JP2007518368A - Ultra wide band scrambler to reduce power spectral density - Google Patents

Abstract

A scrambling method for scrambling ultra-wideband (UWB) data is disclosed. The UWB data (420) having payload data and non-payload data is shifted by shifting the first bit string (412) by the first bit number and shifting the second bit string (410) by the second bit number. Combining the one bit string and the second bit string (411A to 414N), generating (418) scrambler control bits (416) from the shifted and combined first and second bit strings, and these scramblers The UWB data is scrambled by scrambled (424) according to the control bits. According to another aspect, the payload data is scrambled using a pseudo-random number generator having a seed set of seeds with low seed correlation, where each seed in the seed set has a predetermined number of bits, and The UWB data is scrambled by selectively applying random frame inversion (428,430) to the entire non-payload data and / or the entire frame of data.

Description

(Field of Invention)
The present invention relates to an ultra-wide band (ultra-wide band) transmission technology, and more particularly to an ultra-wide band signal scrambling method and an apparatus for reducing power spectral density due to discrete frequency components of an ultra-wide band signal.

(Background of the Invention)
Ultra Wideband (UWB) technology uses a very short duration baseband pulse to spread the energy of the transmitted signal very thinly and widely from near zero to several GHz. UWB technology is currently used in military applications, and techniques for generating UWB signals are well known. The recent decision to permit market research and operation of consumer products with built-in UWB technology, published by the Federal Communications Committee (FCC), will soon enable commercial use.

  The key motivation for FCC decisions to allow commercial use is that UWB transmissions do not require a new communication spectrum, because if the UWB signal is properly configured, This is because it can coexist in the same spectrum as the application signal, and mutual interference can be ignored. However, in order to guarantee negligible mutual interference, the FCC has specified emission limits for UWB applications. For example, the basic requirement of the FCC is that UWB systems do not generate signals that do not interfere with other narrowband communication systems.

Moe et al., “On the Power Spectral Density of Digital Pulse Streams Generated by M-ary Cyclostationary Sequences in the Presence of Stationary Timing Jitter”

  The radiation characteristic of the UWB signal can be determined by examining its power spectral density (PSD). An ideal synchronous data pulse stream based on probability theory is well known. The characterization of “time-hopping spread spectrum” signaling (signal transmission) schemes in the presence of random timing jitter using a stochastic approach is described by Moe et al .: “On the Power Spectral Density of Digital. Pulse Streams Generated by M-ary Cyclostationary Sequences in the Presence of Stationary Timing Jitter ”, IEEE Transactions on Communications, Vol. 46, No. 9, September 1998, pages 1135 to 1145. According to this document, the power spectrum of a UWB signal consists of a continuous component and a discrete component. Generally speaking, the discrete component contributes to PSD more than the continuous component. Thus, discrete components cause more interference to narrowband radio systems than continuous components.

  Data whitening can reduce the presence of discrete components, thereby reducing the PSD of the UWB signal. In a conventional communication system, data whitening, for example, timing recovery and equalization, is generally performed using a scrambler (scrambler). However, the performance of suppressing the PSD of the scrambler is insufficient for UWB. For example, when a block of data is repeated, a strong line spectrum can be generated even if the data is scrambled within the block. This is the case with IEEE 802.15.3a developed by the Institute of Electrical and Electronic Engineers (IEEE). Conventional scramblers have a high pulse repetition frequency (PRF) according to IEEE 802.5.3a, that is, about 100 Mbps to 500 Mbps, and a time division multiple access (TDMA) frame structure of these scramblers. Therefore, it is insufficient to suppress PSD in IEEE 802.5.3a.

  Therefore, there is a need for an improved scrambler to reduce the PSD of UWB signals. The present invention fulfills this need among others.

(Summary of Invention)
The present invention shifts (shifts) a first bit string (bit string) by a first bit number, shifts a second bit string by a second bit number, and shifts the shifted first bit string and second bit string. And an apparatus and method for scrambling UWB data by generating scrambler control bits from the shifted and combined first and second bit strings. At least a part of the UWB data is scrambled according to the generated scrambler control bits.

  According to another aspect of the present invention, a pseudo-random number sequence (random sequence) initialized using a seed selected from a seed set consisting of substantially uncorrelated seed values (an initial value that becomes a seed for generating a random number sequence, etc.) Use to scramble the UWB data by scrambling the payload (net) data. The non-payload (accompanying) portion of the scrambled data is then selectively inverted.

  The present invention is best understood by reading the detailed description of the embodiments with reference to the following drawings. In the drawings, similar components are denoted by the same reference numerals.

ここに、{ａ_n}は、不平衡バイナリ（２進数）の独立、一様分布（ｉ．ｉ．ｄ．：Independent Identically Distributed）の乱数列であり、ｗは、パルス形状及び伝送パワーを決める時間領域パルスである。{ａ_n}は、次式(2)に示す確率関数を有し定常的であるものと仮定する：

(Detailed description of examples)
In order to understand the present invention, it is useful to analyze the PSD of a clocked random number sequence. The digitally controlled signal is assumed to be randomly transmitted in a multiple of the basic clock period _Tc . This signaling technique is modeled as (1) below:

Here, {a _n } is an unbalanced binary (binary number) independent, uniform distribution (iid) random number sequence, and w determines the pulse shape and transmission power. Time domain pulse. {a _n } is assumed to be stationary with the probability function shown in equation (2):

It is known to those skilled in the art that the continuous component S ^c (f) and the discrete component S ^d (f) of the UWB signal can be modeled as shown in the following equations (3a) and (3b), respectively. :

These equations show that PSD is determined by the following three parameters:
W (f)-pulse shape and transmission power;
_Tc -clock period or pulse rate;
· Distribution of p-a _n.
From the above definition, the following conditions are derived:
When p = 0, S ^c (f, 0) = 0 and S ^d (f, 0) = max (S ^d (f));
When p = 1, S ^c (f, 1) = 0 and S ^d (f, 1) = max (S ^d (f)).
(In these cases, S ^d (f) reaches a maximum and all energy goes to discrete components no matter what waveform is used for the pulse.)
When p = 0.5, S ^c (f, 0.5) = max (S ^c (f)) and S ^d (f, 0.5) = 0.
(In this case, S ^c (f) reaches a maximum and all energy goes to a continuous component no matter what waveform is used for the pulse.)
Therefore, if the total PSD is kept constant, the distribution of the random number sequence determines the PSD distribution between the continuous component and the discrete component.

  FIG. 1A shows a power spectrum of one pulse, and FIGS. 1B to 1D show PSDs of clock operation random number sequences each having a different probability distribution. In particular, FIG. 1B shows a PSD with p = 0.25, FIG. 1C shows a PSD with p = 0.5, and FIG. 1D shows a PSD with p = 1.0. These figures show that changing p changes the PSD distribution between the continuous and discrete components, which results in a higher PSD than the continuous component. Apparently, there is only a line spectrum when p = 1.0; there is only a continuous component when p = 0.5; and there are both continuous and discrete components when p = 0.25. To do. These figures confirm the conclusions drawn from equations (3a) and (3b).

Here, the PSD of the time division multiple access (TDMA) system will be described. The IEEE 802.15.3a standard uses an IEEE 802.15.3 media access control (MAC) layer, which is a TDMA system. A line spectrum appears if the pulses in the TDMA frame are not evenly distributed between 1 and -1. Figures 2A and 2B schematically illustrate this problem. FIG. 2A shows the general structure of a TDMA system, where the start points of each frame are separated by T _c . FIG. 2B shows how the bits in the frame form a repetitive pulse train. As shown in FIG. 1C, the line frequency is suppressed only when the pulses are randomly and evenly distributed in the Y direction (ie, between 1 and −1). The repetitive pulse train generates a line spectrum even if the data is scrambled inside the block. Therefore, conventional scramblers are not sufficient to suppress PSD because the Y-direction data is not evenly distributed.

  3A and 3B show the PSD of a TDMA system, where the frame consists of 8 pulses. These figures show that if the pulses are not evenly distributed in the Y direction between 1 and -1, a line spectrum appears even if the pulses are randomly distributed within the frame. FIG. 3A shows the PSD of the original data, and FIG. 3B shows the PSD of the data processed by the prior art method. Assume that the data changes randomly and independently. However, these data are not necessarily evenly distributed in the Y direction, so p is not necessarily equal to 0.5. Since p is not necessarily 0.5, it can be seen that both the data continuous spectrum and the discrete spectrum are generated.

Currently, most of the system proposals for IEEE 802.15.3a use a scrambler with IEEE 802.15.3 MAC and a scrambler with a 15-bit linear feedback shift register (LFSR). A binary (binary number) pseudo random binary sequence (PRBS) is generated. In IEEE 802.15.3, a scrambler is applied to payload data of a data string (sequence) including payload data and non-payload data, and is also applied to some upper layer data. At the beginning of each frame, predetermined values are loaded into the scrambler, and these predetermined values are referred to as initial settings in this specification. For selection as an initial setting, four seeds represented by 2-bit identifiers (b ₁ , b ₀ ) are defined as shown in Table 1 below:

  As shown in Table 1, there is a high correlation between these seed values (ie, only the first two bits of each seed value are unique), and therefore a line spectrum results from the lack of reasonable randomness. obtain. This problem becomes more pronounced for the next generation UWB, which provides a data rate of 4 Gbps.

  Here, the architecture (structure) and method of the scrambler according to the present invention for suppressing PSD in, for example, the IEEE 802.15.3a system will be described. In the preferred embodiment, a two-layer LFSR architecture for payload data is proposed to increase the randomness of the scrambler initialization and thereby increase the randomness in the Y direction.

FIG. 4 shows a preferred two-layer LFSR. The preferred circuit 422 comprises two shift registers 410 and 412, a plurality of exclusive OR (XOR) circuits 414a-414n, a control shift register 416, and an XOR circuit 418. In the preferred embodiment, referring to FIG. 4, the two-layer LFSR operates as follows:
1. At the head of each frame, sign_ctl_org1 (shift register) 410 is shifted right by n1 bits, and sign_ctl_org2 (shift register) 412 is shifted right by n2 bits, where each of n1 and n2 is 1 and the maximum number of shifts, for example, The 15-bit shift register shown in FIG.
2. The XOR circuits 414A to 414N take the exclusive OR (XOR) of the original numbers shifted to the right, that is, sign_ctl_orig = sign_ctl_orig1 (+) sign_ctl_orig2.
3. The sign_ctl_orig is loaded into the sign_ctl_reg (control shift register) 416.
4). Sign_ctl_reg 416 is used to generate sign control bits necessary to scramble payload data.
5). Returning to step 1, the next frame is processed.

  Here, simulation for different initial settings of the scrambler will be described. In these simulations, the payload data is all “1” and there is no non-payload data. The pulse rate is 4 Gbps and the resolution is 100 kHz. Frame sizes of 1024 bytes, 256 bytes, and 64 bytes were used. A frame size of 1024 bytes is required for evaluation in IEEE 802.15.3a, and a frame size of 64 bytes is the minimum frame size in the specification. The simulation results are shown in FIGS. 5A and 5B for a frame size of 1024 bytes, in FIGS. 5C and 5D for a frame size of 256 bytes, and in FIGS. 5E and 5F for a frame size of 64 bytes. 5A, 5C and 5E show the results of LFSR-15 (15-bit LFSR) with the four seeds shown in Table 1, and FIGS. 5B, 5D and 5F have 15 shift bits per layer (ie, 1 The results are shown for a two-layer LFSR (layer LFSR-15). The shift bits for the two-layer LFSR-15 are generated using a 15-bit polynomial generator that generates a binary pseudorandom number sequence (PRBS), and these results indicate that the two-layer LFSR-15 has a PSD of approximately 13 dB each. , 26dB and 34dB are reduced.

  Alternatively, the random number sequence can be generated from collected physical noise (noise) sequences, such as shot noise, Johnson noise, 1 / F noise, sea ambient noise, and photon noise. Other noise sequences can also be used. These noise sequences can be amplified and A / D converted. A bit stream from this digitized sequence can be used as a random number sequence. In order to synchronize the receiver and transmitter, it may be desirable to transmit this random number sequence as a preamble to a message scrambled with the random number sequence.

  The method described above provides an improvement in line spectrum suppression. For this improvement, the scrambler in the transmitter and the scrambler in the receiver are required to be synchronized.

Note that the scrambler seed selection mechanism in IEEE 802.15.3 shown in Table 1 generates highly correlated seeds. In this implementation, LFSR-15 has (2 ¹⁵ -1) states. These seeds can be defined as seed 4 = state (n); seed 2 = state (n + 1); and seed 1 = state (n + 2). Note that there may be an excess of state overlap between these three seeds. For example, if one frame of 1024 bytes uses 2 ¹³ states, 4 frames have 2 ¹³ (seed 3) + (2 ¹³ +2) = 2 ¹⁴ +2 <(2 ¹⁵ −1) or less states. Use. Therefore, this scrambler cannot provide sufficient randomness for all applications.

  This finding suggests that other seed sets that are independent of each other provide better randomness in the scrambler setting than the prior art.

  A preferred seed set with seeds that are nearly uncorrelated with each other is shown in Table 2 below:

These four new seeds can be described as:
Seed 1 = state (n)
Seed 2 = state (mod (n + 1 × 2 ^15-2 , 2 ¹⁵ ))
Seed 3 = state (mod (n + 2 × 2 ^15-2 , 2 ¹⁵ ))
Seed 4 = state (mod (n + 3 × 2 ^15-2 , 2 ¹⁵ ))
With these seeds, there is no state overlap between the four seeds. One frame of 1 Kbyte has 2 ¹³ states, and 4 frames use 4 × 2 ¹³ = 2 ¹⁵ > 2 ¹⁴ +2 states. This analysis shows that the specification of the four uncorrelated seeds increases the randomness of the scramble.

  Here, simulation will be described. All the payload data for the scrambler is “1”, and there is no non-payload data. The pulse rate is 4 Gbps and the resolution is 100 kHz. A frame size of 256 bytes was used. The simulation results are shown in FIGS. 6A and 6B, FIG. 6A shows the results of LFSR-15 with the four correlated seeds shown in Table 1, and FIG. 6B shows the LFSR− with the four uncorrelated seeds shown in Table 2. 15 results are shown. These simulation results show that uncorrelated seeds reduce PSD by about 10 dB.

  As shown in FIG. 6B, although improved, spectral lines still exist. These lines are associated with a relatively small number of seeds. To further reduce the spectral lines, we propose random frame inversion. Other reasons for introducing random frame inversion are explained in the next paragraph.

Since IEEE 802.15.3a uses IEEE 802.15.3 MAC, the maximum frame length is 2K bytes. Assuming that the UWB data rate is 4 Gbps, the minimum frame rate is:
4G / (2K × 8) = 256K
T _f = 4 × 10 ^-6 (4)
Since T _f << 1, non-payload data (eg, sync words) can generate strong spectral lines when viewed at higher resolution. Table 2 lists some values that non-payload data contributes to PSD under the assumption of T _f = 10 ⁻³ , where p _{s +} is the ratio of non-payload data in the frame. This table shows that non-payload data constitutes a small part of the frame, but its contribution to the PSD cannot be ignored due to the high pulse repetition frequency (PRF) of the frame. If a smaller frame length is used in harsh environments to reduce the frame error rate, the same ratio of non-payload data will produce a stronger spectral line, and a PSD larger than those listed in Table 3 below. Is generated.

ここに、ｌ及びｋはそれぞれ、フレーム、及びフレーム内のパルスの指標（インデックス）値である。 In order to further reduce the PSD of the new scrambler and reduce the PSD generated by non-payload data, we propose the following Random Frame Reversion (RFR) method. A framed data stream can be represented as shown in equation (5) below:

Here, l and k are an index value of a frame and a pulse in the frame, respectively.

・そして上記データ列に次式の演算を適用して、新たな列を生成する：

・{ｃ_l,k}は新たなデータ列として伝送される。 RFR can reduce spectral lines generated by non-scrambled non-payload data. In the preferred embodiment, the RFR is implemented as follows:
Generate a random number sequence {b _n } with the uniform distribution function:

And then apply the following formula to the above data sequence to create a new sequence:

{ _{Cl, k} } is transmitted as a new data string.

  A suitable circuit for performing random frame inversion is shown in FIG. 4B. In this figure, payload data is scrambled in circuit 424 in accordance with a scramble sequence supplied by a scramble generator, eg, scramble generator 422 described above. Non-payload data is inserted into the frame by circuit 426 after the payload data is scrambled, and is therefore not scrambled. The output signal of the circuit 426 is supplied to a selection type inverter (logic inverter) 428 that responds to the random number bit generation period 430. Inverter 428 either passes the frame data as received or inverts the frame data in response to a signal from random number generator 430. This circuit can be used to selectively invert the entire frame according to the random number bits supplied by the random number generator 430 or to pass the entire frame without inversion.

  7A and 7B show the simulation results. The pulse rate is 4Gbps and the resolution is 100kHz. A frame size of 256 bytes was used. All payload data for the scrambler is “1”. FIG. 7A shows the PSD of the data processed without RFR by the new seed LFSR-15, and FIG. 7B shows the PSD of the data processed with RFR by the new seed LFSR-15. This result shows that RFR reduces PSD by about 10 dB.

In a preferred alternative, the following method is proposed:
1. Different (ie, more random) seed values for seed selection;
2. Selective random frame inversion (RFR) for at least non-payload data.

  There are four seed values in Table 2, each seed value including 15 bits. These seed values are almost uncorrelated, and therefore the pseudo-random sequence generated using these seed values is almost uncorrelated. The seed sets shown in Table 2 are exemplary only and employ seed sets with seeds with different seed values, seed sets with more or fewer seeds, and seed sets with more or fewer bits per seed. can do. A method for generating an appropriate uncorrelated seed value for use in a seed set will be apparent to those skilled in the art from the description herein. Simulations with seed sets having different numbers of seeds were performed using the same configuration as described above. These simulation results are shown in FIGS. 8A, 8B, 8C and 8D to compare the 8 seed random seed set with the 16 seed random seed set. 8A and 8B show the results for the 8-seed and 16-seed seed sets, respectively, without RFR, and FIGS. 8C and 8D show the results for the 8-seed and 16-seed seed sets, respectively, with RFR. The seeds for simulation shown in FIGS. 8A to 8D are generated by using the rand () function of MATLAB (registered trademark).

This simulation shows the following:
• The new method significantly reduces the PSD of UWB signals;
Without RFR, the 16 seed set is about 5 dB better than the 8 seed set, and the 8 seed set is about 1 dB better than the 4 seed set;
With RFR, the 16 seed set is about 1 dB better than the 8 seed set, and the 8 seed set is about 3 dB better than the 4 seed set. These three seed sets show very similar performance.

Therefore, in conclusion, a simple and effective design of a scrambler for line spectrum control can be as follows:
1. Use seed selection from 4 seed (or more) seed sets with low seed correlation (ie random);
2. Selective RFR is used for non-payload data.

  The above description provides an analysis of the effect of scrambling on the PSD of UWB signals in the IEEE 802.15.3a system and proposes a new whitening method including PSD. Simulations show that the current scrambler proposed in IEEE 802.15.3a is insufficient and that the new whitening method can improve the performance of suppressing PSD. This new method is easy to implement and does not require a MAC layer change. This method can be used to develop an IEEE 802.15.3a scrambler.

  While the components of the present invention have been described with reference to particular components, it is contemplated that one or more components can be implemented with software running on a general purpose computer. In such embodiments, one or more of the functions of the various components can be implemented with software that controls a general purpose computer. The software can be embodied on a computer readable carrier, such as a magnetic or optical disk, a memory card, or an audible frequency, radio frequency, or optical carrier.

  In addition, although the invention is illustrated and described herein with reference to specific embodiments, the invention is not limited to these details. Rather, various modifications can be made to these details without departing from the scope of the invention without departing from the scope of the claims. For example, while the present invention has been described with reference to reducing PSD of UWB signals under IEEE 802.15.3a, it is conceivable to reduce PSD using the present invention in other communication systems and standards. It is done.

2 is a graph showing frequency versus power spectral density (PSD) for one pulse. Figure 6 is a graph showing frequency versus PSD for a random sequence probability of 0.25. Figure 6 is a graph showing frequency versus PSD for a random sequence probability of 0.5. FIG. 6 is a graph showing frequency versus PSD for a random sequence probability of one. 1 shows a general structure of a TDMA system. It is a graph which shows the data in the flame | frame about the continuous flame | frame in a TDMA system. FIG. 6 is a diagram showing a PSD of a frame having a number of original data pulses. FIG. 6 is a diagram showing a PSD of a frame having a large number of pulses processed according to the prior art. 2 is a block diagram of a two-layer linear feedback shift register (LFSR) used in a scrambler according to one embodiment of the present invention. FIG. 1 is a block diagram of a random frame inversion circuit according to a preferred embodiment of the present invention. FIG. FIG. 6 is a diagram showing an initial PSD of a scrambler for four seeds according to the prior art when used with data having a frame size of 1024 bytes. FIG. 5B is a diagram showing an initial PSD of a scrambler in comparison with FIG. 5A for a two-layer LFSR according to the present invention when used with data having a frame size of 1024 bytes. FIG. 7 is a diagram showing an initial PSD of a scrambler for 4 seeds according to the prior art when used with data having a frame size of 256 bytes. FIG. 5C is a diagram showing a scrambler default PSD compared to FIG. 5C for a two-layer LFSR according to the present invention when used with data having a frame size of 256 bytes. FIG. 7 is a diagram showing a PSD of default scrambler settings for four seeds according to the prior art when used with data having a frame size of 64 bytes. FIG. 5B is a diagram showing a scrambler default PSD compared to FIG. 5E for a two-layer LFSR according to the present invention when used with data having a frame size of 64 bytes. FIG. 7 shows a scrambler PSD for a 15-bit two-layer LFSR (LFSR-15) according to the present invention when used with four correlated seeds. FIG. 6D shows the scrambler PSD for a 15-bit two-layer LFSR (LFSR-15) according to the present invention compared to FIG. 6A when used with four uncorrelated seeds. FIG. 6 shows a PSD of a two-layer LFSR-15 according to the present invention used with four uncorrelated seeds without random frame inversion (RFR). FIG. 7B shows a PSD of a two-layer LFSR-15 according to the present invention compared to FIG. 7A, used with four uncorrelated seeds with random frame inversion (RFR). FIG. 2 shows a PSD of a prior art scrambler used with 8 uncorrelated seeds without RFR. 8D shows a PSD of a two-layer LFSR-15 according to the present invention, compared to FIG. 8A, used with 16 uncorrelated seeds without RFR. FIG. FIG. 6 shows a PSD of a prior art scrambler used with 8 uncorrelated seeds with RFR. 8D shows a PSD of a two-layer LFSR-15 according to the present invention, compared to FIG. 8C, used with 16 uncorrelated seeds with RFR. FIG.

Claims (21)

In a scramble method for scrambling ultra-wideband (UWB) data:
Shifting the first bit string by a first number of bits;
Shifting the second bit string by a second number of bits;
Combining the shifted first bit string and the second bit string;
Generating scrambler control bits from the shifted and combined first and second bit strings;
A scramble method for UWB data, comprising: scrambling at least a part of the UWB data in accordance with the generated scrambler control bits.   The method of claim 1, further comprising the step of randomly initializing the first bit string and the second bit string.   The method of claim 2, wherein the first bit string and the second bit string are randomly initialized using a pseudo-random number sequence.   The method of claim 2, wherein the first bit string and the second bit string are randomly initialized using a random number sequence.   The method according to claim 1, wherein a scrambling method for scrambling the UWB data is applied for each frame.   The method of claim 1, wherein the UWB data includes payload data and non-payload data, and the scrambling step scrambles the payload data. The method further includes:
The method of claim 6, comprising selectively applying random frame inversion to the non-payload data. The step of selectively applying random frame inversion to the non-payload data includes:
A sub-step of generating a pseudo-random number sequence by a uniform distribution function;
The method according to claim 7, further comprising a substep of selectively inverting the data sequence in accordance with the pseudo-random number sequence. In a scramble method for scrambling ultra-wideband (UWB) data with payload data and non-payload data:
Scrambling the payload data using a pseudo-random sequence configured for initialization using a seed set of substantially uncorrelated seed values;
And selectively applying random frame inversion to the non-payload data. A method for scrambling UWB data. The step of selectively applying random frame inversion to the non-payload data includes:
A sub-step of generating a random number sequence with a uniform distribution function;
The method according to claim 9, further comprising a substep of selectively inverting the data sequence in accordance with the random number sequence. In a scrambler that scrambles UWB data:
A first shift register that shifts the first bit string by a first number of bits;
A second shift register for shifting the second bit string by a second number of bits;
A coupling circuit coupling the shifted first bit string and the second bit string;
A third shift register for loading the shifted and combined first and second bit strings;
A control circuit for generating scrambler control bits from the shifted first and second bit strings and scrambles at least a part of the UWB data according to the generated scrambler control bits. A scrambler characterized by that.   The scrambler according to claim 11, further comprising a polynomial generator for generating a pseudo-random number sequence for initializing the first bit string and the second bit string.   The scrambler according to claim 12, wherein the random number sequence generated by the polynomial generator is 15 bits or more.   The scrambler according to claim 11, wherein the first shift register and the second shift register are initialized according to a new frame. Each frame of the UWB data includes payload data and non-payload data, and the scrambler further includes:
15. The scrambler according to claim 14, further comprising a selective random frame inversion circuit that selectively inverts at least non-payload data of each scrambled frame. The selective random frame inversion circuit includes:
A random number generator for generating a pseudo-random number sequence by a uniform distribution function;
The scrambler according to claim 15, further comprising an inverter that selectively inverts a data sequence in accordance with the pseudo-random number sequence. In a scrambler system for data whitening to reduce the power spectral density (PSD) of an ultra-wideband (UWB) signal with payload data and non-payload data:
A scrambler configured to scramble the payload data, the scrambler comprising a linear feedback shift register configured for initialization using a seed set of substantially uncorrelated seed values;
A scrambler system comprising a selective random frame inversion circuit configured to selectively invert the non-payload data.   The scrambler system, wherein the seed set includes at least four seed values, and the seed value in the seed set has at least 16 bits. The selective random frame inversion circuit includes:
A random number generator for generating a pseudo-random number sequence by a uniform distribution function;
An scrambler system comprising an inverter that selectively inverts a data sequence in accordance with the pseudo-random number sequence. In a computer readable carrier including software configured to control a computer to implement a scrambling method for scrambling UWB data, the scrambling method includes:
Shifting the first bit string by a first number of bits;
Shifting the second bit string by a second number of bits;
Combining the shifted first bit string and the second bit string;
Generating scrambler control bits from the shifted and combined first and second bit strings;
Scramble at least a portion of the UWB data in response to the generated scrambler control bits. In a computer readable carrier including software configured to control a computer to implement a scrambling method for scrambling UWB data having payload data and non-payload data, the scrambling method includes:
Scrambling the payload data using a pseudo-random sequence configured for initialization using a seed set of substantially uncorrelated seed values;
Selectively applying random frame inversion to the non-payload data,
The random frame inversion steps include:
A sub-step of generating a random number sequence with a uniform distribution function;
A sub-step of selectively inverting the data sequence according to the random number sequence;
A computer-readable carrier comprising: a sub-step for transmitting the data sequence selectively inverted.

Images