Service bundling within enterprises and service cooperation outside enterprises are quite common on cloud. The correlations between services called Alliance Relation (AR) have great impact on the system QoS of service composition. Existing research have not fully and systematically considered the types of AR. Meanwhile, many works suffer from the low searching efficiency of finding the optimal matching AR for composite services. In this paper, we propose a novel approach called Q-ARIGraph-NSGA3, where we establish a multi-granularity optimization model on quotient space and an efficient matching search method in which a partition mechanism is applied to accelerate the generation of the graph. We expand the alliance relation types in service composition with service granulation and AR granulation employed concurrently. Extensive experiments are conducted on a real-world web service dataset, which demonstrate that our approach outperforms the state-of-the-art approaches in terms of effectiveness and efficiency.

The historical data of services usage shows us that both user preference and service quality are dynamic, and service quality has a certain influence on user preference. Due to the dynamic characteristics of both user preference and quality of service (QoS), how to recommend the best suitable services to users has become an urgent problem to be solved. But the most service recommendation approaches neglect the cyclical feature in dynamic preference model, and also neglect the impact of QoS on the user preference. We propose a time-aware recommendation method which considers the dynamic preference, the dynamic QoS and the impact of QoS on user preference comprehensively. Our experiments conducted on the real-world dataset WS-Dream, and the results show that our proposed approach outperforms several classical approaches and state-of-the-art approaches in terms of accuracy, recall, F1-value and Hamming distance.

User preferences are dynamically changing due to a variety of factors. Therefore, how to effectively extract the user's current preferences and accurately predict subsequent user preferences is one of the keys to improving service recommendation. This paper proposes a deep learning-based user preference prediction method, which combines user dynamic preference extraction and preference prediction. Firstly, the time-series LDA model is used to extract the user dynamic preference, then the LSTM neural network method is used to predict the user's preference of next time slice based on the user dynamic preference sequence. Finally, services are recommended to users based on predicted preference. The experimental results on real-world dataset show that the prediction performance of Time-aware-gate LSTM is slightly better than traditional LSTM, and the recommendation performance of the recommendation algorithm that takes preference prediction into account is better than those without preference prediction.

Multiple description coding (MDC) is an error-resilient source coding scheme that creates multiple bitstreams of approximately equal importance. The reconstructed signal based on any single bitstream has an acceptable quality. However, a higher quality reconstruction can be achieved with a larger number of bitstreams. We develop a multiple description video coding scheme based on a three-loop structure. We modify the discrete cosine transform structure to the matching pursuits framework and evaluate performance gain using maximum-likelihood (ML) enhancement when both descriptions are available. We find that ML enhancement works best for low motion sequences and results in gains of up to 1.3 dB in terms of average PSNR. Rate distortion performance is characterized. Performance comparison is made between our MDC scheme and single description coding (SDC) schemes over lossy channels, including two state Markov channels and Rayleigh fading channels. We find that MDC outperforms SDC in bursty slowly varying environments. In the case of Rayleigh fading channels, interleaving helps SDC close the gap and even outperform MDC depending on the amount of interleaving performed, at the expense of additional delay.

Multiple description coding (MDC) is an error resilient source coding scheme that creates multiple bitstreams of approximately equal importance. We develop a 2 description video coding scheme based on the 3 loop structure originally studied in Reibman et al. (1999). We modify the discrete cosine transform structure to the matching pursuits framework and evaluate performance gain using maximum likelihood (ML) enhancement when both descriptions are available. We find that ML enhancement works best for low motion sequences. Performance comparison is made between our MDC scheme and single description coding (SDC) schemes over two-state Markov channels and Rayleigh fading channels. We find that MDC outperforms SDC in bursty slowly varying environments. In the case of Rayleigh fading channels, interleaving helps SDC close the gap and even outperform MDC depending on the amount of interleaving performed, at the expense of additional delay.

We propose a new adaptive modulation technique for simultaneous voice and data transmission over fading channels and study its performance. The proposed scheme takes advantage of the time-varying nature of fading to dynamically allocate the transmitted power between the inphase (I) and quadrature (Q) channels. It uses fixed-rate binary phase shift keying (BPSK) modulation on the Q channel for voice, and variable-rate M-ary amplitude modulation (M-AM) on the I channel for data. For favorable channel conditions, most of the power is allocated to high rate data transmission on the I channel. The remaining power is used to support the variable-power voice transmission on the Q channel. As the channel degrades, the modulation gradually reduces its data throughput and reallocates most of its available power to ensure a continuous and satisfactory voice transmission. The scheme is intended to provide a high average spectral efficiency for data communications while meeting the stringent delay requirements imposed by voice. We present closed-form expressions as well as numerical and simulation results for the outage probability, average allocated power, achievable spectral efficiency, and average bit error rate (BER) for both voice and data transmission over Nakagami-m fading channels. We also discuss the features and advantages of the proposed scheme. For example, in Rayleigh fading with an average signal-to-noise ratio (SNR) of 20 dB, our scheme is able to transmit about 2 bits/s/Hz of data at an average BER of 10/sup -5/ while sending about 1 bit/s/Hz of voice at an average BER of 10/sup -2/.

We determine the bit-error rate (BER) of multilevel quadrature amplitude modulation (M-QAM) in flat Rayleigh fading with imperfect channel estimates, Despite its high spectral efficiency, M-QAM is not commonly used over fading channels because of the channel amplitude and phase variation. Since the decision regions of the demodulator depend on the channel fading, estimation error of the channel variation can severely degrade the demodulator performance. Among the various fading estimation techniques, pilot symbol assisted modulation (PSAM) proves to be an effective choice. We first characterize the distribution of the amplitude and phase estimates using PSAM. We then use this distribution to obtain the BER of M-QAM as a function of the PSAM and channel parameters. By using a change of variables, our exact BER expression has a particularly simple form that involves just a few finite-range integrals. This approach can be used to compute the BER for any value of M. We compute the BER for 16-QAM and 64-QAM numerically and verify our analytical results by computer simulation. We show that for these modulations, amplitude estimation error leads to a 1-dB degradation in average signal-to-noise ratio and combined amplitude-phase estimation error leads to 2.5-dB degradation for the parameters we consider.