A new detection method for data‐dependent superimposed training with equi‐spaced amplitude modulation or equi‐spaced square quadrature amplitude modulation is presented in this study. Symbol and bit error floor of the proposed detection method is analysed. To remove the error floor, a data coding method is also proposed. Analysis and simulation results show that the proposed detection method outperforms the existing methods, and the data detection performance can be further improved via data coding.

In mobile communications, navigation, and radar applications, good periodic binary sequences are normally used with matched-filter technique. However, from the viewpoint of signal processing, there is no need to only consider the filter coefficients as binary. This letter presents a theoretical bound for real-valued mismatched sequence sets with zero correlation zone (ZCZ) and then proposes a method for constructing sets of optimal mismatched ZCZ sequences with respect to the theoretical bound for application in mismatched filtering.

In this paper, based on a belief-propagation decoding strategy, a class of generalized parity-check codes called complex rotary codes is investigated. It is shown that, by using iterative sum-product decoding, the complex rotary codes have a much lower decoding complexity than Turbo codes, but have almost the same performance for the high code rate and short frame case (frame length< 500 bits). It is also shown that the prime block size of complex rotary codes is essential to achieve better performance because of its uniform checking characteristic.

The presence and availability of several independent wired and wireless networks within a communication environment has led to the channel adaptive multiple I/O (CAMIO) protocol being developed as an appropriate and novel technique for data transmission. The proposed system utilises an intelligent multiplexing technique that enables the transmission/reception of multiple data sources via several wired/wireless networks simultaneously. Real-time configuration and adaptation capabilities when data sources/channels are dynamically added/deleted/modified are described. Data prioritisation methods, combined with traffic control and congestion-reactive techniques, are presented. Simulation results demonstrate that the use of the CAMIO protocol over multiple transmission/reception channels increases the overall data transmission rate, handles the delivery of time-sensitive packets and maintains the quality of service, especially in situations where one or more communication channels fail to operate.

Number theoretic transforms, which operate in the ring or field of integers and use modular arithmetic operations, can perform the operations of convolution and correlation efficiently and without round off errors. Thus, they are very useful for the implementation of digital filters. One such a transform is the new Mersenne number transform (NMNT) which solves the problem of the rigid relationship between the transform length and the modulus, and also the lack of fast algorithms in traditional Mersenne number transforms (MNTs). In this paper, a new architecture that implements this transform is presented. The proposed architecture is scalable and easily parameterised. An implementation in an XC2V4000 FPGA chip has shown that this architecture can work at a frequency of up to 114MHz with a throughput rate of 228MS/s.

In the context of orthogonal frequency division multiplexing (OFDM), most papers define OFDM symbol synchronization as frame synchronization. In this paper, a frame is defined as a block of several OFDM symbols in the time dimension. The main contribution of this work is to propose a novel, low-complexity, frame synchronization method for multi-carrier systems that does not require any transmission overhead; this increases the channel capacity, power and bandwidth efficiency. Also, it enables a flexible multiple-access scheme to be implemented in OFDM, whereby each user can have different frame size. The proposed frame synchronization method is compared with a conventional method using a preamble. Power compensation and frame averaging techniques are proposed to improve the performance of the frame synchronizer. The performance of the proposed frame synchronizer is evaluated over a Rayleigh fading channel for orthogonal frequency division multiple-access (OFDMA) and multi-carrier code-division multiple-access (MC-CDMA) systems.

The contribution of this paper is to derive a novel figure of merit that can be measured in practice for symbol synchronisers. It is used to estimate the probability of symbol error, including the effect of symbol slip, for different frame and signaling alphabet sizes. The new figure of merit proves to provide a significantly more accurate characterisation than previously published methods. Specifically, by applying the novel figure of merit, the paper compares the symbol synchronisation performance of a conventional digital phase locked loop (DPLL) symbol synchroniser with that of modulation-derived synchroniser (MDS) under additive white Gaussian noise (AWGN) channel conditions.

It is widely acknowledged that the effect of impulsive noise is a major source of performance degradation within a wide range of communication systems. This is due to the fact that non-Gaussian interference is neglected within the system design philosophy for reasons of complexity and tractability. We directly address this problem using a novel 'de-noising' technique in which significant performance gains are achieved with low-complexity.

Superimposed codes for multiple-access communication in a binary adder channel are analyzed. The superposition mechanism used in this correspondence is ordinary addition. Each user is assigned a codeword from a superimposed code. It is proved that every constant-weight code C of weight w and maximal correlation c corresponds to a subclass of a disjunctive code D of order m>