Sen-Hung Wang

Novel Low-Complexity SLM Schemes for PAPR Reduction in OFDM Systems

The selected mapping (SLM) is a major scheme for peak-to-average power ratio (PAPR) reduction in orthogonal frequency division multiplexing (OFDM) systems. It has been shown that the complexity of the traditional SLM scheme can be substantially reduced by adopting the conversion vectors to replace the inverse fast Fourier transform (IFFT) operations. Each conversion vector is obtained by taking the IFFT of the phase rotation vector. Unfortunately, the corresponding phase rotation vectors of the conversion vectors in do not have equal magnitude, leading to significant degradation in bit error rate (BER) performance. This drawback can be remedied by adopting the perfect sequences as the conversion vectors. This paper presents two novel classes of perfect sequences, which are shown to be compositions of certain base vectors and their cyclic-shift versions. Then, two novel low-complexity SLM schemes are proposed by utilizing the special structures of the perfect sequences. The BER performances of both the proposed schemes are exactly the same as the traditional SLM scheme.

A Novel Low-Complexity Precoded OFDM System With Reduced PAPR

The computational complexity and peak-to-average power ratio (PAPR) of conventional precoded orthogonal frequency division multiplexing (OFDM) systems can be reduced using a T-OFDM precoded system based on the Walsh-Hadamard matrix. The present paper proposes a novel precoding scheme for further reducing the computational complexity and PAPR of T-OFDM. In the proposed scheme, the precoding matrix is combined with an inverse discrete Fourier transform to construct a new transform matrix at the transmitter. Notably, the transform matrix is both unitary and circulant, with each column being a perfect Gaussian integer sequence containing just four non-zero elements of {±1,±j}. A low-complexity receiver is additionally constructed for the proposed precoding scheme. A closed-form expression is derived for the bit error rate (BER) in T-OFDM and the proposed precoding scheme under frequency-selective fading channels. The simulation results for the BER are shown to be in good agreement with the mathematical derivations. In addition, it is demonstrated that T-OFDM and the proposed scheme have an equivalent BER performance when their precoding matrices are designed in such a way as to obtain full frequency diversity. However, the proposed scheme has a better PAPR performance and a lower computational complexity than T-OFDM.

Random access design for clustered wireless machine to machine networks

In this paper, we consider the infrastructure mode wireless machine-to-machine (M2M) network, such as the machine-type communication (MTC) in the LTE-Advanced, where there could be tens of thousands of machines within a macro cell and intending to access the same macro base station. Since most of the M2M communications are of low duty cycle and the machine traffic is usually bursty, the bottleneck for M2M communications is usually at the random access stage. In the LTE-Advanced, the access class barring (ACB) and extended access barring (EAB) have been proposed for the random access of MTC. For the one-shot random access attempts which correspond to the extremely low-duty-cycle scenarios where the network has enough time before the next traffic burst to resolve random access collisions, ACB and EAB have been shown to be sufficient if the machines can tolerate long access delay. However, when the machine traffic is recurrent with higher duty cycle, the ACB and EAB will fail to resolve collisions before the next wave of traffic comes in. To accommodate the M2M traffic and reduce the random access delay under limited random access resource, a clustered network structure is considered for which the machines in a macro cell are divided into clusters, and the machines belonging to a cluster communicate to the cluster head which then aggregates the traffic and relays to the macro base station. This clustered M2M network spatially reuses the random access resource. Thus it can increase the number of machines supported by the network and/or reserve more random access resource for the conventional (human) devices such that their qualities of service will not be affected too much by the M2M communications. Our analysis shows that full reuse of the random access resource among the clusters is feasible. In addition, simulations are conducted to study the random access performance of the clustered network, and to determine the number of clusters that should be formed and the number random access preambles that should be allocated to the machines when the total number of machines is given.

Gaussian Integer Sequences With Ideal Periodic Autocorrelation Functions

A Gaussian integer is a complex number whose real and imaginary parts are both integers. Meanwhile, a sequence is defined as perfect if and only if the out-of-phase value of the periodic autocorrelation function is equal to zero. This paper presents two novel classes of perfect sequences constructed using two groups of base sequences. The nonzero elements of these base sequences belong to the set {±1, ±j}. A perfect sequence can be obtained by linearly combining these base sequences or their cyclic shift equivalents with arbitrary nonzero complex coefficients of equal magnitudes. In general, the elements of the constructed sequences are not Gaussian integers. However, if the complex coefficients are Gaussian integers, then the resulting perfect sequences will be Gaussian integer perfect sequences (GIPSs). In addition, a periodic cross-correlation function is derived, which has the same mathematical expression as the investigated sequences. Finally, the maximal energy efficiency of the proposed GIPSs is investigated.

Perfect Gaussian Integer Sequences of Arbitrary Composite Length

A composite number can be factored into either N=mp or N=2n, where p is an odd prime and m, n ≥ 2 are integers. This paper proposes a method for constructing degree-3 and degree-4 perfect Gaussian integer sequences (PGISs) of an arbitrary composite length utilizing an upsampling technique and the base sequence concept proposed by Hu, Wang, and Li. In constructing the PGISs, the degree of the sequence is defined as the number of distinct nonzero elements within one period of the sequence. This paper commences by constructing degree-3 PGISs of odd prime length, followed by degree-2 PGISs of odd prime length. The proposed method is then extended to the construction of degree-3 and degree-4 PGISs of composite length N=mp. Finally, degree-3 and degree-4 PGISs of length N=4 are built to facilitate the construction of degree-3 and degree-4 PGISs of length N=2n, where n ≥ 3.

Low Complexity Transmitter Architectures for SFBC MIMO-OFDM Systems

Multiple-input multiple-output (MIMO) orthogonal frequency division multiplexing (OFDM) systems with space-frequency block coding (SFBC) have a high computational complexity since the number of inverse fast Fourier transforms (IFFTs) required scales in direct proportion to the number of antennas at the transmitter. This paper proposes to generate the SFBC encoded signals of the various antennas in time domain by exploiting the time-domain signal properties and signal correlations among the various transmitter antennas, achieving a significant reduction in computational complexity. In particular, it is demonstrated that the time domain SFBC encoded signals of the various antennas can be obtained from the time domain signal of the first antenna. Therefore, the proposed scheme requires only one IFFT irrespective of the number of transmission antennas. In addition, a low-complexity peak-to-average power ratio (PAPR) reduction scheme is presented based on the proposed transmitter architectures.

Further results on degree-2 perfect Gaussian integer sequences

A complex number whose real and imaginary parts are both integers is called a Gaussian integer. A Gaussian integer sequence is said to be perfect if it has an ideal periodic autocorrelation function (PACF) where all out-of-phase values are zero. Further, the degree of a Gaussian integer sequence is defined as the number of distinct non-zero Gaussian integers within one period of the sequence. Recently, the perfect Gaussian integer sequences have been found important practical applications as signal processing tools for orthogonal frequency-division multiplexing systems. The present article generalises the authors’ earlier paper by Lee et al. (2015) related to the Gaussian integer sequences with ideal PACFs. By the applications of two-tuple-balanced binary sequences and cyclic difference sets, a number of new degree-2 perfect Gaussian integer sequences with different periods are obtained.

Novel MC-CDMA system using fourier duals of sparse perfect Gaussian integer sequences

Serious multiple access interference (MAI) exists in uplink transmission of a multi-carrier code division multiple access (MC-CDMA) system in frequency-selective fading channels because orthogonality among codes cannot be restored. A novel MC-CDMA system is proposed in this paper to avoid MAI by employing Fourier duals of sparse perfect Gaussian integer sequences (SPGISs) as frequency-domain spreading codes. The SPGISs are obtained by linearly combining four base sequences or their cyclic-shift equivalents using nonzero Gaussian integer coefficients of equal magnitudes. The number of nonzero elements of SPGISs is 16 at most. The Fourier dual is the fast Fourier transform (FFT) of a SPGIS. Thus, when the Fourier duals of SPGISs are employed as frequency-domain spreading codes, the corresponding time-domain spreading codes are SPGISs. Furthermore, the modulated symbols of users should be properly allocated and the receiver architecture should be redesigned to avoid MAI. The computational complexity of the proposed MC-CDMA system is much lower than that of the traditional MC-CDMA system at the transmitter. Simulation results demonstrate both the bit error rate and peak-to-average power ratio performance of the proposed MC-CDMA system outperform those of the orthogonal frequency division multiple access, single-carrier frequency division multiple access, and the traditional MC-CDMA systems.

Gaussian Integer Sequences with Ideal Periodic Autocorrelation Functions

A Gaussian integer is a complex number whose real and imaginary parts are both integers. Meanwhile, a sequence is defined as perfect if and only if it has an ideal periodic autocorrelation function. This paper proposes a method for constructing sparse perfect Gaussian integer sequences (SPGISs) in which most of the sequence elements are zero. The proposed SPGISs are obtained by linearly combining four base sequences or their cyclic-shift equivalents using nonzero Gaussian integer coefficients of equal magnitudes. Each base sequence contains four nonzero elements belonging to the set {±1, ±j}. The number of nonzero elements of the constructed SPGISs depends on the choice of complex coefficients and cyclic shifts. However, each SPGIS has at most 16 nonzero elements, irrespective of the sequence length. A systematic investigation is performed into the properties of the SPGISs and their Fourier dual equivalents. Finally, a general expression is derived for a perfect Gaussian integer sequence (PGIS) of length 4n, where n is any positive integer and most of the sequence elements are nonzero.

A Low-Complexity Architecture for PAPR Reduction in OFDM Systems With Near-Optimal Performance

Selected mapping (SLM) schemes are widely used to reduce the peak-to-average power ratio (PAPR) in orthogonal frequency-division multiplexing (OFDM) systems. Various time-domain approaches have been proposed for reducing the number of inverse fast Fourier transform (IFFT) operations required to generate the candidate signals in traditional SLM schemes. However, the resulting time-domain-generated signals are somewhat correlated, and thus, the PAPR reduction performance is seriously degraded. Accordingly, this paper proposes a novel PAPR reduction method in which frequency-domain phase rotation, cyclic shifting, complex conjugate, and subcarrier reversal operations are all employed to increase the diversity of the candidate signals. Furthermore, to circumvent the multiple-IFFT problem, all of the frequency-domain operations are converted into time-domain equivalents. It is shown that the subcarrier partitioning and reassembling processes are key to realizing low-complexity time-domain equivalent operations. Moreover, it is shown theoretically and numerically that the computational complexity of the proposed scheme is significantly lower than that of the traditional SLM method and that the PAPR reduction performance is within 0.001 dB of that of SLM. Overall, the results indicate that among all of the low-complexity architectures proposed in the literature, the method proposed in this paper most closely approximates the PAPR reduction performance of the traditional SLM scheme.