By precoding multiple data bocks before cyclic-prefixed orthogonal frequency-division multiplexing (CP-OFDM) modulation, the rectangularly-pulsed CP-OFDM frame modulation structure has been adopted to achieve better error performance than CP-OFDM over linear time-varying multipath (LTV) channels. Typical embodiments include cyclic-preflxed orthogonal time frequency space (CP-OTFS), cyclic-preflxed orthogonal time sequency multiplexing (CP-OTSM), cyclic-preflxed affine frequency division multiplexing (CP-AFDM), cyclic-preflxed single carrier (CP-SC), and cyclic-preflxed anchor-aided single carrier (CP-ASC). These CP-OFDM frame modulations show various desirable but not simultaneously existing features, including resistance to double channel dispersion, low peak-to-average power ratio (PAPR), and high spectral compactness. By incorporating anchor-symbol insertion, selected mapping, and uniform data spreading into the CP-OFDM frame modulation structure, cyclic-preflxed anchor-aided time sequency multiplexing (CP-ATSM) is proposed in this paper. CP-ATSM performs comparably with CP-OTFS and CP-OTSM and outperforms CP-AFDM, CP-SC, and CP-ASC in uncoded error performance over LTV channels, while offering reduced PAPR compared to CP-OTFS, CP-OTSM, and CP-AFDM. Besides, CP-ATSM performs comparably with CP-ASC and outperforms CP-OTFS, CP-OTSM, CP-AFDM, and CP-SC significantly in spectral efficiency.

Initial time synchronization (ITS) is essential for the coherent demodulation in orthogonal time frequency space systems operating over the linear time-varying multipath (LTV) channels. Previous studies have developed efficacious preamble-frame-based ITS approaches using various select-the-largest identification rules to estimate the leading channel path delay over LTV channels exhibiting a strong leading path. However, such ITS approaches provide noticeable estimation inaccuracy under LTV channels without a strong leading path. In this paper, a comb-type preamble-frame (CPF)-based ITS system adopting discrete-time oversampling and a new select-the-leading-peak identification rule is proposed to provide improved ITS performance. Particularly, whether LTV channels exhibit a strong leading path or not, the improved CPF-based ITS system outperforms existing preamble-frame-based ITS systems significantly in estimation accuracy and robustness against Doppler spread.

In uplink cellular environments, random access preamble detection (RAPD) is achieved by identifying possibly multiple received requesting orthogonal frequency-division multiplexing (OFDM) signals carrying arbitrarily-chosen constant-amplitude (CA) preamble sequences through threshold comparison. Such threshold-based RAPD is analytically studied in terms of false alarm, outage, and false identification probabilities for the receiver operating under independent Rician/Rayleigh multipath channels and in the presence of time offsets among multiple received requesting OFDM signals. The RAPD system is shown to operate under less influence of inter-sequence interference when the preamble sequence set consists of mutually orthogonal CA sequences and contains as many ϖmin-permissible cyclically-shiftable constant-amplitude (CSCA) sequences as possible, where assigning (MPMA) and Zadoff-Chu sequence families, which aremin-permissible CSCA sequences are cyclically shifted from each other with a minimum cyclic-shift-distance ϖminin time domain. The performance characteristics are thus demonstrated for the RAPD systems adopting the modified phase-model-assigning (MPMA) and Zadoff-Chu sequence families, which are comprised of orthogonal and nonorthogonal, respectively, ϖmin-permissible CSCA sequence subfamilies. While constraining upper false-alarm and outage probability bounds, the system adopting the MPMA sequence family is shown to provide much smaller false-identification probability than the system adopting the Zadoff-Chu sequence family.

In multicell multiuser multiple-input multiple-output orthogonal frequency-division multiplexing (OFDM) cellular environments, simultaneous channel estimation (SCE) is achieved by first separating multiple received sounding OFDM signals carrying different constant-amplitude (CA) pilot sequences from all cells and then estimating channel responses jointly at each base station receiver. Multicell least-square-based SCE is analytically studied herein in terms of average mean-square-error for each base station receiver operating under independent uplink Rician/Rayleigh multipath channels and in the presence of time offsets among multiple received sounding signals. The SCE system is shown to suppress the mutual interference among various received sounding signals effectively when the pilot sequence subset for each cell consists of all CA sequences that are cyclically shiftable with a minimum cyclic shift distance ϖminin time domain (namely, ϖmin-permissible CSCA sequences), and CA sequences from different subsets are mutually orthogonal. Thus, a modified phase-model-assigning (MPMA) sequence family is constructed herein of orthogonal ϖmin-permissible CSCA sequence subfamilies, and compared with Zadoff-Chu and Yu-Lee sequence families which are both comprised of nonorthogonal ϖmin-permissible CSCA sequence subfamilies. The system adopting the MPMA family is demonstrated to provide higher estimation accuracy in SCE than the systems adopting Zadoff-Chu and Yu-Lee families.

Constant-amplitude (CA) sequences are efficacious in the construction of multicarrier sounding waveforms in order to facilitate simultaneous channel estimation (SCE) in uplink multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems. The permissible single-root length-N Zadoff-Chu sequences that are cyclically shiftable with a minimum cyclic shift distance (CSD) ${\varpi _{{\text{min }}}}$ in time domain are widely adopted to enable the separation of simultaneously received sounding OFDM waveforms and thus achieve accurate SCE when at most $\left\lfloor {N/{\varpi _{\min }}} \right\rfloor $ MIMO channels are estimated simultaneously. To enable the large-size SCE of more than $\left\lfloor {N/{\varpi _{\min }}} \right\rfloor $ MIMO channels, nonorthogonal multiple-root Zadoff-Chu sequences have to be adopted and this degrades the SCE performance remarkably due to heavy inter-sequence interference (ISI). In this paper, various CA sequence families are thus investigated for performance prevalence in the large-size SCE over independent Rician multipath channels. It is analytically shown that the large-size SCE can be achieved with less influence from ISI when the adopted CA sequence family is composed of orthogonal sequence subfamilies, each containing as many permissible cyclically-shiftable CA sequences as possible.

In this paper, coarse initial time synchronization (CITS) is investigated for orthogonal time frequency space systems operating over the linear time-varying multipath channel exhibiting real-valued delay and Doppler shifts on the delay-Doppler grid. By acquiring a comb-type preamble frame waveform carrying the discrete-Fourier-transform sequence, a particular CITS system is developed to estimate the start time of a received preamble frame on the leading channel path. Based on the cross-ambiguity between the received signal and local discrete-frequency-shifted comb-type preamble signals for discrete time shifts, the CITS system can operate over a wide timing acquisition range and reach a coarse accuracy scale under the influence of data self-interference. When the channel exhibits a strong leading path and wide ranges of Doppler shifts, the comb-type CITS system is shown to outperform the conventional CITS systems acquiring a linear frequency modulated preamble waveform and a hybrid data/preamble frame waveform, significantly in estimation accuracy and robustness.

Iterative decision-feedback equalization is commonly employed in single-carrier frequency-division-multiplexing to suppress the intersymbol interference (ISI) caused during the equalized symbol decision process while avoiding significant noise enhancement, and thereby provides satisfactory error performance. Conventional algorithms have been designed based on the same equalization criterion in all iterative steps and thus leave the room for performance enhancement in that the reduced levels of residual ISI in ensuing iterative steps are not taken into account. In this paper, an iterative equalization algorithm is proposed based on various equalization criteria in different iterative steps and shown to improve error performance in equalized symbol decision over Rician multipath channels.

Rectangularly-pulsed orthogonal time frequency space (OTFS) can be efficiently signaled by a frame of orthogonal frequency-division multiplexing (OFDM) block signals and outperforms OFDM in error performance over doubly selective channels. However, due to the discontinuity in rectangular pulse, the rectangularly-pulsed OTFS suffers large power spectral sidelobes decaying as $f^{-2}$, with f being the frequency in Hertz. In this paper, spectral precoding is integrated with rectangularlypulsed OTFS to make the compact spectrum rendering small sidelobes decaying as $f^{-2I-2}$ or faster. The characteristics of peak-to-average power ratio, achievable data rate, spectral efficiency, and error performance are also investigated for the spectrally-precoded OTFS.

In practical orthogonal frequency division multiplexing standards, preamble waveforms are adopted to facilitate timing synchronization, frequency synchronization and system parameter acquisition. Due to implementation simplicity, the crosscorrelation-based method is adopted to determine the transmitted preamble sequence in the literature. However, because of the lack of channel estimation, the crosscorrelation-based method suffers severe performance degradation in large delay spread channels. To solve the problem, a preamble sequence detection method based on the generalized likelihood ratio test (GLRT) principle is proposed and analyzed in the paper. Due to embedded channel estimation, the GLRT method is shown to provide the decision error performance robust to channel dispersion and outperform the conventional crosscorrelation-based method significantly in large delay spread channels, at cost of increased implementation complexity.

Cancellation symbols (CSs) are inserted into guard subcarriers to suppress the sidelobe power of the OFDM pilot waveform for channel estimation. Different from the conventional schemes where CSs are designed to suppress the sidelobe power of composite OFDM waveforms conveying data and pilot symbols, the inserted CSs depend on pilot symbols only and are shown to make the spectrally compact pilot waveform at the cost of less cancellation symbol power. The composite OFDM waveforms conveying spectrally precoded data symbols, optimum pilot symbols, and pilot-dependent CSs are shown to exhibit highly compact spectrum, while maintaining the channel estimation optimality on the quasi-static multipath channel.

The spectrally compact pilot waveform that exhibits small power spectral sidelobes and thus high spectral compactness while achieving accurate channel estimation is desirable for orthogonal frequency division multiplexing systems. However, existing pilot waveforms either suffer large power spectral sidelobes or can achieve high spectral compactness only when the length of the carried pilot sequence is limited to a constrained set of non-prime numbers, particularly those containing a large positive integer power of two as a factor. In this paper, new procedures are proposed for designing more pilot sequences with a broader choice of sequence lengths, so that spectrally compact pilot waveforms can be constructed to achieve optimal channel estimation on quasi-static multipath channels.

Selective-transmission (ST) strategy is proposed for non-orthogonal multiple access (NOMA) in downlink coordinated multipoint networks. Based on fixed and adaptive power allocation approaches, various ST-NOMA schemes are investigated and compared with conventional NOMA schemes under joint-transmission (JT) and hybrid joint- transmission and strong-channel-selection (JT+SC) strategies. Adaptive power allocation is also analytically derived to guarantee perfect user fairness. Performance results show that fair ST- NOMA outperforms conventional JT-NOMA and JT+SC- NOMA in user fairness, while providing high system throughput.

In practical orthogonal frequency division multi-plexing systems and standards, preamble waveforms are adopted to facilitate initial timing and frequency synchronization. These preamble waveforms are constructed from properly designed preamble sequences which possess exact or near zero periodic autocorrelation (ZAC) property to achieve robust synchronization and consist of exact or near constant amplitude (CA) symbols to achieve low waveform peak-to-average power ratio (PAPR). Unfortunately, preamble waveforms carrying conventional CAZAC sequences (which have ZAC property and consist of CA symbols) or near-CAZAC sequences result in large power spectral sidelobes and thus are not spectrally compact. In this paper, new near-CAZAC sequences are designed to construct spectrally compact OFDM preamble waveforms with robust timing and frequency synchronization and very low waveform PAPR comparable to those constructed from the conventional CAZAC sequences.

The uplink spectrum of cellular networks is often underutilized due to the asymmetric behavior of wireless data services. To enhance the efficiency, a full-duplex (FD) cellular system is proposed to reuse the uplink spectrum for concurrent uplink and downlink transmissions. By jointly designing transmit beamforming and power control, a weighted sum-interference power minimization problem is formulated for downlink users to manage the self- interference and multiuser interference with the guaranteed QoS of uplink users. An iterative joint transmit beamforming and power control (TB-PC) algorithm is then proposed to solve the joint design problem based on the dual decomposition. The performance is validated by computer simulation, and the proposed FD scheme can potentially improve the spectrum efficiency when a small serving cell is considered for downlink users, as compared with the conventional half- duplex scheme.

Pilot-aided channel estimation is essential to orthogonal frequency division multiplexing systems and standards. Under the system setup that data and pilot waveforms are processed individually for enhancing the spectral efficiency of the composite waveform, pilot patterns are designed here to jointly optimize channel estimation on quasistatic multipath channels and provide extremely high spectral compactness by suppressing spectral sidelobes. Specifically, a general constraint on pilot pattern is developed to endow the pilot waveform with very small spectral sidelobes and thus compact spectrum. Using the constraint, systematic multiple-stage design procedures are proposed to develop the desired pilot patterns achieving Cramér Rao bound for channel estimation on quasistatic multipath channels while providing fast decaying spectral sidelobes. Among the demonstrated pilot patterns achieving optimum channel estimation, the pilot waveforms using AI- and BI-typed pilot patterns developed from the procedures are shown to exhibit higher spectral compactness than conventional pilot waveforms. Besides, CI-typed pilot pattern that exploits extremely high spectral compactness without resort to channel estimation optimization is also developed. By sacrificing a slight loss in channel estimation, the pilot waveforms using CI-typed pilot pattern are shown to exhibit spectrums approaching the Nyquist spectrum.

The performance of Smart Data Pricing (SDP) highly depends on the accuracy and reliability of measuring network bandwidth usage. The existing Internet protocol uses the distributed control, and the transport network protocols run together inside the routers and switches. However, it is hard to manage and retrieve online and precise measurements from the networks due to the large number of traffic flows, and the complexity of deploying measurement infrastructures. In this paper, we propose the Location-based Time-dependent SDP (LTDP) mechanism where based on SDN, we apply different data fusion models for fusing the measured Internet traffic status from different locations. Emulation experiments are conducted to investigate the performance of the proposed LTDP.

In this paper, the orthogonal frequency-division multiplexing signal without inserting guard subintervals (NG-OFDM) is considered and combined with spectral precoding to achieve high spectrum compactness. In order to reject interblock interference among received NG-OFDM blocks over dispersive channels, a guard subinterval in each received block is removed at the receiver and this turns the resultant receiver to be nonorthogonal over subcarrier subchannels and thus subject to artificial intercarrier interference. It is shown that the proposed spectrally precoded NG-OFDM system can provide channel capacity very close to nonprecoded and spectrally precoded OFDM using cyclic prefix (CP-OFDM) when the removal guard ratio is small, and much higher channel capacity per unit bandwidth than spectrally precoded CP- OFDM conveying the same data rate provided that a small out-of-band power proportion is required.

Single-carrier frequency division multiple access with anchor-symbol insertion (ANC-SC-FDMA) is proposed in this paper. Fast sidelobe decaying constraints are developed for both interleaved and localized subcarrier allocation schemes. It is shown that the proposed ANC-SC-FDMA provides much higher spectral compactness than conventional SC-FDMA without trading off system complexity and increasing PAPR characteristics remarkably. By specifying zero anchor-symbol value, ANC-SC-FDMA can further improve error performance.

In this paper, the specific-band power minimization precoding technique is incorporated with spectrally precoded OFDM to suppress near spectral sidelobes more efficiently and thus improve the spectral compactness. When the total number of redundant symbols introduced by both spectral precoder and specific-band power minimization precoder is fixed, the proposed OFDM system can find a balance between near sidelobe power suppression capability and sidelobe decaying capability to meet the desired spectral performance by properly allocating redundancy to the two precoders.

In this paper, distributed detection is investigated in a wireless sensor network (WSN) using the hybrid multiple-access channel (MAC) which is a composite of orthogonal and nonorthogonal MACs. Specifically, a simple ideal sensing (IS) fusion rule is proposed, and the corresponding detection error probability upper bound and approximate system outage probability are analytically derived. Performance results show that the IS fusion rule provides nearly minimum detection error probability under high sensor observation signal-to-noise power ratios, and offers a diversity gain in system outage probability which is equal to the number of orthogonal channels. Moreover, the hybrid MAC WSN with IS fusion rule is shown to provide more performance choices than orthogonal and nonorthogonal MAC WSNs with IS fusion rule in terms of detection error probability and system outage probability.

Spectrally precoded orthogonal frequency-division multiple access (OFDMA) using cyclic prefix and interleaved subcarrier allocation (IOFDMA) is a useful rectangularly pulsed OFDMA signaling format for it exploits high spectral compactness and frequency diversity characteristics. In this paper, two new spectral precoders of the largest precoding rate are developed for IOFDMA in order to provide improved spectral compactness and reduce implementation complexity, respectively. Specifically, a new correlative precoder is developed and shown to minimize the dominant spectral coefficient among all correlative precoders of the largest precoding rate and thus can construct IOFDMA signals providing extremely high spectral compactness. Besides, a multiple-stage spectral precoding technique is proposed to construct a low-complexity orthogonally precoded IOFDMA with power spectral sidelobes decaying as f^{-2L-2} for L>1, where L is the number of stages. The corresponding orthogonally precoded IOFDMA signals conveying independent zero-mean data are shown to provide the same power spectrum as existing orthogonally precoded IOFDMA signals of the largest precoding rate while requiring much less complexity.

In this letter, the orthogonal multirate (MR) modulation signal is constructed by the orthogonal multiplexing of block-precoded component modulation waveforms occupying the Nyquist bandwidth and shown to outperform the conventional nonorthogonal MR modulation signal in error performance, while giving the same bandwidth efficiency.

In this paper, a new spectral precoding scheme is developed for constant-envelope orthogonal frequency-division multiplexing (CE-OFDM) signals to provide very small power spectral sidelobes. It is shown that the proposed spectrally precoded CE-OFDM scheme can provide higher spectral compactness than previously designed spectrally precoded CE-OFDM block scheme and continuous-phase CE-OFDM scheme.

In this paper, the distributed estimation of a deterministic scalar source by power-constrained wireless sensor networks with a fusion center is investigated over a hybrid multiple access channel Specifically, the maximum likelihood estimation is adopted, and the corresponding optimal and sub optimal schemes are developed to allocate sensor powers under a total power constraint. Particularly, a closed-form expression for sensor powers is obtained for suboptimal power allocation and a two-step procedure is proposed to solve the optimal power allocation problem numerically. It is analytically shown that the suboptimal power allocation scheme can offer asymptotically the same mean squared error performance as the optimal power allocation scheme when the number of sensors transmitting through the same coherent channel increases to infinity.

Spectrally precoded orthogonal frequency-division multiplexing with cyclic prefix (SP-CP-OFDM) and SP-CP-OFDM multiple access (SP-CP-OFDMA) are investigated in this paper. A general constraint on spectral precoding is developed for both SP-CP-OFDM and SP-CP-OFDMA signal formats with unconstrained guard ratios and arbitrary input data statistics to ensure very small power spectral sidelobes decaying asymptotically as f-2J-2, with J a predesigned integer-valued parameter. In the SP-CP-OFDM signaling format, new spectral precoders are devised to meet the constraint. It is shown that the proposed SP-CP-OFDM signals can provide much smaller spectral sidelobes than the previously designed SP-CP-OFDM signals with block partitioning and constrained guard ratios. In the SP-CP-OFDMA signaling format, a general procedure is proposed to facilitate the precoder design for SP-CP-OFDMA signals with arbitrary subcarrier allocation for each user and some spectral precoders are accordingly devised. It is shown that the proposed SP-CP-OFDMA signals can suppress sidelobe powers effectively and thus achieve higher spectral compactness than nonprecoded rectangularly-pulsed CP-OFDMA.

In this letter, a spectral sidelobe-decaying property of the Walsh-Hadamard code employed in multicarrier code-division multiple access (MC-CDMA) systems is exploited. It is shown that the sidelobe-decaying characteristic of the MC-CDMA signal can be properly described by the code index of the underlined Walsh-Hadamard codeword.

Spectrally precoded orthogonal frequency-division multiplexing (OFDM) and constant-envelope OFDM (CE-OFDM) are two modified OFDM formats that enhance the signaling features of OFDM, while allowing for the efficient implementation by discrete Fourier transform and the insertion of guard intervals to counteract channel dispersion. The spectrally precoded OFDM signal provides very small power spectral sidelobes decaying asymptotically as f-2J-2 with J integer-valued design parameter, thereby suppressing sidelobe powers effectively and achieving higher spectral efficiency than nonprecoded rectangularly-pulsed OFDM which has power spectral sidelobes decaying asymptotically as f-2. The CE-OFDM signal exhibits a unity peak-to-average power ratio, and thus enables the efficient use of transmitter's power amplifier. In this paper, a specific CE-OFDM block signaling format is proposed to incorporate spectral precoding with an aim to suppressing sidelobe powers more effectively than conventional nonprecoded CE-OFDM. The general constraint on spectral precoding is developed to ensure that the power spectral sidelobes of the spectrally precoded CE-OFDM block signal with arbitrary data statistics decay asymptotically as f-2J-2. Some spectral codes previously found for spectrally precoded OFDM are shown to meet the constraint and thus suited for spectrally precoded CE-OFDM. Specifically, spectral performance is analyzed for the spectrally precoded CE-OFDM block signal with independent and identically distributed data to illustrate the prevailing spectral advantage. The corresponding receiver is also developed to demodulate the spectrally precoded CE-OFDM block signal over dispersive channels.

A constant-power adaptive transmission technique adopting differentially coherent orthogonally multiplexed orthogonal phase modulation (DOMOPM) signals is studied for flat fading channels. Numerical results show that the constant-power adaptive DOMOPM system significantly outperforms the constant-power adaptive noncoherent frequency shift keying system in average spectral efficiency.

Spectrally precoded orthogonal frequency-division multiplexing (OFDM) is a promising rectangularly pulsed OFDM signaling format which can provide very small power spectral sidelobes, while allowing for efficient implementation by fast Fourier transform and guard interval insertion. In this paper, general constraints on spectral precoding are developed for OFDM signals with zero padding (ZP-OFDM) or cyclic prefix (CP-OFDM) to warrant the desirable spectral property that the power spectral sidelobes decay asymptotically as f-2K-2, where K is a preassigned positive integer. In accordance with the constraints, block partitioning is adopted to construct a general signaling format to facilitate the precoder design of the spectrally precoded CP-OFDM that can provide fast decaying sidelobes. New correlative and orthogonal precoders are also devised so that the desirable spectral property of fast sidelobe decaying is achieved with spectrally precoded ZP-OFDM and CP-OFDM signals.

Adaptive transmission technique adopting rectangularly-pulsed orthogonally multiplexed modulation (OMM) signals is studied for the flat Rayleigh fading channel. By choosing spectrally efficient OMM signals in accordance with channel status under an error rate constraint, constant-power adaptive OMM systems are shown to significantly outperform the conventional constant-power adaptive orthogonal frequency division multiplexing systems in average spectral efficiency.

Spectrally precoded orthogonal frequency-division multiplexing (OFDM) is a promising rectangularly pulsed OFDM signaling format which can provide very small power spectral sidelobes, while allowing for efficient implementation by fast Fourier transform and guard insertion. In this paper, a general constraint on spectral precoding is developed for OFDM signals with cyclic prefix (CP-OFDM) to warrant the desirable spectral property that the power spectral sidelobes decay as f-2K-2, with K a preassigned integer. Specifically, block partitioning is adopted to facilitate the precoder design of such spectrally precoded CP-OFDM. New correlative and orthogonal precoders are also devised. It is analytically shown that the proposed spectrally pre- coded OFDMs can provide better choices of spectral efficiency and average error performance than conventional spectrally precoded OFDMs.

An adaptive transmission technique adopting differentially coherent orthogonally multiplexed orthogonal phase modulation (DOMOPM) signals is studied for flat Nakagami-m fading channels. By choosing spectrally efficient DOMOPM signals in accordance with channel status under an error rate constraint, constant-power adaptive DOMOPM systems are shown to significantly outperform the constant-power adaptive noncoherent frequency shift keying systems in average spectral efficiency.

By bilinearly transforming the rectangularly pulsed multicarrier basis signals, new basis sets containing full-response continuous basis signals are developed for constructing spectrally efficient orthogonal frequency-division multiplexing (OFDM) signals with or without cyclic prefix (CP) or zero padding (ZP). Both multiple-product and linear-combination formats are presented for these basis sets. In the multiple-product format, the developed full-response basis signals are shown to be zero-valued at interval edges, and thus can be used to construct continuous-phase OFDM signals. It is analytically shown that the constructed continuous-phase OFDM signals exhibit relatively small power spectral sidelobes which fall off as f-4, and can thus provide much higher spectral efficiency than the traditional rectangularly pulsed OFDM signal. In the linear-combination format, the proposed basis signals are shown to be the linear combinations of rectangularly pulsed multicarrier basis signals, and thereby allow for the efficient realization of fast Fourier transform (FFT). Specifically, a spectrally precoded signaling structure is developed in conjunction with inverse FFT and guard insertion to construct continuous-phase OFDM signals with or without CP or ZP. The corresponding spectral decoders are also developed to enable zero-forcing (ZF) demodulation for spectrally precoded OFDM. With the zero-forcing demodulation, the spectrally precoded OFDM schemes are shown to outperform in average error-rate characteristics the conventional rectangularly pulsed and correlatively coded OFDM schemes on various delay-spread channels

A class of correlative codes is proposed to significantly improve the spectral performance of the rectangularly pulsed orthogonal frequency-division multiplexing (OFDM) signal with or without cyclic prefix or zero padding. It is analytically shown that the correlatively coded OFDM signal can achieve very high spectral efficiency while yielding an extremely small fractional out-of-band power.

In this paper, we deal with noncoherent detection of a digitally phasor block-modulated signal in the additive white Gaussian noise channel when a direct-current (DC) offset is present in the receiver baseband processing. By processing the received baseband signal block by block, a generalized linear transform is used to remove the offset prior to data detection, thereby releasing the succeeding detection process from the threat of DC offset. Operating on transform output blocks, a generalized maximum-likelihood scheme is developed for noncoherent data detection without a priori knowledge of channel amplitude and phase. When all the signaling blocks are confined within the space expanded by the basis vectors obtained from the onset-removal transform matrix, the proposed detection scheme can exploit the advantage of performing data detection and estimation on channel amplitude and phase jointly in the maximum-likelihood sense. It is analytically shown that the block detection scheme provides the bit error performance asymptotically approaching that of the corresponding ideal coherent phase-shift-keyed (PSK) detection in the absence of DC offset when the block length is increased. An iterative detection scheme is also modified from the block detection scheme to simplify the realization complexity. Both block and iterative detection schemes are shown to outperform the conventional training-sequence-aided PSK detection scheme under the same transmission throughput efficiency.

By substituting the on-off-keyed signaling for the orthogonal group signaling adopted in two recently reported modulation formats, namely the orthogonally multiplexed orthogonal amplitude modulation (OMOAM) (Chung, 2002) and the orthogonally multiplexed orthogonal phase modulation (OMOPM) (Chung, 2003), the orthogonally multiplexed on-off-keyed amplitude modulation (OMO/sup 2/AM) family and the orthogonally multiplexed on-off-keyed phase modulation (OMO/sup 2/PM) family are proposed and investigated in this paper. The power spectral density characteristics of both OMO/sup 2/AM and OMO/sup 2/PM signals constructed from typical band-limited and time-limited basis signal sets are studied in terms of fractional power containments. Based on the maximum-likelihood principle, the optimum schemes for coherently demodulating uncoded OMO/sup 2/AM and OMO/sup 2/PM signals on the additive white Gaussian noise channel are developed. Union bounds and approximate upper bounds are analyzed and verified by simulation to evaluate the bit error probability characteristics of the optimum schemes. From numerical results, the OMO/sup 2/AM and OMO/sup 2/PM schemes are found to provide other good choices of power and spectral efficiencies ranging among the efficiencies achieved by the OMOAM and OMOPM families. When constructed from the practically realizable basis set with rectangularly pulsed quadrature carrier signals, improvement over certain orthogonally frequency-division-multiplexing modulation schemes in both power and spectral efficiencies is also achieved by some OMO/sup 2/AM and OMO/sup 2/AM schemes.

Due to the emergence of digital libraries and the fast development of multimedia, more and more people use data hiding techniques to hide annotations or side information in images. We propose a novel data hiding mechanism based on pair-wise logical computation. The proposed mechanism can achieve the benefits of reversible and lossless reconstruction of the hidden data and the host image without utilizing any information from the original host image. It does not degrade the visual quality of the recovered host image after extracting the hidden data. Moreover, satisfactory data hiding capacity can be obtained simultaneously. The proposed data hiding mechanism is suitable for applying to the data hiding of images, scanned texts, figures, and signatures, especially for side information and annotation data embedding

The family of orthogonally multiplexed orthogonal phase modulations (OMOPMs) is investigated in this paper. Both coherent and differentially coherent signaling formats are considered and compared with the recently reported orthogonally multiplexed orthogonal amplitude-modulated (OMOAM) signals. When constructed from the same basis set, both OMOPM and OMOAM signals exhibit the identical power spectral density and, thus, the similar spectral performance characteristics. Based on the maximum-likelihood principle, coherent and differentially coherent detection algorithms are developed. Tight approximate upper bounds are derived and verified by simulation to evaluate the bit error probability characteristics of the maximum likelihood algorithms. It is shown that both coherent and differentially coherent OMOPM can provide a great choice of power and spectral efficiencies ranging among the efficiencies achieved by the classical modulations based on the orthogonal multiplexing technique.

A family of coherent orthogonally multiplexed orthogonally amplitude-modulated (OMOAM) signals is presented. The OMOAM signal is constituted by orthogonally multiplexing AI component signals, each constructed by a data-chosen group of L orthogonal pulse-amplitude-modulated basis signals. A generalized signal format is proposed to model the OMOAM signal in a unified way that the family contains not only classical orthogonally multiplexed modulations as special embodiments, but also a multitude of new modulations. A generalized optimum receiver for coherently demodulating OMOAM signals is developed and analyzed in terms of the bit error probability for the additive white Gaussian noise channel. The spectral characteristics of the OMOAM signals constructed from various time-limited and band-limited basis sets are studied in terms or the fractional out-of-band power containment. Several general trends of error and spectral performance characteristics are exploited.

This paper proposes a simple diversity scheme that makes use of nonlinear block coding, symbol interleaving, and differential phase modulation. The resultant signaling mechanism merits intrinsic time diversity within each codeword, which is resistant to correlated fading. The maximum-likelihood noncoherent receiver which operates over correlated Rayleigh fading is developed. Exact and approximate upper bounds are derived to assess the codeword-error probability characteristics of the proposed system.

This paper proposes a simple diversity scheme which operates on the frequency-nonselective correlated Rayleigh fading channel. The proposed transmitter consists of a nonlinear block encoder, a symbol interleaver and a differential phase modulator. The resultant signaling mechanism merits intrinsic time diversity within each codeword, which is resistant, to rapid channel fading. The maximum-likelihood noncoherent receiver is developed to detect the received diversity signal codeword by codeword. A union-bound expression is derived to assess the codeword error probability characteristics of the proposed diversity scheme and particularly used as the code merit figure for searching for good codes. It is analytically shown that the diversity systems employing the searched short-blocklength codes can offer remarkable gains in error performance on the correlated Rayleigh channel.

This paper proposes a pilot-symbol-assisted (PSA) signaling format that enables the differential detection of the differentially phase-encoded quadrature frequency/phase modulated (NDQFPM) signal using N orthogonal frequencies. The PSA-NDQFPM signal is transmitted on a frame-by-frame structure. The channel considered is an additive white Gaussian noise channel where the channel phase varies so slowly that it is assumed constant over one frame duration. Each frame consists of J NQFPM pilot symbols and I data blocks, each containing L NDQFPM symbols, with the first symbol differentially phase encoded with respect to pilot symbols. Based on the maximum-likelihood sequence estimation principle, the maximum-likelihood differential-detection (ML-DD) algorithm is developed to detect the PSA-NDQFPM signal on a block-by-block basis. Tight upper and asymptotic bounds are derived and verified by simulation to evaluate the bit error performance of the ML-DD algorithm. Both bandwidth and power efficiencies of PSA-NDQFPM with ML-DD are compared to NQFPM with coherent detection, constant-envelope NQFPM with coherent detection, constant-envelope NDQFPM (also known as NFSK/4DPSK) with ML-DD, and conventional 2DQFPM (also known as DQ/sup 2/PSK) with ML-DD. It is analytically shown that PSA-NDQFPM with ML-DD has lower power efficiency than NQFPM with coherent detection and NFSK/4DPSK with ML-DD, but still can provide noticeably higher power efficiency than DQ/sup 2/PSK with ML-DD when a medium frame length is used. Due to the use of pilot symbols, PSA-NDQFPM with a medium frame length has marginally lower bandwidth efficiency than its coherent counterpart NQFPM, but still yields significantly higher bandwidth efficiency than constant-envelope NQFPM and NFSK/4DPSK.

A noncoherent detection scheme for differentially encoded amplitude- and phase-shift-keyed data transmission, assisted with differentially phase-shift-keyed preamble signals, is presented and analyzed for an additive white Gaussian noise channel. It is analytically shown that, despite a controllable reduction in bandwidth efficiency, the proposed scheme yields an error performance approaching that of the quadrature amplitude modulation scheme with ideal coherent detection within a loss of approximately 1 dB.

This paper studies the differentially amplitude and phase-encoded (DAPE) quadrature amplitude modulation (QAM) transmission over correlated Rayleigh channels with diversity reception. Operating over two successive received symbols, the optimum and an asymptotic maximum-likelihood (AML) differentially coherent receiver are developed and compared with a conventional switched diversity combining (SDC) grid receiver. It is shown that the AML and SDC grid receivers are much simpler in complexity than the optimum receiver in that no channel side information is required in their realization. An exact expression of the bit-error probability (BEP) is obtained for the SDC grid receiver. Based on a union bound argument, a BEP upper bound for the AML receiver is also derived and verified by simulation. Numerical results on 16- and 64-point constellations show that the AML receiver exhibits an almost optimum performance and the SDC grid receiver with a small level of diversity is nearly optimum. It is also shown by simulation that the conventional equal-gain diversity receiver is almost optimal for demodulating a 16-point DAPE QAM signal.

We present the performance analysis of a fast frequency-hopped (FH) binary orthogonal frequency-shift keying acquisition receiver for communication against adverse environments. The receiver employs noncoherent, noise-normalized, matched-filtered (MF) correlation detection for rapid acquisition in the search mode. Our analysis includes four types of communication environments, namely additive white Gaussian noise (AWGN) channel, AWGN channel with partial-band noise jamming, fading channels, and fading channels with partial-band noise jamming. The considered fading channels include Nakagami-m, Rician, and Rayleigh amplitude models. Based on Beaulieu's (see ibid., vol.38, no.9, p.1463, 1990) convergent series approach, efficient analytical formulas are developed for performance evaluation. Example performance results for various environments are presented in terms of two acquisition probabilities, namely the detection probability and the false alarm probability of the noise-normalized MF detector. It is analytically shown that with a short MF correlation length and with a sufficiently large ratio of signal power to noise power the fast FH diversity combining yields noticeable performance improvement for environments with strong fading. When the MF correlation is lengthened, this improvement tends to fade away and the diversity combining results in performance loss.

A novel differentially coherent (DC) detection scheme for rapid code acquisition is proposed for a direct sequence (DS) spread-spectrum system being used to communicate over a fast Rayleigh fading mobile radio channel. The scheme processes the received signal at baseband using a complex differential detector with one-chip time delay and then performs a coherent partial correlation of the detector output with the product of a local DS code and its one-chip delayed phase to form a test sample. While still preserving the pseudo-noise attribute at its output, the differential detector is able to effectively suppress slowly fluctuating phase components due to fading and carrier frequency offset, thereby enhancing the in-sync correlation at the following correlator. The performance of this detection scheme is analysed based on a central-limit-theorem argument, and verified by simulation. Both serial and parallel acquisition systems employing the proposed DC detector are compared with the conventional parallel I-Q system by mean acquisition time performance. It is shown that the serial DC acquisition system significantly outperforms the parallel I-Q system in very fast Rayleigh fading environments.>

Multiple-hop autocorrelation (MHAC) processing is considered as a means of estimating the parameters of frequency-hopping signals in the presence of broadband thermal noise. It is shown that power sampling in the MHAC domain suppresses the dependence on the hopping frequency, carrier phase, and hop timing, while maintaining the hop-rate information. Moreover, when the signal's hop-rate is known a priori or can be estimated reliably, the multiple-hop combining of single-hop autocorrelation (SHAC) estimates can be employed to extract hop-timing information. Taking advantage of these features, we propose and analyze a hop-rate estimator and an epoch estimator, based on maximum-likelihood arguments. The conditional mean and variance of both estimators are derived analytically and confirmed by extensive simulation.>