CN1305240C - Method and apparatus for producing given bandwidth and power spectral density noise - Google Patents

Abstract

The present invention relates to the simulation of baseband signals for generating a certain normalization signal-to-noise ratio requirement in a spread spectrum digital communication system, particularly to a method and a device for generating given bandwidth noise and power spectrum density noise. The device comprises a given bandwidth noise module, a power coefficient computation module and a power correction module, wherein the given bandwidth noise module comprises a white noise module, a filter module and a power normalization module. White noise generated in the white noise module is input into a filter module to be filtered; the power of output signals is input into the power normalization module to be normalized. The power coefficient computation module figures out required noise power coefficients; the power noise power coefficients and the white noise normalized by the normalization module output white noise conforming to power spectrum density requirements via the power correction module. The method is suitable for simulating spread spectrum communication systems of IS-95, CDMA 2000, WCDMA, etc. by a conventional platform.

Description

Produce the method and the device of given bandwidth and power spectral density noise

Technical field

The invention belongs to the communication signal process field, relate to the emulation that produces the baseband signal of certain Normalized Signal/Noise Ratio requirement in the spread spectrum digital communication system, particularly a kind of method and device that produces given bandwidth and power spectral density noise.

Background technology

In spread-spectrum communication system emulation, need to produce and satisfy certain Normalized Signal/Noise Ratio E _b/ N ₀Data source.E _bBe signal bit energy, N ₀Be noise power spectral density.Signal is a deterministic signal, and it is the parameter of determining noise that Normalized Signal/Noise Ratio requires actual.Noise refers generally to white noise.In conventional simulation software, the generation module of white noise is provided, its power is 1.Through after the front end filter, the bandwidth of noise changes in the actual channel.In spread-spectrum communication system, baseband signal bandwidth is more than 2 times of spreading code bandwidth.At this moment, just need to produce the noise source of given bandwidth and power spectral density.Directly adopt the bandwidth and the pairing power spectral density of the white noise of white noise module generation to meet the demands.

When spread-spectrum communication system emulation, the generality that often adopt at random, repeated experiments guarantees simulation result, this point require simulation data source can be flexibly, produce easily, make every effort to travel through the multiple situation that the meaning represented is arranged.At present, the bandwidth of the white noise module that provides in the conventional simulation software and power spectral density can not directly be set according to the needs of oneself simultaneously.Possible alternative can adopt some softwares that the base band data that satisfies Normalized Signal/Noise Ratio directly is provided, as COSSAP etc.Except software itself costs an arm and a leg, also exist simultaneously and the problem of the two platform collaborative simulations of Matlab, no matter be to adopt data sharing, still module invokes seems flexibly, freely always shared between two different platforms is not so good as a platform.Another way adopts instrument to produce, and except instrument cost an arm and a leg, data sample length was because memory device is subjected to bigger restriction, and especially under the situation that bandwidth is bigger, memory device can't be finished at a high speed, data acquisition for a long time.Directly write the white noise that algorithm routine is realized certain bandwidth and power spectral density in addition, this mode does not make full use of existing module, needs extra expense to develop same module, and development efficiency is low.

Summary of the invention

The objective of the invention is to disclose a kind of method and device that produces given bandwidth and power spectral density noise, under the conventional software environment,, utilize its white noise module that carries to produce the noise of given bandwidth and power spectral density as C language, Matlab language.

The present invention discloses a kind of method that produces given bandwidth and power spectral density noise, comprises the steps:

Step 1: the white noise module produces the white noise sample;

Step 2: according to bandwidth requirement, designing filter, this filter can design as required, as long as bandwidth meets the demands;

Step 3: with white noise sample input filter, and with the filter output signal power normalization;

Step 4: according to the requirement of Normalized Signal/Noise Ratio, obtain required white noise power, determine power coefficient.The power coefficient computing module is according to the power of spreading factor, bandwidth, baseband digital signal, the noise power that Normalized Signal/Noise Ratio is determined needs.Specifically be calculated as follows:

Chip period: T _cChip amplitude: A; Sample rate: f _s

Noise power spectral density: N ₀

Spreading factor: SF

Signal bit energy: E _b

Because,

$\frac{E_b}{N_0}=\frac{{\mathrm{SFA}}^2{T}_c{f}_s}{{2\mathrm{\σ}}^2}$

So, power coefficient

${\mathrm{\σ}}^2=\frac{{\mathrm{SFA}}^2{T}_c{f}_s}{2(E_b/N_0)}$

Step 5: according to power coefficient given bandwidth white noise is carried out power correction, σ is multiply by in the output of step 3 get final product.

The invention also discloses a kind of device that produces given bandwidth and power spectral density noise, described device is made up of 3 modules: given band noise module, power coefficient computing module, power correction module, wherein given band noise module comprises white noise module, filter module, power normalization module, the white noise that produces in the white noise module, the input filter module is carried out filtering, and output signal power input power normalization module is carried out normalization; The power coefficient computing module calculates the noise power coefficient that needs; The normalized white noise of this power noise power coefficient and normalization module is through the power correction module, and output meets the white noise that power spectral density requires.

Described filter module adopts the root raised cosine filter that is positioned at the base station end in the WCDMA system emulation.

Described power coefficient computing module is according to the power of spreading factor, bandwidth, baseband digital signal, the noise power factor sigma that Normalized Signal/Noise Ratio calculates needs:

${\mathrm{\σ}}^2=\frac{{\mathrm{SFA}}^2{T}_c{f}_s}{2(E_b/N_0)}$

Wherein: T _c--chip period, A--chip amplitude, f _s--sample rate,

N ₀--noise power spectral density, SF--spreading factor, signal bit energy: E _b

Described power correction module is the normalized white noise that the normalization module is exported to be multiply by σ get final product.

Method uniqueness of the present invention, novelty do not retrieve in present patent and pertinent literature.This method has following characteristics:

The first, adopt conventional simulation Software Platform, utilize existing module to transform, avoid redesigning algorithm, improved simulation efficiency greatly.

The second, program adopts modularized design, and input parameter is set flexibly.

The 3rd, be applicable to that all produce the noise process of given bandwidth and power spectral density in the signal processing field.

Description of drawings

Further specify the present invention below in conjunction with drawings and Examples.

Fig. 1 has illustrated that the present invention is applied to the enforcement block diagram of WCDMA system up-link emulation;

Fig. 2 has illustrated structured flowchart of the present invention.

Embodiment

In this example, the emulation of WCDMA up link is divided into four parts according to function: up reception simulation data source 11, Radio Frequency Subsystem equivalent model 12, baseband signal processing module 13, proof of algorithm 14.The data source that meets the WCDMA agreement is produced by 3GPP data source module 1111, at first enter yard formed filter 1112, here the transfer function of sign indicating number formed filter meets the root raised cosine characteristic, the processing procedure of analogue mobile station side, the signal of output is through channel model module 1113, the space transmission course of analog signal, signal arrives base station side and at first enters array antenna received, obtain vector data by 1114 processing of Array Model module, again through 1115 outputs of sign indicating number formed filter, the processing procedure of anolog base station side, the transmission characteristic of two sign indicating number formed filters is synthesized a raised cosine filter.The signal imitation multi-user data that a plurality of up reception simulation data sources 11 produce.The multi-user data that enters Radio Frequency Subsystem equivalent model 12 at first carries out the multi-user through multi-user's synthesis module 121 and synthesizes, add the noise data that meets the signal to noise ratio requirement that produces by interchannel noise module 126 again, the composite signal of summation module 122 outputs enters the characteristic of channel module 123 of simulation actual channel characteristic, enter automatic gain control module 124 and sampling module 125 again, the radio-frequency channel of finishing base station side receives to be handled.So far obtained the emulation baseband signal.The emulation and the assessment of algorithm when this signal process baseband signal processing module 13 and proof of algorithm 14 are finished sky.

Interchannel noise module 126 is exactly the noise module that produces certain bandwidth and power spectral density.This module has adopted module of the present invention exactly, introduces in detail below, shown in figure two.

Method provided by the invention is made up of 3 modules: given band noise module 21, power coefficient computing module 22, power correction module 23.

Step 1 utilizes white noise module 211 to produce white noise, the randn among the Matlab;

Step 2 designs the filter 212 of given bandwidth according to bandwidth requirement, adopts the filter in the actual channel, and this filter adopts the root raised cosine filter that is positioned at the base station end in the WCDMA system emulation, and parameter satisfies protocol requirement;

Step 3 is obtained output signal power, and rate of doing work normalization 213 then;

Step 4 is according to the power of spreading factor, bandwidth, baseband digital signal, Normalized Signal/Noise Ratio is determined needs by power coefficient computing module 22 noise power;

Specifically be calculated as follows:

Chip period: T _cChip amplitude: A; Sample rate: f _s

Noise power spectral density: N ₀

Spreading factor: SF

Signal bit energy: E _b, noise power spectral density N ₀

Because,

${\mathrm{\σ}}^2=\frac{{\mathrm{SFA}}^2{T}_c{f}_s}{2(E_b/N_0)}$

So, power coefficient

${\mathrm{\σ}}^2=\frac{{\mathrm{SFA}}^2{T}_c{f}_s}{2(E_b/N_0)}$

Step 5 is handled the noise that step 3 produces according to the result of step 4, and 23 pairs of normalization noises of power correction module are revised, and σ is multiply by in the output of step 3 get final product.Generation meets the white noise that power spectral density requires.

In a word, adopt method provided by the invention to go up the generation of realization noise at conventional platform (C language, Matlab language) according to bandwidth in the broadband communication and power spectral density needs, method is simple, adopts ready-made module to build, reduce the complexity of algorithm greatly, improve operation efficiency; Adopt conventional simulation Software Platform, saved system cost, improved the algorithm applicability, system is constituted oversimplify, technical difficulty reduces, and is an algorithm very with practical value.This method is applicable to utilizes conventional platform (as Matlab, C etc.) to carry out the emulation of spread-spectrum communication systems such as IS-95, cdma2000, WCDMA.

Claims (4)

1. a method that produces given bandwidth and power spectral density noise is characterized in that, comprises the steps:

Step 1: the white noise module produces the white noise sample;

Step 2: according to bandwidth requirement, designing filter;

Step 3: with white noise sample input filter, and with the filter output signal power normalization;

Step 4: according to the requirement of Normalized Signal/Noise Ratio, obtain required white noise power, determine power coefficient σ;

Described noise power factor sigma calculates according to power, the Normalized Signal/Noise Ratio of spreading factor, bandwidth, baseband digital signal:

${\mathrm{\σ}}^2=\frac{{\mathrm{SFA}}^2{T}_c{f}_s}{2\left({E_b/N}_0\right)}$

Wherein: T _c--chip period, A--chip amplitude, f _s--sample rate,

N ₀--noise power spectral density, SF--spreading factor, E _bBe the signal bit energy;

Step 5: multiply by power coefficient σ by output numerical value and come given bandwidth white noise is carried out power correction, produce the white noise that meets the power spectral density requirement with step 3.

2. the method for given bandwidth of generation as claimed in claim 1 and power spectral density noise is characterized in that, described filter adopts the root raised cosine filter that is positioned at the base station end in the WCDMA system emulation.

3. device that produces given bandwidth and power spectral density noise, it is characterized in that, described device comprises: given band noise module, power coefficient computing module, power correction module, wherein given band noise module comprises white noise module, filter module, power normalization module, the white noise that produces in the white noise module, the input filter module is carried out filtering, and power output signal input power normalization module is carried out normalization; The power coefficient computing module calculates the noise power coefficient that needs, this noise power coefficient and described normalization module are exported normalized white noise through described power correction module, the normalized white noise of normalization module output be multiply by power coefficient σ, thereby output meets the white noise that power spectral density requires;

Described power coefficient computing module is according to the power of spreading factor, bandwidth, baseband digital signal, the noise power factor sigma that Normalized Signal/Noise Ratio calculates needs:

${\mathrm{\σ}}^2=\frac{{\mathrm{SFA}}^2{T}_c{f}_s}{2(E_b/N_0)}$

Wherein: T _c--chip period, A--chip amplitude, f _s--sample rate,

N ₀--noise power spectral density, SF--spreading factor, E _bBe the signal bit energy.

4. the device of given bandwidth of generation as claimed in claim 3 and power spectral density noise is characterized in that, described filter module adopts the root raised cosine filter that is positioned at the base station end in the WCDMA system emulation.

Images