Search

This article presents and analyzes fading measurements of optical signal in a terrestrial free space optical communication link using a commercial system. Unlike previous work, the link distance is very short (around 91 meters). Still, it was possible to observe significant scintillation index variations over a day. This behaviour agrees with previous results obtained in the literature for much longer links.

In orbital angular momentum (OAM) based free-space optical (FSO) communication systems, a CCD camera can be used at the reception side to capture images of the laser beam carrying the transmitted OAM modes. The tasks of extracting features from these images and identifying the transmitted modes are studied in this paper. The ability of machine learning algorithms to perform these tasks is explored. Laguerre-Gauss beams and turbulent channels are considered. Different modulation alphabets formed by using sets of superposed and multiplexed OAM modes are investigated. Appropriate choice of these alphabets can increase data rate transmission.

In this letter, channel capacity of dimmable visible light communication systems with M-ary pulse amplitude modulation (M-PAM) channel input alphabets is investigated. Mutual information is maximized considering both the probability input distribution and the M-PAM levels simultaneously. Schemes with non-equally spaced M-PAM levels are obtained as a result of this maximization process and they lead to gains in capacity when compared with previous schemes proposed in the literature.

A universal homophonic coding scheme is introduced which combines properties of differential encoding and homophonic symbol interleaving. The use of a uniform random interleaver is discussed and it is argued that asymptotically, when the interleaver length grows to infinity, the homophonic coding scheme generates a sequence of independent and uniformly distributed binary digits. For practical reasons, an interleaver with a complexity that grows linearly with its length is proposed. As a consequence, both the generation and the recovery of the digits produced by the proposed universal homophonic coding scheme are easy to implement. Validation tests for this new universal homophonic coding scheme were performed using the statistical test suite IR 6483 available from the National Institute of Standards and Technology, USA, and the corresponding computer simulation results are presented. The statistical tests demonstrate that the performance of the proposed universal homophonic coding scheme is at least as good as that of other published schemes, however with a significant reduction in plaintext expansion.

In this paper we show how a graph-based systemcan be designed for an N-FSK T-user multiple access channel withnoise. The channel is a frequency selective fast fading channelwith additive white Gaussian Noise. The system does not knowor attempt to determine the fading incurred at a given moment.The system is composed of a multiuser detector and T differentRepeat-Accumulate (RA) decoders. The decoders are optimizedas a set and consequently have the same parameters. A structuralvariation between the users’ encoders/decoders is sufficient toallow for multiple access and channel error correction, whileproviding rates close to channel capacity.

This paper considers a Gaussian two-user multiple-input multiple-output (MIMO) multiple access channel (MAC) with cooperating encoders. After defining a form of cooperation strategy, we derive achievable rate bounds by applying the results obtained by Willems for the MAC with generalised feedback. Then, we characterise an achievable rate region and also provide a lower bound for this region, which are very close. Although the achievable rate bounds are described by non-concave functions, the boundary of the achievable rate region could be generated by maximising a weighted sum of the rates. Numerical results obtained generally exceed the capacity region of a MIMO MAC without cooperation. Copyright © 2012 John Wiley & Sons, Ltd.

This paper addresses a practical aspect of coded on-off keying free-space optical communication systems: quantization of demodulator outputs in receiver. Receivers with and without knowledge of channel state information (CSI) are considered. Proposed quantization schemes for soft decision decoding with CSI are uniform with quantization intervals proportional to channel gains. The schemes are studied and compared under the criterion of the cut-off rate of the discrete channel created by quantization.

A method of analysis for comparing the performance of time-division-coded modulation and superposition-coded modulation schemes with unequal error protection has been proposed. The performance of practical channel codes was modeled heuristically by using a signal-to-noise ratio loss factor /spl lambda/ and it was shown that the superiority of superposition-coded modulation depends on the degree of inequality in protection prescribed. A different approach, using the channel cutoff rate R/sub 0/ instead of /spl lambda/, is proposed here. Results of the analysis of the performance of both approaches are compared, and they are also compared to those results obtained from practical design examples.

This paper analyzes a weighted symbol-flipping (SF) algorithm for nonbinary low density parity-check (NB-LDPC) codes over binary input AWGN (BI-AWGN) channels. The weighted SF algorithm has two main parts: the symbol-flipping function and the new candidate symbol rule. First, it is demonstrated that a rule for choosing the new candidate symbol based on absolute values of observed channel outputs is equivalent to a rule based on Euclidean distances. Then, it is verified that the weighting factor of flipping function has negligible impact on algorithm performance. Motivated by this fact, a SF decoding algorithm is proposed whose flipping function requires only syndrome values and flips symbols in parallel. It is observed that SF decoding outperforms WSF for q-ary codes with large q.

This article describes the construction of block codes derived from trellis termination of a binary catastrophic convolutional encoder. Two construction methods are considered: the Zero Tail (ZT) method and a modified Tail Biting (TB) method. 4-PSK block coded modulation (BCM) schemes are designed based on the constructed codes. These schemes outperform similar schemes derived from the best non-catastrupltic encoder.

The construction of block multidimensional M-PSK modulation codes with unequal error protection (UEP) is considered. An UEP profile for the Rayleigh fading channel is obtained by partitioning four-dimensional constellations that are Cartesian products of non-uniform 8-PSK modulations or are subsets of a product of uniform modulations. New codes are constructed by using the multilevel coding technique with very simple binary component codes. These codes are matched to the VSELP codec described in the IS-54 mobile system.

Desde que Gallager introduziu o algoritmo de decodificacao Bit-Flipping (BF) com decisao abrupta para codigos Low Density Parity Check (LDPC), outras duas variantes foram propostas por Sipser e Spielman para os codigos conhecidos como "Expander Codes". Posteriormente, uma versao da decodificacao BF por decisao suave conhecida como decodificacao Modified Weighted BF (MWBF), foi investigada. Esta tese propoe versoes modificadas dos algoritmos de Sipser e Spielman. Resultados de simulacoes para codigos Low Density Generator Matrix (LDGM) sistematicos, com comprimento longo mostraram um melhor desempenho da versao proposta. Adicionalmente, para um comprimento medio dos codigos LDGM, resultados de simulacoes mostraram um desempenho similar a decodificacao MWBF com a vantagem de nao ser necessario o uso de operacoes em ponto flutuante. Abstract

In this paper, we investigate the sum capacity of a noisy multiple access channel (MAC): the T-user N-frequency noiseless MAC in cascade with a noisy channel. The noisy channel includes frequency-selective Rayleigh fading and Additive White Gaussian Noise (AWGN). The receiver may use multiuser detection (MUD) or single user detection (SUD). We show that when using SUD, the uniform distribution achieves sum capacity. Considering MUD and assuming that all users have the same input distribution, we show through numerical optimisation, that, for most relevant values of T and N, a product of uniform input distributions maximises a tight upper bound on sum capacity. Numerical results were obtained and it was observed that, for fixed N, there is a value of T which maximises sum capacity. As expected, the results also show that there is a significant gain in sum capacity when MUD is used. Copyright © 2009 John Wiley & Sons, Ltd.

This paper presents a study about q-ary product turbo codes applied to FFH-CDMA systems. A method to implement the q-ary soft decoding is proposed. First, an adaptation of Chase decoding algorithm to satisfy q-ary symbols requirements is developed. The adaptation does not significantly increase decoding complexity when compared to the binary algorithm. Second, a modification of iterative (turbo) decoding allows appropriate feedback of decoded q-ary symbols and their reliabilities in each decoder iteration. Simulation results indicate a good performance/complexity trade-off of the q-ary turbo coded system when compared to other previous proposed systems.

For pt.I see ibid., p.Z000319-23, (2006). In part I of this paper, the channel capacity per dimension of a fast frequency-hoping code division multiple-access (FFH-CDMA) with unquantized detector outputs was evaluated and it showed that hard limiting of outputs can significantly degrade capacity. In this paper, a coded FFH-CDMA system whose maximum likelihood receiver works with unquantized outputs is considered. With a matched bit-to-symbol conversion in transmitter, it is shown that a proper design criterion is to maximize symbol Hamming distance. Based on this new criterion, a search algorithm was developed and many good convolutional codes were found. Analytical and simulation results corroborated the effectiveness of the proposed design criterion

This paper considers a Fast Frequency-Hopping Code Division Multiple Access (FFH-CDMA) system used in a mobile radio environment. In the receiver, Maximum Likelihood (ML) detection with un-quantized detector outputs implies the use of a nonlinearity of difficult implementation. To approximate this nonlinearity, we propose to use an unique perceptron. The system performance with this approximation is almost identical to ML detection. A system with un-quantized detector outputs can support a considerable greater number of simultaneous active users then a system with hard quantization of outputs. By calculating the capacity (per dimension) and the cutoff rate of the FFH-CDMA systems, we show that hard limiting of detector outputs can also significantly degrade system performance or reduce the number of active users.

Capacity of M-ary Amplitude and Phase-Shift Keying(M-APSK) over an Additive White Gaussian Noise(AWGN) channel that also introduces an unknown carrier phase rotation is considered. The phase remains constant over a block of L symbols and it is independent from block to block. Aiming to design codes with equally probable symbols, uniformly distributed channel inputs are assumed. Based on results of Peleg and Shamai for M-ary Phase Shift Keying(M-PSK) modulation, easily computable upper and lower bounds on the effective M-APSK capacity are derived. For moderate M and L and a broad range of Signal-to-Noise Ratios(SNR's), the bounds come close together. As in the case of M-PSK modulation, for large L the coherent capacity is approached., 4 pages, 5 figures, to appear in Proceedings of the 2005 IEEE International Symposium on Information Theory, Adelaide, Australia, September 4-9, 2005

This paper addresses the design of coded FFH-CDMA systems in a mobile radio environment. We show that hard limiting of detector outputs can significantly degrade the capacity of these systems. We also show that the main system design criterion is maximization of symbol Hamming distance. Analytical and simulation results corroborated the effectiveness of the proposed design criterion.

We present a construction of new differentially encoded/decoded rotationally invariant modulation schemes. The principle of this construction is the representation of a codeword from a rotationally invariant (RI) code in decomposed form. >

This paper proposes a coded modulation scheme for transmission of two classes of information on a broadcast AWGN channel. The scheme is based on multilevel coding with a nonstandard partition of a 32-QAM constellation. The QAM constellation consists of four groups of 8-PSK subconstellations instead of 8-QAM subconstellations. This results in a constellation with reduced peak-to-average power ratio. As component codes of the multilevel construction only one binary convolutional code and a punctured version of it are used. The nonstandard partition allows parallel decoding of the component codes in a multistage decoding algorithm. The error performance of the proposed scheme is obtained. In comparison with Wei's scheme with similar parameters, the proposed scheme achieves higher coding gains for the most important class at the cost of a small degradation in performance for the less important class. Although only results for a specific constellation are presented, the design principles can be easily extended to other QAM constellations.