

### Claim 3 (depends on 1)

The filter of claim 1 having approximately 21 coefficients.

### Claim 4 (depends on 1)

The filter of claim 1 having a systolic architecture.

### Claim 5 (depends on 1)

The filter of claim 1 having one output signal per input clock cycle.

### Claim 6 (depends on 1)

The filter of claim 1 including a plurality of processing units, each unit including a first delay element, and a multiply-accumulate circuit.

### Claim 7 (depends on 6)

The filter of claim 6 including a second delay element to provide a delay approximately equal to the processing time of said multiply-accumulate circuit.

### Claim 8

A wide band code division multiple access transceiver comprising:
an analog processing section; and
a digital processing section coupled to said analog processing section, said digital processing section including an odd, symmetric, finite impulse response filter, said filter including a plurality of processing units to implement a systolic architecture, said filter to produce an output signal in response to each input clock signal, and said filter constructed to result in an self-convolution that exhibits coefficients approximately equal to zero for even integer sample numbers.

### Claim 9 (depends on 8)

The transceiver of claim 8 wherein said filter includes a plurality of processing units, each unit including a first delay element and a multiply-accumulate circuit.

### Claim 10 (depends on 9)

The transceiver of claim 9 including a second delay element to provide a delay approximately equal to the processing time of said multiply-accumulate circuit.

### Claim 11 (depends on 10)

The transceiver of claim 10 including a sequencing circuit so that a single adder may be used to implement two different additions.

### Claim 12

A method of implementing a code division multiple access receiver comprising:
receiving a code division multiple access signal;
generating symmetric finite impulse response filter coefficients;
generating an output ripple signal having a magnitude at a frequency of one-half of the chip rate of about three decibels less than the normalized gain; and
causing the self-convolution of the filter to have coefficients that are approximately equal to zero for even integer sample numbers.

### Claim 13 (depends on 12)

The method of claim 12 further including setting the edge of the stopband frequency to have ninety-nine percent of half of the occupied bandwidth.

### Claim 14 (depends on 12)

A method of claim 12 including processing said received signal through a first filter having a systolic architecture.

### Claim 15 (depends on 12)

The method of claim 12 further including causing the filter coefficients to be non-zero for even integer sample numbers.

### Claim 16 (depends on 1)

A discrete-time square-root-raised-cosine filter as defined in claim 1, further comprising:
a first delay element having an output;
a second delay element having an input coupled to the output of the first delay element and having an output;
a first processor having a first input coupled to the output of the second delay element and having first and second outputs;
a second processor having a first input coupled to the output of the first delay element, a second input coupled to the first output of the first processor, and an output; and
an adder tree having a first input coupled to the second output of the first processor, a second input coupled to the output of the second processor and an output to provide a filtered signal.

### Claim 17 (depends on 16)

A discrete-time square-root-raised-cosine filter as defined in claim 16, wherein the processors are constructed to store coefficients of the filter.

### Claim 18 (depends on 1)

A discrete-time square-root-raised-cosine filter as defined in claim 1, further comprising:
a first processor having an input to couple to an input signal and having a first output and a second output; and
a second processor having a first input to couple to the input signal, a second input coupled to the first output of the first processor, and a third input coupled to the second output of the first processor.

### Claim 19 (depends on 18)

A discrete-time square-root-raised-cosine filter as defined in claim 18, wherein the second processor comprises:
a first adder having a first input to couple to the input signal and a second input coupled through a delay element to the first output of the first processor;
a multiplier having a first input coupled to an output of the first adder, a second input to couple to a clock signal, and an output; and
a second adder having a first input coupled to the output of the multiplier and a second input coupled to the second output of the first processor.

### Claim 20 (depends on 18)

A discrete-time square-root-raised-cosine filter as defined in claim 18, wherein the second processor comprises:
a first multiplexer having an input to couple to the input signal;
a second multiplexer having a first input coupled though a delay element to the first output of the first processor and a second input coupled to the second output of the first processor; and
an adder having a first input coupled to an output of the second multiplexer.

### Claim 21 (depends on 20)

A discrete-time square-root-raised-cosine filter as defined in claim 20, wherein the second processor further comprises:
a multiplier having a first input coupled to an output of the adder and a second input to couple to a clock signal.

### Claim 22 (depends on 21)

A discrete-time square-root-raised-cosine filter as defined in claim 21, wherein the second processor further comprises:
a third multiplexer having a first input coupled to an output of the multiplier and a second input coupled to the output of the adder.

### Claim 23 (depends on 22)

A discrete-time square-root-raised-cosine filter as defined in claim 22, wherein the first multiplexer and the third multiplexer are constructed to be coupled to a source of a first select signal and the second multiplexer is constructed to be coupled to a source of a second select signal, the first and second select signals having respectively different phases.
