WO2012065451A1 - 适用于跳频及功率波动系统的削峰系数更新方法及装置 - Google Patents
适用于跳频及功率波动系统的削峰系数更新方法及装置 Download PDFInfo
- Publication number
- WO2012065451A1 WO2012065451A1 PCT/CN2011/076534 CN2011076534W WO2012065451A1 WO 2012065451 A1 WO2012065451 A1 WO 2012065451A1 CN 2011076534 W CN2011076534 W CN 2011076534W WO 2012065451 A1 WO2012065451 A1 WO 2012065451A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- carrier
- filter coefficient
- power
- prototype filter
- coefficient
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B1/00—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission
- H04B1/69—Spread spectrum techniques
- H04B1/713—Spread spectrum techniques using frequency hopping
Definitions
- the present invention relates to the field of communications, and in particular, to a method and apparatus for updating a peak clipping coefficient suitable for frequency hopping and power fluctuation systems. Background technique
- the Global System for Mobile Communications (GSM) and Code Division Multiple Access (CDMA) systems are typical communication systems that use frequency hopping technology.
- GSM Global System for Mobile Communications
- CDMA Code Division Multiple Access
- the pulse cancellation peak clipping algorithm based on the combined path has its incomparable advantages in resources and implementation structure, and has been widely used in various chips and communication systems.
- a big problem with this structure is that the shaping filter coefficients of CPG (Cancellation Pulse Generators) cannot be updated in real time with fluctuations in frequency or carrier power, thus being applied to frequency hopping and power fluctuation systems. There is a problem of spectrum leakage or poor peak performance.
- CPG Cancellation Pulse Generators
- the prior application with the application number 200910192407.8 proposes a problem of solving the frequency hopping application by an adaptive look-up table method, which traverses according to a known frequency hopping sequence, and calculates the filter coefficients of each set of frequency points, according to The frequency hopping sequence number completes the switching of the CPG filter coefficient table.
- This method can only eliminate the problem of spectrum leakage caused by the CPG filter coefficient changing with the frequency point.
- the prior application with the application number 200810216615.2 proposes a method of peak clipping processing based on multi-carrier power over-matching in the GSM system.
- This multi-carrier peak clipping method for multi-carrier system carrier power over-matching, combined with GSM The slot-level power control of the system.
- a non-uniform peak clipping between carriers is used to ensure that the system exceeds the rated configuration when the power exceeds the rated configuration. Best performance.
- This patent is only for a peak clipping device that guarantees system performance when the power exceeds the rated configuration. It does not solve the application problem that the peak performance index will be poor when the power level changes rapidly during frequency hopping applications. Summary of the invention
- the invention provides a method and a device for updating a peak clipping coefficient suitable for frequency hopping and power fluctuation systems, which can enable CPG filter coefficients to be updated in real time as the frequency or carrier power fluctuates.
- a method for updating a peak clipping coefficient suitable for frequency hopping and power fluctuation systems comprising the steps of: generating an offset pulse shaping filter coefficient update enable signal when a carrier frequency change occurs or a relative change in carrier power is greater than a set value;
- the offset pulse shaping filter coefficients currently used by the system are updated. Further, the formula for determining a carrier power weighting correction factor value of each carrier according to a change in carrier power is:
- ch wr — wevi ⁇ is the currently obtained carrier average power value
- c/2 jwr — o1 ⁇ 2 is the carrier reference power value
- A the prototype filter stage number
- the pulse shaping filter coefficient, ⁇ ( ) is the modified prototype filter coefficient of each carrier, M is the carrier number, A is the prototype filter order, and N is the length of the prototype filter coefficient, which is the frequency of each carrier. For the signal sampling rate.
- a peak clipping coefficient updating device suitable for frequency hopping and power fluctuation systems, comprising: an update control unit, configured to generate a cancellation pulse shaping filter when a carrier frequency change occurs or a relative change in carrier power is greater than a set value a coefficient update enable signal is sent to the cosine signal generating unit and the carrier power weighting correction factor generating unit;
- a prototype filter coefficient storage unit configured to store coefficients of a prototype filter of each carrier
- a carrier power weight correction factor generating unit configured to determine a carrier power weighting value of each carrier according to a change in carrier power
- a cosine signal generating unit configured to generate a cosine signal corresponding to each carrier frequency point according to the frequency control word and the phase control word of each input carrier;
- a canceling pulse shaping filter coefficient calculating unit configured to perform power weighting calculation on the prototype filter coefficients of each carrier, to obtain a modified prototype filter coefficient of each carrier; and the synchronized cosine signal and the corrected
- the prototype filter coefficients are subjected to time domain multiplication and accumulation processing to obtain offset pulse shaping filter coefficients;
- a normalization, gain compensation, and quantization processing unit for normalizing, gaining, and quantizing the cancellation pulse shaping filter coefficients.
- the carrier power weighting correction factor generating unit determines a carrier power weighting correction factor value of each carrier according to a change in carrier power:
- ch wr — wevi ⁇ is the currently obtained carrier average power value
- c/2 wr — o1 ⁇ 2 is the carrier reference power value
- J 1, .. ., M
- M is the number of carriers.
- the cancellation pulse shaping filter coefficient calculation unit needs to adjust the maximum amplitude value and the cosine signal of the modified prototype filter coefficient before performing the time domain multiplication and accumulation processing on the synchronized cosine signal and the modified prototype filter coefficient.
- the zero phase value is aligned.
- cancellation pulse shaping filter coefficient calculation unit performs a time domain multiplication and accumulation process on the synchronized cosine signal and the corrected prototype filter coefficient as follows:
- the pulse shaping filter coefficient, ⁇ ( ) is the modified prototype filter coefficient of each carrier, M is the carrier number, A is the prototype filter order, and N is the length of the prototype filter coefficient, which is the frequency of each carrier. For the signal sampling rate.
- the normalization, gain compensation, and quantization processing unit formulates the normalization, gain compensation, and quantization processes for canceling the pulse shaping filter coefficients:
- the invention provides a method and a device for updating a peak clipping coefficient in a system with frequency hopping, carrier power fluctuation and carrier number variation.
- the invention ensures that the peak clipping performance index is equivalent to the static peak clipping performance index. Therefore, the performance indicators of the system when applied to different scenarios are guaranteed.
- FIG. 1 is a schematic diagram of an implementation of a combined carrier pulse cancellation peak clipping device in the prior art
- FIG. 2 is a frequency spectrum diagram of a CPG filter coefficient constructed by a prototype filter coefficient ⁇ ( ) in the present invention
- FIG. 3 is a flow chart of a method for updating a peak clipping coefficient applicable to a frequency hopping and power fluctuation system according to the present invention
- Fig. 4 is a block diagram showing the structure of a peak clipping coefficient updating apparatus suitable for frequency hopping and power fluctuation systems in the present invention. detailed description
- FIG. 1 is a schematic diagram of an implementation of a combined carrier pulse cancellation peak clipping device in the prior art.
- the CPG filter 20 in FIG. 1 needs to be updated in real time when the frequency and carrier power change.
- the update of the CPG filter coefficients is mostly done by the central processing unit (CPU), which makes it difficult to adjust the CPG filter coefficients in real time.
- CPU central processing unit
- A is the prototype filter order
- N is the length of the prototype filter coefficient
- / s is the signal sampling rate.
- the CPG filter coefficients are calculated by the digital filter configuration of the prototype filter and the system.
- the CPG filter coefficients are a set of filter coefficients for spectral shaping of the peak cancellation pulse according to the spectral characteristics of the signal, which are stored in the RAM table.
- a set of parameters, the cancellation pulse after CPG filtering can ensure that the adjacent channel leakage power ratio (ACPR) of the signal meets the system requirements after peak clipping.
- FIG. 2 is a spectrum diagram of CPG filter coefficients constructed by prototype filter coefficients.
- the abscissa is a normalized digital angular frequency (Normalized Frequency), and the ordinate is a signal spectrum suppression degree (Magnitude). ).
- the construction of this set of filter coefficients is related to the digital frequency of the carrier. In the frequency hopping system application, the change of the carrier frequency is real-time and fast, so the CPG filter coefficients are also required to be updated in real time. It is guaranteed that the canceling pulse energy after peak clipping will not leak out of the band due to the rapid change of the carrier frequency, thereby causing interference to the adjacent channel signal.
- the carrier number and carrier power of the system will change rapidly.
- the carrier power fluctuation causes the power imbalance between carriers
- the carrier demodulation index with lower carrier power after peak clipping will deteriorate.
- the power difference between carriers reaches a certain level
- the carrier with lower power will not meet the system requirements after the peak clipping.
- the calculation of the CPG filter coefficient needs to be corrected to the following formula (2):
- M the number of carriers
- A the prototype filter order
- N the length of the prototype filter coefficients
- W the prototype filter coefficient.
- the calculation of the completion formula (2) is only part of the calculation of the CPG filter coefficient.
- normalization, gain compensation and quantization processing are also required, so that the CPG filter coefficients can be applied by the hardware logic.
- the normalization is normalized according to the maximum value of the CPG filter coefficient /zf ⁇ , and the quantization is performed according to the bit width defined by the quan-num parameter.
- cpg cpg
- the normalization process needs to perform the division operation.
- the division operation is more complicated for hardware logic calculation. If the divisor is a fixed constant, the division operation can be converted into multiplication operation.
- the logic operation is simple and fast to complete the operation. However, the value of (L /2" + l) here varies with the number of carriers and the ratio of carrier power.
- the operation of dividing in hardware is complicated, which makes the calculation of this step become the bottleneck of hardware logic implementation. . How to find a simple method to realize the normalized operation of CPG filter coefficients becomes the key link to complete the calculation of CPG filter coefficients by hardware logic.
- the present invention gives a corresponding processing method for this problem.
- Equation (4) only one parameter 3 ⁇ 4 is a variable, so that the change and correction of ⁇ can correspond to formula (4).
- the weighting of the correction factor can be non-continuous, so the weighting value of the ⁇ factor can also be obtained by using the lookup table.
- the index address of the table is the sum value, and the content of the table can be changed according to the total power. Fix the granularity to determine.
- the Kn table corresponding to the requirements of different system applications can be determined according to the actual situation.
- the flowchart of the method for updating the peak clipping coefficient of the frequency hopping and power fluctuation system specifically includes the following steps:
- the center frequency of one or more carriers in a multi-carrier frequency hopping communication system changes according to a frequency hopping sequence with time, and the power of the carrier is also adjusted according to service changes and quality of service, in frequency hopping and power.
- the CPG filter coefficient update is initiated according to the change of power and frequency points.
- a register of carrier reference power and reference frequency needs to be separately set, which is used for obtaining with the current system. The power is compared with the frequency point information.
- Ch wr — ⁇ / ⁇ and c/zj ⁇ — ⁇ / ⁇ correspond to the reference power and reference frequency register values
- ch _pwr_new ⁇ ch e — «ew corresponds to the currently obtained carrier average power value and carrier frequency point information value.
- the conditions for generating the update start signal are:
- the carrier power weighting positive factor Ki needs to be determined according to the current carrier power variation, and the carrier power weighting positive factor corresponding to the carrier power variation is calculated as shown in the formula (5):
- Equation (5) round D is rounded off the value in parentheses, "for fixed point"
- the bit width here is discrete with respect to power, which is implemented by using a lookup table.
- the granularity and length of the table are determined according to the dynamic range of power of each system carrier and the change of power does not affect the CFR performance index.
- the conventional value is to positively calculate the CPG coefficient by the power variation granularity of 2 dB.
- the input of the table is the currently obtained carrier power value or power gear number. After the determination, the corresponding power of the prototype filter coefficients of each carrier is performed.
- the prototype filter coefficients of the modified carriers are prototype filter coefficients of each carrier, ⁇ is the number of carriers, and ⁇ is the prototype filter order.
- the coefficients need to be normalized, gain compensated, and quantized by the data.
- the current Kn factor is determined according to the sum of the previously obtained factors as the index of the ⁇ factor lookup table.
- the CPG filter coefficient is the CPG filter coefficient
- quan-num is the quantization bit width of the 4th pulse-eliminating filter
- cordic ain is the gain brought by the CORDIC operation method used in the peak cancellation pulse extraction, if the peak cancellation pulse is extracted
- the cordic ain takes a value of 1
- A is the prototype filter order
- N is the length of the prototype filter coefficient
- ⁇ is the offset pulse.
- the normalized gain factor is determined by the total power variation granularity of the carrier.
- the calculated new CPG filter coefficients are downloaded to the CPG filter coefficient RAM area.
- the CPG filter coefficient RAM area is divided into the main and standby tables, and the newly obtained CPG filter coefficients are downloaded into the standby table, and the download is successfully performed. After that, the back table is switched to the main table and the related parameter register values are updated.
- FIG. 4 is a structural block diagram of a device for updating a peak clipping coefficient suitable for frequency hopping and power fluctuation systems according to the present invention.
- the carrier power weighting correction factor in FIG. 4 is a countermeasure for power fluctuation or carrier number change.
- the device updates the CPG filter coefficients when the carrier power changes. Considering that frequently updating the CPG filter coefficients may cause the frequency to glitch, the CPG filter coefficient update that normally accounts for carrier power variation is discrete. In progress, the CPG filter coefficient update enable signal is generated when the power change reaches a certain magnitude. The magnitude of this change in different system applications can be defined according to the needs of the respective system.
- the apparatus of the present invention includes an update control unit 301, a prototype filter coefficient storage unit 302, a cosine signal generation unit 303, a carrier power weighting correction factor Ki generation unit 304, a CPG filter coefficient calculation unit 305, normalization, gain compensation, and quantization processing.
- Unit 306 wherein the specific functions of each unit are as follows:
- the update control unit 301 is configured to generate a CPG filter coefficient update enable signal to the cosine signal generation unit and the carrier power weight correction factor generation unit when the carrier frequency changes or the carrier power relative change amount is greater than the set value.
- the prototype filter coefficient storage unit 302 is configured to store coefficients of the prototype filter of each carrier, and according to the signal system, a set of prototype filter coefficients (single mode system) or multiple sets of prototype filter coefficients (mixed mode system) can be stored.
- the cosine signal generating unit 303 is configured to generate a cosine signal corresponding to each carrier frequency point according to the input frequency control word and the phase control word of each carrier.
- the CPG filter coefficient calculation unit 305 is configured to perform power weight calculation on the prototype filter coefficients of each carrier to obtain a modified prototype filter coefficient of each carrier; and the synchronized cosine signal and the corrected
- the prototype filter coefficients are subjected to time domain multiplication and accumulation processing to obtain CPG filter coefficients; the maximum amplitude value of the modified prototype filter coefficients is required before performing time domain multiplication and accumulation processing on the synchronized cosine signals and the modified prototype filter coefficients. Align with the zero phase value of the cosine signal;
- the CPG filter coefficient calculation unit 305 formulates the time domain multiplication and accumulation process for the synchronized cosine signal and the modified prototype filter coefficient as follows:
- the pulse shaping filter coefficient, ⁇ ( ) is the prototype filter coefficient of each carrier after correction, M is the carrier number, A is the prototype filter order, and N is the length of the prototype filter coefficient, which is the frequency of each carrier. For the signal sampling rate.
- the normalization, gain compensation and quantization processing unit 306 is configured to perform normalization, gain compensation, and quantization processing on the CPG filter coefficients.
- the specific formula is:
- quan_num is to cancel the quantization bit width of the pulse shaping filter
- cordic ⁇ ain is the gain brought by the CORDIC operation mode used in the peak cancellation pulse extraction, if the peak cancellation pulse is extracted.
- the cordic ain takes a value of 1
- A is the prototype filter order
- N is the length of the prototype filter coefficient
- ⁇ is the offset pulse forming normalized gain factor, which is determined by the carrier's total power variation granularity.
- the CPG filter coefficients can be weighted according to the change of the total power, if the influence of the total power variation on the peak clipping performance is not considered, the determined carrier configuration type is determined. , which is a constant, which is a factor after normalizing and quantizing the result of the CPG filter coefficient calculation.
- the update control unit 301 includes a frequency control comparison unit, a carrier power fluctuation comparison unit, and a storage unit for storing carrier power absolute level information, and the storage unit can be configured by the software at initialization, depending on the difference with each system application, and the frequency control is compared.
- the CPG filter coefficient update enable signal is generated when any one of the unit or carrier power fluctuation comparison units is changed.
- the prototype filter coefficient storage unit 302 is a set of RAM storage units. When supporting single-mode applications, a set of quantized prototype filter coefficients stored in one system is stored. When multi-mode system applications are supported, multiple formats are stored. The quantized prototype filter coefficients are stored in blocks of the length of the filter coefficients of the longest order, so as to perform multiplexing processing of the prototype filter coefficients according to the configuration of the carrier type.
- the cosine signal generating unit 303 includes a frequency and phase control word multiplexing unit and an NCO generating unit.
- the frequency and phase control word multiplexing unit performs multiplexing of the multi-carrier frequency point and the phase control word
- the NCO generating unit performs a cosine signal of each carrier required for inputting each carrier frequency control word and phase control word information.
- the frequency control word controls the frequency of the output signal.
- Frequency control word (output frequency / system clock) * ( 2 ⁇ ⁇ ), ⁇ is the number of bits of the frequency control word. This formula can be used for the frequency.
- the control word is set to get a sample of the specific frequency you want.
- the phase of the output cosine sample is accumulated by the frequency control word as a step amount, and the phase control word is used to set the phase of the first output number.
- the carrier power weighting correction factor generating unit 304 includes a storage unit of a carrier absolute power gear position, a power comparison determining unit, and a factor storage unit, and the power value of each carrier is compared with the current carrier absolute power gear storage unit according to the arrival of the CPG update enable signal. Performing comparison to obtain gear position information corresponding to the current carrier power, and corresponding to the search factor storage unit according to the gear position information A carrier power weighting correction factor corresponding to each carrier power fluctuation is obtained.
- the CPG filter coefficient calculation unit 305 includes a weight multiplier unit and a multiply-accumulate unit for calculating CPG filter coefficients, and the weight multiplier unit performs weighting calculation on the prototype filter coefficients of each carrier to obtain a prototype filter of each carrier after correction.
- the multiply and accumulate unit of the CPG filter coefficient is used to perform the time domain multiplication and accumulation process on the cosine signal of the NCO output and the modified prototype filter coefficient to obtain the CPG filter coefficient, and the corrected prototype must be guaranteed before the multiply accumulate
- the maximum amplitude value of the filter coefficients is strictly aligned with the zero phase value of the cosine signal.
- the normalization, gain compensation and quantization processing unit 306 includes a multiplier and a total power weighting factor generation unit that requires normalization processing and CORDIC gain compensation within the peak clipping process.
- the normalization process is performed according to the amplitude normalization.
- the gain compensation is the gain brought by the two CORDIC processes in the original offset pulse of the compensation.
- the normalized gain compensation process is performed according to the following formula: f , Mkr2 cpg - quan * Cordic gain. , 1 ⁇ 1
- the CPG filter coefficient quanium is the quantization bit width of the 4th pulse-eliminating filter
- cordic ain is the gain brought by the CORDIC operation method used for peak cancellation pulse extraction, if the peak cancellation pulse is extracted Without the CORDIC algorithm, the cordic ain takes a value of 1, A is the prototype filter order, and N is the length of the prototype filter coefficients.
- the invention can ensure the peak clipping performance in the frequency hopping and power fluctuation system.
- the standard is equivalent to the static peaking performance index, thus ensuring the performance optimization of the system in various scenarios.
- the spirit and scope of the invention Thus, it is intended that the present invention cover the modifications and the modifications of the invention.
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Transmitters (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Description
适用于跳频及功率波动系统的削峰系数更新方法及装置 技术领域
本发明涉及通信领域, 尤其涉及一种适用于跳频及功率波动系统的削 峰系数更新方法及装置。 背景技术
全球移动通讯系统 ( GSM , Global System for Mobile Communications ) 和码分多址( CDMA, Code Division Multiple Access ) 系统是典型的应用跳 频技术的通信系统。 目前, 基于合路后实现的脉冲抵消削峰算法在资源及 实现结构上有其不可比拟的优势, 已广泛应用于各个芯片与通信系统中。 而该结构存在一个很大的问题是: 抵消脉冲 ( CPG , Cancellation Pulse Generators )成型滤波器系数不能随着频点或载波功率的波动进行实时更新, 这样, 在应用于跳频及功率波动系统中时会出现频谱泄露或削峰性能变差 的问题。
跳频及功率波动系统具有如下特点:
1 )各载波的中心频点随时间按跳频序列发生变化;
2 )各载波的功率随着业务情况变化会发生功率的快速调整;
3 )载波数会随着频点的变化而减少或增加。
现有技术中, 针对上述问题提出了一些解决方案, 但是都不能同时兼 顾频点实时变化及功率波动两方面的应用。 比如:
申请号为 200910192407.8的在先申请提出了一种通过自适应查表法来 解决跳频应用的问题, 该方法是按照已知的跳频序列遍历, 计算出每组频 点的滤波器系数, 根据跳频序列号完成对 CPG滤波器系数表的切换, 这种 方法仅仅能杜绝 CPG滤波器系数随着频点的变化而发生频谱泄露的问题, 在应对功率波动的时候存在一定的缺陷, 缺少因载波功率等级的变化对削
峰性能指标恶化的补偿措施, 这样当功率变化较大时可能会因解调指标随 功率降低严重恶化而降低削峰性能。
申请号为 200810216615.2的在先申请提出了一种基于 GSM系统多载 波功率超配情况下削峰处理的方式, 这种针对多载波系统载波功率超配情 况下的多载波削峰方式, 结合了 GSM系统的时隙级功率控制, 对非 BCCH (广播控制信道) 的削峰根据功率变化进行控制时, 釆用的是一种载波间 非均匀削峰的方式来保证系统在功率超出额定配置时的最佳性能。 该专利 仅仅针对的是在功率超过额定配置时如何保证系统性能的一种削峰装置, 没有解决跳频应用时功率等级快速变化时削峰性能指标会较差的应用问 题。 发明内容
本发明提供一种适用于跳频及功率波动系统的削峰系数更新方法及装 置, 能够使得 CPG滤波器系数随着频点或载波功率的波动进行实时更新。
本发明技术方案包括:
一种适用于跳频及功率波动系统的削峰系数更新方法, 包括步骤: 在任意载波频点发生变化或其载波功率相对变化量大于设定值时产生 抵消脉冲成型滤波器系数更新启动信号;
根据载波功率的变化确定各载波的载波功率加权修正因子值, 以此对 各载波的原型滤波器系数进行功率加权计算, 得到修正后的各载波的原型 滤波器系数;
根据输入的各载波的频率控制字及相位控制字产生各载波频点对应的 余弦信号, 对同步后的余弦信号和修正后的原型滤波器系数进行时域乘累 加处理得到抵消脉冲成型滤波器系数;
对抵消脉冲成型滤波器系数进行归一化、 增益补偿及量化处理后, 以 此对系统当前釆用的抵消脉冲成型滤波器系数予以更新。
进一步地, 所述根据载波功率的变化确定各载波的载波功率加权修正 因子值的公式为:
K. = round(l0(ch-pwr-new '- ch-pwr-old') o x2")
其中, 为载波功率加权修正因子值, ch wr— wevi^为当前获得的载波 平均功率值, c/2 jwr— o½为载波参考功率值,《为定点化的位宽, = 1, ..., Μ , M为载波数。
进一步地, 所述对各载波的原型滤波器系数进行功率加权计算的公式 为:
gi (k) = K^ gi (k) i = l,...,M 其中, 为修正后的各载波的原型滤波器系数, 为各载波的原 型滤波器系数, A为原型滤波器阶数。
进一步地, 所述对同步后的余弦信号和修正后的原型滤波器系数进行 时域乘累加处理前需要将修正后的原型滤波器系数的最大幅度值与余弦信 号的零相位值对 3C 进一步地, 所述对同步后的余弦信号和修正后的原型滤波器系数进行 时域乘累加处理的公式为:
M
h{k) = ^ round{gi{k) * cos(2^-( - [Λ /2 J) (f; + fs/4)/fs ) ) k = 0, 1 . 其中, 为抵消脉冲成型滤波器系数, ^( )为修正后的各载波的原 型滤波器系数, M为载波数, A为原型滤波器阶数, N为原型滤波器系数的 长度, 为各载波的频点, 为信号釆样率。
进一步地, 所述对抵消脉冲成型滤波器系数进行归一化、 增益补偿及 量化处理的公式为:
(k^ * 7 quan - mrdir σηι'η
cpg coef{k) = round{ } ~ - g ) = round(Z^) * Kn) k = 0,l,..., N
/?(LN / 2」 + 1)
其中, 为抵消脉冲成型滤波器系数, quan—num为抵消脉冲成型滤 波器的量化位宽, cordic_gain为峰值抵消脉冲提取时釆用的 CORDIC运算 方式所带来的增益, 若在峰值抵消脉冲提取时未釆用 CORDIC 算法, cw fc 'w取值为 1 , A为原型滤波器阶数, N为原型滤波器系数的长度, Kn为抵消脉冲成型归一化增益因子, 由载波的总功率变化粒度决定。
一种适用于跳频及功率波动系统的削峰系数更新装置, 包括: 更新控制单元, 用于在任意载波频点发生变化或其载波功率相对变化 量大于设定值时产生抵消脉冲成型滤波器系数更新启动信号发送给余弦信 号产生单元和载波功率加权修正因子产生单元;
原型滤波器系数存储单元, 用于存储各载波的原型滤波器的系数; 载波功率加权修正因子产生单元, 用于根据载波功率的变化确定各载 波的载波功率加权^ ί'爹正因子值;
余弦信号产生单元, 用于根据输入的各载波的频率控制字及相位控制 字产生各载波频点对应的余弦信号;
抵消脉冲成型滤波器系数计算单元, 用于对各载波的原型滤波器系数 进行功率加权计算, 得到修正后的各载波的原型滤波器系数; 以及对同步 后的所述余弦信号和所述修正后的原型滤波器系数进行时域乘累加处理得 到抵消脉冲成型滤波器系数;
归一化、 增益补偿及量化处理单元, 用于对抵消脉冲成型滤波器系数 进行归一化、 增益补偿及量化处理。
进一步地, 所述载波功率加权修正因子产生单元根据载波功率的变化 确定各载波的载波功率加权修正因子值的公式为:
Κ. = round(l0(ch-pwr-new '- ch-pwr-old')/20 x2")
其中, 为载波功率加权修正因子值, ch wr— wevi^为当前获得的载波 平均功率值, c/2 wr— o½为载波参考功率值,《为定点化的位宽, J = 1, ..., M ,
M为载波数。
进一步地, 所述抵消脉冲成型滤波器系数计算单元对同步后的余弦信 号和修正后的原型滤波器系数进行时域乘累加处理前需要将修正后的原型 滤波器系数的最大幅度值与余弦信号的零相位值对齐。
进一步地, 所述抵消脉冲成型滤波器系数计算单元对同步后的余弦信 号和修正后的原型滤波器系数进行时域乘累加处理的公式为:
M
h{k) = ^ round{gi{k) * cos(2^-( - [Λ /2 J) (f; + fs/4)/fs ) ) k = 0, 1 . 其中, 为抵消脉冲成型滤波器系数, ^( )为修正后的各载波的原 型滤波器系数, M为载波数, A为原型滤波器阶数, N为原型滤波器系数的 长度, 为各载波的频点, 为信号釆样率。
进一步地, 所述归一化、 增益补偿及量化处理单元对抵消脉冲成型滤 波器系数进行归一化、 增益补偿及量化处理的公式为:
(k^ * 7 quan - mrdir σηι'η
cpg coef{k) = round{ } ~ - g ) = round(Z^) * Kn) k = 0,l,..., N
/?(LN / 2」 + 1) 其中, 为抵消脉冲成型滤波器系数, quan—num为抵消脉冲成型滤 波器的量化位宽, cordic_gain为峰值抵消脉冲提取时釆用的 CORDIC运算 方式所带来的增益, 若在峰值抵消脉冲提取时未釆用 CORDIC 算法, cw fc 'w取值为 1 , A为原型滤波器阶数, N为原型滤波器系数的长度, Kn为抵消脉冲成型归一化增益因子, 由载波的总功率变化粒度决定。
本发明有益效果如下:
本发明提供了一种应用于存在跳频、 载波功率波动以及载波数变化情 况的系统中的削峰系数更新方法及装置, 通过本发明, 保证了削峰性能指 标与静态削峰性能指标相当, 从而使系统应用于不同场景时的性能指标得 到了保证。
附图说明
图 1为现有技术中合载波脉冲抵消削峰装置的实现原理图;
图 2为本发明中由原型滤波器系数 ^( )构建的 CPG滤波器系数 频 谱图;
图 3 为本发明中适用于跳频及功率波动系统的削峰系数更新方法的流 程图;
图 4为本发明中适用于跳频及功率波动系统的削峰系数更新装置的结 构框图。 具体实施方式
下面结合各个附图对本发明的具体实现过程予以进一步详细的说明。 请参阅图 1 ,图 1为现有技术中合载波脉冲抵消削峰装置的实现原理图, 图 1 中的 CPG滤波器 20在频点和载波功率发生变化时需要进行实时的更 新, 目前系统中 CPG 滤波器系数的更新多由中央处理器 (CPU, Central Process Unit ) 完成, 这样很难做到实时性的调整 CPG滤波器系数。
目前, 合载波脉冲抵消削峰装置中 CPG滤波器系数的计算如公式 (1):
M
Kk) = X g(k) *∞ 27T(k― LN/2J) (J + fs /4)/fs ) k = 0,1,2...N - 1
'=ι (1) 其中, 为 CPG滤波器系数, M为载波数, 为原型滤波器系数,
A为原型滤波器阶数, N为原型滤波器系数的长度, 为载波频点, /s为信 号釆样率。 CPG滤波器系数是由原型滤波器和系统的数字频点配置计算得 到的, CPG滤波器系数是对峰值抵消脉冲按照信号频谱特性进行频谱成型 的一组滤波器系数, 是存储在 RAM表中的一组参数, 抵消脉冲经过 CPG 滤波后可以保证信号的临近信道泄漏功率比( ACPR )指标在削峰后满足系 统要求。
请参阅图 2 ,图 2为由原型滤波器系数 构建的 CPG滤波器系数 频谱图, 图 2中, 横坐标为归一化的数字角频(Normalized Frequency ), 纵 坐标为信号频谱抑制度( Magnitude )。这组滤波器系数的构建是与载波的数 字频点相关的, 在跳频系统应用中, 载波频点的变化是实时和快速的, 因 此要求 CPG滤波器系数也要进行实时的更新, 这样才能保证削峰后的抵消 脉冲能量不会由于载波频点的快速变化而泄漏到带外, 从而造成对邻道信 号的干扰。
另一方面, 在系统应用中, 系统的载波数及载波功率会发生快速变化, 当载波功率波动造成载波间功率不平衡时, 削峰后载波功率较低的那个载 波解调指标会恶化, 当载波间功率差达到一定程度时, 会造成功率较低的 载波在削峰后解调指标满足不了系统要求指标。 这时 CPG滤波器系数的计 算需修正为下面的公式 (2 ):
M
h{k) = j round(Ki * gi {k) *€θ 5(2π(Α - J) ( + fs /4)/fs ) ) k = 0, 1, 2 (2) 其中, 为 CPG滤波器系数, round ( )为对括号中的值进行四舍 五入取整处理, M为载波数, A为原型滤波器阶数, N为原型滤波器系数的 长度, W为原型滤波器系数, 为考虑了载波功率波动情况下的载波功 率加权 ^^正因子, 为载波频点, /为信号釆样率。
完成公式(2 ) 的计算只是 CPG滤波器系数计算的一部分工作, 对于 公式 (2 )计算出的结果还需要进行归一化、 增益补偿及量化处理, 这样 CPG滤波器系数才能被硬件逻辑应用。 其中, 归一化是按照 CPG滤波器系 数 /zf^的最大值进行归一化处理的, 量化是按照 quan—num参数定义的位宽 进行的。 如下面公式 (3)描述: cpg
其中, quan—num为 CPG滤波器的量化位宽, cordic _gain为 4氏消脉冲
提取时釆用的 CORDIC运算方式所带来的增益, k为原型滤波器阶数, N 为原型滤波器系数的长度。
上面公式( 3 )中, 若在提取峰值抵消脉冲时未釆用 CORDIC算法计算 时, 将该参数省略或定义为 1即可。
上面公式(3 ) 中归一化过程需要进行除法的运算, 除法的运算对于硬 件逻辑计算来说实现起来比较复杂, 如果除数为一固定的常数则可以将除 法运算转化为乘法运算, 就可以用逻辑运算简单并快速的完成该运算过程 了。 但是这里的 (L /2」 + l)值是随着载波数及载波功率配比的变化而变化 的, 而在硬件中实现除法的运算较为复杂, 使得这一步的计算成为硬件逻 辑实现的瓶颈。 如何寻找一种简单的方法实现 CPG滤波器系数的归一化运 算成为用硬件逻辑完成 CPG滤波器系数更新计算的关键环节, 本发明对这 一问题给出了相应的处理方法。
对于除法运算, 当除数因子是一个变量时, 为将除数的运算对应到乘 法运算, 如果将这个参数的提取方式进行抽象化或能够用已知的简单参量 来表征, 问题就可以得到很好的解决, 下面给出了具体实现方法。
这里, ^是随着 A(LA / 2」+ I)的改变而变化的, 而 A(LA / 2」+ I)与原 型滤波器系数是有对应关系的, 以四载波系统为例进行说明:
(LAV2」+I) =^ /2」+1)*( 式中的 、 κ2、 κ3、 κ4因子表征的是各载波功率变化的载波功率加权修正因子, 与各载波功 率的对应关系为
, 则上式变化为:
,其中的 ¾I 如果与载波的总功率变化^ 存在一定的近似 对应关系问题就会变得比较简单和可实现了。 这样上面的公式就可以近似 等价于下式: N/2」 + l) = g(LN/2」 + l)*( + 2+ 3+ 4)-g (LN/2」 + l)*
V尸额总 (4)
通过构造功率不平衡的用例, 分别按照实际公式计算 Kr 和等价到 总功率后的近似公式计算 发现从计算的结果来看, 两者的理论计算 值比较接近, 如表 1所示:
表 1
通过上面这种映射, 在具体实现上可以很好的简化 CPG系数计算时的 归一化问题, (4)式中只有一个参量 ¾是变量, 这样 ^的变化及修正可以 对应到公式(4) 中的总功率的变化。 根据仿真来看 ^修正因子的加权可 以是非连续的, 因此这里同样可以釆用查找表的方式获得 ^ 因子的加权 值, 表格的索引地址为 的和值, 表格的内容可以根据总功率的变化及修 正粒度来确定。不同系统应用的需求对应的 Kn表格可以根据实际情况来确 定。
下面对照系统中信号处理的流程进行说明。请参阅图 3, 图 3为本发明
中适用于跳频及功率波动系统的削峰系数更新方法的流程图, 具体包括以 下步骤:
首先, 多载波跳频通信系统中一个或多个载波的中心频率会随着时间 按照跳频序列进行变化, 同时载波的功率也会随着业务变化及服务质量而 进行调节, 在跳频及功率波动的多载波通信系统中, CPG滤波器系数更新 是根据功率及频点的变化情况来启动的, 对于每个载波需要分别设置一个 载波参考功率和参考频率的寄存器, 用于与当前系统获得的功率与频点信 息进行比较判断。 ch wr— ο/ί 和 c/zj ^— ο/ί对应参考功率和参考频率寄存 器值, ch _pwr_new ^ ch e— «ew对应当前获得的载波平均功率值和载波 频点信息值。 更新启动信号的产生条件为:
Absich _pwr_old-ch _pwr_new)> A.P ch req_old≠ ch reqjiew 其中 ΔΡ为载波功率相对变化量设定门限值。 上述两个条件满足任一条 件时,即产生 CPG滤波器系数更新启动信号 cpg— update— en ,作为启动 CPG 重新计算及更新的启动控制信号。
在产生了 CPG滤波器系数更新启动信号后, 需要依据当前的载波功率 变化确定载波功率加权爹正因子 Ki, 对应载波功率变化的载波功率加权爹 正因子 的计算如公式 (5)所示:
K. = round{\0{ch-pwr -" eWi-ch-pwr-oldi )l20 2n) ^ 公式 (5)中, round D为对括号中的值进行四舍五入取整处理, 《为定 点化的位宽。 这里 相对于功率的加权是离散的, 是釆用查找表来实现的, 表格的粒度及长度是根据各个系统载波功率的动态范围及功率的变化不影 响 CFR性能指标来确定的。 常规值是以 2dB的功率变化粒度对 CPG系数 进行^ ^正。 表格的输入为当前获得的载波功率值或功率档位号。 确定了 后, 对各路载波的原型滤波器系数进行对应的功率加权计算, 公式如下: gl(k) = K^ gi(k) i = \,...,M
其中, 为修正后的各载波的原型滤波器系数, 为各载波的原 型滤波器系数, Μ为载波数, Α为原型滤波器阶数。
同时, 还需要根据输入的各载波的频率控制字及相位控制字产生各载 波频点对应的余弦信号。
( 6 )
其中, 为 CPG滤波器系数, round D为对括号中的值进行四舍五 入取整处理, M为载波数, ^为修正后的各载波的原型滤波器系数, k 为原型滤波器阶数, N为原型滤波器系数的长度, 为各载波的频点, 为 信号釆样率。 需要注意的是, 对同步后的余弦信号和修正后的原型滤波器 系数进行时域乘累加处理前需要将修正后的原型滤波器系数的最大幅度值 与余弦信号的零相位值对 3C
d. 完成了 CPG滤波器系数的计算后, 需要对该系数进行归一化、 增 益补偿及数据的量化处理, 首先根据前面得到的 因子之和作为 ^ 因子 查找表的索引确定当前的 Kn因子,然后按照下式 (7)进行乘法运算即获得最 终更新后的 CPG系数, 对 ^进行归一化增益加权计算的公式如下: cpg _ coef(k) = round ( H{k) * - " ) = round( z( :) * Kn) k = 0,1,..., N
( 7 )
其中, 为 CPG滤波器系数, quan—num为 4氏消脉冲成型滤波器的 量化位宽, cordic ain为峰值抵消脉冲提取时釆用的 CORDIC运算方式所 带来的增益, 若在峰值抵消脉冲提取时未釆用 CORDIC算法, cordic ain 取值为 1 , A为原型滤波器阶数, N为原型滤波器系数的长度, ^为抵消脉
冲成型归一化增益因子, 由载波的总功率变化粒度决定。
然后, 将计算得到的新 CPG滤波器系数下载到 CPG滤波器系数 RAM 区, 通常 CPG滤波器系数 RAM区是分成主备表的, 将新得到的 CPG滤波 器系数下载到备表中, 成功下载后将背表切换为主表并进行相关参数寄存 器值的更新。
请参阅图 4,图 4为本发明中适用于跳频及功率波动系统的削峰系数更 新装置的结构框图, 图 4中的载波功率加权修正因子 即是针对功率波动 或载波数变化的应对处理, 该装置在载波功率发生了变化时会对 CPG滤波 器系数进行更新 ,考虑到频繁的更新 CPG滤波器系数可能会使频语起毛刺 , 因此通常应对载波功率变化的 CPG滤波器系数更新是离散进行的, 当功率 变化达到一定的量级才会产生 CPG滤波器系数更新启动信号。 在不同系统 应用中这个变化的量值可以根据各自系统的需要来定义。
本发明装置包括更新控制单元 301、 原型滤波器系数存储单元 302、 余 弦信号产生单元 303、 载波功率加权修正因子 Ki产生单元 304、 CPG滤波 器系数计算单元 305、 归一化、 增益补偿及量化处理单元 306, 其中, 各个 单元的具体作用如下:
更新控制单元 301 ,用于在任意载波频点发生变化或其载波功率相对变 化量大于设定值时产生 CPG滤波器系数更新启动信号发送给余弦信号产生 单元和载波功率加权修正因子产生单元。
原型滤波器系数存储单元 302 , 用于存储各载波的原型滤波器的系数 , 根据信号的制式可以存放一组原型滤波器系数(单模系统)或多组原型滤 波器系数(混模系统)。
余弦信号产生单元 303 ,用于根据输入的各载波的频率控制字及相位控 制字产生各载波频点对应的余弦信号。
载波功率加权修正因子产生单元 304,用于根据载波功率的变化确定各 载波的载波功率加权修正因子值, 具体公式为:
K. = round(l0(ch-pwr-new '- ch-pwr-old') o x2") 其中, 为载波功率加权修正因子值, Ch ^?W_«£?H^为当前获得的载波 平均功率值, 为载波参考功率值,《为定点化的位宽, ' = l,...,iki , M为载波数。
CPG滤波器系数计算单元 305 , 用于对各载波的原型滤波器系数进行 功率加权计算, 得到修正后的各载波的原型滤波器系数; 以及对同步后的 所述余弦信号和所述修正后的原型滤波器系数进行时域乘累加处理得到 CPG滤波器系数; 对同步后的余弦信号和修正后的原型滤波器系数进行时 域乘累加处理前需要将修正后的原型滤波器系数的最大幅度值与余弦信号 的零相位值对齐;
CPG滤波器系数计算单元 305对同步后的余弦信号和修正后的原型滤 波器系数进行时域乘累加处理的公式为:
M
h{k) = ^ round{gi{k) * cos(2^-( - [Λ/2 J) (f; + fs/4)/fs ) ) k = 0,1 . 其中, 为抵消脉冲成型滤波器系数, ^ ( )为修正后的各载波的原 型滤波器系数, M为载波数, A为原型滤波器阶数, N为原型滤波器系数的 长度, 为各载波的频点, 为信号釆样率。
归一化、 增益补偿及量化处理单元 306, 用于对 CPG滤波器系数进行 归一化、 增益补偿及量化处理, 具体公式为:
(k^ * 7 quan - mrdir σηι'η
cpg coef{k) = round{ } ~ - g ) = round(Z^) * Kn) k = 0,l,..., N
/?(LN / 2」 + 1)
其中, 为抵消脉冲成型滤波器系数, quan_num为抵消脉冲成型滤波器 的量化位宽, cordic^ain为峰值抵消脉冲提取时釆用的 CORDIC运算方式 所带来的增益,若在峰值抵消脉冲提取时未釆用 CORDIC算法, cordic ain 取值为 1 , A为原型滤波器阶数, N为原型滤波器系数的长度, ^为抵消脉 冲成型归一化增益因子, 由载波的总功率变化粒度决定。
在归一化、 增益补偿及量化处理单元 306 中, 可以根据总功率的变化 对 CPG滤波器系数进行加权 , 若不考虑总功率的变化对削峰性能带来的影 响, 针对确定的载波配置类型, 其为一常数, 该常数是对 CPG滤波器系数 计算的结果进行归一化及量化处理后的一个因子。
下面对上述各个单元的组成及功能予以进一步详细的说明。
更新控制单元 301 包括频率控制比较单元、 载波功率波动比较单元和 用于存储载波功率绝对等级信息的存储单元, 该存储单元可依赖与各系统 应用的差别由软件在初始化时进行配置, 频率控制比较单元或载波功率波 动比较单元任意一个发生了变化即产生 CPG滤波器系数更新启动信号。
原型滤波器系数存储单元 302为一组 RAM存储单元,当支持单模应用 时所存储的是一组一个制式的量化后的原型滤波器系数, 当支持多模系统 应用时存储的是多个制式的量化后的原型滤波器系数, 以阶数最长那个制 式的滤波器系数长度为单位分块存储, 以便于根据载波类型的配置情况进 行原型滤波器系数的复用处理。
余弦信号产生单元 303包括频率和相位控制字复用单元和 NCO产生单 元。 其中频率和相位控制字复用单元完成对多路载波频点和相位控制字的 复用, NCO产生单元完成对输入的各载波频率控制字和相位控制字信息产 生需要的各载波的余弦信号。 频率控制字控制输出信号的频率, 两者之间 的关系为: 频率控制字 = (输出频率 /系统时钟) * ( 2ΛΝ ), Ν为频率控制字 的位数, 通过这个公式可以对频率控制字进行设置, 以得到自己想要的特 定频率的样本。 输出余弦样本的相位以频率控制字为步进量进行累加, 相 位控制字用于设定第一个输出数的相位。
载波功率加权修正因子产生单元 304 包括载波绝对功率档位的存储单 元、 功率比较判决单元和 因子存储单元, 根据 CPG更新启动信号的到来 将各载波的功率值与当前的载波绝对功率档位存储单元进行比较以获得当 前载波功率对应的档位信息, 根据该档位信息对应查找 因子存储单元以
获得对应每个载波功率波动后的载波功率加权修正因子 。
CPG滤波器系数计算单元 305包括 加权乘法器单元和用于计算 CPG 滤波器系数的乘累加单元, 加权乘法器单元完成对各载波原型滤波器系 数的加权计算, 得到修正后的各载波的原型滤波器系数; CPG滤波器系数 的乘累加单元用于完成对 NCO输出的余弦信号与修正后的原型滤波器系数 进行时域乘累加处理得到 CPG滤波器系数, 在乘累加之前必须保证修正后 的原型滤波器系数的最大幅度值与余弦信号的零相位值严格对齐。
归一化、 增益补偿及量化处理单元 306 包括乘法器及总功率加权因子 产生单元, 需要进行归一化处理和削峰处理内部的 CORDIC增益补偿。 归 一化处理是按照幅度归一进行的, 增益补偿是补偿的原始抵消脉冲中两次 CORDIC处理带来的增益, 按照下面的公式进行归一化的增益补偿处理: f ,Mkr2cpg-quan *cordic gain. , 1 ΛΓ 1
cpg coef = round (-^—^ , , ―—— ) = 1,2 „,N + 1
A(LN/2」 + l)
其中, 为 CPG滤波器系数, quan ium为 4氏消脉冲成型滤波器的 量化位宽, cordic ain为峰值抵消脉冲提取时釆用的 CORDIC运算方式所 带来的增益, 若在峰值抵消脉冲提取时未釆用 CORDIC算法, cordic ain 取值为 1, A为原型滤波器阶数, N为原型滤波器系数的长度。
上述变量中, 除了 ^W/^ + i)不是常数外, 其它都是常数, 若定义 c/?g_ a«为 16bit,
.646Ί , «加权因子的量化位宽 N釆用 1 Obit, 当釆用硬件方式实现除法较为复杂, 希望将公式转化为乘法运算, 从而简 化设计, 基于这一想法在实际实现过程中上式等价与下面的计算:
、 ν 2Λ16*2Λ10
COQ coef= round (h(k). *Kn) , 这里 Kn=
FS- J 7 " max(A(n))*1.6…^^
467A2 其中 为上一步计算的系数结果,经过增益补偿、 归一化处理及量化 处理后得到最终存放在 CPG系数 RAM区的系数。
综上所述, 在跳频及功率波动系统中釆用本发明可以保证削峰性能指
标与静态削峰性能指标相当, 从而保证系统在各种场景下的性能优化。 本发明的精神和范围。 这样, 倘若本发明的这些修改和变型属于本发明权 利要求及其等同技术的范围之内, 则本发明也意图包含这些改动和变型在 内。
Claims
1、一种适用于跳频及功率波动系统的削峰系数更新方法,其特征在于, 包括:
在任意载波频点发生变化或其载波功率相对变化量大于设定值时, 产 生抵消脉冲成型滤波器系数更新启动信号;
根据载波功率的变化确定各载波的载波功率加权修正因子值, 并对各 载波的原型滤波器系数进行功率加权计算, 得到修正后的各载波的原型滤 波器系数;
根据输入的各载波的频率控制字及相位控制字产生各载波频点对应的 余弦信号, 对同步后的余弦信号和修正后的原型滤波器系数进行时域乘累 加处理得到抵消脉冲成型滤波器系数;
对抵消脉冲成型滤波器系数进行归一化、 增益补偿及量化处理后, 对 系统当前釆用的抵消脉冲成型滤波器系数予以更新。
2、 如权利要求 1所述的削峰系数更新方法, 其特征在于, 所述根据载 波功率的变化确定各载波的载波功率加权修正因子值的公式为:
K. = round(\ 0{ch-pwr -" ew- - ch-pwr-oldi )l20 x 2n ) , 其中, 为载波功率加权修正 因子值, ch w_«£m¾为当前获得的载波平均功率值, ch j w—oldi为氣波参 考功率值, 《为定点化的位宽, = 1,..., Μ , Μ为载波数。
3、 如权利要求 2所述的削峰系数更新方法, 其特征在于, 所述对各载 波的原型滤波器系数进行功率加权计算的公式为:
gi (k) = Ki * gi (k) i = l,...,M , 其中, ^为修正后的各载波的原型滤波 器系数, 为各载波的原型滤波器系数, A为原型滤波器阶数。
4、 如权利要求 1所述的方法, 其特征在于, 所述对同步后的余弦信号 和修正后的原型滤波器系数进行时域乘累加处理前, 该方法还包括: 将修 正后的原型滤波器系数的最大幅度值与余弦信号的零相位值对
5、 如权利要求 1或 4所述的削峰系数更新方法, 其特征在于, 所述对 同步后的余弦信号和修正后的原型滤波器系数进行时域乘累加处理的公式 为:
M
h{k) = jfound(gi( )*cos(2^ -lN/2])(i].+i 4)/is) ) k = ,\,2...N , 其中, 为抵消脉冲成型滤波器系数, ^( )为修正后的各载波的原型滤 波器系数, Μ为载波数, Α为原型滤波器阶数, N为原型滤波器系数的长度, 为各载波的频点, 为信号釆样率。
6、 如权利要求 1所述的削峰系数更新方法, 其特征在于, 所述对抵消 脉冲成型滤波器系数进行归一化、 增益补偿及量化处理的公式为:
(k^*7 quan - mrdir σηι'η
cpg coef{k) = round{ } ~ — g ) = round(h(k)* Kn) k = 0,1,..., N ,
/?(LN/2」 + 1) 其中, 为抵消脉冲成型滤波器系数, quan_num为抵消脉冲成型滤波器 的量化位宽, cordic^ain为峰值抵消脉冲提取时釆用的 CORDIC运算方式 所带来的增益,若在峰值抵消脉冲提取时未釆用 CORDIC算法, cordic ain 取值为 1, A为原型滤波器阶数, N为原型滤波器系数的长度, ^为抵消脉 冲成型归一化增益因子, 由载波的总功率变化粒度决定。
7、一种适用于跳频及功率波动系统的削峰系数更新装置,其特征在于, 包括:
更新控制单元, 用于在任意载波频点发生变化或其载波功率相对变化 量大于设定值时产生抵消脉冲成型滤波器系数更新启动信号发送给余弦信 号产生单元和载波功率加权修正因子产生单元;
原型滤波器系数存储单元, 用于存储各载波的原型滤波器的系数; 载波功率加权修正因子产生单元, 用于根据载波功率的变化确定各载 波的载波功率加权^ ί'爹正因子值; 余弦信号产生单元, 用于根据输入的各载波的频率控制字及相位控制 字产生各载波频点对应的余弦信号;
抵消脉冲成型滤波器系数计算单元, 用于对各载波的原型滤波器系数 进行功率加权计算, 得到修正后的各载波的原型滤波器系数; 以及对同步 后的所述余弦信号和所述修正后的原型滤波器系数进行时域乘累加处理得 到抵消脉冲成型滤波器系数;
归一化、 增益补偿及量化处理单元, 用于对抵消脉冲成型滤波器系数 进行归一化、 增益补偿及量化处理。
8、 如权利要求 7所述的削峰系数更新装置, 其特征在于, 所述载波功 率加权修正因子产生单元根据载波功率的变化确定各载波的载波功率加权 修正因子值的公式为:
K. = round(l0(ch-pwr-new '- ch-pwr-old')/20 x2")
其中, 为载波功率加权修正因子值, Ch ^?W_«£?H^为当前获得的载波 平均功率值, c z 为载波参考功率值,《为定点化的位宽, J = 1, ..., M ,
M为载波数。
9、 如权利要求 7所述的削峰系数更新装置, 其特征在于, 所述抵消脉 冲成型滤波器系数计算单元, 在对同步后的余弦信号和修正后的原型滤波 器系数进行时域乘累加处理前, 还用于将修正后的原型滤波器系数的最大 幅度值与余弦信号的零相位值对齐。
10、 如权利要求 7或 9所述的削峰系数更新装置, 其特征在于, 所述 抵消脉冲成型滤波器系数计算单元对同步后的余弦信号和修正后的原型滤 波器系数进行时域乘累加处理的公式为:
M
h{k) = ^ round{gi{k) * cos(2^-( - [Λ /2 J) (f; + fs/4)/fs ) ) k = 0, 1 . 其中, 为抵消脉冲成型滤波器系数, ^( )为修正后的各载波的原 型滤波器系数, M为载波数, A为原型滤波器阶数, N为原型滤波器系数的 长度, 为各载波的频点, 为信号釆样率。
11、 如权利要求 7 所述的削峰系数更新装置, 其特征在于, 所述归一 化、 增益补偿及量化处理单元对抵消脉冲成型滤波器系数进行归一化、 增 益补偿及量化处理的公式为:
(k^ * 7 quan - mrdir σηι'η
cpg coef(k) = round{ } ~ — g ) = round(h(k) * Kn) k = 0,1,..., N ,
/?(LN / 2」 + 1) 其中, 为抵消脉冲成型滤波器系数, quan_num为抵消脉冲成型滤波器 的量化位宽, cordic_gain为峰值抵消脉冲提取时釆用的 CORDIC运算方式 所带来的增益,若在峰值抵消脉冲提取时未釆用 CORDIC算法, cordic_gain 取值为 1 , k为原型滤波器阶数, N为原型滤波器系数的长度, Kn为抵消 脉冲成型归一化增益因子, 由载波的总功率变化粒度决定。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201010550673.6A CN101986636B (zh) | 2010-11-19 | 2010-11-19 | 适用于跳频及功率波动系统的削峰系数更新方法及装置 |
CN201010550673.6 | 2010-11-19 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2012065451A1 true WO2012065451A1 (zh) | 2012-05-24 |
Family
ID=43710939
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2011/076534 WO2012065451A1 (zh) | 2010-11-19 | 2011-06-28 | 适用于跳频及功率波动系统的削峰系数更新方法及装置 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN101986636B (zh) |
WO (1) | WO2012065451A1 (zh) |
Families Citing this family (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101986636B (zh) * | 2010-11-19 | 2014-07-16 | 中兴通讯股份有限公司 | 适用于跳频及功率波动系统的削峰系数更新方法及装置 |
CN202276359U (zh) * | 2011-10-20 | 2012-06-13 | 中兴通讯股份有限公司 | 削峰装置及移动通信系统 |
CN102664658B (zh) * | 2012-04-17 | 2014-05-28 | 北京大学 | 一种基于扩频系统的跳频gsm干扰抑制方法及其系统 |
CN103701748B (zh) * | 2013-12-31 | 2017-07-07 | 京信通信系统(中国)有限公司 | 通信系统的削峰方法和系统 |
CN109194307B (zh) * | 2018-08-01 | 2022-05-27 | 南京中感微电子有限公司 | 数据处理方法及系统 |
CN109660277B (zh) * | 2018-12-28 | 2020-06-16 | 北京理工大学 | 一种混合扩频信息生成装置及方法 |
CN109889692B (zh) * | 2019-02-21 | 2021-05-04 | 无锡中感微电子股份有限公司 | 一种电视信号重采样处理方法及装置 |
CN110417702B (zh) * | 2019-07-23 | 2021-06-15 | 三维通信股份有限公司 | 滤波器系数生成方法、系统和降低信号峰均比的系统 |
CN117835118B (zh) * | 2024-03-06 | 2024-05-17 | 深圳市橙子数字科技有限公司 | 一种音频均衡方法及装置 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1538650A (zh) * | 2003-04-16 | 2004-10-20 | 华为技术有限公司 | 一种多载波信号削波装置及方法 |
US20070140376A1 (en) * | 2005-12-21 | 2007-06-21 | Lg-Nortel Co., Ltd. | Apparatus and method for crest factor reduction in a communication system |
CN101651653A (zh) * | 2009-09-17 | 2010-02-17 | 京信通信系统(中国)有限公司 | 多载波跳频通信系统削峰方法 |
CN101848182A (zh) * | 2009-03-25 | 2010-09-29 | 大唐移动通信设备有限公司 | 一种实现自适应削峰的方法和装置 |
CN101986636A (zh) * | 2010-11-19 | 2011-03-16 | 中兴通讯股份有限公司 | 适用于跳频及功率波动系统的削峰系数更新方法及装置 |
-
2010
- 2010-11-19 CN CN201010550673.6A patent/CN101986636B/zh active Active
-
2011
- 2011-06-28 WO PCT/CN2011/076534 patent/WO2012065451A1/zh active Application Filing
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1538650A (zh) * | 2003-04-16 | 2004-10-20 | 华为技术有限公司 | 一种多载波信号削波装置及方法 |
US20070140376A1 (en) * | 2005-12-21 | 2007-06-21 | Lg-Nortel Co., Ltd. | Apparatus and method for crest factor reduction in a communication system |
CN101848182A (zh) * | 2009-03-25 | 2010-09-29 | 大唐移动通信设备有限公司 | 一种实现自适应削峰的方法和装置 |
CN101651653A (zh) * | 2009-09-17 | 2010-02-17 | 京信通信系统(中国)有限公司 | 多载波跳频通信系统削峰方法 |
CN101986636A (zh) * | 2010-11-19 | 2011-03-16 | 中兴通讯股份有限公司 | 适用于跳频及功率波动系统的削峰系数更新方法及装置 |
Also Published As
Publication number | Publication date |
---|---|
CN101986636A (zh) | 2011-03-16 |
CN101986636B (zh) | 2014-07-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2012065451A1 (zh) | 适用于跳频及功率波动系统的削峰系数更新方法及装置 | |
CN102576542B (zh) | 从窄频带信号确定上频带信号的方法和设备 | |
CN103854649B (zh) | 一种变换域的丢帧补偿方法及装置 | |
KR101693280B1 (ko) | 오디오 데이터 처리 방법, 장치 및 시스템 | |
US20070299655A1 (en) | Method, Apparatus and Computer Program Product for Providing Low Frequency Expansion of Speech | |
EP1439524A1 (en) | Audio decoding device, decoding method, and program | |
US20120278068A1 (en) | Voice activity detection method and apparatus, and electronic device | |
US9424850B2 (en) | Method and apparatus for allocating bit in audio signal | |
CN102549659A (zh) | 抑制音频信号中的噪声 | |
EP2664161A1 (en) | Loudness maximization with constrained loudspeaker excursion | |
TW200838176A (en) | Signal-to-noise ratio (SNR) determination in the time domain | |
CN113035207B (zh) | 音频处理方法及装置 | |
CN101202730A (zh) | 在电路中处理信号的方法和系统 | |
KR102201791B1 (ko) | 오디오 신호를 처리하기 위한 방법 및 장치 | |
RU2491656C2 (ru) | Устройство декодирования звукового сигнала и способ регулирования баланса устройства декодирования звукового сигнала | |
US20180096694A1 (en) | Linear predictive coding apparatus, linear predictive decoding apparatus, and methods, programs and recording medium therefor | |
CN102413543B (zh) | 一种lte系统的初始agc方法及设备 | |
CN116259327A (zh) | 一种音频信号自适应均衡方法、系统、设备及存储介质 | |
US12101613B2 (en) | Bass enhancement for loudspeakers | |
JP3522986B2 (ja) | ノイズキャンセラおよびこのノイズキャンセラを使用した通信装置 | |
CN100361401C (zh) | 移动通信终端中的信号噪声比改进方法及其移动通信终端 | |
JP2002076960A (ja) | ノイズ抑制方法及び携帯電話 | |
US20230238009A1 (en) | Speech coding method and apparatus, speech decoding method and apparatus, computer device, and storage medium | |
CN117079657B (zh) | 压限处理方法、装置、电子设备及可读存储介质 | |
US8582700B2 (en) | Reducing switching noise |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11842357 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 11842357 Country of ref document: EP Kind code of ref document: A1 |