WO2006073371A1

WO2006073371A1 - Method and apparatus for pulse mapping in impulse radio modulation

Info

Publication number: WO2006073371A1
Application number: PCT/SG2005/000001
Authority: WO
Inventors: Yew Soo Eng; Zhan Yu
Original assignee: Matsushita Electric Industrial Co., Ltd.
Priority date: 2005-01-03
Filing date: 2005-01-03
Publication date: 2006-07-13

Abstract

A method and apparatus for constructing a pulse mapping table for Impulse radio pulses modulated with respect to n pulse parameters, n≥1, are disclosed. The method comprises identifying all possible pulses for a given modulation scheme; and assigning m binary bits, m≥2, to each pulse in a manner such that a Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is larger than 1.

Description

Method and Apparatus for Pulse Mapping in impulse Radio

Modulation

Field of the invention

The present invention relates broadly to a method of constructing a pulse mapping table for impulse radio pulses, to an impulse radio transmitter, and to an impulse radio receiver.

Background

Recent advances in communications technology, such as the availability of highspeed switching semiconductor devices, have enabled the use of transmitting and receiving a sequence of short-duration radio frequency (RF) pulses, where the pulse duration is typically less than a nanosecond. This is often referred to as Impulse Radio (IR) communications. The IR technology has been proposed to some communication systems, such as IR ultra-wideband (UWB) system and IR millimeter wave system.

Using short-duration RF pulses, the IR technology can provide signal transmission over an ultra wide frequency bandwidth. The average power spectral densities, depending on the pulse repetition frequency and pulse amplitude levels, are in a low region, for example, of 10^"11 watts per Hertz. This low power emission minimizes interference with other wired or wireless systems operating in the same frequency band. Many advantages can be found in IR technology, like the large bandwidth can offer high capacity communication at short distances and also can provide rich multipath diversity to be collected with low-complexity RAKE reception.

There are various data modulation techniques for IR technology, including pulse- polarity modulation, pulse-amplitude modulation (PAM), phase modulation, frequency modulation, pulse-position modulation (PPM) (also referred to as time-shift or pulse-interval modulation) and M-ary versions of these. Another implementation to data modulation is known as transmitted-reference (TR) scheme. A typical TR scheme is described in US Patent Application Publication No. 2001/0053175 A1, "Ultra-wideband Communications System". In the TR scheme, there is a reference pulse transmitted before each data pulse. The reference pulse has a fixed amplitude level or time position (or called pulse position). The modulation can be carried out on the polarity, amplitude or time position of data pulse relative to the reference pulse. For example, a TR pulse interval amplitude modulation (PIAM) has been proposed to modulate both polarity and time position of data pulse relative to the reference pulse, as e.g. described in T. Zasowski, F. Althaus, and A. Wittneben, "An energy efficient transmitted-reference scheme for ultra-wideband communications," International Workshop on Ultra Wideband Systems (IWUWBS) Joint with IEEE Conference on Ultra Wideband Systems and Technologies (UWBST), pp. 146-150, Japan, 18-21 May 2004.

For a modulation with more than one bit per symbol, the mapping technique becomes essential. The conventional Gray mapping technique, described e.g. in Nils Rinaldi, Sven Zeisberg, Manuel Pezzin, Lucille Rouault, Axel Schmidt, Holger Hosel, Rainer Moorfeld, Sebastien de Rivaz, Benoϊt Denis, "U.CAN.'s Ultra Wide Band System: Baseband Algorithm Design," IWUWBS, Finland, 2-5 June 2003, is designed to have only one digital bit changing between a desired pulse and its neighboring pulses. The neighboring pulse is defined as the pulse that has the smallest difference in amplitude, pulse position, frequency or phase from the desired pulse. For example, the neighboring pulse in PPM has the smallest time position difference from the desired pulse. If a 4-PPM has four pulse positions t₀, t_1} t₂ and t₃ in the modulation, where t_o < t₁ < t₂ < t₃, a desired pulse at t₂ has two neighboring pulses at t_1t and J₃. Due to the channel distortion or interference, detection errors may occur at the receiver by detecting other pulses instead of the desired pulse. The neighboring pulses of the desired pulse have the largest probability to be selected at the detection errors.

Forward Error Correction (FEC) codes are commonly used to correct such detection errors. The error control coding is an important technique in digital communications, especially wireless digital communications. Error control for a one-way communication system must be accomplished using FEC codes, that is, by employing error-correcting codes that automatically correct errors detected at the receiver.

However, the conventional Gray mapping does not explore the error correction ability of FEC codes, since the Hamming distance between the desired pulse and its neighboring pulses is only one. A larger Hamming distance can support FEC codes to correct more detection errors at the receiver. In the TR-PIAM, although the Hamming distance between the neighboring pulse with opposite polarity is maximized, the Gray mapping is still used between the neighboring pulses with different time positions. Hence, the TR-PIAM cannot provide large Hamming distance when the detection errors occurred by time position distortion, which commonly happen in dense multipath or interference environments.

In the literature, Quadrature Amplitude Modulation (QAM) and Trellis-Coded Modulation (TCM) can be found to maximize the Hamming distance and Euclidean distance between neighboring symbols. However, both modulations are designed for bandwidth-constrained communications, which are quite different from IR communications. IR communications use ultra short-duration RF pulses and can be treated as time-constrained communications. Therefore, there exists a need to have a better pulse mapping technique than Gray mapping to maximize the Hamming distance between neighboring pulses in IR communications.

Summary

In accordance with a first aspect of the present invention there is provided a method of constructing a pulse mapping table for impulse radio pulses modulated with respect to n pulse parameters, /?>1 , the method comprising identifying all possible pulses for a given modulation scheme; and assigning m binary bits, m≥2, to each pulse in a manner such that a Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is larger than 1.

The assigning of the m binary bits may comprise arranging all possible pulses into a /7-dimensional matrix structure; partitioning the pulses in the n-dimensional matrix structure into sets; reordering the pulses from the sets based on respective rules associated with the respective sets into a series; and assigning m binary bits to each pulse in the series. The assigning of the m binary bits to each pulse in the series from the first pulse to the last pulse in the series may be in ascending order following the sequential increment of the binary number formed by the m binary bits.

The assigning of the m binary bits to each pulse in the series from the first pulse to the last pulse in the series may be in descending order following the sequential decrement of the binary number formed by the m binary bits.

The n pulse parameter may comprise one or a combination of any two or more of a group consisting of amplitude, time position, frequency and phase of the impulse radio pulses.

The number of all possible pulses may be the product of the number of different pulses based on one pulse parameter and the number of different pulses based on another pulse parameter.

A Transmitted Reference (TR) scheme may be employed, and the pulse mapping table may be applied to the modulation of a data pulse with reference to a reference pulse.

The pulse mapping table may function as an adaptive mapping table comprising a full bit rate mapping table portion and at least one lower bit rate mapping portions.

The Hamming distance between neighbouring pulses with respect to each of the n pulse parameters may be at least (m - 1).

In accordance with a second aspect of the present invention there is provided an impulse radio transmitter for transmitting impulse radio pulses modulated with respect to n pulse parameter, n≥λ , the transmitter comprising a data base having stored therein a pulse mapping table; and an impulse radio modulator utilising the pulse mapping table in modulating the impulse radio pulses; wherein the pulse mapping table is constructed by identifying all possible pulses for a given modulation scheme, assigning m binary bits, m≥2, to each pulse in a manner such that a Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is larger than 1. The impulse radio transmitter may further comprise a channel encoder for encoding a binary data input stream prior to the impulse radio Modulator.

The channel encoder may apply an FEC code to the binary data input stream.

The impulse radio transmitter may further comprise an interleaver for spreading the binary data input stream prior to the impulse radio modulator.

The impulse radio transmitter may further comprise a pulse generator for generating the impulse radio pulses based on an output from the impulse radio modulator.

The pulse generator may generate the impulse radio pulses further based on an output from a code generator for encoding.

The assigning of the m binary bits may comprise arranging all possible pulses into a /7-dimensional matrix structure; partitioning the pulses in the n-dimensional matrix structure into sets; reordering the pulses from the sets based on respective rules associated with the respective sets into a series; and assigning m binary bits to each pulse in the series.

The assigning of the m binary bits to each pulse in the series from the first pulse to the last pulse in the series may be in ascending order following the sequential increment of the binary number formed by the m binary bits.

The n pulse parameter may comprise one or any combination of two or more of a group consisting of amplitude, time position, frequency and phase of the impulse radio pulses. The number of all possible pulses may be the product of the number of different pulses based on one pulse parameter and the number of different pulses based on another pulse parameter.

A transmitted reference (TR) scheme may be employed, and the impulse radio modulator may apply the pulse mapping table to the modulation of a data pulse with reference to a reference pulse.

The pulse mapping table may function as an adaptive mapping table comprising a full bit rate mapping portion and at least one lower bit rate mapping portion.

In accordance with a third aspect of the present invention there is provided a impulse radio receiver for receiving impulse radio pulses modulated with respect to n pulse parameter, n≥λ, the transmitter comprising a data base having stored therein a pulse mapping table; and an impulse radio demodulator utilising the pulse mapping table in demodulating the impulse radio pulses; wherein the pulse mapping table is constructed by identifying all possible pulses for a given modulation scheme, assigning m binary bits, m≥2, to each pulse in a manner such that a Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is larger than 1.

Brief Description of the Drawings

Embodiments of the invention will be better understood and readily apparent to one of ordinary skill in the art from the following written description, by way of example only, and in conjunction with the drawings, in which:

Figure 1 is an exemplary block diagram of an IR transmitter comprising an

IR modulator with pulse mapping table according to an example embodiment.

Figure 2a illustrates all four possible pulses of a representative 4-PPM according to an example embodiment. Figure 2b illustrates all four possible pulses of a representative 4-PAM according to an example embodiment.

Figure 3 illustrates all eight possible pulses of a representative combination of 4-

PPM and 2-PAM according to an example embodiment.

Figure 4 illustrates a representative mapping table for a combination of 4-PPM and 2-PAM according to an example embodiment.

Figure 5 is an exemplary flow chart for the pulse mapping of a combination of M- ary PPM and Λ/-ary PAM according to an example embodiment.

Figure 6 illustrates the partition of pulses in a representative combination of 8- PPM and 4-PAM to two sets with two rows and eight columns according to an example embodiment.

Figure 7a illustrates a representative bit assignment of pulses at subset 634 with two rows and eight columns according to an example embodiment.

Figure 7b illustrates a representative bit assignment of pulses at subset 636 with two rows and eight columns according to an example embodiment.

Figure 8 is an exemplary flow chart for the pulse mapping of /W-ary PPM-only according to an example embodiment.

Figure 9 illustrates a representative bit assignment of pulses for 16-PPM according to an example embodiment.

Detailed Description

The example embodiments described provide a method and apparatus for a pulse mapping technique in IR communications such as UWB-based wireless or wired communications and millimeter-wave based wireless or wired communications. In IR communications, information bits are carried by the modulated amplitude, time position, polarity, frequency or phase of the pulses. The example embodiments of the present invention can maximize the Hamming distance between all neighboring pulses in order to minimize detection errors at the receiver. The discussion on the embodiments of the present invention as follows will make apparent that the pulse mapping technique adopted may be simpler and faster than the conventional QAM and TCM, and is suitable for time-constrained communications.

For a modulation scheme with more than one bit per symbol, the mapping issue becomes essential. Due to the channel distortions and interference, a problem of detection errors may occur at the receiver due to detection of other pulses instead of the desired pulse. Within said other pulses, the neighboring pulses of the desired pulse have the largest probability to be selected. FEC codes are commonly used to correct such detection errors.

The example embodiments of the present invention explore the error correction ability of FEC codes by maximizing the Hamming distance between neighboring pulses. The embodiments to be discussed later present a pulse mapping method of grouping the neighboring pulses by set partitioning and assigning digital bits to the pulses in each group. This pulse mapping method in one embodiment of the present invention comprises a combination of /W-ary PPM and Λ/-ary PAM, and in another embodiment comprises /W-ary PPM-only or Λ/-ary PAM-only.

The example embodiments of the present invention can maximize the Hamming distance between neighboring pulses. As a result, the larger Hamming distance between neighboring pulses can improve the error correction ability of FEC codes, thus more detection errors at the receiver due to channel distortions and interference can be corrected. The overall positive effect on the IR system is a smaller bit error rate.

Figure 1 shows an exemplary block diagram of an IR transmitter comprising an IR modulator with a pulse mapping table 100. In a first example embodiment of the present invention, a pulse mapping table 108 is constructed for a combination of M-ary PPM and N- ary PAM. In a second example embodiment of the present invention, a pulse mapping table 108 is constructed for /W-ary PPM or Λ/-ary PAM. Both pulse mapping methods in the example embodiments are designed to maximize the Hamming distance between neighboring pulses.

Figure 1 comprises a channel encoder 104, an interleaver 106, an IR modulator with pulse mapping table 100, pulse generator 112, code generator 118 and transmitting antenna 116. The binary data stream 102 represents input messages. The messages may include any one or combination of text, video, images, audio, etc.

A binary data stream 102 is sent to the channel encoder 104. All kinds of FEC codes can be adopted in the channel encoder 104, for example block codes and convolutional codes. The interleaver 106 may be utilised after the channel encoder 104 to spread the coded bits, so that burst errors within a code word appear to be independent at the receiver. The interleaver 106 is optional and can be removed from the IR transmitter 120 in different embodiment.

The IR modulator with pulse mapping table 100 comprises an IR modulator 110 and a pulse mapping table 108. PPM and PAM are two common modulation schemes for IR systems.

In PPM, the time position of pulses is modulated based on the input bits 107. The order of M-ary PPM, M, is the number of possible time positions for modulation. M should be an integer power of 2, for example 2, 4, 8, ... etc. Figure 2a illustrates all four possible pulses of a representative 4-PPM with M = 4. As shown in Figure 2a, the pulses 200, 202, 204 and 206 have different time positions, t₀ + T1 , t₀ + T2, t₀ + T3 and t₀ + T4 respectively. t₀ is the initial time position, and T1 , T2, T3 and T4 are the predetermined time delay. In PPM, all the pulses should have the same predetermined amplitude, such as A.

In PAM, the amplitude of pulses is modulated based on the input bits 107. The order of Λ/-ary PAM, N, is the number of possible amplitudes for modulation. Λ/ should be an integer power of 2, for example 2, 4, 8, ... etc. Figure 2b illustrates all four possible pulses of a representative 4-PAM with N = 4. As shown in Figure 2b, the pulses 208, 210, 212 and 214 have different predetermined amplitudes, A₁, A₂, A₃ and A₄, respectively. The amplitude can be negative, such as A₃ and A₄. The amplitude polarity depends on the amplitude of the first half cycle of pulse. In PAM, all the pulses should have the same predetermined time position, such as t-,.

The time position and amplitude of pulses simultaneously can be modulated by combining both M-ary PPM and Λ/-ary PAM. Figure 3 illustrates a representative combination of 4-PPM and 2-PAM. The pairs (300, 302), (304, 306), (308, 310) and (312, 314) have different predetermined time positions, t₀ + Tλ, t_o + T2, t₀ +T3 and t₀ + T4 respectively, while the two pulses at each pair have different predetermined amplitudes but the same time positions.

As mentioned previously, for a modulation with more than one bit per symbol, the mapping issue becomes essential. The purpose of the pulse mapping table 108 (Figure 1) provides the predetermined time position, amplitude, frequency or phase of pulses to the modulator 110 (Figure 1) for modulating the input bits 107 (Figure 1). Figure 4 illustrates a representative mapping table 400 for the example in Figure 3, which is for a combination of 4-PPM and 2-PAM. For example, if a group of input bits "1 0 0" arrives, the amplitude -A and time delay T2 should be used in the pulse modulation.

Furthermore, an adaptive mapping table may be proposed based on the pulse mapping method presented in this invention. If the chip rate is fixed, different bit rates lead to different M levels in the /W-ary PPM or other modulations. For example, if the full bit rate reduces to a half rate, the 16-PPM becomes a 4-PPM. The adaptive mapping table may include full bit rate mapping and various lower bit rate mappings. A user or system can thereby select different pulse mapping from the adaptive mapping table when the bit rate is changed.

In Figure 1, the signal 122 is sent by the IR modulator 110 to the pulse mapping table 108 to inform the change of bit rate. After receiving the signal 122, the pulse mapping table 108 responds by sending a new pulse mapping rule 120 to the IR modulator 110 for mapping adjustment. The code generator 118 is employed to reduce interference. The codes employed may be either time hopping Pseudo Noise (PN) codes and/or polarity PN codes. Time hopping PN codes adjust the time delays of pulses and polarity PN codes adjust the polarities of pulse amplitudes. The code generator 118 is optional and can be removed in different embodiments.

The output signals from IR modulator 100 and code generator 118 are combined together and sent to the pulse generator 112. The generated pulses in the example embodiments may comprise time spaced pulses or time spaced bursts. The combination of both outputs is used to decide the final time position and amplitude of pulses. Another function of the pulse generator 112 is to generate a predetermined pulse shape, for example sinusoidal monopulse, Gaussian monopulse, Gaussian derivative monopulse, etc. In the example embodiments of the present invention, the sinusoidal monopulse is adopted in the figures for illustration, however, the pulse shape is not limited to sinusoidal monopulse in different embodiments. Finally, the modulated and coded pulses 114 are emitted by the transmitting antenna 116.

It will be appreciated by a person skilled in the art that a corresponding IR demodulator utilises the pulse mapping table 108 in demodulating the received Impulse Radio pulses, in example embodiments of the present invention.

The following example embodiment describes in detail the pulse mapping method for IR modulation that is comprised of a combination of M-ary PPM and Λ/-ary

PAM. However, this embodiment is not restricted only to the combination of pulse position and pulse amplitude. The pulse mapping method can be employed using a combination of any two or more of pulse amplitude, pulse position, pulse frequency and pulse phase. If TR scheme is involved, the amplitude, time position, frequency and phase of the data pulse relative to the reference pulse may be used in the combination.

In this first embodiment, the pulse mapping table 108 is constructed for a combination of M~ary PPM and Λ/-ary PAM. The pulse mapping method is designed to maximize the Hamming distance between neighboring pulses. Figure 5 is an exemplary flow chart of pulse mapping method. After startup, the system checks the orders of Λ/-ary PAM S500 and M-ary PPM in step S502, If N = 1, the mapping method for M-ary PPM-only is carried out in step 504 and processing terminates. If M = 1 , the mapping method for Λ/-ary PAM-only is carried out in step 504. The mapping methods for M-ary PPM-only and /V-ary PAM-only can be identical and combined together in the same block S504. The mapping method for Λf-ary PPM-only or Λ/-ary PAM-only will be described in more detail in the second embodiment below.

At S506, the M and N are checked again. If both M and N are integer powers of 2, for example 2, 4, 8, ... etc, the mapping operation continues. Otherwise the operation goes to end S528.

At S508, all possible waveforms are arranged into N rows and M columns according to the requirements. Firstly, all the pulses with same amplitude are grouped in the same row. There are in total N rows in the matrix, since the amplitude of pulse can have N values predetermined by /V-ary PAM. These N rows are arranged from top to bottom in a decreasing order of amplitude, such that A₁ > A₂ > ... > AN , where A,- (/ = 1 ,2, ... , Λ/) is the amplitude of the i-th row from the top. For example, the order of amplitude may be 2 > 1 > -1 > -2, if A₁ = 2, A₂ = 1 , A₃ = -1 and A₄ = -2.

Secondly, there are exactly M pulses with different time positions in each row. M different time positions are predetermined by M-ary PPM. In each row, the pulses are arranged from left to right in ascending order of time delay, such that T₁ < T₂ < ... < T_M , where 7}(/ = 1 , 2, ..., M) is the time delay of the j-th column from the left. Finally, a matrix of pulses with N rows and M columns is built up.

At S510, a comparison between M and N is made. If M ≥N, the partition on N rows is carried out. On the other hand, if M < N, the partition on M columns is carried out. The purpose of this comparison is to minimize the operational complexity. Since the partition stops when either N S512 or M S518 reaches 2, the number of partitions can be reduced by carrying out the partition on the smaller value of M or N.

If M > N and N > 2, N rows are partitioned equally into two halves in the example embodiments or into sets at step S514. A first prefix bit "0" is assigned to the upper set and "1" is assigned to the lower set. Then the value N is replaced by N/2 at step S516, since the number of rows in each set is N/2. If the new value of N is still larger than 2, the second partition of rows at the upper set and lower set are carried out separately, and a second prefix bit "0" or "1" is assigned to four subsets. In addition, the second prefix bit should be added just behind the first prefix bit. For example, the upper and lower subsets of the upper set, have two prefix bits "00" and "01 " respectively. The partition keeps running until N = 2.

The similar partition and prefix bit assignment are carried out on M columns at S520. The only difference is to divide M columns into left and right halves or sets, and assign prefix bit "0" to the left set and "1" to the right set. The partition keeps running until M = 2.

After the partition stops, all the remaining sets have either 2 rows and M columns (when M > N) or 2 columns and N rows (when M < N). For both cases, the same partition method is carried out to each set at S524, in the example embodiment.

An example in Figure 6 illustrates the partition of pulses in a representative combination of 8-PPM and 4-PAM. However, the pulse partition is not restricted only to this example. A matrix of pulses 632 with eight columns and four rows is constructed. The amplitudes of pulses in the same row are identical, and the time delays of pulses in the same column are identical. The pulse amplitudes follow the rule A₁ > A₂ > A₃ > A₄, and the pulse time delays follow the rule T₁ < T₂ < ... < T₈. Since M > N (8 > 4), the partition is carried out on the four rows. The matrix 632 is partitioned into two sets, where the set 634 contains the upper two rows and the set 636 contains the lower two rows. A prefix bit "0" is assigned to set 634 and "1" is assigned to set 636. After partitioning, the value of N is replaced by N/2. Since the new N is equal to 2, the partition stops.

Now the mapping method for set 634 is illustrated in Figure 7a. The same mapping method can be applied to set 636 in Figure 7b. Furthermore, the mapping method in Figure 7a is applicable to any set with either 2 rows and arbitrary M columns (when M > N) or 2 columns and arbitrary N rows (when M < N).

As shown in Figure 7a, the set 634 is partitioned into two groups 700 and 702. All the pulses in 700 and 702 are obtained, respectively, from the odd and even columns of set 634.

The group 700 is further divided into two subgroups 704 and 706 in a criss-cross manner as shown in Figure 7. The two pulses 600 and 602 from the upper left corner of group 700 are combined with the two pulses 612 and 614 from the bottom right corner of group 700 to become a subgroup 704. On the other hand, the two pulses 604 and 606 from the upper right corner of group 700 are combined with the two pulses 608 and 610 from the bottom left corner of group 700 to become a subgroup 706. The subgroup 704 is located on the left side of subgroup 706.

Similarly, the group 702 is further divided into two subgroups 708 and 710 in a criss-cross manner. The pulses 601, 603, 613 and 615 from upper left and bottom right corners are combined together to become a subgroup 710. The pulses 609, 611 , 605 and 607 from bottom left and upper right comers are combined together to become a subgroup 708. The subgroup 708 is located on the left side of subgroup 710.

At 712, all sixteen pulses are listed in one row. The first four pulses in 712 are obtained from the subgroup 704. The order of these four pulses remains the same as in subgroup 704. The second four pulses in 712 are obtained from the subgroup 706, but the order of these four pulses is reversed. The third four pulses in 712 are obtained from the subgroup 708 with the same order, and the last four pulses in 712 are obtained from the subgroup 710 with a reverse order.

Finally, log₂(2M) = 4 bits are assigned to each pulse of the row 712 from left to right in ascending order, that is "0000", "0001", ..., and "1111" as shown in Figure 7a or inversely "1111", "1110" , ..., and "0000". These two bit assignments can also be replaced periodically to suppress certain demodulating errors. In both bit assignments, the Hamming distance between two neighboring pulses in the same column is always 3, and the

Hamming distance between two neighboring pulses in the same row varies between 3 and 4. Therefore, the minimum Hamming distance is "3", which is larger than the

Hamming distance, "1", in conventional Gray mapping.

The mapping method presented in Figure 7a is also applied to set 636. It can be seen in Figure 7b that the same bit assignment appears in set 636. Following the operation at S526 in Figure 5, the bits assigned to each pulse in Figure 7a and 7b are grouped, respectively, with the prefix bit "0" and "1" assigned to set 634 and 636 in Figure 6. Finally, the pulse mapping table 108 for the example in Figure 6 is completely constructed. The pulse mapping method discussed above is not restricted only to the combination of pulse position and pulse amplitude. The method may be employed as a combination of any two of pulse amplitude, pulse position, pulse frequency and pulse phase. If TR scheme is involved, the amplitude, time position, frequency and phase of data pulse related to the reference pulse also may be used in the combination.

The following example embodiment describes pulse mapping for /W-ary PPM-only or Λ/-ary PAM-only. However, this embodiment is not restricted only to PPM and PAM. The pulse mapping method may also be extended to pulse frequency modulation or pulse phase modulation. If TR scheme is involved, the amplitude, time position, frequency and phase of data pulse relative to the reference pulse may be used.

The second embodiment of the invention is presented by constructing the pulse mapping table 108 for /W-ary PPM-only or Λ/-ary PAM-only S504. As mentioned in the first embodiment, the pulse mapping method is designed to maximize the Hamming distance between neighboring pulses. Figure 8 is an exemplary flow chart of the pulse mapping method for /W-ary PPM. The flow chart in Figure 8 is not restricted only to Λ/-ary

PPM and can be applied to the pulse mapping of Λ/-ary PAM and also other modulations, such as W-ary PPM with TR and Λ/-ary PAM with TR.

After startup S800, the system checks the order M of /W-ary PPM at S802 and S804. At S802, if M is of integer power 2, for example 2, 4, 8, ... etc, the mapping operation continues. Otherwise the operation goes to the end at step S814. At step S804, if /W = 2, a simple mapping method for 2-PPM-only is carried out at step S806, where a bit "0" is assigned to the pulse with a smaller time delay and "1" is assigned to the pulse with a larger time delay.

If /W > 2, all M pulses in the /W-ary PPM are arranged into one row with M columns at S808. These /W pulses have different time delays that are predetermined by /W-ary PPM.

The pulses are arranged from left to right in ascending order of the time delay, such that T_f < T₂ < ... < TM , where 7} (J = 1 , 2, ..., /W) is the time delay of the j-th column from left. The M pulses are partitioned into two sets at step S810 by grouping the odd columns of pulses as a set and the even columns of pulses as another set. Furthermore at step

S812, these two sets are partitioned into a few subsets, where each subset contains of just four pulses. After the partitioning, log₂ (M) bits are assigned to each pulse. The partitioning method and bit assignment at step S812 are explained in detail in Figure 9.

Figure 9 illustrates a representative bit assignment of pulses for an example 16- PPM. However, the flow chart in Figure 8 and bit assignment in Figure 9 are not restricted only to this example.

A row of pulses with 16 different time delays 916 is constructed at 917. The amplitudes of pulses in this row are fixed to A₁, but the time delays of pulses follow the rule T₁ < T₂ < ... < T₁₆. The row of pulses 917 is partitioned into two sets, where the set 918 contains all odd columns and the set 920 contains all even columns. Furthermore, the set 918 is partitioned into two subsets 922 and 924. The subset 922 contains the first four columns of set 918, and the subset 924 contains the last four columns of set 918. Similarly, the set 920 is partitioned into another two subsets 926 and 928. The subsets 926 and 928, contain the first four columns and last four columns of the set 920 respectively. In general, the set partitioning is done to divide the sets 918 and 920 into a few exclusive subsets in an order from left to right such that each subset contains exactly four pulses.

Next, the four pulses in the subset 922 are reordered and substituted into the row 930 as shown in Figure 9. Similarly, the four pulses in subsets 924, 926 and 928 are substituted into the row 930 with the same reordering manner. The first eight pulses of the row 930 are listed in 932 without change to their order. However, the last eight pulses of the row 930 are listed in 932 in a reverse order.

Finally, log₂ (M) = 4 bits are assigned to each pulse of the row 932 from left to right in ascending order at 934 that is "0000", "0001", ..., and "1111" as in Figure 9, or inversely "1111", "1110" , ..., and "0000". These two bit assignments can also be replaced periodically to suppress certain demodulating errors.

It will be appreciated by a person skilled in the art that numerous variations and/or modifications may be made to the present invention as shown in the specific embodiments without departing from the spirit or scope of the invention as broadly described. The present embodiments are, therefore, to be considered in all respects to be illustrative and not restrictive.

Claims

Claims:

1. A method of constructing a pulse mapping table for impulse radio pulses modulated with respect to n pulse parameters, n>1 , the method comprising: identifying all possible pulses for a given modulation scheme; and assigning m binary bits, m>2, to each pulse in a manner such that a Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is larger than 1.

2. The method as claimed in claim 1 , wherein the assigning of the m binary bits comprises: arranging all possible pulses into a n-dimensional matrix structure; partitioning the pulses in the n-dimensional matrix structure into sets; reordering the pulses from the sets based on respective rules associated with the respective sets into a series; and assigning m binary bits to each pulse in the series.

3. The method as claimed in claim 2, wherein the assigning of the m binary bits to each pulse in the series from the first pulse to the last pulse in the series is in ascending order following the sequential increment of the binary number formed by the m binary bits.

4. The method as claimed in claim 2, wherein the assigning of the m binary bits to each pulse in the series from the first pulse to the last pulse in the series is in descending order following the sequential decrement of the binary number formed by the m binary bits.

5. The method as claimed in any one of the preceding claims, wherein the n pulse parameter comprises one or a combination of any two or more of a group consisting of amplitude, time position, frequency and phase of the impulse radio pulses.

6. The method as claimed in claim 5, wherein the number of all possible pulses is the product of the number of different pulses based on one pulse parameter and the number of different pulses based on another pulse parameter.

7. The method as claimed in any one of the preceding claims, wherein a Transmitted Reference (TR) scheme is employed, and the pulse mapping table is applied to the modulation of a data pulse with reference to a reference pulse.

8. The method as claimed in any one of the preceding claims, wherein the pulse mapping table functions as an adaptive mapping table comprising a full bit rate mapping table portion and at least one lower bit rate mapping portions.

9. The method as claimed in any one of the preceding claims, wherein the

Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is at least (m - 1).

10. An impulse radio transmitter for transmitting impulse radio pulses modulated with respect to n pulse parameter, n>1 , the transmitter comprising: a data base having stored therein a pulse mapping table; and an impulse radio modulator utilising the pulse mapping table in modulating the impulse radio pulses; wherein the pulse mapping table is constructed by identifying all possible pulses for a given modulation scheme, assigning m binary bits, m≥2, to each pulse in a manner such that a Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is larger than 1.

11. The impulse radio transmitter as claimed in claim 10, further comprising a channel encoder for encoding a binary data input stream prior to the impulse radio Modulator.

12. The impulse radio transmitter as claimed in claim 11 , wherein the channel encoder applies an FEC code to the binary data input stream.

13. The impulse radio transmitter as claimed in any one of claims 10 to 12, further comprising an interleaver for spreading the binary data input stream prior to the impulse radio modulator.

14. The impulse radio transmitter as claimed in any one of claims 10 or 13 further comprising a pulse generator for generating the impulse radio pulses based on an output from the impulse radio modulator.

15. The impulse radio transmitter as claimed in claim 14, wherein the pulse generator generates the impulse radio pulses further based on an output from a code generator for encoding.

16. The impulse radio transmitter as claimed in any one of claims 10 to 15, wherein the assigning of the m binary bits comprises: arranging all possible pulses into a n-dimensional matrix structure; partitioning the pulses in the n-dimensional matrix structure into sets; reordering the pulses from the sets based on respective rules associated with the respective sets into a series; and assigning m binary bits to each pulse in the series.

17. The impulse radio transmitter as claimed in claim 16, wherein the assigning of the m binary bits to each pulse in the series from the first pulse to the last pulse in the series is in ascending order following the sequential increment of the binary number formed by the m binary bits.

18. The impulse radio transmitter as claimed in claim 16, wherein the assigning of the m binary bits to each pulse in the series from the first pulse to the last pulse in the series is in descending order following the sequential decrement of the binary number formed by the m binary bits.

19. The impulse radio transmitter as claimed in any one of claims 10 to 18, wherein the n pulse parameter comprises one or any combination of two or more of a group consisting of amplitude, time position, frequency and phase of the impulse radio pulses.

20. The impulse radio transmitter as claimed in claim 19, wherein the number of all possible pulses is the product of the number of different pulses based on one pulse parameter and the number of different pulses based on another pulse parameter.

21. The impulse radio transmitter as claimed in any one of claims 10 to 20, wherein a transmitted reference (TR) scheme is employed, and the impulse radio modulator applies the pulse mapping table to the modulation of a data pulse with reference to a reference pulse.

22. The impulse radio transmitter as claimed in any one of claims 10 to 21 , wherein the pulse mapping table functions as an adaptive mapping table comprising a full bit rate mapping portion and at least one lower bit rate mapping portion.

23. The impulse radio transmitter as claimed in any one of claims 10 to 22, wherein the Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is at least {m - 1).

24. An impulse radio receiver for receiving impulse radio pulses modulated with respect to n pulse parameter, n>1 , the transmitter comprising: a data base having stored therein a pulse mapping table; and an impulse radio demodulator utilising the pulse mapping table in demodulating the impulse radio pulses; wherein the pulse mapping table is constructed by identifying all possible pulses for a given modulation scheme, assigning m binary bits, m≥2, to each pulse in a manner such that a Hamming distance between neighbouring pulses with respect to each of the n pulse parameters is larger than 1.